Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

Summary

The manuscript by the Khmelinskii group reports that they have successfully constructed two conditional degron libraries of budding yeast for almost all proteins. For this purpose, the authors employed an improved auxin-inducible degron (AID2). Initially, they constructed yeast libraries by fusing HaloTag to the N- or C-terminus of proteins and found that C-terminal tagging is less likely to affect the location and function of proteins (Fig. 1). Based on this finding, the authors fused mNG-AID-3Myc or AID-3Myc (AID-v1 or AID-v2 library, respectively) to more than 5600 proteins and found that 4079 proteins were significantly depleted when cells were treated with 5-Ph-IAA (Fig. 2). A fitness defect was observed for over 60% of essential proteins, indicating the target depletion showed the expected phenotype in many cases (Fig. 3). Finally, the authors screened proteins required for maintaining viability in the presence of MMS, CPT and HU, and identified common proteins involved in DNA repair (such as RAD52 epistasis proteins) and other proteins specific for MMS, CPT or HU resistance (Fig. 4). Furthermore, the authors revealed that an ER membrane protein, Gsf2, is required for HU resistance, which was not found in previous studies with the YKO library because gsf2∆ cells in the YKP library had aquired a suppressor mutation (Fig. 4e).

Major comments

• In Figure S2a, the authors initially checked the growth of yeast cells expressing OsTIR1(F74G) under the GAL1 promoter, saying that "expression of OsTir1(F74G) from the strong galactose-inducible GAL1 promoter had a negligible impact on yeast fitness (page 3)". To me, the OsTIR1(G74G) expressing cells showed slightly slower growth compared to the control cells. Moreover, the cells expressing it under the very strong GPD promoter showed apparent slow growth, suggesting that OstIR1(F74G) overexpression caused a side effect. The authors should carefully evaluate the cells with GAL1-OsTIR1(F74G).
• Given the possibility that OsTIR1(F74G) overexpression might cause a growth problem, it is not appropriate to compare OsTIR1+ and OsTIR1- conditions for evaluating growth fitness (Fig. 2). As shown in Fig. S4b, it is more appropriate to compare the +/- 5-Ph-IAA conditions. Additionally, the 5-Ph-IAA concentration used in this study was not clearly mentioned in the method section and figure legends.
• The authors found that fitness defects were observed for over 60% of essential proteins (Fig. 3). In other words, depletion of the remaining 40% was not enough to induce growth defects. The authors should discuss how the current AID library can be improved to achieve better target depletion. Previous literature reported various possibilities, such as using a tandem degron tag and combining AID with the Tet promoter system (PMID 25181302, 26081484). Although optional, it would be wonderful if the authors would generate an improved library.

Minor comments

• 5-Ph-IAA is not auxin because it does not induce the auxin responses in plants (PMID 29355850). Therefore, the authors should be careful when they refer to 5-Ph-IAA and should not call it auxin.
• The availability of the HaloTag and AID libraries should be indicated.
• Page 3: "Finally, the extent of AID-dependent degradation varied with protein abundance, in that highly expressed proteins were more likely to be only partially degraded compared to lowly expressed ones (Fig. 2e, Fig. S2e)". Fig. S2e should be Fig. S2d, shouldn't it?

Significance

This paper is technically robust and well-conducted. It presents a comprehensive study showcasing the effectiveness of the conditional degron library. The HaloTag libraries will also be useful. The yeast libraries presented in this study will be invaluable for future screenings and studies across all aspects of yeast biology.

Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

Learn more at Review Commons

---

Reply to the reviewers

Response to Reviewer #1

Major comments:

1. * The connection between vB12 and MMA is weak, and the attempt to connect these pieces to PPI seems somehow forced. For instance, the authors do not convincingly demonstrate that MMA causes the PPI deficit. Furthermore, vB12 may rescue PPI independently of MMA. The authors should be more transparent about the lack of connection or causality between changes in metabolism and behavior.* We appreciate the reviewer's comment and acknowledge that we have not demonstrated causal relationships between increased MMA, PPI deficits in Tbx1+/- mice and their rescue by vB12. They are associations.

In the revised manuscript, we have clarified this in the Discussion, para.4, by adding the following phrase. “The results of our study do not prove a causal relationship between elevated brain MMA and PPI impairment, nor do they tell us whether rescue of the PPI impairment by vB12 occurs by reducing MMA”.

Regarding the comment of a weak connection between vitamin B12 and MMA, we respectfully disagree.

The biochemistry underlying the connection is outlined clearly in the Introduction, page 4, para. 2.

Patients with vitamin B12 deficiency typically exhibit elevated levels of MMA and administration of vitamin B12 (cobalamin) helps to normalize MMA values (Robert & Brown, 2003). Furthermore, several animal models with genetic alterations in the vitamin B12 pathway exhibit high levels of MMA. For instance, mice lacking the cobalamin transporter have increased MMA (Bernard et al., 2018). Additionally, mice lacking the mutase (Mut), which requires vitamin B12 as a cofactor for the conversion of methylmalonyl CoA to succinyl-CoA for entry into the Krebs cycle, demonstrate elevated levels of MMA and are unresponsive to vitamin B12 treatment (Peters et al. 2006). In the revised manuscript, we have cited these references (Introduction, page 4, para. 2).

Throughout the manuscript, an important control is missing: WT+ vB12 group. Data from this group should be added to Figures 1, 3, 4, 6, and Supplemental Figures 3 and 4 to show the effect of vB12 on WT mice.

All of the experiments reported in the original manuscript included this control group although it was not always included in the data analysis and therefore in the figures, as observed by the reviewer, In the revised manuscript all of the relevant figures and tables now include these data.

In Figures 1C and 3, data from the respective WT controls in the Df1 and Tbx1 cohorts should be shown.

• *

The wild type (WT) animals serve as the control group for both Tbx1+/- and Df/+ mice because they were littermates, obtained from matings between Df1/+ and Tbx1+/- mice. This has been clarified in the Materials and Methods, and in a new cartoon which has been added to Supplementary Figure 1 (1C) showing all of the animal groups used for the various studies (NMR, transcriptomics, behavior).

The Supplemental Table S1 includes 17 WT controls for Tbx1+/- and 6 WT controls for Df1/+, but Figure 1C includes only one group of 11 WT controls. For which group were those 11 WT controls?

Here are several examples of inconsistency in the data: "For this, we first performed a preliminary metabolome analysis using isolated whole brains of male and female Tbx1+/- (n= 18) and WT (n= 10) mice between one and two months of age. A set of metabolites was quantified in brain extracts by liquid chromatography tandem mass spectrometry (LC-MS/MS) (Supplementary Table 1)." Again, the number of mice the authors note in the manuscript does not match that shown in Supplementary Table 1 (Tbx1+/- (n=14) and WT (N=17). "We analyzed whole brain tissue isolated from Df1/+ and WT (control) mice (n = 5 per genotype)." Again, the numbers of mice do not match those in Supplementary Table 1, which notes 6 Df1/+ mice and 6 WT mice.

We apologize for these errors and inconsistencies in the text and tables, all of which have been corrected in the revised manuscript. In addition, we have added the aforementioned cartoon (Supplementary Figure 1C) and we have improved the presentation of the data (genotypes and treatments) in Supplementary Tables 1 and 2. We hope that these changes provide the expected clarity to the data.

MMA is the only metabolite that is similarly changed between Tbx1+/- and Df1/+ brains. This is an interesting observation. However, there is no other overlap in metabolic changes between these two mutants. This is a concern that requires clarification

We appreciate the reviewer's comment. The observation is not altogether surprising considering that the Df1 deletion includes at least 9 genes involved in metabolic pathways (cited refs (Maynard, 2008;Meechan, 2011; Devaraju and Zakharenko, 2017)) any of which might counteract or compensate for changes caused by Tbx1 mutation alone. In addition, heterozygosity for other genes in the deleted region (Df1 encompasses over 20 genes) might affect metabolic processes indirectly. In the revised manuscript we have added the following phrase to the Discussion, para.1 “Thus, even though the two mutants are genetically and metabolically very different, in Df1/+ mice, the MMA phenotype is not affected by heterozygosity of other genes in the deletion.

The authors mention that MMA is not changed in pre-term Tbx1+/- embryos, but no data are provided. What about MMA levels in Df1/+ embryos?

In pre-term Tbx1+/- and WT embryos MMA was undetectable. This is now stated in the final paragraph of the first section of the Results.

We did not measure MMA in Df1/+ embryos. It was not a goal of the study to compare the metabolome of these two genetically very different mutants. The MMA data on Df1/+ mice are presented because they show the potential relevance of this phenotype for the human disease, and they justify the use of the single gene (Tbx1+/-) mutants for studies into metabolism-related disease mechanisms. See also response to point 12

* * In some cases, the differences in metabolites (e.g., glutamine, glutamate, phosphoethanolamine, taurine, leucine, myo-inositol, and niacinamide) between the WT and Tbx1+/- mice is very minimal (Supplemental Figure 2). The y-axis scale should start at 0.

We have changed the y-axis settings where necessary

The vB12 is administered via two different regimens: 1) every 3 days for 28 days, at 4-8 weeks of age for metabolic measurements, and 2) twice a week for 2 months for PPI behavioral testing. Is there any reason the authors chose different protocols?

We apologize for the confusion, which was due to an oversight in the Materials and Methods section that has been corrected. The weekly injection regimen was the same for mice used in the behavioral and metabolic studies, but the treatment time was shorter for the metabolic studies, for practical reasons beyond our control; mice received vB12 injections twice a week, beginning at 4 wks of age and continuing until 8 wks or 12 wks of age for metabolic and behavioral studies respectively.

* * The authors should add the following references to the study: Long et al., Neurogenetics (2006), which shows no change in PPI in Tbx1+/- mice. This discrepancy compared with the current study results and those of Paylor et al., Proc Natl Acad Sci U S A (2006) should be discussed.

We have not cited the study by Long et al. because there are no obvious reasons for the discrepancy (age, mouse strain, sex) that could be discussed. Beyond this of course we cannot comment on data generated by another research group. Nevertheless, the presence of the PPI deficit in Tbx1+/- mice has been confirmed in two different Tbx1 alleles Tbx1 lacZ/+ and Tbx1ΔE5/+, by two different investigators, Dr. Richard Paylor using Tbx1 lacZ/+ mice (Paylor et al. 2001) and Dr. Elvira De Leonibus (co-author of this manuscript) using Tbx1ΔE5/+ mice, in two different countries (USA and Italy) in a rederived colony of mice.

• Figure 6B is a concern. The PPI decrease in the Tbx1+/- group appears to be driven by results from 3-4 mice. First, are those data statistical outliers? *

With all due respect, this is not the case. Eight Tbx1+/- mice, i.e., >50% of those tested had PPI values below the minimum observed in WT mice. The behavioural data were checked for the presence of outliers in each group using the Grubbs test, which yielded negative results. Our finding of PPI deficits in Tbx1+/- mice are in line with previously published data in Tbx1+/- and other animal models of 22q11.2 microdeletion (Paylor et al., 2006; Paylor and Lindsay, 2006; Stark et al., 2008), as well as in humans (Sobin et al., 2005).

Second, experiments in the same mice would be more informative. Do PBS-treated mutants recover PPI if they are treated with vB12 and vice versa? If the authors are concerned about the age difference, they also should include age-dependent effects on PPI.

We decline to perform the proposed experiment for the reasons described in section 4 of the Revision Plan

• *

Because vB12 treatment completely rescued the MMA level in Df1/+ mice (Figure 3), the authors should include a figure showing PPI test results in Df1/+ mice.

Vitamin B12 treatment fully rescued the MMA phenotype in both mutants (Figure 3). Whether it rescues the PPI defect in Df1/+ mice is not important for this study. We used Df1/+ mice as an entry point, in order to give validity to the pursuit of the MMA phenotype in the single gene mutant (Tbx1+/-), in which we expect that it will be easier to find disease mechanisms. For this reason, we focused our attention on identifying metabolic alterations in adult Tbx1+/- mice.

See also response to point 7.

*Figure 1A and B table: Did the authors mean Log2FC instead of FC? The authors also should present the *

*source data by adding supplemental tables that include raw data and normalized conversion, etc., as described in the multivariate statistical data analysis of the LC-MS/MS data. *

• *

The new Figures 1A, Figure 3 and the accompanying tables now state Log2FC. New Supplementary Table 1 presents the raw data that were normalized on the basis of the amount of protein in the samples, described and referenced in the Materials and Methods

"...we identified a new metabolic phenotype that was associated with reduced sensorimotor gating deficits in Tbx1+/- mice". Although the authors showed the PPI rescue by treating Tbx1+/- mice with vB12, that result alone does not prove the association of metabolic phenotype with sensorimotor-gating deficit; other supporting data are needed.

• *

This is perhaps a question of semantics; by associated we mean that the two phenotypes, metabolic alterations and reduced PPI were observed together

The authors stated, "Results showed that there were very few differentially expressed genes in Tbx1+/- vs WT brains, (n=22 out of 14535 expressed genes (Fig. 5 and Supplementary Tab. 2)". However, they described how 3 differentially expressed genes are involved in mitochondrial activity in the Discussion. The authors should describe those 3 genes and their relation to the metabolic change.

The results that the reviewer refers to have changed in the revised manuscript due to the inclusion of the control group WT +vB12 in the data analysis. The transcriptome analysis revealed that vB12 had a stronger impact than genotype, and as a consequence, the statistical analysis of all groups did not highlight minor differences between the two genotypes.

Figure 5B: The authors claimed that they detected similar transcription profiles between WT+vB12 vs. Tbx1+/-+vB12, comparing Tbx1+/-+PBS vs. Tbx1+/-+vB12. This is based on 947 genes being downregulated and 834 being upregulated, which is not appropriate. The authors should normalize those data to the numbers of genes upregulated and downregulated in WT+PBS vs. WT+vB12 respective groups.

We said that we detected similar transcription profiles in PBS-treated WT and Tbx1+/- brains; a WT+vB12 group was not present. The latter group is included in revised manuscript and the data reanalyzed comparing all groups.

See also response to points 2 and 15.

Minor comments

1. Supplementary Table S1 shows the identical MMA concentration "0.2" for 6 controls. Is this correct? This was an error that has been corrected; the value is 0.00 (not detectable).

* Remove the callout for Figure 1C at the end of the second paragraph in Results*.

This figure is no longer present

*There are multiple typos throughout the manuscript. *

Here are several examples:

1. * Fig1B graph- Df/+ => Df1/+* Figure changed in revised manuscript

2. "Together, the hydrophilic and lipophilic results revealed a group of 6 compounds that characterized the brain metabolic differences between Tbx1+/- and WT mice (Figure 2B, 2C)". Figure 2A should be included also. Corrected

3. "In support of this notion, is the finding that...(remove) Removed

4. Remove double periods: "The pathways found are depicted in Figure 2C' which reports the impact of each pathway versus p values.." Corrected

5. Panel labels in all figures are misplaced. Panel labels are aligned correctly

We have performed a spelling and grammar check on the text

"In support of this notion... at least nine orthologs are involved in mitochondrial metabolism". What are those 9 mitochondrial genes? Kolar et al., Schizophr Bull (2023) indicates that there are 8 mitochondrial genes within the 22q11.2 locus. The authors need to list these genes.

This reference, which is a review, has been cited in the Introduction, para.3 along with the genes.

The review presents nine mitochondrial genes which the authors divide into two groups, 1) Genes expressed in mitochondria (SLC25A1, TXNRD2, MRPL40, PRODH, and COMT) and 2) Genes that have been shown to have an impact on mitochondrial function (TANGO2, ZDHHC8, UFD1L, and DGCR8). In the abstract they mention only eight genes, the ninth gene COMT is mentioned in the text.

Reviewer #2 (Significance (Required)):

*The manuscript titled "Tbx1 haploinsufficiency causes brain metabolic and behavioral anomalies in adult mice which are corrected by vitamin B12 treatment" by Caterini et al. presents a comparative metabolomics study in the brains of mouse models carrying a heterozygous mutation in the transcription factor Tbx1. This mutation is contrasted with a chromosomal deficiency encompassing Tbx1, among other gene loci, known as Df1/+, which serves as a mouse model for 22q11 microdeletion syndrome. The primary and most significant finding of the study is that Tbx1 heterozygosity alone induces broad metabolomic alterations in the entire brain parenchyma, despite Tbx1 expression being confined to vascular endothelial cells. The authors leverage this observation to investigate the effects of dietary supplementation with vitamin B12, which alters the metabolome in a manner interpreted by the authors as rescuing or reversing the Tbx1 heterozygosity phenotype. This study holds promise for understanding the individual gene contributions to the penetrant behavioral phenotypes observed in Df1/+ and 22q11 affected subjects. This potential arises from the clear and consequential metabolic phenotypes described, notably the accumulation of methylmalonic acid.

However, despite the intriguing metabolic phenotypes, there are significant issues hindering incontrovertible conclusions. *

Response to Reviewer #2

Major comments

1. Despite the intriguing metabolic phenotypes, there are significant issues hindering incontrovertible conclusions. Chief among these problems is the experimental design's nature, where the effects of genotype and a pharmacological intervention, vitamin B12, are assessed. The current design overlooks the effects of vitamin B12 on wild-type animals in metabolic and behavioral measures, thus precluding the attribution of the effects of vitamin B12 to a rescue. See response to Reviewer 1 (point 2) who made the same criticism. This group is now included in the data analysis of the relevant experiments.

*An alternative explanation, consistent with the measurements, is that vitamin B12 modifies metabolites and transcripts irrespective of genotype. A suggestion of this possibility is the observed effect of B12 lowering glutamate levels in Tbx1 mutant tissue below those in wild-type brain tissue (Fig. 4C). *

This might be true for some metabolites. Indeed, we found 5 metabolites that responded similarly to vB12 in both WT and Tbx1 +/- mice. In contrast, three metabolites responded to vB12 in both WT and Tbx1+/- mice, but the response was more pronounced in Tbx1+/- mice. Finally, a group of eight metabolites was altered exclusively in Tbx1+/- mice after vB12 treatment, including inosine, glutamate and short-chain fatty acids (SCFAs), Figure 4 and Supplementary Figure 6. Thus, overall, our data suggest that with only a few exceptions, the metabolic response to vB12 treatment is genotype-dependent.

• *

This experimental design issue is exacerbated by the multitude of analytes measured by metabolomics, all collectively assumed to change as part of a common genotype-B12 interaction mechanism. This interpretation is feasible only if none of the analytes were to respond to B12 in wild-type animals.

• *

As specified above, the response to vB12 was genotype-dependent. The inclusion of the vB12-treated WT dataset should address this point.

A second major issue arises from the assertion that Tbx1 is exclusively expressed in mouse brain endothelial cells and not in brain parenchyma. A significant unresolved question is how a gene expressed solely in endothelial cells can alter the brain parenchyma metabolome and transcriptome. This issue remains unaddressed and is not sufficiently discussed. If this assertion holds true, then the observations bear great importance in understanding how Df1/+ causes brain phenotypes and, by extension, in human 22q11.

There are quite a lot of published data from the mouse demonstrating the brain endothelial-specific expression of Tbx1 and the lack of expression in other brain cell types. These include studies using reporter genes (Paylor, 2006; Cioffi, 2014), Tbx1Cre based cell fate mapping (Cioffi, 2014, Cioffi, 2022) and single cell whole genome transcriptions (Ximerakis et al., 2019); (https://portals.broadinstitute.org/single_cell/study/aging-mouse-brain). All are cited in the manuscript.

HOW the mutation of Tbx1 alters the brain metabolome and transcriptome will be the object of future studies, Currently, we do not have any data. At the reviewer’s request, we have extended the discussion of this point in the revised manuscript (Discussion, para. 4).

In this vein, the authors should consider that Tbx1 is not expressed in brain endothelial cells in humans and is minimally expressed in fetal astrocytes (see https://brainrnaseq.org/).

https://brainrnaseq.org provides a tool to evaluate the gene expression in the fetal brain. The sequencing was performed on fetal human brain tissue after elective pregnancy termination (4wks-9wks, it is not clear). Our analysis focuses on adult mice, which may contribute to observed differences.

Moreover, Yi et al. (2010) generated a gene expression atlas of human embryogenesis spanning from 4 to 9 weeks of gestational age, revealing downregulation of TBX1 during this timeframe. Conversely, in the normal adult human brain, TBX1 expression is identified in endothelial cells, as indicated in "The Human Brain Cell Atlas v1.0" presented for visualization and data mining through the Chan Zuckerberg Initiative’s CellxGene application, referring to the atlas ontology in Ding et al. (2016). * 3. A third major concern pertains to the general poor quality of the figures. Many figures appear to be directly exported from the software used for data analysis without proper curation. They are inadequately labeled, lack color codes to clarify differences (e.g., volcano plots), feature lettering fonts that are difficult to discern, and have lettering panels placed in awkward positions. Fig. 1 would benefit by the addition of a pathway diagram showing which metabolites are changing. Figure tables/spreadsheets either have sheets labeled in Italian or are empty. Collectively, the manuscript needs more careful data curation and presentation*.

Many of the figures and tables have been modified with respect to the original manuscript and issues of clarity and quality have been improved where necessary.

Other points for consideration are listed below. • The abstract results section does not mention the Df1 mutants at all, and overall the description of the results should be improved

Corrected • The abstract would benefit from defining vB12 before using the abbreviation

Corrected • The section of the Results describing MMA accumulation in the brain would benefit from

• *explaining the choice of 1 month of age for terminal experiments and the choice to use whole brains (are there particularly brain regions suspected to be affected?), * The majority of animals were 2 months of age at sacrifice (age and sex of individual animals are indicated in Supplementary Table 1). Young adult mice were the object of the study for the reason described in the first paragraph of the Results section, namely “Human studies of brain metabolism have mainly been conducted on children and adolescent patients. Therefore, in order to determine whether similar anomalies were present in the mouse models, we performed our studies on young mice between 1 and 2 months of age (Dutta et al., 2016)”.

This is also the age at which the behavioural phenotype has been demonstrated (Paylor et al., 2006), and therefore could, potentially be rescued by vB12 treatment. We do not have regional information pertaining to the adult brain.

2) describing any sex effects for Tbx1 mutants (and clarifying what data points for Tbx1 animals correspond to which sex), and 3) including what sex was used for Df1 experiments.

In preliminary experiment we analyzed males and females’ mice, before electing to use only males. To obtained reliable information about the impact of gender on metabolism and transcription we would have to use much larger numbers of animals. In Supplementary Table 1 pertaining to males and females are now indicated.

• The authors demonstrate that vB12 rescues PPI but use no other behavioral paradigms. It is possible that these mutations and/or vB12 could be impacting anxiety-like behaviors or other behavioral phenotypes. By only including PPI, the authors limit the interpretation of the "rescue" of this phenotype by vB12. * Reduced PPI was the only behavioral anomaly identified in Tbx1+/- mice that were subjected to a standard battery of behavioral tests (Paylor et al., 2006). *

* *Reduced PPI was the only behavioral anomaly identified in Tbx1+/- mice that were subjected to a standard battery of behavioral tests (Paylor et al., 2006).

Reviewer #3 (Evidence, reproducibility and clarity (Required)):

This is a terrific paper looking at influences of Tbx1 heterozygosity on metabolic phenotypes in mice. A weakness is that the locus on effects of B12 is totally unclear--could be neurovascular or even peripheral, but correcting this weakness might include study of Tbx1 conditional mutants, beyond the scope of this study.

• *

Reviewer #3 (Significance (Required)):* good significance

Only two minor suggestions. 'We selected to study primarily Tbx1 single gene mutants because it is the primary candidate disease gene". What is the basis for this statement? Mouse +/- mutants in Mrpl40, Txnrd2, ProhD, and probably others have shown brain phenotypes.*

• *

The basis for TBX1 being considered as the primary candidate disease gene is the finding of TBX1 point mutations in patients who have the full spectrum of clinical phenotypes associated with 22q11.2 deletion syndrome without the chromosomal deletion, namely, congenital heart defects, immune defects, facial dysmorphism, learning defects and developmental delay. Similarly, in the mouse, Tbx1 mutation recapitulates the phenotype observed in multigene deletion mutants, such as Df1/+ mice.

We do not say (or think) that heterozygosity of other genes from del22q11.2 does not contribute to the disease, but mutations of other genes have not been found in individuals with a 22q11.2DS phenotype but without the chromosomal deletion.

In the discussion, the authors could close the loop on low glutamine could result in lower gaba in inhibitory interneurons, and its correction with B12 could restore gaba levels.

Discussion, para. 3. We thank the reviewer for comment. However, the GABA concentration is not altered in Tbx1 haploinsufficient brains; it is only upregulated by Vitamin B12. Therefore, this assumption may be very speculative. Due to differences in the release and reabsorption rates of the three compounds (glutamine, glutamate, and GABA), correctly evaluating the glutamine-glutamate cycle requires separating astrocytes from neurons. We have only discussed the upregulation of glutamate and the GABA response to Vitamin B12, which may counteract the excess of glutamate.

1. __4__Description of analyses that authors prefer not to carry out Please include a point-by-point response explaining why some of the requested data or additional analyses might not be necessary or cannot be provided within the scope of a revision. This can be due to time or resource limitations or in case of disagreement about the necessity of such additional data given the scope of the study. Please leave empty if not applicable.

---

Reviewer 1, point 11. The PPI decrease in the Tbx1+/- group appears to be driven by results from 3-4 mice. First, are those data statistical outliers?

Second, experiments in the same mice would be more informative. Do PBS-treated mutants recover PPI if they are treated with vB12 and vice versa? If the authors are concerned about the age difference, they also should include age-dependent effects on PPI.

We are unable to perform this experiment because, as stated in the manuscript, the mice were sacrificed at the end of the experiment and the brains preserved for histological analysis (not part of this study). The generation of mice for new experiments would take about one year. With all due respect, we do not believe that the data that would be obtained are sufficiently important to justify, ethically and economically, this work.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #3

Evidence, reproducibility and clarity

This is a terrific paper looking at influences of Tbx1 heterozygosity on metabolic phenotypes in mice. A weakness is that the locus on effects of B12 is totally unclear--could be neurovascular or even peripheral, but correcting this weakness might include study of Tbx1 conditional mutants, beyond the scope of this study.

Only two minor suggestions.

• 'We selected to study primarily Tbx1 single gene mutants because it is the primary candidate disease gene". What is the basis for this statement? Mouse +/- mutants in Mrpl40, Txnrd2, ProhD, and probably others have shown brain phenotypes.

• In the discussion, the authors could close the loop on low glutamine could result in lower gaba in inhibitory interneurons, and its correction with B12 could restore gaba levels.

Significance

good significance

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

The manuscript titled "Tbx1 haploinsufficiency causes brain metabolic and behavioral anomalies in adult mice which are corrected by vitamin B12 treatment" by Caterini et al. presents a comparative metabolomics study in the brains of mouse models carrying a heterozygous mutation in the transcription factor Tbx1. This mutation is contrasted with a chromosomal deficiency encompassing Tbx1, among other gene loci, known as Df1/+, which serves as a mouse model for 22q11 microdeletion syndrome. The primary and most significant finding of the study is that Tbx1 heterozygosity alone induces broad metabolomic alterations in the entire brain parenchyma, despite Tbx1 expression being confined to vascular endothelial cells. The authors leverage this observation to investigate the effects of dietary supplementation with vitamin B12, which alters the metabolome in a manner interpreted by the authors as rescuing or reversing the Tbx1 heterozygosity phenotype. This study holds promise for understanding the individual gene contributions to the penetrant behavioral phenotypes observed in Df1/+ and 22q11 affected subjects. This potential arises from the clear and consequential metabolic phenotypes described, notably the accumulation of methylmalonic acid.

• However, despite the intriguing metabolic phenotypes, there are significant issues hindering incontrovertible conclusions. Chief among these problems is the experimental design's nature, where the effects of genotype and a pharmacological intervention, vitamin B12, are assessed. The current design overlooks the effects of vitamin B12 on wild-type animals in metabolic and behavioral measures, thus precluding the attribution of the effects of vitamin B12 to a rescue. An alternative explanation, consistent with the measurements, is that vitamin B12 modifies metabolites and transcripts irrespective of genotype. A suggestion of this possibility is the observed effect of B12 lowering glutamate levels in Tbx1 mutant tissue below those in wild-type brain tissue (Fig. 4C). This experimental design issue is exacerbated by the multitude of analytes measured by metabolomics, all collectively assumed to change as part of a common genotype-B12 interaction mechanism. This interpretation is feasible only if none of the analytes were to respond to B12 in wild-type animals. Given that this experimental design recurs across a substantial portion of the paper, it may constitute a factor that undermines the strength of the conclusions.

• A second major issue arises from the assertion that Tbx1 is exclusively expressed in mouse brain endothelial cells and not in brain parenchyma. A significant unresolved question is how a gene expressed solely in endothelial cells can alter the brain parenchyma metabolome and transcriptome. This issue remains unaddressed and is not sufficiently discussed. If this assertion holds true, then the observations bear great importance in understanding how Df1/+ causes brain phenotypes and, by extension, in human 22q11. In this vein, the authors should consider that Tbx1 is not expressed in brain endothelial cells in humans and is minimally expressed in fetal astrocytes (see https://brainrnaseq.org/).

• A third major concern pertains to the general poor quality of the figures. Many figures appear to be directly exported from the software used for data analysis without proper curation. They are inadequately labeled, lack color codes to clarify differences (e.g., volcano plots), feature lettering fonts that are difficult to discern, and have lettering panels placed in awkward positions. Fig. 1 would benefit by the addition of a pathway diagram showing which metabolites are changing. Figure tables/spreadsheets either have sheets labeled in Italian or are empty. Collectively, the manuscript needs more careful data curation and presentation.

Other points for consideration are listed below:

1) The abstract results section does not mention the Df1 mutants at all, and overall the description of the results should be improved

2) The abstract would benefit from defining vB12 before using the abbreviation

3) The section of the Results describing MMA accumulation in the brain would benefit from 1) explaining the choice of 1 month of age for terminal experiments and the choice to use whole brains (are there particularly brain regions suspected to be affected?), 2) describing any sex effects for Tbx1 mutants (and clarifying what data points for Tbx1 animals correspond to which sex), and 3) including what sex was used for Df1 experiments.

4) The authors demonstrate that vB12 rescues PPI but use no other behavioral paradigms. It is possible that these mutations and/or vB12 could be impacting anxiety-like behaviors or other behavioral phenotypes. By only including PPI, the authors limit the interpretation of the "rescue" of this phenotype by vB12.

Significance

The manuscript titled "Tbx1 haploinsufficiency causes brain metabolic and behavioral anomalies in adult mice which are corrected by vitamin B12 treatment" by Caterini et al. presents a comparative metabolomics study in the brains of mouse models carrying a heterozygous mutation in the transcription factor Tbx1. This mutation is contrasted with a chromosomal deficiency encompassing Tbx1, among other gene loci, known as Df1/+, which serves as a mouse model for 22q11 microdeletion syndrome. The primary and most significant finding of the study is that Tbx1 heterozygosity alone induces broad metabolomic alterations in the entire brain parenchyma, despite Tbx1 expression being confined to vascular endothelial cells. The authors leverage this observation to investigate the effects of dietary supplementation with vitamin B12, which alters the metabolome in a manner interpreted by the authors as rescuing or reversing the Tbx1 heterozygosity phenotype. This study holds promise for understanding the individual gene contributions to the penetrant behavioral phenotypes observed in Df1/+ and 22q11 affected subjects. This potential arises from the clear and consequential metabolic phenotypes described, notably the accumulation of methylmalonic acid.

However, despite the intriguing metabolic phenotypes, there are significant issues hindering incontrovertible conclusions.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

Summary:

Using mass spectrometry, nuclear magnetic resonance spectroscopy, and transcriptomics analyses, Caterino et al. identified metabolic changes in the brains of young adult Df1/+ mouse models of 22q11.2 deletion syndrome (22q11DS). They observed similar changes in Tbx1+/- mice, as Tbx1 is one of the genes encoded inside the 22q11.2 microdeletion. Among the metabolites identified, methylmalonic acid (MMA) showed the most prominent increase. Systemic vitamin B12 (vB12) administration rescued the MMA increase and had strong effects on the brain metabolomes and transcriptomes in Tbx1+/- and Df1/+ mice. Finally, the pre-pulse inhibition (PPI) deficit observed in Tbx1+/- mice was partially rescued by vB12 treatment. As such, this study provides a description of metabolic changes caused by Tbx1 haploinsufficiency and suggests vB12 as a therapeutic avenue in 22q11DS. Although intriguing, this study suffers from several shortcomings that should be corrected or clarified before publication.

Major comments:

1. The connection between vB12 and MMA is weak, and the attempt to connect these pieces to PPI seems somehow forced. For instance, the authors do not convincingly demonstrate that MMA causes the PPI deficit. Furthermore, vB12 may rescue PPI independently of MMA. The authors should be more transparent about the lack of connection or causality between changes in metabolism and behavior.

2. Throughout the manuscript, an important control is missing: WT+ vB12 group. Data from this group should be added to Figures 1, 3, 4, 6, and Supplemental Figures 3 and 4 to show the effect of vB12 on WT mice.

3. In Figures 1C and 3, data from the respective WT controls in the Df1 and Tbx1 cohorts should be shown.

4. The Supplemental Table S1 includes 17 WT controls for Tbx1+/- and 6 WT controls for Df1/+, but Figure 1C includes only one group of 11 WT controls. For which group were those 11 WT controls?

5. Here are several examples of inconsistency in the data: "For this, we first performed a preliminary metabolome analysis using isolated whole brains of male and female Tbx1+/- (n= 18) and WT (n= 10) mice between one and two months of age. A set of metabolites was quantified in brain extracts by liquid chromatography tandem mass spectrometry (LC-MS/MS) (Supplementary Table 1)." Again, the number of mice the authors note in the manuscript does not match that shown in Supplementary Table 1 (Tbx1+/- (n=14) and WT (N=17). "We analyzed whole brain tissue isolated from Df1/+ and WT (control) mice (n = 5 per genotype)." Again, the numbers of mice do not match those in Supplementary Table 1, which notes 6 Df1/+ mice and 6 WT mice.

6. MMA is the only metabolite that is similarly changed between Tbx1+/- and Df1/+ brains. This is an interesting observation. However, there is no other overlap in metabolic changes between these two mutants. This is a concern that requires clarification.

7. The authors mention that MMA is not changed in pre-term Tbx1+/- embryos, but no data are provided. What about MMA levels in Df1/+ embryos?

8. In some cases, the differences in metabolites (e.g., glutamine, glutamate, phosphoethanolamine, taurine, leucine, myo-inositol, and niacinamide) between the WT and Tbx1+/- mice is very minimal (Supplemental Figure 2). The y-axis scale should start at 0.

9. The vB12 is administered via two different regimens: 1) every 3 days for 28 days, at 4-8 weeks of age for metabolic measurements, and 2) twice a week for 2 months for PPI behavioral testing. Is there any reason the authors chose different protocols?

10. The authors should add the following references to the study: Long et al., Neurogenetics (2006), which shows no change in PPI in Tbx1+/- mice. This discrepancy compared with the current study results and those of Paylor et al., Proc Natl Acad Sci U S A (2006) should be discussed.

11. Figure 6B is a concern. The PPI decrease in the Tbx1+/- group appears to be driven by results from 3-4 mice. First, are those data statistical outliers? Second, experiments in the same mice would be more informative. Do PBS-treated mutants recover PPI if they are treated with vB12 and vice versa? If the authors are concerned about the age difference, they also should include age-dependent effects on PPI.

12. Because vB12 treatment completely rescued the MMA level in Df1/+ mice (Figure 3), the authors should include a figure showing PPI test results in Df1/+ mice.

13. Figure 1A and B table: Did the authors mean Log2FC instead of FC? The authors also should present the source data by adding supplemental tables that include raw data and normalized conversion, etc., as described in the multivariate statistical data analysis of the LC-MS/MS data.

14. "...we identified a new metabolic phenotype that was associated with reduced sensorimotor gating deficits in Tbx1+/- mice". Although the authors showed the PPI rescue by treating Tbx1+/- mice with vB12, that result alone does not prove the association of metabolic phenotype with sensorimotor-gating deficit; other supporting data are needed.

15. The authors stated, "Results showed that there were very few differentially expressed genes in Tbx1+/- vs WT brains, (n=22 out of 14535 expressed genes (Fig. 5 and Supplementary Tab. 2)". However, they described how 3 differentially expressed genes are involved in mitochondrial activity in the Discussion. The authors should describe those 3 genes and their relation to the metabolic change.

16. Figure 5B: The authors claimed that they detected similar transcription profiles between WT+vB12 vs. Tbx1+/-+vB12, comparing Tbx1+/-+PBS vs. Tbx1+/-+vB12. This is based on 947 genes being downregulated and 834 being upregulated, which is not appropriate. The authors should normalize those data to the numbers of genes upregulated and downregulated in WT+PBS vs. WT+vB12 respective groups.

Minor comments:

1. Supplementary Table S1 shows the identical MMA concentration "0.2" for 6 controls. Is this correct?

2. Remove the callout for Figure 1C at the end of the second paragraph in Results.

3. There are multiple typos throughout the manuscript. Here are several examples:

a) Fig1B graph- Df/+ => Df1/+

b) "Together, the hydrophilic and lipophilic results revealed a group of 6 compounds that characterized the brain metabolic differences between Tbx1+/- and WT mice (Figure 2B, 2C)". Figure 2A should be included also.

c) "In support of this notion, is the finding that...(remove)

d) Remove double periods: "The pathways found are depicted in Figure 2C' which reports the impact of each pathway versus p values.."

e) Panel labels in all figures are misplaced.

f) "In support of this notion... at least nine orthologs are involved in mitochondrial metabolism". What are those 9 mitochondrial genes? Kolar et al., Schizophr Bull (2023) indicates that there are 8 mitochondrial genes within the 22q11.2 locus. The authors need to list these genes.

Significance

Caterino et al. provide the evidence of a change in brain metabolism associated with 22q11.2 deletion syndrome (22q11DS), which is a major risk factor for neuropsychiatric disease, especially schizophrenia. The authors attributed these changes to haploinsufficiency of Tbx1, which is located inside the 22q11.2 genomic locus, as Tbx1+/- mice and 22q11DS models (Df1/+ mice) demonstrated similar metabolic changes. Among these changes, the authors detected an abnormally high accumulation of methylmalonic acid (MMA), a toxic metabolite that negatively affects mitochondrial activity and glutamate uptake. The authors also showed that systemic vitamin B12 (vB12) administration in Tbx1+/- mice and 22q11DS mice returned MMA to its normal levels in the brains of both strains. Finally, administration of vB12 partially rescued a behavioral phenotype in Tbx1+/- mice. Specifically, vB12 improved pre-pulse inhibition (PPI) of the acoustic startle in these mutants. This study may potentially motivate a new direction of research toward elucidating the Tbx1-dependent metabolic mechanism and its connection to abnormal behavioral phenotypes in 22q11DS.

• This is the first thorough characterization of metabolic changes in the brain of mouse models of 22q11DS. This work may advance our understanding of the mechanism of this disease and schizophrenia in general. Previous studies mostly focused on cell-autonomous mechanisms in 22q11DS, but because Tbx1 is not expressed in the adult brain, this study strongly argues that non-cell-autonomous mechanisms are involved in the pathogenicity of 22q11DS. This description is the strongest part of the manuscript. However, as it stands, this study does not causally connect the metabolite abnormalities with aberrant behavior in 22q11DS models. The rescue of metabolic changes by vB12 is interesting, but the interpretation of the results is a bit speculative. There are also several missing controls and other limitations that preclude this reviewer from a more enthusiastic assessment of this manuscript. The manuscript has the potential to be much stronger if the authors add more experiments and analyze their data in a more appropriate manner.

• This manuscript describes metabolic abnormalities in 22q11DS models that are driven by the haploinsufficiency of the Tbx1 gene, which is located within the 22q11DS-related genomic region. This is an incremental advance in the 22q11DS field. The authors identified multiple metabolic abnormalities that would be of interest to researchers in the fields of psychiatry and translational and basic neuroscience. They also showed that some metabolic abnormalities are improved by vB12 treatment. They argue that vB12 improves PPI, the abnormality of which is associated with many psychiatric diseases, including schizophrenia. If confirmed, this will be a conceptual advance in the translational neuroscience field. However, making such a statement, exciting as it may be, would be speculative if based solely on the data presented in this manuscript..

• This reviewer has expertise in basic and translational neuroscience with focus on neurodevelopmental psychiatric diseases.

Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

Learn more at Review Commons

---

Reply to the reviewers

Reviewer1

I recommend combining Figures 1 and 3 either before or after the results shown in Figure 2, so the reader's expectation for quantification is immediately satisfied.

Response：

Thank you for your suggestion. In the revised manuscript, images of GFP::POP-1 in compound mutants are moved to Figure 3. The schematic diagram of the gonad (previously Fig. 1A) and GFP::POP-1 images in wild type are kept in Figure 1, as they are described in Introduction.

Major comments:

Delving into the figure legend of Fig. 3 and the normalization procedure described in the Methods "Quantification of POP-1 asymmetry in the Z1 and Z4 division" raised concerns. The method therein described is overly complicated but also neglects background subtraction. My first question about this method: what range of distances between daughters is measured in Z? These distances are not discussed in absolute terms, and this is important for our understanding of how much correction for tissue depth might be necessary, as L1s are very thin.

To check my understanding, the authors use as a control a nuclear-localized GFP driven in the somatic gonad precursors in otherwise wild-type worms by the sys-1 promoter. They observe that the regression on a log scale of anterior:posterior (and vice versa) Z1 and Z4 daughter fluorescence over the distance between the daughters in the Z plane is fit by y = −0.034x + 0.0148, which is practically a slope of 0 and an intercept of 0. This means that they observed an ~1:1 ratio (as log(1)=0) of fluorescence in the anterior and posterior daughters of otherwise wild-type worms, at least across the range of very small X values of relevant distances between daughters (again, the relevant range of distances really matters and should be presented), making the normalization seem unnecessary.

Response:

Normalization is essential to compare POP-1 signals between daughter cells since the signal intensities depend on the depth of cells. Depth differences between SGP daughter cells range from 0 to 7.5 micrometers. For example, when we input the maximum difference (7.5) into our correction equation y = −0.034x + 0.0148 (the logarithmically transformed linear regression equation), we get:

y = −0.034 * 7.5 + 0.0148 = -0.2402

To interpret this on the original scale, we apply the inverse logarithmic transformation:

10^(-0.2402) ≈0.575

This result indicates that even if GFP::POP-1 expression is the same in both cells, the depth difference alone can cause approximately a 1.74-fold (1/0. 575) difference in fluorescence intensity.

Similarly, if we use a median value of 3.5 micrometers as the depth difference, we get: y = -0.1042. After the inverse logarithmic transformation, this corresponds to a 0.787 or 1.27 (1/0.787) fold difference in fluorescence intensity.

Without normalization, we risk misinterpreting such differences in expression levels when in reality the expression is the same. Conversely, actual differences in GFP::POP-1 signal could be masked or overestimated due to the depth effect.

In the revised manuscript, examples of depth differences between SGP daughters are shown in Fig. 2S which is added in response to the comment of reviewer 2, asking images of lin-17 mom-5 animals.

In the revised manuscript, we explained the depth effects in the legend of Fig. 3 as follows.

"Since SGP daughter cells are often present at distinct focal planes, we normalized the depth effects on fluorescence intensities (see Materials and Methods for details) for the quantification shown in (B). The images in (A) and (C) are from animals with SGP daughters at similar depths."

Then based on this regression and 95% CI, the authors predict values that reflect true equivalence of fluorescence of POP-1::GFP in the two SGP daughters, compare the observed values to these predictions, and ultimately display in violin plots these differences of observed and expected. Correct?

Response:

Yes, your understanding is correct.

Is this complicated treatment the only way to detect differences in polarity of anterior and posterior daughters of Z1 and Z4? What happens if the authors measure GFP::POP-1 and calculate the following?

Z1p(MGV - background control from same focal plane)

---

Z1a(MGV - background control from same focal plane)

If this straightforward analysis shows asymmetric signal in the control that is made symmetrical or reversed in the mutants, the hypothesis would seem to be supported with a much more straightforward method. Samples could be analyzed separately in two bins by worm body position, which affects which cell is superficial in the sample. As it is, the Figure 3 Y axis label is hard to interpret without reading the methods at length, diminishing its impact.

Response:

Thank you for the suggestion. Your suggested calculation would be simple if we could assume that control signals (sys-1p::GFP::NLS or sys-1p::GFP::POP-1 in the same wild-type cell) on the same focal plane are the same among animals. However, since there are apparent variations in expression levels among individuals, your suggested method is not appropriate for evaluating differences in sys-1p::GFP::POP-1 intensities between the SGP daughter cells of the same animal.

Missing control: The sys-1 promoter-driven NLS-tagged fluorescent protein as a control to compare to the GFP::POP-1 is analyzed only in the wild-type, and apparently not in the mutants under consideration. Phillips et al. (2007) show that sys-1p transcriptional activity is equivalent between the SGP daughters in wild-type worms, but neither those results nor the method of normalizing to a sys-1p::GFP::NLS signal in this paper address the question of whether sys-1 promoter activity is equivalent in these cells in mutants upstream in the Wnt pathway. If the current method of normalization is to be used, it seems important to normalize to the sys-1p::GFP::NLS regression in each mutant background.

Response:

Thank you for your suggestion. We used sys-1p::GFP::NLS as a control to normalize depth effects, which should be the same across all genotypes because the GFP molecules in SGPs should be equally distributed between SGP daughter cells, not because sys-1 promoter activities are similar among them. Since SGP daughters divide within a short time (about 2 hours), it is likely that the fluorescence of newly synthesized GFP (maturation time of about 1 hour) in SGP daughters is neglectable compared to GFP inherited from the SGP cells. Similarly, sys-1p::GFP::POP-1 signals in SGP daughters reflect the distribution of GFP::POP-1 from SGPs rather than the transcriptional activities of the sys-1 promoter in the daughter cells. sys-1p::GFP::POP-1 or sys-1p::GFP::SYS-1 has been widely used to evaluate polarity of asymmetric divisions in a number of studies, none of which consider transcriptional differences of the sys-1 promoter in the daughter cells.

1. How was lin-17(mn589) generated? if this is the first report of this allele, full information on what the lesion is and how it was derived should to be reported.

Response:

Thank you for your question regarding the lin-17(mn589) allele. We would like to point out that the information about this allele is provided in the Methods section of the original manuscript as follows.

"lin-17(mn589) (gifted by Mike Herman) carries a mutation in the seventh cysteine residue of the CRD domain (C104Y). mn589 exhibits 47% Psa phenotype (indicating T cell polarity defects)."

The methods section lacks a description of how the mes-1 experiments were done, in terms of timing, duration, and temperature; mes-1(bn7) is a temperature sensitive allele.

Response:

Thank you for pointing out the lack of detailed methodology for the mes-1 experiments. The germless phenotype of mes-1 mutants is partial even at high temperatures. We have not performed temperature shifts to observe the phenotype. As per your suggestion, we added the following text to the Strains section:

"mes-1(bn7) is a temperature-sensitive allele with higher penetrance of the germless phenotype at 25{degree sign}C than at 15{degree sign}C, and was grown at 22.5{degree sign}C. The germless phenotype of mes-1(bn7) was observed by the absence of the mex-5::GFP::PH signal through direct observation of epifluorescence."

Minor comments

1. The paper lacks a discussion of precedent in the literature for Wnt-independent Frizzled activity; this is a major finding that is being undersold in the current version of the manuscript.

Response:

Thank you very much for appreciating out finding. We have added the following paragraph to the Discussion section:

"Wnt-independent functions of Frizzled receptors

We have shown that lin-17/Fzd functions in a Wnt-independent manner to control SGP polarity, since the missing DTC phenotype of lin-17; cwn-2 and lin-17 mom-5 was completely rescued by ΔCRD-LIN-17. In addition, SGP polarity is normal in the quintuple Wnt mutant that has mutations in all the Wnt genes (Yamamoto et al., 2011). In seam cells, Wnt receptors including LIN-17/Fzd and MOM-5/Fzd appear to have Wnt-independent functions for cell polarization, as seam cells are still mostly polarized in the quintuple Wnt mutants, while they are strongly unpolarized in the triple receptor mutants (lin-17 mom-5 cam-1/Ror) (Yamamoto et al., 2011). In Drosophila, Fz/Fzd has been primarily considered to function Wnt-independently to coordinate planar cell polarity (PCP) between neighboring cells (Lawrence et al., 2007), though Fz function can still be regulated by Wnt, as PCP orientation can be directed by ectopically expressed Wnt proteins (Wu et al., 2013).

          In Drosophila, Fz regulates PCP by interacting with other PCP components including Van Gogh (Vang). In C. elegans, we found that vang-1/Vang does not appear to function with LIN-17/Fz, since most vang-1 single mutants and cwn-1 cwn-2 vang-1 triple mutants have two gonadal arms (215/216 and 58/58, respectively). As Fz interacts with Disheveled (DSH) in Drosophila PCP regulation, in C. elegans, the Disheveled homologs DSH-2 and MIG-5 regulate SGP polarity (Phillips et al., 2007). Therefore, LIN-17 might regulate the DSH homologs in a Wnt-independent manner. "

Added Reference：

1. Lawrence PA, Struhl G, Casal J. (2007). Planar cell polarity: one or two pathways? Nat Rev Genet. 8, 555-563.
2. Wu, J., Roman, A.C., Carvajal-Gonzalez, J.M., & Mlodzik, M. (2013). Wg and Wnt4 provide long-range directional input to planar cell polarity orientation in Drosophila. Nature Cell Biology, 15(9), 1045-1055.

Important: I think "Fig. 6 Germ cell independent migration of germ cells" title is a typo; should be "Germ cell independent migration of DTCs"

Response:

Thank you for pointing out the typo. We corrected it in the revised manuscript.

This is a very important experiment! I think a greater description of the mes-1 phenotype would be helpful, since loss of germline was not 100% penetrant in mes-1(bn7) hermaphrodites in Strome et al., 1995. The legend says "Germless mes-1 phenotype was confirmed by the absence of the mex-5::GFP::PH signal in the gonad." Consider adding a few sentences to the results (or methods, from which the mes-1 experiments are currently missing) describing that only mes-1 animals that lacked germline fluorescence were analyzed for DTC migration.

Response:

Thank you for providing the context. To address the concerns, we made the following changes to our manuscript:

1. In the Results section, we revised the sentence "We found that 84% of DTCs (n = 90) in germless mes-1 animals..." to "Among mes-1 animals that lack germ cells, we found 84% of DTCs (n = 90)...".
2. We also modified the sentence "We noticed that some germless mes-1 animals..." to "We noticed that some mes-1 animals that lack germ cells...".

Please correct "secreting the Notch ligand LAG-2" this is a membrane-bound, not secreted ligand

Response:

Thank you for your comment. In the revised manuscript, we modified the relevant sentence in the Introduction section as follows:

"Firstly, DTCs function as niche cells for germline stem cells, inhibiting their entry into meiosis by expressing the Notch ligand LAG-2 (Henderson et al., 1994)."

Fig 1. The qualitative loss of polarity would be better depicted with

a grayscale image instead of green-on-black.

Response:

Thank you for your suggestion. The GFP::POP-1 images are raw images of the green channel of the confocal microscopy. We believe that SGP polarity is clearly depicted by them.

Fig. 3 the presentation of these violin plots is confusing. The central text that reads "normal polarity, loss of polarity, reversed polarity" with arrows looks like a second Y axis label attached to the Z4 plot. I recommend rearranging. Consider shading the top, bottom, and central regions and explaining the meaning of the shading in the legend.

Response:

Thank you for your suggestions regarding the presentation of Figure 3. In response to your feedback, we have made the following modifications:

First, we moved the text and arrows from the center to the right side of the figure, creating a clearer layout. As you recommended, we applied shading to the top, bottom, and central regions of the violin plots. Additionally, to explain the meaning of the shading, we added a new explanation to the figure legend. Specifically, we included the following text:

"Values within the 95% CI (between the red lines; light green regions) indicate symmetric localization. Values below the lower red line (light blue regions) indicate reversed localization, while values above the upper red line (light red regions) indicate normal localization."

We applied the same modification to Supplemental Fig. 1.

Reviewer 2

Major comments

1- Are the effects of combining the different Wnts with the lin-17 allele specific to the n3091 allele? It would be important to test another allele, for example the sy277 allele has a similar phenotype and is available at CGC. A null would be even better if it is viable. Alternatively, lin-17(RNAi) could instead be used if efficient enough. This is important since the n3091 allele could differentially alter the binding to the various Wnts, resulting in their distinct phenotypes in that background. However, these distinct phenotypes may not be relevant in a wild-type context.

Response:

Thank you for your insightful comment. The lin-17(n3091) allele contains a nonsense mutation at the 35th codon, located between the second and third cysteine residues in the CRD domain (Wnt binding domain) (Sawa et al 1996). Therefore, it is highly unlikely that the N-terminal protein of 34 amino acids produced in lin-17(n3091) can bind to Wnts. In the revised manuscript, we added the missing-DTC phenotype of lin-17(n671) cwn-2 animals, which show a similar phenotype to lin-17(n3091) cwn-2. n671 is a reference allele in WormBase and has a nonsense mutation. Although sy277 has a deletion in the N-terminal region, its phenotype is weaker than that of n3091 and n671 (Sawa et al 1996).

In the revised manuscript, we described lin-17(n671) cwn-2, in the Table 1, Table S1 and added the following sentence.

"We observed a similar phenotype in lin-17(n671); cwn-2 double mutants, confirming that this genetic interaction is not allele-specific."

2- In the lin-17; mom-5 double mutant which lacks DTCs, are Z1 and Z4 there but they do not express DTC markers, or are they never born? A lineage analysis should be presented. Also, are Z2 and Z3 still there on their own? Please show images.

Response:

Thank you for your comments. We quantified sys-1p::GFP::POP-1 signals in Z1 and Z4 daughter cells of lin-17 mom-5 and have not observed any animals lacking Z1, Z4 or germ cells. In the revised manuscript, as Fig. S2, we added images of sys-1p::GFP::POP-1 localizations in SGP daughters, along with germ cells in lin-17 mom-5 as well as in lin-17 cwn-1 egl-20 cwn-2, both of which were not shown in the original manuscript. In response to Reviewer 1's comment, we also included examples of depth effects on fluorescence intensities in Fig. S2 with images of different focal planes.

Fig. S2 is quoted it at the end of the following sentence.

"Then, we quantified the ratios (on a logarithmic scale) of sys-1p::GFP::POP-1 signal intensities proximal to distal daughter cells in various genotypes (Fig. 3A and Fig. S2)."

The loss of polarity phenotype of lin-17 mom-5 has been described in Phillips et al. We missed to cite this in the original manuscript. We added the citation in the revised manuscript.

"These asymmetries were strongly disrupted and weakly affected in lin-17 mom-5 double and lin-17 single mutants, respectively, as described previously (Phillips et al., 2007; Siegfried et al., 2004)."

Minor comments

1- The term "mirror-symmetry" is redundant. Consider using "symmetry"

or "symmetrical polarity".

Response:

As noted in the cross-comment by Reviewer 1, we believe that "mirror-symmetry" is the appropriate term.

We think that "symmetry" implies the same lineage, whereas the relationship between the Z1 and Z4 lineages is not. "Mirror symmetry" was also used in Herman & Horvitz (1994) to describe the defect in the F lineage in lin-44/Wnt mutants as follows.

"we observed division patterns that were mirror symmetric to those of the wild type (Fig. 2). One plausible explanation is that the polarity of the first asymmetric cell division was reversed, causing the polarities of all subsequent asymmetric cell divisions also to be reversed."

2- "... they are permissively pushed distally by germ cells while proliferating" is confusing as it is unclear what proliferating cell you are referring to - germ cells or the DTC? proliferating? sense. Replace by: "they are pushed distally by proliferating germ cells"

Response:

Thank you for your helpful comment. We agree with your suggestion and modify the sentence as follows:

Original: "... they are permissively pushed distally by germ cells while proliferating" Revised: "... they are pushed distally by proliferating germ cells"

3- Fig. 2 is cited in the text before Fig. 1.

Response:

Thank you for pointing this out. Figure1 is mentioned in the Introduction before Figure 2 is referenced in the Result section in the original manuscript. We think the reviewer might be confused, as the POP-1 localization defect was shown in Figure 1. In response to the reviewer 1's comment, we moved the POP-1 localization images of the compound mutants to Figure 3. In addition, we noticed that in the original manuscript, Figure 1B was mentioned before Figure 1A in the Introduction. Therefore, we have modified the sentences in the Introduction.

The original sentence was:

"In the gonad, at the L1 stage, somatic gonadal precursor cells (SGPs), Z1 and Z4 have LH and HL polarity, respectively (Siegfried et al., 2004) (Fig. 1B). This mirror-symmetric polarity creates their mirror-symmetric lineages producing distal tip cells (DTCs) from the distal daughters (Z1.a and Z4.p) (Fig. 1B)."

The revised sentence now reads:

"In the gonad, at the L1 stage, somatic gonadal precursor cells (SGPs), Z1 and Z4 have LH and HL polarity, respectively, creating their mirror-symmetric lineages producing distal tip cells (DTCs) from the distal daughters (Z1.a and Z4.p) (Siegfried et al., 2004) (Fig. 1A and 1B)."

4- The results also suggest that MOM-5/Frizzled might be the receptor for Wnts regulating DTC production, as lin-17 mom-5 double mutants completely lack DTCs." Table 1 results rather suggest that lin-17 and mom-5 are the two frizzled receptor involved in DTC specification and that they are largely redundant.

Response:

As the reviewer noted, lin-17 and mom-5 function redundantly in DTC specification (SGP polarization). However, their functions are clearly different in terms of genetic interactions with Wnt genes (e.g. lin-17 cwn-2 but not mom-5 cwn-2 show the DTC-missing phenotype). We propose that MOM-5 but not LIN-17 functions as a receptor for Wnts.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

This is an interesting paper where authors study how DTC specification and migration are genetically controlled. The results show a complex interaction between the various Wnt components with partially redundant roles for the ligands and receptors. This helps understanding how the two DTCs are specified and symmetrically polarized in vivo. In addition, authors convincingly show that DTCs can significantly migrate in the absence of any push by GSC proliferation.

Major

1. Are the effects of combining the different Wnts with the lin-17 allele specific to the n3091 allele? It would be important to test another allele, for example the sy277 allele has a similar phenotype and is available at CGC. A null would be even better if it is viable. Alternatively, lin-17(RNAi) could instead be used if efficient enough. This is important since the n3091 allele could differentially alter the binding to the various Wnts, resulting in their distinct phenotypes in that background. However, these distinct phenotypes may not be relevant in a wild-type context.
2. In the lin-17; mom-5 double mutant which lacks DTCs, are Z1 and Z4 there but they do not express DTC markers, or are they never born? A lineage analysis should be presented. Also, are Z2 and Z3 still there on their own? Please show images.

Minor

1. The term "mirror-symmetry" is redundant. Consider using "symmetry" or "symmetrical polarity".
2. "... they are permissively pushed distally by germ cells while proliferating" is confusing as it is unclear what proliferating cell you are referring to - germ cells or the DTC? proliferating? sense. Replace by: "they are pushed distally by proliferating germ cells"
3. Fig. 2 is cited in the text before Fig. 1.
4. "The results also suggest that MOM-5/Frizzled might be the receptor for Wnts regulating DTC production, as lin-17 mom-5 double mutants completely lack DTCs." Table 1 results rather suggest that lin-17 and mom-5 are the two frizzled receptor involved in DTC specification and that they are largely redundant.

Referees cross-commenting

Dear reviewer #1, with all due respect, I do do understand your point as the Fig. 1 in Bowerman shows lineages that are left-right "asymmetrical", not "non-mirror symmetrical"? In the paper we are reviewing, the Z1-Z4 lineages are anterior-posterior symmetrical. But this is a minor issue and we can wait to see what the authors are going to reply...

Significance

The paper is interesting but in its current form, the genetics results around DTC specification are somewhat complex and difficult to interpret. The results linked to DTC migration are easier to take home.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

Summary

This manuscript illuminates the molecular regulation of mirror-image-symmetry in the C. elegans hermaphrodite somatic gonad precursor cells, reporting that a diversity of Wnt ligands and a Wnt receptor with Wnt-independent activity together polarize the two progenitors of the somatic gonad in opposite ways. The Wnt ligand genes act nonredundantly with respect to the polarity of the SGPs.

The conclusions arise from the careful interpretation of compound genetic mutants in the Wnt pathway, including the surprising finding of a Wnt-independent effect of lin-17/Frizzled on cell polarity that is bolstered by a new allele and construct of lin-17 that affect specifically its Wnt-interacting domain.

Major comments

While I find the genetic analyses compelling, I think the presentation of the data undermines the strength of the quantitative analysis of expression. Recommendations follow.

Genetic analyses: Major strength. Complex compound mutants are presented with ample sample sizes and careful interpretation of genetic interactions, presented clearly in tables that demonstrate large sample sizes. The new allele of lin-17 and delta-CRD LIN-17 construct reveal the surprising Wnt-independence of this Frizzled-type receptor. While of course I am interested in this Wnt-independent role, it is not necessary for the current manuscript to determine its mechanism.

Statistical analysis: When I first saw Fig. 1 B and C and read the associated results sections, I thought the qualitative polarity results reported in Fig. 1 would benefit from quantification, including at a minimum the sample size and penetrance of the phenotypes reported in Fig1B and 1C; it would be even better to quantify expression levels in the daughter cells. Then I discovered this appears to be what is reported in Figure 3. I think the presentation of these data are confusing. I recommend combining Figures 1 and 3 either before or after the results shown in Figure 2, so the reader's expectation for quantification is immediately satisfied.

Delving into the figure legend of Fig. 3 and the normalization procedure described in the Methods "Quantification of POP-1 asymmetry in the Z1 and Z4 division" raised concerns. The method therein described is overly complicated but also neglects background subtraction. My first question about this method: what range of distances between daughters is measured in Z? These distances are not discussed in absolute terms, and this is important for our understanding of how much correction for tissue depth might be necessary, as L1s are very thin.

To check my understanding, the authors use as a control a nuclear-localized GFP driven in the somatic gonad precursors in otherwise wild-type worms by the sys-1 promoter. They observe that the regression on a log scale of anterior:posterior (and vice versa) Z1 and Z4 daughter fluorescence over the distance between the daughters in the Z plane is fit by y = −0.034x + 0.0148, which is practically a slope of 0 and an intercept of 0. This means that they observed an ~1:1 ratio (as log(1)=0) of fluorescence in the anterior and posterior daughters of otherwise wild-type worms, at least across the range of very small X values of relevant distances between daughters (again, the relevant range of distances really matters and should be presented), making the normalization seem unnecessary.

Then based on this regression and 95% CI, the authors predict values that reflect true equivalence of fluorescence of POP-1::GFP in the two SGP daughters, compare the observed values to these predictions, and ultimately display in violin plots these differences of observed and expected. Correct?

Question: Is this complicated treatment the only way to detect differences in polarity of anterior and posterior daughters of Z1 and Z4? What happens if the authors measure GFP::POP-1 and calculate the following?

Z1p(MGV - background control from same focal plane)

Z1a(MGV - background control from same focal plane)

If this straightforward analysis shows asymmetric signal in the control that is made symmetrical or reversed in the mutants, the hypothesis would seem to be supported with a much more straightforward method. Samples could be analyzed separately in two bins by worm body position, which affects which cell is superficial in the sample. As it is, the Figure 3 Y axis label is hard to interpret without reading the methods at length, diminishing its impact.

Missing control: The sys-1 promoter-driven NLS-tagged fluorescent protein as a control to compare to the GFP::POP-1 is analyzed only in the wild-type, and apparently not in the mutants under consideration. Phillips et al. (2007) show that sys-1p transcriptional activity is equivalent between the SGP daughters in wild-type worms, but neither those results nor the method of normalizing to a sys-1p::GFP::NLS signal in this paper address the question of whether sys-1 promoter activity is equivalent in these cells in mutants upstream in the Wnt pathway. If the current method of normalization is to be used, it seems important to normalize to the sys-1p::GFP::NLS regression in each mutant background.

Missing methods: How was lin-17(mn589) generated? if this is the first report of this allele, full information on what the lesion is and how it was derived should to be reported.

The methods section lacks a description of how the mes-1 experiments were done, in terms of timing, duration, and temperature; mes-1(bn7) is a temperature sensitive allele.

Minor comments

The paper lacks a discussion of precedent in the literature for Wnt-independent Frizzled activity; this is a major finding that is being undersold in the current version of the manuscript.

Important: I think "Fig. 6 Germ cell independent migration of germ cells" title is a typo; should be "Germ cell independent migration of DTCs"

This is a very important experiment! I think a greater description of the mes-1 phenotype would be helpful, since loss of germline was not 100% penetrant in mes-1(bn7) hermaphrodites in Strome et al., 1995. The legend says "Germless mes-1 phenotype was confirmed by the absence of the mex-5::GFP::PH signal in the gonad." Consider adding a few sentences to the results (or methods, from which the mes-1 experiments are currently missing) describing that only mes-1 animals that lacked germline fluorescence were analyzed for DTC migration.

Please correct "secreting the Notch ligand LAG-2" this is a membrane-bound, not secreted ligand

Fig 1. The qualitative loss of polarity would be better depicted with a grayscale image instead of green-on-black.

Fig. 3 the presentation of these violin plots is confusing. The central text that reads "normal polarity, loss of polarity, reversed polarity" with arrows looks like a second Y axis label attached to the Z4 plot. I recommend rearranging. Consider shading the top, bottom, and central regions and explaining the meaning of the shading in the legend.

Referees cross-commenting

I have no comments to add to the points raised by Reviewer 2, other than to share my opinion that "mirror symmetry" is helpful terminology, as there are other possible developmental symmetries other than the "HL LH" mirror symmetry seen in the SGPs. For an example of a non-mirror developmental symmetry, see Fig. 1 of this dispatch by Bowerman 2006 https://www.sciencedirect.com/science/article/pii/S0960982206024481

Significance

Strengths: Rigorous genetics

Limitations: Data presentation, expression level analysis, communication of results.

Advance: Wnt signaling has highly conserved, ancient roles in the earliest polarity events in animal embryos, so learning how the Wnt signaling pathway dictates cell polarity, including a new Wnt-independent role for a major receptor, has broad implications for understanding cell polarity via this pathway in animal development. Another major advance is the finding that a conserved developmental signaling pathway with members that are redundant in certain contexts nonetheless elicits specific responses from certain cells during development, thus generating a diversity cell types.

Audience: These findings will be of interest to a broad audience interested in animal development, symmetry-breaking, complexity, and the Wnt-signaling pathway.

Additionally, a worm-specific audience will be interested in the finding that the DTCs are capable of germline-independent migration, as the new paradigm of Agarwal et al. ascribes DTC propulsion to germline pushing forces. How the DTC migrates is a longstanding question in our field.

Reviewer: I study developmental genetics and cell biology in the C. elegans hermaphrodite gonad. I have sufficient expertise to evaluate all of the experiments presented here. In my opinion, this is a highly valuable developmental genetics study with significant but easily addressable flaws in presentation.

Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

Learn more at Review Commons

---

Reply to the reviewers

'The authors do not wish to provide a response at this time.'

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #3

Evidence, reproducibility and clarity

Summary:

DNA double-strand breaks are harmful to cells. The authors used CRISPR-Cas9 to create DSBs in repetitive elements (Ty transposons) in the Saccharomyces cerevisiae genome. This method builds on previous ones that used the HO endonuclease or prokaryotic restriction enzymes. They used Cas9-based DSBs to assess the role of Tel1 (ATR) in DSB sensing, comparing it with the DSB-generating xenobiotic zeocin. The first part of the paper is fine, but it lacks information about the yield. The system is not as efficient as it is claimed to be, and it is slow. The second part of the paper is not well connected to the first half, and several controls are needed to make it sound.

Major comments:

1. A significant portion of the manuscript posits that the CRISPR-Cas9 system is capable of creating as many DSBs as the theoretical number of Ty targets. Nevertheless, it is clear that this is not the case. The authors can determine the number of breaks per construction. A plot of the frequency of average DSBs in chromosomes IV and III appears to be feasible based on the Southern blots presented, although this could give an underestimate, given that some broken chromosomes may become trapped in the well during DSB processing. However, the percentage of zero DSBs for a given chromosome is readily quantifiable, allowing the frequency of DSBs per cell to be determined through straightforward calculations. Once this has been achieved, the authors should revisit their assessments of Rfa1/Rad52/Tel1 foci formation and number.

2. OPTIONAL: It seems plausible that a saturation of DSBs per cell may occur when the actual number of DSBs is plotted against the theoretical maximum. In such a scenario, the introduction of additional Cas9 molecules could enhance the efficiency of DSB generation. This could be achieved by increasing the number of copies of the CAS9 gene.

3. The majority of the discussion chapter is dedicated to the second part of the paper (Tel1), yet no mention is made of the subject I have just commented on. Similarly, a comparison should be made between the Cas9-based system and previous methodologies for creating single and multiple DSBs (HO, I-SceI, restriction enzymes, radiation, radiomimetic drugs, etc.) in terms of efficiency, time to DSB, cell response, etc. It is surprising that the Cas9 system takes so long to generate DSBs and that the cell cycle profile indicates very little arrest in G2/M. This should also be discussed.

4. OPTIONAL: Since the lack of a strong G2/M arrest is intriguing, it would be good to do a Western blot of Rad53 to learn more about Cas9-based DSB sensing and the DNA damage checkpoint. Comparing it to both the well-established HO system (even a single HO cut) and radiation/radiomimetics would be ideal.

5. OPTIONAL: Given the lack of the G2/M arrest when Cas9 is expressed in asynchronous cultures, the authors may try to synchronize cells in G1 and G2/M before Cas9 is induced. This could help them find out if the G1 and G2/M peaks in Figure 2A are caused by DSBs leading to both a G1 and a G2/M arrest. Alternatively, they could film the cells after Cas9 is added and check microcolony formation.

6. The timeframes of Rad52 foci in Figure 5 indicate an anomalous pattern, which raises concerns about the validity of the experiment setup. The selected cells did not change their morphology throughout the 140-minute observation period. The unbudded cell remained unbudded, and the bud did not grow in six out of seven cells. For cells with small buds (cells 2 to 5), the expectation was that the bud would grow until the cell either became a dumbbell (indicative of a G2/M arrest) or divided its nucleus. Furthermore, no re-budding events were observed. Is this pattern real? Why is it so? Once this issue is addressed, could the duration of the Rad52 foci be quantified? Was a single z-plane taken? If so, some Rad52 foci that appear and/or disappear could reflect their migration to an on-focus or out-of-focus plane.

7. OPTIONAL: It would be beneficial to conduct a double labeling with known DSB factors that coexist with Tel1, or are downstream of it, in order to enhance the data shown in Figures 6 and 7. This would also address some of the queries raised about Tel1 clustering and location on the nuclear periphery.

8. In the experiment with the condensin mutant smc2-8, a temperature shift from 24 to 37 ºC is carried out, which represents a significant physiological change. Therefore, it is essential to include a parallel control with a WT strain to rule out the possibility that the unclustering of Tel1 foci is due to the temperature shift.

9. The sentence in the Discussion about condensin's role in maintaining Tel1 clustering is misleading. It could be interpreted as suggesting that condensin actively gathers Tel1 foci after DSBs (lines 650-651), when in fact condensin's function is simply to maintain Ty element clustering prior to DSBs, as the authors themselves cite in the text.

10. Taking into account the importance of Cas9/sgRNA plasmid constructions, PFGE and Southern blot in this work, the author should make the effort to describe them all in much more detail in M&M.

Minor comments:

1. In the Figure 7 legend, panel A is missing, and the text for the other panels is consequently misplaced.

2. The functional verification of the yEGFP-Tel1 construction (Fig. 6A & B) would be better presented as a supplementary figure. The same rationale applies to Figure 8.

3. Please include G2/M in the abbreviations.

4. Please clarify what is meant by "5h40". Is this 5 hours and 40 minutes? If so, please use alternative nomenclature.

5. Some sections of the text appear superfluous. For example, the definitions of mean, SEM and SD, and the rationale for choosing SEM over SD (lines 211 to 215), as well as the information about the purpose of PFGE in a figure legend (line 322).

Referees cross-commenting

After reading the comments of the other reviewers, I agree with them, and some of them raise the same concerns as I do.

Significance

General assessment: The study addresses the generation of multiple DSBs by Cas9 when an increasing number of targets are incorporated for cutting. The strategy to create an increasing dose of DSBs based on the different Ty elements is an innovative approach, although it is constrained by the nature of the target. The assessment of DSBs by PFGE plus Southern blot is well planned, although there is potential to exploit the obtained results further to assess the actual vs theoretical DSBs, saturation effect, etc. They then sought to use their dose-dependent system to examine the role of Tel1 in the DSB response, comparing Cas9 with zeocin. However, a comparison between the two strategies is challenging without prior knowledge of the number of DSBs present in each treatment. Overall, the study represents a promising effort to use Cas9 for the generation of multiple DSBs in yeast. Nevertheless, the system is constrained for further mechanistic studies on the DSB response due to its slow kinetics, low yields, and lack of expected DNA damage responses, particularly the G2/M checkpoint.

Advance: Despite such a dose-response assessment for the Cas9-based DSBs have not been performed in yeast, similar studies with sequence-specific DSBs have been done before in Lorraine Symington's lab (and others). Perhaps the Cas9-based system is simpler than the one that relies on multiple insertions of the HO cutting sequence, but it appears neither simpler than the one based on the expression of restriction enzymes nor more efficient. The Cas9 system has several limitations that make it less suitable for studying how cells react against multiple DSBs. It is slow, probably saturates after a few DSBs, and does not render a full DNA damage response. However, there is still value in understanding and making predictions about this important gene-editing method.

Audience: This paper is intended for a specialized audience, including those involved in basic research on DSB sensing and repair, particularly in yeast.

Field of expertise of the reviewer: Cell and molecular biology of S. cerevisiae, with a particular interest in the DNA damage response.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

Summary.

The authors have developed a system to generate a variable number of double-strand breaks (DSBs) - specifically 1, 15, or 59 - in the genome of Saccharomyces cerevisiae in a galactose-inducible manner. This system relies on the galactose-inducible expression of Cas9 nuclease and the use of guide RNAs (gRNAs) targeting different classes of Ty elements. The efficiency of DSB induction was assessed through several methods, including monitoring cell viability in media containing galactose versus glucose, pulsed-field gel electrophoresis (PFGE), PFGE combined with Southern blotting using probes specific to different chromosomes, and the formation of foci by repair and checkpoint proteins. Specifically, the system was utilized to examine foci of the apical checkpoint kinase Tel1 and the repair factors RPA and Rad52.

From their experiments, the authors conclude that the system is capable of inducing DSBs in a controlled and dose-dependent manner, and that the DSBs are indeed induced at the chromosomal loci where the Ty elements reside.From microscopy experiments, the authors conclude that some DSBs are more persistent, while many are repaired via homologous recombination (HR). New DSBs can form as a result of the continuous expression of Cas9. Finally, Tel1 appears to form multiple foci after DSB induction and these foci localize to the nuclear periphery.

Major comments.

In my opinion, some experiments lack appropriate quantification, particularly those evaluating the dose-response effect and the efficiency of DSB induction. Additionally, some conclusions appear to be overestimated in relation to the experimental data. Further experiments and more detailed analyses would be beneficial to fully support the claims and conclusions presented. The data and methods are presented in a manner that should allow for reproducibility. The experiments are adequately replicated, but the statistical analyses are lacking in most of the figures. For example, in Figures 4B, 4C, and 6D, it would be helpful to indicate when the increase in cells containing repair foci is significant.

Additional experiments and modifications suggested:

• Figure 1B: Quantify the effect of DSB induction on cell viability with a dose-response curve. No differences are observed between the 0.5% gal and 1% gal conditions. The greatest effect on viability occurs in 2% gal and in the absence of sucrose, which might be due to the absence of sucrose. This should be addressed or discussed.
• Figure 1B and Figure 2: A quantification of galactose-induced Cas9 expression in the presence of different doses of galactose or over time, respectively, is a necessary control (mRNA or protein expression).
• Figure 2A: From the cell cycle analysis using FACS, the authors suggest that Cas9 induction causes checkpoint activation. This can be easily confirmed by using more direct methods for checkpoint activation, such as Rad53 phosphorylation. In addition, checkpoint activation should be monitored in synchronized cells released in galactose in order to evaluate whether Cas9-induced DSBs in Ty elements can trigger checkpoint in the first cell cycle or require more time. This reflects the timing of DSB induction.
• Figure 3: In addition to the restriction analysis, kinetics of DSB formation at specific Ty loci by classical Southern blot or qPCR is, in my opinion, necessary to demonstrate the effectiveness and efficiency of DSB induction over time, especially in relation to Cas9 expression.
• Figure 5: In this case as well, I believe that a quantitative analysis of the number of cells in the different conditions illustrated in the figure is necessary to understand the dynamics of repair and the formation of new DSBs. Some conclusions are somewhat strong, particularly the correlation between cell cycle phase and repair kinetics (long-lasting Rad52 foci versus short-lived Rad52 foci) and would require quantitative and statistical analysis. In addition, it is not clear to me why cells in which the Rad52 focus disappears do not proceed through the cell cycle (e.g., cells represented in rows 8 and 9 of the figure).
• Figure 7E and F: It would be interesting to see if the number of Rad52 foci per cell changes in the smc2-8 mutant to understand if RAD52 is required to form the repair center or if this depends on the clustering of Ty elements at the nuclear periphery.

Minor comments.

The authors have cited relevant literature to support their methodology, findings, and conclusions. The text and figures are clear. The descriptions in the text are precise and well-articulated, making the data easy to understand. The figures are well-designed and clearly labeled.

Specific suggestions:

• Figure 2C: Please change the order of the panel: place chromosome III at the top and chromosome IV at the bottom, as described in the text.
• Figure 4: The dose-response effect is not evident when monitoring repair foci. It appears that the proportion of cells showing RPA and Rad52 foci is generally low, especially after the formation of 59 DSBs. This is particularly concerning given that the strain in which 59 DSBs are induced already has 20% of cells with RPA foci at time = 0. The authors attribute the lower presence of foci to improved repair caused by the clustering of multiple lesions, but could it simply be due to lower Cas9 efficiency when there are many target sites?
• Related to Figure 4:The statement on line 400: "the proportion of nuclei displaying Rfa1 foci was consistently double than that of nuclei bearing Rad52 foci, probably reflecting the increased residence time of resected filaments in comparison with the process of homology search" is somewhat strong, considering that the proteins are labeled with different fluorophores, which might experience different rates of photobleaching.

I believe the proposed experiments can be completed in about 6 months. The request to monitor the kinetics of DSB formation at specific Ty elements in more detail might take more time if the PCR or Southern blot techniques need to be optimized. However, the authors may have other methods in mind that are more familiar to them for evaluating these kinetics, which could be equally valid.

Referees cross-commenting

I have carefully read the comments on the manuscript from the other reviewers. I noticed that many of our opinions coincide, and I am convinced that all the requests are appropriate.

Significance

My field of expertise centers on checkpoint regulation and homologous recombination in Saccharomyces cerevisiae. I have extensively utilized the galactose-inducible HO endonuclease system for inducing DSBs. I believe that developing additional systems to induce one or more localized DSBs in specific genomic regions is crucial for addressing unresolved questions regarding DSB response. An ideal system would also operate independently of galactose. Based on my experience, an effective system for DSB induction should induce breaks rapidly and simultaneously to produce innovative and reproducible results. From the data presented, I am uncertain whether the system developed in this work meets these criteria.

The development of a Cas9-based system capable of forming multiple DSBs could be an important tool for studying the DSB response, although I have doubts about how much it will truly enhance our understanding of damage repair. Other systems, cited by the authors, have been previously developed to produce multiple DSBs with different nucleases and using TY elements. Although the Ty-HO system to induce 1, 7, or 10 DSBs was developed in L. Symington's laboratory in 2004, the use of this system has been very limited, likely because it is not easy to monitor what happens to individual DSBs and because the efficiency of simultaneously inducing multiple DSBs may decrease with the increase in the number of target sites. I am concerned that the tool developed in this research article may have limited applicability, being relevant primarily to a small niche within the scientific community focused on DSB response, and thus might generate only a narrow interest. However, the latter part of the paper, which addresses the localization of Tel1, seems promising, despite being preliminary. Additionally, the development of a fluorescent variant of Tel1 is intriguing. This new variant appears to retain the protein's functions and forms well-visible foci within the cell, which seem to be brighter and more intense compared to those obtained with the variant previously developed by M. Lisby and R. Rothstein.

Strengths:

• The study presents an innovative system to induce a variable number of double-strand breaks (DSBs) in the genome of Saccharomyces cerevisiae using a galactose-inducible Cas9 and specific gRNAs for Ty elements.
• The authors use a variety of methods to evaluate DSB induction, including cell viability assays, PFGE, Southern blotting, and foci formation of repair and checkpoint proteins. This comprehensive approach provides sufficient evidence that DSBs are induced.
• The latter part of the article, focusing on the formation and localization of Tel1 foci, is particularly important. It sheds new light on the functions of Tel1 and its role in the DSB response. This finding is a significant preliminary indication that warrants further development, as the authors suggest in the discussion. I find the development of a tool to effectively visualize Tel1, which is a low-abundance protein in the cell, to be innovative and important for the community working in DSB repair and checkpoint.

Limitations:

• Some experiments lack appropriate quantification. For instance, a dose-response curve quantifying the effect of DSB induction on cell viability is missing. Additionally, quantification of galactose-induced Cas9 expression over time (mRNA or protein) is necessary.
• Statistical analyses are lacking in most figures. It is important to indicate when increases in cells containing repair foci are significant, particularly in Figures 4B, 4C, and 6D.
• The suggestion of checkpoint activation from cell cycle analysis using FACS should be validated with more direct methods, such as Rad53 phosphorylation, and monitored in synchronized cells to evaluate the timing of checkpoint activation.
• Further analysis is needed to demonstrate the effectiveness and efficiency of DSB induction over time, especially in relation to Cas9 expression. Monitoring the kinetics of DSB formation at specific Ty loci by classical Southern blot or qPCR would be beneficial.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

The manuscript entitled "A CRISPR-Cas9-based system for the dose-dependent study of DNA double-strand breaks sensing and repair" by Coiffard and colleagues reports on a novel system to monitor the response to multiple DSBs in yeast by targetting Cas9 nuclease to Ty elements. The data is clearly presented, but there are several technical concerns and questions the conclusions of the manuscript. First of all, it is not clear how many DSB cells contain at any given time. The kinetics of DSB formation appears to be slow (hours) relative to the formation and progression of DNA repair foci (minutes). Tel1 foci are found primarily at the nuclear pheriphery, but given the slow kinetics and low number of foci, it is unclear whether the peripheral localization of Tel1 represents physiologically relevant repair and aberrant events where repair failed. Unfortunately, there is not assessment of the outcome of repair. Is the repair error-free or do the cells accummulate indels and/or structural changes at the Cas9 cleavage sites? Without this information, the assay will be of limited use for the scientific community.

Other major issues:

1. The introduction should reflect on the chemical structure of DSB ends and the fact that Cas9 remains bound to DNA after cleavage, which may delay repair.
2. Different Ty elements may be cut with different kinetics. The authors should compare their system to (Gnügge and Symington 2020).
3. The introduction should reflect on the physiological relevance of 15/59 DSBs.
4. What is the copy number of the Cas9 and gRNA plasmids? Could cell-to-cell variation in copy number explain the variation in the number of Tel1.
5. Do the authors observe mutations caused by leaky expression of gRNAs in the uninduced state?
6. Line 413: I don't think the data in figure 4 completely warrants the conclusion the "DSBs induced by Cas9 engage into HR in a dose-dependent manner". More foci with 15 DSBs than with 59 DSBs? Interference or other explanations? I think this discrepancy warrants additional discussion.
7. What is observed with a 1h pulse of Cas9 induction followed by glucose? Can the assay monitor the kinetics of repair?
8. The interpretation of foci in figure 5 is difficult to follow given that only 1 focus is observed while many DSBs are induced. Without further experimentation, these speculations should be moved to the Discussion.
9. Tel1 foci were always observed at the nuclear periphery (figure 6C). This information should be quantified and compared to Rad52, given that Rad52 foci have previously also been observed at the nuclear periphery for persistent DNA lesions (Whalen and Freudenreich, 2020; Nagai et al., 2008; Lisby et al., 2010). Do these foci perhaps reflect a small subset of Cas9-induced DSBs that are not repair?
10. Line 699: the notion that Tel1 senses DSBs is novel and does not fit well with the literature given that Tel1 is recruited to foci downstream of Mre11 (Lisby et al., 2004). Unless the authors can provide additional evidence, I suggest to rather write that Tel1 is an early transducer of the DNA damage response.

Minor suggestions:

1. The manuscript could benefit from correction of English grammar.
2. Figure 7F: please also include cells with 1 focus in the graph.
3. Figure 7: the labels A-F do not follow the legend.
4. Figure 3, legend: it should be stated in the legend, how long Cas9 was induced in this experiment.
5. Line 673: it could be noted that in mammalian cells, many more foci are observed, which is probably due to the larger size of the nucleus.
6. The conclusion that Tel1 behaves exceptionally in terms of the number of foci that are formed is perhaps an overstatement, since only three proteins were analyzed and the occurrence of 8 foci was rare (1:1000 cells).
7. Line 685: the word "form" indicates that the DSB are already at the nuclear periphery, when bound by Tel1. How do the authors exclude that Tel1 foci form all over the nucleus and then subsequently relocalize to the nuclear periphery? Time-lapse microscopy would be able to reveal where the Tel1 foci form.

Referees cross-commenting

I have read and agree with the comments of the other reviewers.

Significance

The manuscript entitled "A CRISPR-Cas9-based system for the dose-dependent study of DNA double-strand breaks sensing and repair" by Coiffard and colleagues reports on a novel system to monitor the response to multiple DSBs in yeast by targetting Cas9 nuclease to Ty elements. The data is clearly presented, but there are several technical concerns and questions the conclusions of the manuscript. First of all, it is not clear how many DSB cells contain at any given time. The kinetics of DSB formation appears to be slow (hours) relative to the formation and progression of DNA repair foci (minutes). Tel1 foci are found primarily at the nuclear pheriphery, but given the slow kinetics and low number of foci, it is unclear whether the peripheral localization of Tel1 represents physiologically relevant repair and aberrant events where repair failed. Unfortunately, there is not assessment of the outcome of repair. Is the repair error-free or do the cells accummulate indels and/or structural changes at the Cas9 cleavage sites? Without this information, the assay will be of limited use for the scientific community.

Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

Learn more at Review Commons

---

Reply to the reviewers

The authors do not wish to provide a response at this time.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #3

Evidence, reproducibility and clarity

Protein kinases play an important role in regulating cell division and the dynamics of the actomyosin ring both in yeast and mammalian cells. While this raises the possibility that regulated protein dephosphorylation may also affects the division dynamics, few experimental studies have been reported. In the current study, Chrupcala and Moseley screen 9 non-essential protein phosphatases and show that the conserved protein phosphatase PP2A-B56 regulates division plane positioning in fission yeast. The authors show that Par1 localizes to the division site and that cells lacking Par1 have impaired levels of regulators of actomyosin ring positioning, including the anillin-like protein Mid1. Interestingly, restoring wildtype Mid1 levels partially suppresses the division defects observed in the par1∆ mutant.

The study is well-executed with attention to detail and careful phenotype quantification and analyses. The conclusions of the study are largely supported by experiments to yield the novel insights. This reviewer commends the authors on reporting the negative results regarding Mid1-3xSLiM mutant (Results section) and failure to detect physical interaction between Par1 and Mid1 (Discussion section). The authors provide an interesting discussion. Furthermore, they describe very relevant additional experiments but they restrict their execution to future studies that complement the current manuscript

Major concern

The authors show that par1∆ cells have impaired levels of Mid1, Cdc15 and Rga7, yet propose that Mid1 reduction is the principal cause of cytokinetic defects in par1∆ cells. Even though increase in Mid1 levels rescues par1∆ defects, it is possible that increased Mid1 levels suppress another primary defect (e.g. Rga7 increased levels). Thus, it would be interesting to perform the following two experiments: 1. The authors in the discussion say "we found that cytokinesis defects arise from decreased Mid1 levels" but this is not formally shown other than in par1∆ cells. Thus I would suggest monitor cytokinesis in cells with Mid1 levels directly reduced to levels comparable to those observed in par1∆ cells (as quantified in Fig 3B). Monitoring the effects of reduced Mid1 levels on Rga7 and Cdc15 would also be interesting. 2. Monitoring the levels of Rga7 and Cdc15 in par1∆ cells rescued by the second copy Mid1.

Minor concern

Even though definitive evidence for Par1 mechanism of Mid1 regulation might be difficult to obtain, the authors may choose to strengthen their work on the role of dephosphorylation at the actomyosin ring. For example, using the GFP-GFPnanobody pairing to force interaction between Mid1 and Par2 in par1∆ cells may provide support for dephosphorylation playing a more direct role in actomyosin ring positioning.

Referees cross-commenting

Consultation regarding Review 1:

Reviewer 3 shares the interest in points 1,3,5.

Relating to point 2: Not sure what the reviewer specifically wishes here.

Relating to point 4: Extensive analyses of many cytokinetic proteins in par1∆ cells, and importance of their levels for cytokinesis, would be interesting but perhaps beyond the scope of this study. I believe it would suffice to monitor Cdc15, Bgs1 and Rga7 in the 2xMid1 rescue that authors performed.

Consultation regarding Review 2: Reviewer 3 shares the interest in major points 1 (though authors do leave open the possibility that PP2A acts indirectly),3,4,5. Regarding point 2: This might be difficult to ascertain directly and instead it might suffice to show that Mid1 levels reduced to those observed in par1∆ phenocopy the par1 mutant's division defects.

Significance

The study is well-executed with novel findings on regulation of the cell division and the physiological roles of phosphatases. The study will benefit cell polarity and cell division research fields as well as researchers interested in roles of protein phosphatases.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

Summary:

In this study, the authors investigates the functional relationship of Mid1, which plays a central role in the positioning of the contractile ring (CAR) in fission yeast, with protein phosphatases. Previous studies have shown that Mid1 is phosphorylated and modified by several kinases, but its relationship with phosphatases is not well understood except of Clp1. Therefore, the authors examined whether phosphorylase gene disruption strains were able to divide symmetrically. The authors found that cells lacking par1, which encodes the regulatory subunit of PP2A, frequently divided asymmetrically. Furthermore, they found that the par1 deletion mutant showed reduced protein levels of Mid1 in the cells. Furthermore, they showed that increasing the expression level of Mid1 can partially compensate for the mitotic abnormalities of the par1 strain. Since Par1 localizes to the division site, the authors also investigated the possibility of binding with Mid1, focusing on its SLiM motif. However, at this time, it is unclear whether direct binding of Par1 to Mid1 is necessary for Mid1 to exert its cellular functions, since no conspicuous cytokinesis abnormalities were observed in Mid1 with the mutation in the SLiM motif.

The focus of the study to be interesting and the clarity of the overall argument of the manuscript to be almost adequate. However, the authors should investigate or mention the following points. Also, a more appropriate and convincing way of presenting experimental results is required.

Major comments:

1. Does Par1 (PP2A) work in the dephosphorylation of Mid1? Looking at the band pattern of the western blot of Mid1 in Fig. 3E, there are several bands. This band shift probably indicates phosphorylation of Mid1, but is there any difference in the band shift between the wild-type strain and the par1 gene disrupted strain? Mid1 is phosphorylated in a cell cycle-dependent manner. So, it would be interesting to examine the cell cycle-dependent phosphorylation pattern using synchronous culture.
2. In Fig. 5A, par1- and ppa2-deficient strains show little nuclear localization of Mid1 in interphase cells. If little or no Mid1 enters the nucleus, could it be possible that the mitotic abnormality is not due to a reduction in Mid1 protein levels, but is directly due to an effect on the shuttling of Mid1 between the cell nucleus and cell surface?
3. The photographs showing cellular localization of proteins, such as Figs. 3, 4, 5, 6B, and 6C, are not convincing. The authors should show a typical picture of the cell population as supplemental figures, rather than trimming a single cell.
4. Is the Mid1-deficient strains impaired in the localization of Par1 to the division plane? As shown in Fig. 3A, the fluorescence of Cdc15 and Rga7 is predominantly enhanced in the par1 deletion strain (Fig. S2 also shows a protein with a significant difference in localization). The authors should consider these points and make more comprehensive discussion.
5. Fig. 3D and part of Fig. 5B also use the same photograph. This raises the suspicion that these data were taken only once. If so, the authors should do replicated experiments to assure the authenticity of the data. In addition, perhaps it is a careless mistake, but this way of presenting data should not be allowed unless specifically mentioned in the manuscript.

Minor point.

Typos. p. 11 Mid1-13my levels,.

Lack of uniformity in description p. 13 Methyl-2-benzimidazole and p. 15 methyl-2-benzimidazole In the figure of the paper Mid1-mNeonGreen and mid1-mNeonGreen

Significance

The importance of this study is that it deepens our knowledge of Mid1 protein of fission yeast, an important model organism for the study of cytokinesis. Although the possibility that Par1 is involved in the maintenance of Mid1 protein levels has been demonstrated, the molecular mechanism has not been clarified. It might be expected that anilline, which is similar to Mid1 in animal cells, might have a similar relationship to protein phosphatases, but there is no evidence for this. Therefore, at this moment, the effect of these findings may be limited to the fission yeast research community.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

The study by Chrupcala and Moseley explores the role of the protein phosphatase PP2A-B56 in cytokinesis, particularly focusing on its regulatory subunit Par1 and its interaction with Mid1 in fission yeast (S. pombe). The authors report that in cells lacking Par1 (par1d), the contractile rings are mispositioned, and the levels of multiple cytokinetic proteins, including Mid1, are altered, leading to defects in cytokinesis. They suggest that the reduction in Mid1 levels is responsible for the observed cytokinetic defects and show that restoring Mid1 levels can partially rescue these defects. The study further demonstrates that Mid1 shuttling between the nucleus and cell membrane is impaired in par1d mutants. Thus, par1d seems to cause a range of defects, from altered protein levels to specifically impairing Mid1 functions. However, the study does not demonstrate a direct interaction between Par1 and Mid1, leaving the exact mechanism by which Par1 regulates Mid1 levels unclear.

The manuscript is well-written and investigates the lesser-studied role of protein phosphatases in cytokinesis, which is of general interest. However, a major criticism is the lack of evidence supporting a direct role of PP2A-B56 and its regulatory subunit Par1 in cytokinesis. It is possible that the phenotype observed in par1d cells results from a global impairment of protein regulation rather than a specific effect on Mid1. Indirect factors, such as changes in cell size/shape, could also produce similar phenotypes.

To strengthen the manuscript, the authors should consider the following points:

1. Changes in cell size/shape can influence Mid1 levels and nucleus positioning. Many par1d cells shown in the manuscript have bulgy morphologies. The manuscript should address whether cell size/morphology differences contribute to the observed phenotypes.
2. The quantification of nucleus positioning needs further consideration, as the positioning of the nucleus has been shown to be a function of cell size, etc.
3. The manuscript proposes that Par1 acts by reducing Mid1 levels and that restoring Mid1 levels can rescue the defects. However, given the known mechanisms of Mid1 localization, it is unclear why a reduction in Mid1 levels alone would lead to cytokinetic defects. To establish that the defects are due to Mid1 reduction, the authors should demonstrate that lowering Mid1 levels (e.g., using a low-expression promoter) phenocopies the cytokinesis defects observed in par1d cells.
4. As shown in par1d cells, other cytokinesis proteins have impaired levels. It should be investigated whether changes in the levels of other proteins (e.g., Cdc15, Rga7, Myp2, etc.) in par1d cells contribute to the cytokinetic defects. In general, it is not discussed how the par1d mutant can lead to differential levels of many cytokinesis proteins. Does the reduction of Mid1 explain this shift? It is conceivable that changes in the levels of other proteins could also produce similar cytokinesis defects.
5. The manuscript would benefit from additional functional rescue experiments. For example, expressing a phosphatase-dead version of Par1 in par1d cells could help determine if its phosphatase activity is necessary for cytokinesis.

Significance

The study by Chrupcala and Moseley explores the role of the protein phosphatase PP2A-B56 in cytokinesis, particularly focusing on its regulatory subunit Par1 and its interaction with Mid1 in fission yeast (S. pombe). The authors report that in cells lacking Par1 (par1d), the contractile rings are mispositioned, and the levels of multiple cytokinetic proteins, including Mid1, are altered, leading to defects in cytokinesis. They suggest that the reduction in Mid1 levels is responsible for the observed cytokinetic defects and show that restoring Mid1 levels can partially rescue these defects. The study further demonstrates that Mid1 shuttling between the nucleus and cell membrane is impaired in par1d mutants. Thus, par1d seems to cause a range of defects, from altered protein levels to specifically impairing Mid1 functions. However, the study does not demonstrate a direct interaction between Par1 and Mid1, leaving the exact mechanism by which Par1 regulates Mid1 levels unclear.

The manuscript is well-written and investigates the lesser-studied role of protein phosphatases in cytokinesis, which is of general interest. However, a major criticism is the lack of evidence supporting a direct role of PP2A-B56 and its regulatory subunit Par1 in cytokinesis. It is possible that the phenotype observed in par1d cells results from a global impairment of protein regulation rather than a specific effect on Mid1. Indirect factors, such as changes in cell size/shape, could also produce similar phenotypes.

Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

Learn more at Review Commons

---

Reply to the reviewers

Reply to Reviewers

We would like to thank all the reviewers for their thorough reading and helpful comments. Below, please find our point-by-point response. The reviewer comments received through ReviewCommons have not been altered except for formatting.

Reviewer #1 (Evidence, reproducibility and clarity (Required)):

The authors extended the existing recombination-induced tag exchange (RITE) technology to show that they can image a subset of NPCs, improving signal-to-noise ratios for live cell imaging in yeast, and to track the stability or dynamics of specific nuclear pore proteins across multiple cell divisions. Further, the authors use this technology to show that the nuclear basket proteins Mlp1, Mlp2 and Pml39 are stably associated with "old NPCs" through multiple cell cycles. The authors show that the presence of Mlp1 in these "old NPCs" correlates with exclusion of Mlp1-positive NPCs from the nucleolar territory. A surprising result is that basket-less NPCs can be excluded from the non-nucleolar region, an observation that correlates with the presence of Nup2 on the NPC regardless of maturation state of the NPC. In support of the proposal that retention of NPCs via Mlp1 and Nup2 in non-nucleolar regions, simulation data is presented to suggest that basket-less NPCs diffuse faster in the plane of the nuclear envelope.

However, there are some points that do need addressing:

Major Points 1. Taking into account that the Nup2 result in Figure 4B forms the basis for one half of the proposed model in Figure 6 regarding the exclusion of NPCs from the nucleolar region of the NE, there is a relatively small amount of data in support of this finding and this proposed model. For example, the only data for Nup2 in the manuscript is a column chart in Figure 4B with no supporting fluorescence microscopy examples for any Nup2 deletion. Further, the Nup60 deletion mutant will have zero basket-containing NPCs, whereas the Nup2 deletion will be a mixture of basket-containing and basket-less NPCs. The only support for the localization of basket-containing NPCs in the Nup2 deletion mutant is through a reference "Since Mlp1-positive NPCs remain excluded from the nucleolar territory in nup2Δ cells (Galy et al., 2004), the homogenous distribution observed in this mutant must be caused predominantly by the redistribution of Mlp-negative NPCs into the nucleolar territory."

As suggested by the reviewer, we have added fluorescence microscopy examples for the Nup2 deletion to new Figure 4D. In addition, we have added data on Nup1 as suggested by reviewer 3. Since we observed a significant effect on nucleolar NPC density also upon depletion of Nup1 (new Figure 4A), we have overall revised the text and model to now reflect the shared role of Nup1 and Nup2.

We have also localized Mlp1-GFP in a nup2Δ background as well as in the Nup60ΔC background where Nup2 can no longer bind to the NPC. In both strains, Mlp1-containing NPCs remain excluded from the nucleolus as now shown in the new Figure 4E. Although we also observed partial Mlp1 mislocalization to a nuclear focus in the nup2Δ strain, such mislocalization was only minimal in the strain with the Nup2-binding domain in Nup60 deleted (nup60ΔC), supporting our conclusion that Nup2 contributes to nucleolar exclusion of NPCs independent of Mlp1. Similarly, Mlp1-positive NPCs remained excluded from the nucleolar territory in cells depleted of Nup1 (new Figure 4B).

1. The authors could consider utilizing this opportunity to discuss their technological innovations in the context of the prior work of Onischenko et al., 2020. This work is referenced for the statement "RITE can be used to distinguish between old and new NPCs" Page 2, Line 43. However, it is not referenced for the statement "We constructed a RITE-cassette that allows the switch from a GFP-labelled protein to a new protein that is not fluorescently labelled (RITE(GFP-to-dark))" despite Onischenko et al., 2020 having already constructed a RITE-cassette for the GFP-to-dark transition. The authors could consider taking this opportunity to instead focus on their innovative approach to apply this technology to decrease the number of fluorescently-tagged NPCs by dilution across multiple cell divisions and to interpret this finding as a measure of the stability of nuclear pore proteins within the broader NPC.

We apologize for this imprecise citation. We have modified the text to indicate that our RITE cassette was previously used in two publications. It now reads: "We used a RITE-cassette that allows the switch from a GFP-labelled protein to a new protein that is not fluorescently labelled (RITE(GFP-to-dark)) (Onischenko et al., 2020, Kralt et al., 2022)." Together with additional changes to the text throughout, we hope that our new manuscript version more clearly highlights the innovation of our approach relative to previous use cases.

1. The authors could also consider taking this opportunity to discuss their results in the context of the Saccharomyces cerevisiae nuclear pore complex structures published e.g. in Kim et al., 2018, Akey et al., 2022, Akey et al., 2023 in which the arrangement of proteins in the nuclear basket is presented, and also work from the Kohler lab (Mészáros et al., 2015) on how the basket proteins are anchored to the NPC. There is additional literature that also might help provide some perspective to the findings in the current manuscript, such as the observation that a lesser amount of Mlp2 to Mlp1 observed is consistent with prior work (e.g. Kim et al., 2018) and that intranuclear Mlp1 foci are also formed after Mlp1 overexpression (Strambio-de-Castillia et al., 1999).

Following the reviewer's suggestion, we extended our discussion of basket Nup stoichiometry and organization in the discussion section including most of the citations mentioned as well as the recent articles on the nuclear basket structure and organization (Stankunas & Köhler 2024 1038/s41556-024-01484-x, Singh et al. 2024 10.1016/j.cell.2024.07.020)

Minor Points 1. What is the "lag time" of the doRITE switching? Do the authors believe that it is comparable to the approximate 1-hour timeframe following beta-estradiol induction as shown previously in Chen et al. Nucleic Acids Research, Volume 28, Issue 24, 15 December 2000, Page e108, https://doi.org/10.1093/nar/28.24.e108

We thank the reviewer for suggesting we analyze the kinetics of RITE switching. We carried out quantitative real-time PCR on genomic DNA and found that the half-time of switching is below 20 min. The majority of the population is switched after 1 hour, similar to the results in Chen et al. This data is now included in Supplemental Figure 1A.

1. The authors could consider a brief explanation of radial position (um) for the benefit of the reader, in Figures 1E (right panel) and 2B (right panel), perhaps using a diagram to make it easier to understand the X-axis (um).

To address this, we have now included a diagram and refer to it in the figure legend and the text.

1. In Figure 1G, would the authors consider changing the vertical axis title and the figure legend wording from "mean number of NPCs per cell" to "mean labeled NPC # per cell" to reflect that what is being characterized are the remaining GFP-bearing NPCs over time?

Thank you for spotting this inaccuracy. We have changed the label to "mean # of labeled NPCs per cell".

1. In Figure 2C, the magenta-labeled protein in the micrographs is not described in the figure or the legend.

A description has been added in figure and legend.

1. In Figure S2A, there is an arrow indicating a Nup159 focus, but this is not described in the figure legend, as is done in Figure 2C.

A description has been added to the legend.

1. In Figure S3C, the figure legend does not match the figure. Was this supposed to be designed like Figure 3C and is missing part of the figure? Or is the legend a typographical error?

We apologize for this error and thank the reviewer for spotting it. The legend has been corrected (now Figure S4B).

1. In Figure S4B, the spontaneously recombined RITE (GFP-to-dark) Nup133-V5 appears in the western blot as equally abundant to pre-recombined Nup133-V5-GFP. In the figure legend, this is explained as cells grown in synthetic media without selection to eliminate cells that have lost their resistance marker from the population. In Cheng et al. Nucleic Acids Res. 2000 Dec 15; 28(24): e108, Cre-EBD was not active in the absence of B-estradiol, despite galactose-induced Cre-EBD overexpression. Would the authors be able to comment further on the Cre-Lox RITE system in the manuscript?

We note that also in the cited publication, cells are grown in the presence of selection to select (as stated in this publication) "against pre-excision events that occur because of low but measurable basal expression of the recombinase". Although the authors report that spontaneous recombination is reduced with the b-estradiol inducible system (compared to pGAL expression control of the recombinase only), they show negligible spontaneous recombination only within a two-hour time window. Indeed, we also observe low levels of uninduced recombination on a short timeframe, but occasional events can become significant in longer incubation times (e.g. overnight growth) in the absence of selection. It should be noted that in our system, Cre expression is continuously high (TDH3-promoter) and not controlled by an inducible GAL promoter. We have added the information about the promoter controlling Cre-expression in the methods section.

1. In Figure 6, the authors may want to consider inverting the flow of the cartoon model to start from the wild type condition and apply the deletion mutations at each step to "arrive" at the mutant conditions, rather than starting with mutant conditions and "adding back" proteins.

Following the suggestions of this reviewer as well as reviewer 3, we have modified our model to smore clearly represent the contributions of the different basket components.

Reviewer #1 (Significance (Required)):

Recent work has drawn attention to the fact that not all NPCs are structurally or functionally the same, even within a single cell. In this light, the work here from Zsok et al. is an important demonstration of the kind of methodologies that can shed light on the stability and functions of different subpopulations of NPCs. Altogether, these data are used to support an interesting and topical model for Nup2 and nuclear-basket driven retention of NPCs in non-nucleolar regions of the nuclear envelope.

We thank the reviewer for this positive assessment of our work.

Reviewer #2 (Evidence, reproducibility and clarity (Required)):

In this study, Zsok et al. develop innovative methods to examine the dynamics of individual nuclear pore complexes (NPCs) at the nuclear envelope of budding yeast. The underlying premise is that with the emergence of biochemically distinct NPCs that co-exist in the same cell, there is a need to develop tools to functionally isolate and study them. For example, there is a pool of NPCs that lack the nuclear basket over the nucleolus. Although the nature of this exclusion has been investigated in the past, the authors take advantage of a modification of recombination induced tag exchange (RITE), the slow turnover of scaffold nups, the closed mitosis of budding yeast, and extensive high quality time lapse microscopy to ultimately monitor the dynamics of individual NPCs over the nucleolus. By leveraging genetic knockout approaches and auxin-induced degradation with sophisticated quantitative and rigorous analyses, the authors conclude that there may be two mechanisms dependent on nuclear basket proteins that impact nucleolar exclusion. They also incorporate some computational simulations to help support their conclusions. Overall, the data are of the highest quality and are rigorously quantified, the manuscript is well written, accessible, and scholarly - the conclusions are thus on solid footing.

We thank the reviewer for this assessment.

Reviewer #2 (Significance (Required)):

I have no concerns about the data or the conclusions in this manuscript. However, the significance is not overly clear as there is no major conceptual advance put forward, nor is there any new function suggested for the NPCs over nucleoli. As NPCs are immobile in metazoans, the significance may also be limited to a specialized audience.

We respectfully disagree with this assessment. First, our work demonstrates the use of a novel approach in the application of RITE that can be useful for other researchers in the field of NPC biology and beyond. For example, doRITE could be applied to study the properties of aged NPCs, an area of considerable interest due to links between the NPC and age-related neurodegenerative diseases.

Second, we characterize the interaction between conserved nuclear components, the NPC, the nucleolus and chromatin. While the specific architecture of the nucleus varies between species, many of these interactions are conserved. For example, Nup2's homologue Nup50 also interacts with chromatin in other systems, including mammalian cells, and thus may contribute to regulating the interplay between the nuclear basket and adjoining chromatin. This adds to our understanding of the multiple pathways and interactions that contribute to nuclear organization. Therefore, although the depletion of NPCs from the nucleolar territory in budding yeast may not be of direct importance, understanding the relationships between NPCs and their environment provide insight about nuclear organization throughout different eukaryotic lineages.

In the revised manuscript, we attempt to better highlight and discuss these aspects.

Reviewer #3 (Evidence, reproducibility and clarity (Required)):

The manuscript of Zsok et al. describes the role of nuclear basket proteins in the distribution and mobility of nuclear pore complexes in budding yeast. In particular, the authors showed that the doRITE approach can be used for the analysis of stable and dynamically associated NUPs. Moreover, it can distinguish individual NUPs and follow the inheritance of individual NPCs from mother to daughter cells. The author's findings highlight that Mlp1, Mlp2, and Pml39 are stably associated with the nuclear pore; deletion of Mlp1-Mlp2 and Nup60 leads to the higher NPC density in the nucleolar territory; and NPCs exhibit increased mobility in the absence of the nuclear basket components.

The manuscript contains most figures supporting the data, and data supports the conclusions. However, authors need to include better explanations for figures in the text and figure legends. Lack of detailed explanation can pose challenges for non-experts. In addition, the authors jump over figures and shuffle them through the manuscript, which disrupts the flow and coherence of the manuscript.

We thank the reviewer for pointing this out. In response to the detailed comments given below, we have moved some figures and added more explicit explanations to the text to improve the flow and make it easier to follow. In addition, we have modified the figure legends throughout the manuscript to make them more accessible to the reader.

Major comments: - The nuclear basket contains Nup1, Nup2, Nup60, Mlp1, and Mlp2 in yeast. Nup60 works as a seed for Mlp1/Mlp2 and Nup2 recruitment and plays a key role in the assembly of nuclear pore basket scaffold (PMID: 35148185). Logically, the authors focused primarily on Nup60 in the current manuscript. However, NUP153 has another ortholog of yeast - Nup1, which has not been studied in this work. I recommend adjusting the title of the manuscript to: Nup60 and Mlp1/Mlp2 regulate the distribution and mobility of nuclear pore complexes in budding yeast. I also suggest discussing why work on Nup1 was not included/performed in the manuscript.

We thank the reviewer for suggesting we should test the role of Nup1. Although we had originally not considered it, since we were focusing on the interactors of Mlp1/2, we found that indeed Nup1 also contributes to nucleolar exclusion. We have therefore changed the title to "Nuclear basket proteins regulate the distribution and mobility of nuclear pore complexes in budding yeast".

• Figure 2B: I suggest choosing a more representative image for Pml39. It looks not like a stable component but rather dynamic as NUP60 or Gle1 based on figure showed in Figure 2B.

We thank the reviewer for pointing out this poor choice of panel. We selected a panel for the 14h timepoint that more clearly shows that individual foci can still be seen for Pml39 after this time. Due to its lower copy number, the foci are dimmer for Pml39 than the other stable Nups. Nevertheless, at both the 11 and 14 h timepoint, clear dots can be detected for Pml39, while e.g. Nup116 in the same figure exhibits a more distributed signal and the signal for Nup60 and Gle1 is no longer visible.

• Depletion of AID-tagged proteins needs to be supported by Western blot analysis with protein-specific antibodies, and PCR results should be included in supplementary data to demonstrate the homozygosity of the strains.

The correct genomic tagging of the depleted proteins by AID was confirmed by PCR. We include this PCR analysis for the reviewer below. Since we are working with haploid yeast cells, all strains only carry a single copy of the genes. Unfortunately, we do not have protein-specific antibodies against the depleted proteins. However, other phenotypes support the successful depletion of the protein: Mlp1-mislocalization upon Nup60 depletion, reduced transcript production in Pol II depletion (characterized previously: PMID: 31753862, PMID: 36220102), growth defect upon Nup1 depletion.

• Figure 5B: Snapshots of images from the movie are required. There are no images, only quantifications.

We have replaced the supplemental movie with a movie showing the detection by Trackmate as well as overlaid tracks. As requested, a snapshot of this movie was inserted in figure 5B. We have also moved the example tracks from the supplement to the main figure. Furthermore, we will deposit the tracking dataset in the ETH Research Collection to make it available to the community.

Description of figure legends is more technical than supporting/explaining the figure. For example, below my suggestions for Figure 1D. Please, consider more detailed explanation for other figures. (D) Left: Schematic of the RITE cassette. NUP of interest is tagged with V5 tag and eGFP fluorescent protein where LoxP sites flank eGFP. Before the beta-estradiol-induced recombination, the old NPCs are marked with eGFP signal, whereas new NPCs lack an eGFP signal after the recombination. ORF: open reading frame; V5: V5-tag; loxP: loxP recombination site; eGFP: enhanced green fluorescent protein. Right: doRITE assay schematic of stable or dynamic Nup behavior over cell divisions in yeast after the recombination.

We have modified the figure legends throughout the manuscript to make them more explanatory and helpful for the reader.

In addition, I recommend highlighting the result in the title of the figures. Please, re-consider titles for Figure S3.

We have split this figure to better group related results. The new figures S4 and S5 are entitled: " A RITE(dark-to-GFP) cassette to visualize newly assembled NPC. " and "Mlp1 truncations localize predominantly to non-nucleolar NPCs."

Minor: P.1 Line 31. Extra period symbol before the "(Figure 1A)".

Fixed

P.2 Line 10. Inconsistent writing of PML39 and MLP1. Both genes are capitalized. The same for P.4 Line 16. In some cases all letters are capitalized in other only the first one.

We are following the official yeast gene nomenclature by spelling gene names in italicized capitals and protein names with only the first letter capitalized. We are sorry that this can be confusing for readers more familiar with other model systems.

P.2 Line 18-22. The sentence is too long and hard to read. I recommend splitting it into two sentences.

We agree and have fixed this.

P.2-3 Line 46-47. The sentence is unclear. Suggestion: We expected that successive cell divisions would dilute the signal of labelled and stably associated with the NPC nucleoporins. By contrast, ...

We have modified the sentence to read: "When tagging a Nup that stably associates with the NPC, we expected that successive cell divisions would dilute labelled NPCs by inheritance to both mother and daughter cells leading to a low density of labelled NPCs. By contrast,..."

P.4 Line 17-21. Please, consider adding extra information and clarifying lines 19-21. For example, in Line 19 Figure 2B you can add that the reader needs to compare row 1 and row 4.

Thank you, we have fixed this as suggested.

P. 5 Line 15. When a number begins a sentence, that number should always be spelled out. You can pe-phrase the sentence to avoid it. Also, I recommend adding an explanation/hypothesis of why new NPCs are less frequently detected in nucleolar territory.

We have formatted the text. Interestingly, new NPCs are more frequently detected in the nucleolar territory than old NPCs. We have reformulated this section to make it clearer, also in response to the next comment.

P.5 Line 17-22. I recommend re-phrasing these two sentences. Logically, it is clear that Mlp1/Mlp2 loss mimics "old NPCs" to look more like "new NPCs", and for that reason, they are more frequently included in the nucleolar territory, but it is not clear when you read these two sentences from the first time.

We have reformulated this section to make it clearer.

P6. Line 16. No figure supporting data on graph (Figure 3B).

We have added fluorescent images of the nup2Δ strain to the figure (new Figure 4D).

P.7 Line 10-13. The sentence is unclear.

We have shortened the sentence and moved part of the content to the discussion in the next paragraph.

P.13,14 etc. If 0h timepoint has been used for normalization, why is it present on the graph?

The 0h timepoint is shown for comparison and to illustrate the standard deviation in the data.

P.15. Line 32-33. There is no image here. Potentially wrong description of the figure.

Thank you for spotting this. This was fixed (new Figure S4B).

Figures: - Inconsistent labeling of figures. For example, Fig.1, Fig.1S, Figure 2 etc.

Thank you, this has been corrected.

• Inconsistent labeling of figures. For example, Fig.1 G "mean number of NPCs per cell" - no capitalization of the first letter. Fig.1S "Fraction in population" is capitalized. In general, titles of axis should be capitalized.

Thank you for spotting this. This was fixed.

Suggestions for Figure 1D and Figure 6 are attached as a separate file.

We thank the reviewer for their suggestions to improve these figures. We have taken their recommendation and revised the figures accordingly (see also response to reviewer 1, minor point 8).

Reviewer #3 (Significance (Required)):

Zsok et al. used the recombination-induced tag exchange (RITE) approach, which is an interesting and powerful method to follow individual NUPs over time with respect to their localization and abundance. This approach has been used before in PMID: 36515990 to distinguish pre-existing and newly synthesized Nup2 populations and has been extended to other basket NUPs in this work. Using this method, the authors support the earlier data on basket nucleoporins and highlight new insights on how basket nucleoporins regulate NPCs distribution and mobility. Overall, the manuscript provides new details on the stability of nucleoporins in yeast and how these data align with the mass spectrometry and FRAP data performed earlier in other studies. The limitation of this study is the absence of data on Nup1. It was unclear why these data were not present. Additional data can be included on the dynamics of Pml39, for example, using the FRAP method. The dynamic of Pml39 at the pore was shown only using the doRITE method.

As suggested, we have tested the role of Nup1 (see above).

Unfortunately, we are not able to provide orthologous data for the dynamics of Pml39. As we discuss in the manuscript, FRAP is not suitable for the analysis of the dynamics of most nucleoporins in yeast due to the high lateral mobility of NPCs in the nuclear envelope and has previously generated misleading results for Mlp1. Furthermore, the low expression levels of Pml39 will make it difficult to obtain reliable FRAP curves for this protein. We therefore do not think that adding FRAP experiments with Pml39 will provide valuable insight.

However, in addition to the Pml39 doRITE result itself, our observation that the Pml39-dependent pool of Mlp1 exhibits stable association with the NPC supports the interpretation of Pml39 as a stable protein as well.

In general, this study represents a unique research study of basic research on nuclear pore proteins that will be of general interest to the nuclear transport field.

Field of expertise: nuclear-cytoplasmic transport, nuclear pore, inducible protein degradation. I do not have sufficient expertise in ExTrack.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #3

Evidence, reproducibility and clarity

The manuscript of Zsok et al. describes the role of nuclear basket proteins in the distribution and mobility of nuclear pore complexes in budding yeast. In particular, the authors showed that the doRITE approach can be used for the analysis of stable and dynamically associated NUPs. Moreover, it can distinguish individual NUPs and follow the inheritance of individual NPCs from mother to daughter cells. The author's findings highlight that Mlp1, Mlp2, and Pml39 are stably associated with the nuclear pore; deletion of Mlp1-Mlp2 and Nup60 leads to the higher NPC density in the nucleolar territory; and NPCs exhibit increased mobility in the absence of the nuclear basket components.

The manuscript contains most figures supporting the data, and data supports the conclusions. However, authors need to include better explanations for figures in the text and figure legends. Lack of detailed explanation can pose challenges for non-experts. In addition, the authors jump over figures and shuffle them through the manuscript, which disrupts the flow and coherence of the manuscript.

Major comments:

• The nuclear basket contains Nup1, Nup2, Nup60, Mlp1, and Mlp2 in yeast. Nup60 works as a seed for Mlp1/Mlp2 and Nup2 recruitment and plays a key role in the assembly of nuclear pore basket scaffold (PMID: 35148185). Logically, the authors focused primarily on Nup60 in the current manuscript. However, NUP153 has another ortholog of yeast - Nup1, which has not been studied in this work. I recommend adjusting the title of the manuscript to: Nup60 and Mlp1/Mlp2 regulate the distribution and mobility of nuclear pore complexes in budding yeast. I also suggest discussing why work on Nup1 was not included/performed in the manuscript.
• Figure 2B: I suggest choosing a more representative image for Pml39. It looks not like a stable component but rather dynamic as NUP60 or Gle1 based on figure showed in Figure 2B.
• Depletion of AID-tagged proteins needs to be supported by Western blot analysis with protein-specific antibodies, and PCR results should be included in supplementary data to demonstrate the homozygosity of the strains.
• Figure 5B: Snapshots of images from the movie are required. There are no images, only quantifications.
• Description of figure legends is more technical than supporting/explaining the figure. For example, below my suggestions for Figure 1D. Please, consider more detailed explanation for other figures. (D) Left: Schematic of the RITE cassette. NUP of interest is tagged with V5 tag and eGFP fluorescent protein where LoxP sites flank eGFP. Before the beta-estradiol-induced recombination, the old NPCs are marked with eGFP signal, whereas new NPCs lack an eGFP signal after the recombination. ORF: open reading frame; V5: V5-tag; loxP: loxP recombination site; eGFP: enhanced green fluorescent protein. Right: doRITE assay schematic of stable or dynamic Nup behavior over cell divisions in yeast after the recombination.

In addition, I recommend highlighting the result in the title of the figures. Please, re-consider titles for Figure S3.

Minor:

P.1 Line 31. Extra period symbol before the "(Figure 1A)".

P.2 Line 10. Inconsistent writing of PML39 and MLP1. Both genes are capitalized. The same for P.4 Line 16. In some cases all letters are capitalized in other only the first one.

P.2 Line 18-22. The sentence is too long and hard to read. I recommend splitting it into two sentences.

P.2-3 Line 46-47. The sentence is unclear. Suggestion: We expected that successive cell divisions would dilute the signal of labelled and stably associated with the NPC nucleoporins. By contrast, ...

P.4 Line 17-21. Please, consider adding extra information and clarifying lines 19-21. For example, in Line 19 Figure 2B you can add that the reader needs to compare row 1 and row 4.

P. 5 Line 15. When a number begins a sentence, that number should always be spelled out. You can pe-phrase the sentence to avoid it. Also, I recommend adding an explanation/hypothesis of why new NPCs are less frequently detected in nucleolar territory.

P.5 Line 17-22. I recommend re-phrasing these two sentences. Logically, it is clear that Mlp1/Mlp2 loss mimics "old NPCs" to look more like "new NPCs", and for that reason, they are more frequently included in the nucleolar territory, but it is not clear when you read these two sentences from the first time.

P6. Line 16. No figure supporting data on graph (Figure 3B).

P.7 Line 10-13. The sentence is unclear.

P.13,14 etc. If 0h timepoint has been used for normalization, why is it present on the graph?

P.15. Line 32-33. There is no image here. Potentially wrong description of the figure.

Figures:

• Inconsistent labeling of figures. For example, Fig.1, Fig.1S, Figure 2 etc.
• Inconsistent labeling of figures. For example, Fig.1 G "mean number of NPCs per cell" - no capitalization of the first letter. Fig.1S "Fraction in population" is capitalized. In general, titles of axis should be capitalized.

Suggestions for Figure 1D and Figure 6 are attached as a separate file.

Significance

Zsok et al. used the recombination-induced tag exchange (RITE) approach, which is an interesting and powerful method to follow individual NUPs over time with respect to their localization and abundance. This approach has been used before in PMID: 36515990 to distinguish pre-existing and newly synthesized Nup2 populations and has been extended to other basket NUPs in this work. Using this method, the authors support the earlier data on basket nucleoporins and highlight new insights on how basket nucleoporins regulate NPCs distribution and mobility. Overall, the manuscript provides new details on the stability of nucleoporins in yeast and how these data align with the mass spectrometry and FRAP data performed earlier in other studies. The limitation of this study is the absence of data on Nup1. It was unclear why these data were not present. Additional data can be included on the dynamics of Pml39, for example, using the FRAP method. The dynamic of Pml39 at the pore was shown only using the doRITE method.

In general, this study represents a unique research study of basic research on nuclear pore proteins that will be of general interest to the nuclear transport field.

Field of expertise: nuclear-cytoplasmic transport, nuclear pore, inducible protein degradation. I do not have sufficient expertise in ExTrack.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

In this study, Zsok et al. develop innovative methods to examine the dynamics of individual nuclear pore complexes (NPCs) at the nuclear envelope of budding yeast. The underlying premise is that with the emergence of biochemically distinct NPCs that co-exist in the same cell, there is a need to develop tools to functionally isolate and study them. For example, there is a pool of NPCs that lack the nuclear basket over the nucleolus. Although the nature of this exclusion has been investigated in the past, the authors take advantage of a modification of recombination induced tag exchange (RITE), the slow turnover of scaffold nups, the closed mitosis of budding yeast, and extensive high quality time lapse microscopy to ultimately monitor the dynamics of individual NPCs over the nucleolus. By leveraging genetic knockout approaches and auxin-induced degradation with sophisticated quantitative and rigorous analyses, the authors conclude that there may be two mechanisms dependent on nuclear basket proteins that impact nucleolar exclusion. They also incorporate some computational simulations to help support their conclusions. Overall, the data are of the highest quality and are rigorously quantified, the manuscript is well written, accessible, and scholarly - the conclusions are thus on solid footing.

Significance

I have no concerns about the data or the conclusions in this manuscript. However, the significance is not overly clear as there is no major conceptual advance put forward, nor is there any new function suggested for the NPCs over nucleoli. As NPCs are immobile in metazoans, the significance may also be limited to a specialized audience.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

The authors extended the existing recombination-induced tag exchange (RITE) technology to show that they can image a subset of NPCs, improving signal-to-noise ratios for live cell imaging in yeast, and to track the stability or dynamics of specific nuclear pore proteins across multiple cell divisions. Further, the authors use this technology to show that the nuclear basket proteins Mlp1, Mlp2 and Pml39 are stably associated with "old NPCs" through multiple cell cycles. The authors show that the presence of Mlp1 in these "old NPCs" correlates with exclusion of Mlp1-positive NPCs from the nucleolar territory. A surprising result is that basket-less NPCs can be excluded from the non-nucleolar region, an observation that correlates with the presence of Nup2 on the NPC regardless of maturation state of the NPC. In support of the proposal that retention of NPCs via Mlp1 and Nup2 in non-nucleolar regions, simulation data is presented to suggest that basket-less NPCs diffuse faster in the plane of the nuclear envelope.

However, there are some points that do need addressing:

Major Points

1. Taking into account that the Nup2 result in Figure 4B forms the basis for one half of the proposed model in Figure 6 regarding the exclusion of NPCs from the nucleolar region of the NE, there is a relatively small amount of data in support of this finding and this proposed model. For example, the only data for Nup2 in the manuscript is a column chart in Figure 4B with no supporting fluorescence microscopy examples for any Nup2 deletion. Further, the Nup60 deletion mutant will have zero basket-containing NPCs, whereas the Nup2 deletion will be a mixture of basket-containing and basket-less NPCs. The only support for the localization of basket-containing NPCs in the Nup2 deletion mutant is through a reference "Since Mlp1-positive NPCs remain excluded from the nucleolar territory in nup2Δ cells (Galy et al., 2004), the homogenous distribution observed in this mutant must be caused predominantly by the redistribution of Mlp-negative NPCs into the nucleolar territory."
2. The authors could consider utilizing this opportunity to discuss their technological innovations in the context of the prior work of Onischenko et al., 2020. This work is referenced for the statement "RITE can be used to distinguish between old and new NPCs" Page 2, Line 43. However, it is not referenced for the statement "We constructed a RITE-cassette that allows the switch from a GFP-labelled protein to a new protein that is not fluorescently labelled (RITE(GFP-to-dark))" despite Onischenko et al., 2020 having already constructed a RITE-cassette for the GFP-to-dark transition. The authors could consider taking this opportunity to instead focus on their innovative approach to apply this technology to decrease the number of fluorescently-tagged NPCs by dilution across multiple cell divisions and to interpret this finding as a measure of the stability of nuclear pore proteins within the broader NPC.
3. The authors could also consider taking this opportunity to discuss their results in the context of the Saccharomyces cerevisiae nuclear pore complex structures published e.g. in Kim et al., 2018, Akey et al., 2022, Akey et al., 2023 in which the arrangement of proteins in the nuclear basket is presented, and also work from the Kohler lab (Mészáros et al., 2015) on how the basket proteins are anchored to the NPC. There is additional literature that also might help provide some perspective to the findings in the current manuscript, such as the observation that a lesser amount of Mlp2 to Mlp1 observed is consistent with prior work (e.g. Kim et al., 2018) and that intranuclear Mlp1 foci are also formed after Mlp1 overexpression (Strambio-de-Castillia et al., 1999).

Minor Points

1. What is the "lag time" of the doRITE switching? Do the authors believe that it is comparable to the approximate 1-hour timeframe following beta-estradiol induction as shown previously in Chen et al. Nucleic Acids Research, Volume 28, Issue 24, 15 December 2000, Page e108, https://doi.org/10.1093/nar/28.24.e108
2. The authors could consider a brief explanation of radial position (um) for the benefit of the reader, in Figures 1E (right panel) and 2B (right panel), perhaps using a diagram to make it easier to understand the X-axis (um).
3. In Figure 1G, would the authors consider changing the vertical axis title and the figure legend wording from "mean number of NPCs per cell" to "mean labeled NPC # per cell" to reflect that what is being characterized are the remaining GFP-bearing NPCs over time?
4. In Figure 2C, the magenta-labeled protein in the micrographs is not described in the figure or the legend.
5. In Figure S2A, there is an arrow indicating a Nup159 focus, but this is not described in the figure legend, as is done in Figure 2C.
6. In Figure S3C, the figure legend does not match the figure. Was this supposed to be designed like Figure 3C and is missing part of the figure? Or is the legend a typographical error?
7. In Figure S4B, the spontaneously recombined RITE (GFP-to-dark) Nup133-V5 appears in the western blot as equally abundant to pre-recombined Nup133-V5-GFP. In the figure legend, this is explained as cells grown in synthetic media without selection to eliminate cells that have lost their resistance marker from the population. In Cheng et al. Nucleic Acids Res. 2000 Dec 15; 28(24): e108, Cre-EBD was not active in the absence of B-estradiol, despite galactose-induced Cre-EBD overexpression. Would the authors be able to comment further on the Cre-Lox RITE system in the manuscript?
8. In Figure 6, the authors may want to consider inverting the flow of the cartoon model to start from the wild type condition and apply the deletion mutations at each step to "arrive" at the mutant conditions, rather than starting with mutant conditions and "adding back" proteins.

Significance

Recent work has drawn attention to the fact that not all NPCs are structurally or functionally the same, even within a single cell. In this light, the work here from Zsok et al. is an important demonstration of the kind of methodologies that can shed light on the stability and functions of different subpopulations of NPCs. Altogether, these data are used to support an interesting and topical model for Nup2 and nuclear-basket driven retention of NPCs in non-nucleolar regions of the nuclear envelope.

Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

Learn more at Review Commons

---

Reply to the reviewers

Reviewer #1

Evidence, reproducibility, and clarity

Singh et al. analyze the expression and putative contribution of TEs in CD4+ T cells in HIV elite controllers. Through re-analysis of existing datasets, the authors describe broad differences in expression of TEs in ECs through analysis of RNA-seq and ATAC-seq data and come up with convincing examples where differentially-expressed innate immune genes correlate with increased accessibility of proximal TEs. Overall, the authors' conclusions are appropriately measured, though the manuscript text should be re-organized for clarity and a few further analyses are needed to support the main message of the paper.

Major comments

The manuscript would benefit from a re-organization of the figures to focus on TEs - in particular, Fig 1B, Fig 2, and Fig 3 reproduce known transcriptional differences between ECs and HCs and serve as quality controls for the authors' computational analysis. Conversely, Supplementary Fig 6 contains very interesting data on KZNF expression and should be included in the main figures.

Authors: Thank you for the suggestion. We agree that Figure S6 should be featured more prominently in the manuscript. Accordingly, we have now incorporated it into the main text as Figure 6. The TE-KZNF correlation plots, previously Figure 5C, have been relocated to this new figure to provide a cohesive presentation of all KZNF-related data within the same figure.

We’ve chosen to keep Figures 1B, 2, and 3 in their original places. We contend that they provide a foundational view of transcriptional variances in gene expression between patient groups, encompassing both previously identified and novel DEGs, which we believe warrants their placement in the main text. Furthermore, they serve as robust quality control measures for subsequent TE-centric transcriptional analyses. Given that there is no limitation in the number of figures in Genome Biology articles, we think it’s adequate to retain them as main figures.

It remains unclear whether differences in TE expression described are specific to ECs or to EC-like CD4+ T cell states. As there are plenty of datasets available that compare the transcriptome of naïve, activated, exhausted, and regulatory CD4+ T cells, the authors should compare the TE expression patterns observed in ECs to activated CD4+ T cells, particularly those with a Th1 and cytotoxic phenotype analogous to those observed in ECs, from healthy donors.

Authors: We thank the reviewer for this constructive suggestion to further study the foundations of HIV-1 elite control. In our initial study, we demonstrate that PBMCs from elite controllers (ECs) exhibit a heightened proportion of activated CD4+ T cells compared to PBMCs of healthy controls (HCs) and a heightened proportion of macrophages, naïve CD4+ T cells, and NK cells compared to PBMCs of treatment-naïve viremic progressors (VPs) (Figure 2D). Additionally, through clustering analysis of deconvoluted CD4+ T cell samples from elite controllers, we ascertain that the clustering pattern is not predicated on the CD4+ T cell subtype (Figure 3B). To further explore the reviewer’s inquiry, we compared the TE expression profile of ECs with that of unstimulated and stimulated CD4+ T cell subsets from HCs (data source: PMID 31570894), integrated into the revised manuscript as Figure S3B.

“Unsupervised clustering of these samples shows that the TE expression pattern of ECs is most similar to that of Th2 progenitor cells, which are associated with HIV-1-specific adaptive immune responses (61). Still, we observed that, for the majority of families, TE expression was higher on average in all EC CD4+ T cell subsets than in CD4+ T cell subsets from HCs, regardless of stimulation (Figure S3B). While a subset of TE families exhibited an expression pattern in ECs similar to that of activated CD4+ T cells of HCs (e.g., high expression of L1s and THE1B), multiple TE families appear to be upregulated in an EC-specific way (e.g., LTR12C and LTR7). Together, these findings underscore the unique immune cell composition, transcriptome, and retrotranscriptome of ECs.” [pg.13-14, L226-235]

While these observations are interesting, pursuing this question further falls beyond the scope of our study, as we note in the Discussion of the revised manuscript. We believe the reviewer’s inquiry pertains to a distinct research question, namely whether the potential for elite control of HIV-1 infection manifests as a detectable phenotype pre-infection within healthy CD4 T cell subsets (i.e., EC-like CD4+ T cell states) or is a unique phenotype that emerges solely after HIV-1 infection.

“Another outstanding question is whether the gene and TE signatures revealed by our analysis of ECs exist in the general population independent of HIV-1 infection or if they are driven by the initial infection. While this inquiry is beyond the scope of this study, we have presented here evidence of common TE signatures between EC CD4+ T cells and Th2 progenitors from HCs (Figure S3B) and established that ECs possess a unique CD4+ T cell retrotranscriptome with potential implications for natural HIV-1 control. Future studies designed to assess elite control prediction should explore whether these TE profiles can serve as predictive variables for whether an individual displays enhanced viral control.” [pg. 38, L663-671]

Therefore, while we appreciate the reviewer's suggestion and offer the addition of these preliminary findings, we believe that further investigation would be better suited for future studies specifically designed to address that question. Our manuscript aims to provide insight into the retrotranscriptome dynamics in ECs and their potential implications for natural HIV-1 control.

In Fig 1, the authors demonstrate differential expression of both innate immune genes and TEs, but the link between the two is unclear. Is there any enrichment in differential expression for TEs located proximal to innate immune genes? This type of analysis should be possible using the authors' own software to map TE expression to specific genomic loci.

__Authors: __Thank you for this excellent question. To answer this inquiry, we used the paired ATAC-seq and RNA-seq datasets for from ECs and HCs (used in Figures 1 and 4) to produce a new list of TE-gene pairs on which we could perform gene set enrichment analysis, the results of which have been integrated into the revised manuscript as Figure 4A.

“We used paired ATAC-seq – which measures chromatin accessibility – and RNA-seq datasets for ECs (n=4) and HCs (n=4) to create a list of TE-gene pairs where the TE locus and gene show increased accessibility and expression, respectively, in ECs compared to HCs (Table S7, see Methods for details). These loci and genes were paired based on proximity, with a maximum distance of 10kb between the TE locus and the gene’s transcription start site, to increase the likelihood of a direct cis-regulatory influence of the TE over the nearby gene. Subsequent gene set enrichment analysis revealed that these genes were predominantly involved in cellular activation, cytokine production, and immune response regulation (Figure 4A). The enrichment for differential accessibility of TE loci near genes involved in these pathways suggests that the distinct TE landscape observed in ECs may contribute significantly to a unique immune regulome in these individuals.” [pg. 21, L357-368]

Thus, we conclude that yes, there is an enrichment for immune-related genes with higher expression in ECs, proximal to differentially accessible TEs. We highlight six of these TE-gene pairs in Figure 4B-C. While we have high confidence in our analyses, future experimental validation is needed to confirm these regulatory relationships.

Optional: In Fig 3, the authors cluster CD4+ T cells based on transcriptomic profiles. It would be interesting to re-cluster these samples based on TE expression alone, given the differences in TE expression described in Fig 5.

__Authors: __Thank you for the suggestion. We agree that it would be valuable to assess how the EC clustering is altered when considering TE expression alone, as opposed to combining gene and TE family expression. To address this, we used the same graph-based k-nearest neighbors method to re-cluster the EC CD4+ T cell RNA-seq samples based only on locus-level TE expression, integrated into the revised manuscript as Figure S7.

“To further explore locus-level expression patterns, we re-clustered the same EC samples (n=128) using only locus-level TE expression. This again resolved four EC clusters (Figure S7A), which interestingly appeared even more distinct than those identified by gene and TE family expression (Figure 3A). The TE locus-based clusters (TL-Cs) aligned well with the gene and TE family clusters (GT-Cs), with an average 70% overlap in samples between each GT-C and its corresponding TL-C (Figure S7B), indicating high consistency (Table S8). The remaining 30% of samples that shifted between clusters did so consistently within individuals, not cohorts, maintaining heterogeneous TL-C compositions similar to the GT-Cs (Figures S7C & S5A). An exception to this heterogeneity was TL-C4, comprising 22 samples from GT-C1 that were almost entirely from the CD4+ T cell subsets of only four participants in the Jiang cohort (Figure S7C, Table S8). No other samples from the Jiang cohort shifted to this cluster from other GT-Cs, suggesting that these patterns reflect individual variation rather than cohort bias. Like the GT-Cs, each TL-C included samples from all five CD4+ T cell subsets and was largely heterogeneous (Figure S7C). Notably, TL-C2 mirrored corresponding GT-C3 in its overrepresentation of EM and TM cells, while TL-C1 uniquely showed an overrepresentation of naïve CD4+ T cells. Beyond sample composition, each TL-C was characterized by a unique pattern of expressed TE loci (Figure S7D). These signatures were heterogeneous across families, with subsets of variable loci from one TE family marking separate clusters (Figure S7E), some of which did not reach the threshold of significance in earlier analyses when analyzed at the family-level, like SVA-D. Many families maintained their cluster-specific signatures, like THE1B (a marker of GT-C2), for which the majority of variable loci were found in corresponding TL-C1. However, some TE families, like the L1s that marked GT-C1, showed more heterogeneous signatures with variable loci marking multiple TL-Cs. These findings underscore the need for future locus-level investigations with high-depth sequencing to fully capture the complexity of TE expression.” [pg. 27-28, L462-488]

We believe these findings not only validate the distinct clustering patterns observed but also highlight the potential of locus-level TE analysis to reveal additional layers of retrotranscriptomic diversity in EC CD4+ T cells.

Significance

The manuscript by Singh et al. describes for the first time the role of TEs in HIV elite controllers, suggesting that TEs may be co-opted for cis-regulatory function. This study builds off prior work demonstrating that HIV-infected CD4+ T cells activate LTR elements that may regulate the expression of interferon-inducible genes, demonstrating that ECs show further upregulation of innate immune genes. While these findings will need to be experimentally validated, this study constitutes a useful resource and adds to the growing body of evidence implicating TEs in cis-regulatory control of immune genes. This study will be of interest to basic scientists interested in genetic mechanisms of HIV control, and if further developed may comprise a useful source of biomarkers to predict viral kinetics in HIV-infected individuals. My expertise is in immunology, TE biology, and viral infection.

Authors: We greatly appreciate this positive evaluation of our manuscript and recognition of its significance in uncovering novel evidence of TE co-option for immune regulatory function in HIV-1 elite control, as well as the suggestion of promising avenues for future research in this field.

Reviewer #2

Evidence, reproducibility and clarity

The authors have re-analyzed published RNA-Seq data from CD4 T cells isolated from HIV elite controllers and reference cohorts, including HIV negative persons, viremic progressors and ART-treated persons. Their main finding is that in some of their comparisons, EC have higher levels of interferon-stimulated genes (ISG), paired with distinct expression patterns of transposable elements. The authors suggest that expression of transposable elements may induce altered expression of ISG, presumably due to immune recognition of TE. They also suggest that reduced expression of KZNF genes, which encode for transcription factors that can suppress TE, may be responsible for enhanced expression of TE. I have the following comments:

1. All data included in this manuscript derive from previously published data. A new dataset, specifically designed to focus on a high-resolution analysis of TE expression, would be better suited to address the proposed questions.

Authors: We agree that a new dataset tailored specifically for high-resolution analysis of TE expression would be optimal for addressing the proposed inquiries, and we emphasize this point in the Discussion of the revised manuscript.

“We found that distinct sets of innate immunity genes and restriction factors are upregulated in different EC clusters even in the absence of active viremia, suggesting that elevated basal expression of these factors plays a previously underappreciated role in the EC phenotype. Further studies will be necessary to cement this idea and would especially benefit from the integration of single-cell omics to dissect TE regulation and clustering in deconvoluted CD4+ T cells of ECs. We also acknowledge that our study is limited by the small number of EC individuals with available omics data, which likely limited our ability to identify significant relationships between transcriptome clustering and available participant metadata (Figure S5). While the rarity of ECs in the seropositive population makes it challenging to study this phenotype, the transcriptomic heterogeneity revealed by our analyses underscores the need for surveying larger and more diverse EC cohorts.” [pg. 37-38, L651-662]

Regrettably, we do not have access to elite controller samples (which are exceedingly rare), and as such the addition of a novel dataset was not feasible within the scope of this revision. Nevertheless, we assert that the publicly available sequencing data analyzed here is robust and suitable for locus- and family-level TE analysis. All sequencing runs were paired-end and of high depth, ensuring proper alignment to and high coverage of TEs at a locus-specific resolution. Additionally, we use in-house pipelines curated for TE analysis, to optimize the accuracy and quantity of TE-assigned reads (see Methods and our GitHub Repository for more details).

Authors: We agree that a new dataset tailored specifically for high-resolution analysis of TE expression would be optimal for addressing the proposed inquiries, and we emphasize this point in the Discussion of the revised manuscript.

“We found that distinct sets of innate immunity genes and restriction factors are upregulated in different EC clusters even in the absence of active viremia, suggesting that elevated basal expression of these factors plays a previously underappreciated role in the EC phenotype. Further studies will be necessary to cement this idea and would especially benefit from the integration of single-cell omics to dissect TE regulation and clustering in deconvoluted CD4+ T cells of ECs. We also acknowledge that our study is limited by the small number of EC individuals with available omics data, which likely limited our ability to identify significant relationships between transcriptome clustering and available participant metadata (Figure S5). While the rarity of ECs in the seropositive population makes it challenging to study this phenotype, the transcriptomic heterogeneity revealed by our analyses underscores the need for surveying larger and more diverse EC cohorts.” [pg. 37-38, L651-662]

Regrettably, we do not have access to elite controller samples (which are exceedingly rare), and as such the addition of a novel dataset was not feasible within the scope of this revision. Nevertheless, we assert that the publicly available sequencing data analyzed here is robust and suitable for locus- and family-level TE analysis. All sequencing runs were paired-end and of high depth, ensuring proper alignment to and high coverage of TEs at a locus-specific resolution. Additionally, we use in-house pipelines curated for TE analysis, to optimize the accuracy and quantity of TE-assigned reads (see Methods and our GitHub Repository for more details).

1. As the authors acknowledge, the described investigations are exploratory, and do not allow to draw firm conclusions. Mechanistic experiments are recommended to address the authors' hypotheses.

Authors: We agree and have duly acknowledged throughout the Discussion the exploratory nature of our investigations and the need for future mechanistic experiments to validate our model. Below are passages from the revised manuscript which we’ve added to emphasize these points.

“These findings underscore the need for future locus-level investigations with high-depth sequencing to fully capture the complexity of TE expression.” [pg. 28, L486-488]

“Each step in the model will require experimental work to be validated. First and foremost, it will be important to confirm that the TEs exhibiting increased transcript levels and accessibility in ECs are indeed boosting the innate immune response and control of HIV-1 in these individuals.” [pg. 34, L583-586]

“CRISPR-Cas9 editing was used in cell lines to demonstrate that a subset of MER41 elements function as enhancers driving the interferon-inducibility of several innate immune genes. However, the specific MER41 loci we identified here as differentially active in ECs have not been tested experimentally for enhancer activity. Thus, further work is warranted to confirm the regulatory function of these loci under the control of STAT1 or other immune TFs, as well as other TE families identified as targets of immune-related TFs (Figure S8).” [pg. 35, L594-600]

“Overall, our results reinforce the concept that TEs are important players in the human antiviral response (25,93) and uncover specific candidate elements for boosting cellular defenses against HIV-1 in ECs. We acknowledge that these associations are drawn from correlative patterns and manipulative experiments are needed to infer causality between chromatin changes at these TEs and increased expression of nearby immunity genes.” [pg. 36, L618-623]

“Further work is needed to validate TE-KZNF regulatory interactions in T cells, probe their connection to epigenetic variation at individual TE loci, and explore their repercussions on gene expression variation in CD4+ T cells, with and without HIV-1 infection.” [pg. 40, L715-718]

Thus, while we appreciate and agree with the suggestion of experimental validation, we contend that these experiments fall beyond the scope of the present study, which is a computational investigation providing insight into the EC retrotranscriptome and its potential implications for natural HIV-1 control.

1. An important limitation is that virological data of EC are not considered. For example, I believe it is a lot more likely that the upregulation of ISG in EC relates to ongoing low-level viral replication. The authors could analyze cell-associated HIV RNA and DNA levels and determine how they associate with ISG expression.

Authors: Thank you for bringing up this important consideration. It's worth noting that the public datasets used in our study reported undetectable viremia in the EC volunteers (PMIDs 30964004, 29269040, 32848246, 27453467). Nonetheless, we sought to address this limitation and explore the potential association between ISG expression and viremia as recommended by the reviewer. These analyses were integrated into the revised manuscript as Figure S6.

“To exclude the possibility that these gene expression signatures in ECs are associated with viremia, we quantified HIV-1 transcript levels in deconvoluted CD4+ T cell RNA-seq samples from ECs and ART-treated PLWH for comparison. In the original studies, all samples were reported to have undetected viremia by blood tests (9,37-39). Consistent with this, we found that the vast majority of the EC and ART samples taken from PBMCs exhibited very low HIV-1 transcript levels, with TPM values generally below 1. However, in samples originating from the lymph nodes of EC individuals (n = 22) (37), we detected HIV-1 expression in some subsets (Figure S6A&B). In agreement with the corresponding study (37), we found elevated HIV-1 transcript levels in germinal center and non-germinal center T follicular helper cells (GC Tfh & nGC Tfh, not included in our clustering analyses) -- and to a lesser extent in T effector memory (EM) cells (Figure S6A, average TPM This added analysis confirms that the increased expression of ISGs in ECs is not correlated with virological transcription and is therefore likely not to be driven by viremia.

1. KZNF genes seem downregulated in EC. Can the authors propose a reason/mechanism for that?

Authors: There is the possibility that KZNF regulatory loops are the cause of their transcriptional downregulation, which has been documented in embryogenesis (PMID 31006620) and cancer (PMID 33087347). We’ve incorporated this hypothesis into the Discussion as an additional consideration for the reader.

“These observations suggest that interindividual variation in KZNF expression in CD4+ T cells could explain why certain TEs are variably expressed and accessible across ECs. But what are the mechanisms underlying variation in ZNF expression? It is possible that TE-KZNF regulatory loops are involved, in which a copy of the TE family targeted by a KZNF is inserted near and regulates the KZNF gene, thereby introducing a negative feedback loop. This phenomenon has been documented in prior studies of KZNF activity in embryogenesis (51) and cancer (115).” [pg. 39-40, L705-711]

While we believe this is a viable hypothesis, it requires further experimentation to confirm the existence of this phenomenon and its impacts in the context of immune cells.

Significance

Overall, I think this is an interesting manuscript that proposes distinct and potentially important mechanisms that may contribute to immune control of HIV. My suggestions to improve the manuscript are complex and cannot be easily addressed through experimental work. I believe a possible option would be to publish the present manuscript without my proposed modifications but highlight the weaknesses of the current paper more clearly; mechanistic studies could then be deferred to a future study.

Authors: We appreciate the reviewer's positive assessment of our manuscript and their recognition of its significance in elucidating novel TE-derived mechanisms that may contribute to natural HIV-1 control. We agree that mechanistic studies are required to test our predictions. As the reviewer suggests, these would be complex experiments that we feel fall beyond the scope of this study. With the additions detailed above in response to the reviewer’s point #2, we believe that we have clearly highlighted the limitations of our work and emphasized the need for future experimentation to validate our findings.

Reviewer #3

Evidence, reproducibility, and clarity

Summary: This manuscript presents an analysis of published gene expression (RNA-seq and ATAC-seq) data from a couple of cohorts of HIV-infected elite controllers (EC), as compared to uninfected controls, (HC), virological progressors (VP). The authors report that HIV elite controllers may exhibit 4 distinct patterns of TE (and gene) expression and suggest that TE expression may drive some form of antiviral gene expression. Further, they show that heterogeneous TE expression may be determined by differential KZNF gene activity among the different clusters of elite controllers. These results are very interesting, even though the conclusions are very preliminary. It presents intriguing correlations between expression of certain TE groups of LINES and HERVs, and the clustering into 4 gene expression groups in EC and is a novel finding. That said, correlation is not causation, and the authors need to be more cautious in presenting their highly preliminary model in Figure 6.

Authors: We are grateful for the reviewer's insightful assessment of our manuscript, acknowledging the novelty and interest of our findings regarding TE expression patterns in HIV-1 elite controllers. We also appreciate their constructive feedback regarding the cautious interpretation of preliminary conclusions. In the revised manuscript, we have underscored the exploratory nature of our investigations and the need for future mechanistic experiments to validate our model.

“These findings underscore the need for future locus-level investigations with high-depth sequencing to fully capture the complexity of TE expression.” [pg. 28, L486-488]

“Each step in the model will require experimental work to be validated. First and foremost, it will be important to confirm that the TEs exhibiting increased transcript levels and accessibility in ECs are indeed boosting the innate immune response and control of HIV-1 in these individuals.” [pg. 34, L583-586]

“CRISPR-Cas9 editing was used in cell lines to demonstrate that a subset of MER41 elements function as enhancers driving the interferon-inducibility of several innate immune genes. However, the specific MER41 loci we identified here as differentially active in ECs have not been tested experimentally for enhancer activity. Thus, further work is warranted to confirm the regulatory function of these loci under the control of STAT1 or other immune TFs, as well as other TE families identified as targets of immune-related TFs (Figure S8).” [pg. 35, L594-600]

“Overall, our results reinforce the concept that TEs are important players in the human antiviral response (25,93) and uncover specific candidate elements for boosting cellular defenses against HIV-1 in ECs. We acknowledge that these associations are drawn from correlative patterns and manipulative experiments are needed to infer causality between chromatin changes at these TEs and increased expression of nearby immunity genes.” [pg. 36, L618-623]

“Further work is needed to validate TE-KZNF regulatory interactions in T cells, probe their connection to epigenetic variation at individual TE loci, and explore their repercussions on gene expression variation in CD4+ T cells, with and without HIV-1 infection.” [pg. 40, L715-718]

We hope these passages provide sufficient caution and clarity in the presentation of our scientific inquiry.

Major comments:

Overall, although preliminary, as the authors note, the results are interesting and worthy of follow-up. At this point, however, a number of issues arise that need further clarification and analysis before I would consider this study complete.

First, the analyses shown in Figures 3-5 based on data from studies on EC of CD4 cells are apparently motivated by the differential TE expression in total PBMCs shown in Fig 1 and 2. Yet, the TE groups (please don't use taxonomic terms like "subfamily") identified in Fig 2 and Fig 4 are completely different, with no overlap. This discrepancy underscores the possibility that the differential expression observed is, at least in part, due to the differences among the groups or clusters in cell type composition, as seen in Fig 2D and 3B which, themselves, could be a consequence of HIV infection and elite control (which has been shown to involve ongoing, albeit low-level, virus replication). This issue must be addressed.

Authors: Thank you for the suggestion. First, we’d like to clarify that the data used in Figures 1 and 2 were not both derived from PBMCs. Figures 1 and S1 examine the differential expression of TEs in EC CD4+ T cells compared to HCs and ART-treated PLWH, respectively. Figure 2 examines differential expression of TEs in EC PBMCs compared to treatment-naïve VPs. Second, regarding Figure 4B-C, the TE loci that we chose to highlight were not based on our results from the PBMC analysis in Figure 2, which is why there is no overlap in the TE families presented. Instead, we selected those TE-gene pairs based on 1) known function of the genes in immunity and/or HIV-1 restriction, 2) known contribution of the TE families to immunity, and 3) differential accessibility and expression of the TEs and genes respectively in ECs compared to HCs. Thus, Figure 4B-C represents select examples that we deemed particularly relevant to the EC phenotype. We have revised the manuscript to better explain the process of TE-gene pair identification and the rationale behind our selection for Figure 4B-C.

“We used paired ATAC-seq – which measures chromatin accessibility – and RNA-seq datasets from the CD4+ T cells of ECs (n=4) and HCs (n=4) (39) to create a list of TE-gene pairs where the TE locus and gene show increased accessibility and expression, respectively, in ECs compared to HCs (Table S7, see Methods for details). These loci and genes were paired based on proximity, with a maximum distance of 10kb between the TE locus and the gene’s transcription start site, to increase the likelihood of a direct cis-regulatory influence of the TE over the nearby gene.” [pg. 21, L357-363)

“In Figure 4B & 4C, we have highlighted six of the TE-gene pairs from Table S7 based on the gene’s function in HIV-1 restriction and the TE family’s known contribution to immune gene regulation.” [pg. 21, L369-371]

Regarding cell type composition, we acknowledge that the differences observed in the proportion of immune cell subtypes may contribute to the differential expression between ECs, VPs, and HCs (Figures 2D and S3A). However, we provide evidence that cell type composition cannot be the sole driver for the clustering of deconvoluted CD4+ T cell RNA-seq samples (Figure 3B and S5D). Cell subtype alone could not explain the observed clustering of EC samples by gene and TE family expression. Clusters 1 and 2, for example, had nearly identical subtype compositions, but were clearly separated on the UMAP (Figures 3A, 3B, and S5D). We remark on this in the Results of the revised manuscript.

“[W]e visualized the samples by cellular subtype, as identified in the original studies, to assess whether the clustering could be explained by CD4+ T cell subtype composition (Figure S5D). Clusters 1 and 2 were essentially indistinguishable in cell type composition, whereas Clusters 3 and 4 showed an overrepresentation of TM/EM and naïve/CM cell types, respectively (Figure 3B). Thus, cell subtype composition could only partially explain the clustering.” [pg. 16, L271-276]

The EC CD4+ T cell clusters also had unique gene ontology, gene & TE expression, and TE accessibility profiles (Figures 3C, 3D, 5). Moreover, while we do not have parallel RNA- and ATAC-seq data from similarly deconvoluted CD4+ T cells of ECs like those used in the clustering analysis (PMIDs 32848246 & 27453467), the original article from which we sourced the parallel RNA- and ATAC-seq data used in Figures 1 and 4 reported that these samples are predominantly effector memory CD4+ T cells (PMID 30964004). If new deconvoluted, multi-omic datasets from ECs become available, we would be interested in further exploring the contribution of cell type composition. However, the current data indicate that it is not a major contributor to the differential TE expression identified in our analyses.

Regarding the impact of ongoing HIV-1 replication upon the unique expression patterns in the EC participants, it's worth noting that the public datasets used in our study reported undetectable viremia in the EC volunteers (PMIDs 30964004, 29269040, 32848246, 27453467). Nonetheless, we sought to address this by quantifying HIV-1 transcription and exploring its potential association with interferon-stimulated gene (ISG) expression, a group of genes that we know would be reactive to active viremia. These analyses were integrated into the revised manuscript as Figure S6.

“To exclude the possibility that these gene expression signatures in ECs are associated with viremia, we quantified HIV-1 transcript levels in deconvoluted CD4+ T cell RNA-seq samples from ECs and ART-treated PLWH for comparison. In the original studies, all samples were reported to have undetected viremia by blood tests (9,37-39). Consistent with this, we found that the vast majority of the EC and ART samples taken from PBMCs exhibited very low HIV-1 transcript levels, with TPM values generally below 1. However, in samples originating from the lymph nodes of EC individuals (n = 22) (37), we detected HIV-1 expression in some subsets (Figure S6A&B). In agreement with the corresponding study (37), we found elevated HIV-1 transcript levels in germinal center and non-germinal center T follicular helper cells (GC Tfh & nGC Tfh, not included in our clustering analyses) -- and to a lesser extent in T effector memory (EM) cells (Figure S6A, average TPM Based on these results, we have concluded that the differential expression of genes and TEs in the EC clusters are not a consequence of low-level viral transcription in ECs.

Finally, a remark on TE nomenclature: The reviewer suggests that we use the term “TE groups” as opposed to taxonomic terms such as TE subfamily or TE family. We respectfully disagree. This nomenclature of TEs has been well defined (PMIDs 26612867, 26612867, 17984973) and is widely used in TE literature. Throughout the manuscript, we have conformed to the nomenclature used to annotate the human genome. One can debate the way TE families and subfamilies have been classified in Dfam (the database through which repetitive elements in the human genome have been annotated), but it is outside the scope of this study to revisit that nomenclature.

Similarly, of the 12 DE TE groups in EC in Fig 5A, only 3 overlap with the 16 in EC Fig S1.

Authors: This is correct, but we don’t believe it’s concerning. In Figure 5A, we are comparing the expression of TE families between separate EC clusters. In Figure S1, we are comparing the expression of TE families in ECs compared to ART-treated PLWH. These are fundamentally different comparisons and thus the differences in the identified DE-TEs between the two figures reflect the distinct biological contexts being investigated in each analysis.

Second, the introduction points out the strongly supported association between elite control and immunogenetic determinants, most notably specific HLA-B types, but also innate immunity factors. This cries out for inclusion of these factors in the analyses of this manuscript, in the format of Figure S4, for example, but none is to be found. The relevant genotypes are likely available in the metadata in the references cited, but, if not, could be inferred from the RNA-seq data.

Authors: Thank you for the recommendation. While our project’s primary focus is on the transcriptomic and epigenomic signatures, we agree that studying the HLA-B genotypes of all EC participants could provide valuable context for understanding the clustering of elite controllers. To explore this, we inferred the HLA-B alleles in each EC participant whose RNA-seq data was included in the clustering analysis, utilizing the arcasHLA tool (PMID: 31173059) on the total CD4+ T cell samples. We then validated these inferred HLA-B alleles against the available metadata from one of the source studies (PMID 27453467) and found that they matched for all participants. This strengthened our confidence in the accuracy of the HLA-B genotype inferences for the other samples where comprehensive HLA-B data was not provided.

In order to assess how these protective HLA-B alleles segregated between the four EC clusters derived from gene and TE family expression, we chose to visualize three of the most common alleles associated with HIV-1 elite control: HLA-B*27:03, *57:01, and *57:03 (PMIDs 30964004, 25119688, 21051598) (Figure R1, available in the Response to Reviewers PDF).

Our analysis revealed that these major protective alleles were not significantly overrepresented in any particular cluster. Consequently, we believe that HLA-B genotype does not have a major impact on the clustering observed in Figure 3.

It would also be very useful to present the KZNF data in Figure 5 the same way, since, looking at Fig 5C, the correlation of high and low KZNF expression, while clearly correlated with a that of few groups of elements, with clustering into specific groups does not appear to be well supported.

Authors: Thank you for the insightful suggestion. While the KZNF genes are included in the gene set used for the clustering analysis in Figure 3, we agree that clustering based solely on KZNF expression and displaying it as we have in Figures 3A and S5 could provide valuable insights. However, when we attempted to cluster the EC RNA-seq samples using only KZNF expression data, we were limited by the relatively low number of KZNF genes that showed sufficient variability across samples (n = 120). For robust statistical power, we require at least 200 features to reliably cluster the 128 EC CD4+ T cell samples. We believe this limitation does not diminish the relevance of KZNFs in the observed clustering patterns but rather highlights the nuanced role each KZNF plays in the regulation of the transcriptome. Each individual KZNF is responsible for the regulation of hundreds to thousands of TE loci (PMID 37730438). Thus, while a clustering approach based solely on KZNF expression may not be feasible, the integral role of KZNFs in modulating the transcriptome through TE regulation remains evident and supports their inclusion in Figure 6 of the revised manuscript.

In general, other than the cell type composition differences, there is no presentation of evidence for any biologically important feature associated with the clusters found.

Authors: We agree that the root cause of the transcriptomic differences between the EC clusters is hard to pin down but we do identify several distinctive features of the clusters that we believe are biologically significant. First, having extracted the lists of genes whose differential expression defined the four EC clusters, gene set enrichment analysis revealed that the clusters were functionally distinct, each characterized by a unique list of top GO terms (Figure 3C). Second, we provide evidence that KZNFs expressed in CD4+ T cells significantly bind to the candidate TE families whose expression defines each of these clusters (Figure 6D) and have significantly decreased expression in ECs compared to VPs (Figure 6C). This is corroborated by pairwise correlation analysis that revealed cluster-specific anticorrelation patterns between these KZNFs and their target TEs (Figure 6A). We present this data in support of our hypothesized KZNF-based mechanism for TE co-option in viral immunity. We do not yet have data indicative of the mechanism by which KZNF expression is in turn regulated. However, we speculate that negative feedback loops may be contributing to changes in KZNF expression.

“These observations suggest that interindividual variation in KZNF expression in CD4+ T cells could explain why certain TEs are variably expressed and accessible across ECs. But what are the mechanisms underlying variation in ZNF expression? It is possible that TE-KZNF regulatory loops are involved, in which a copy of the TE family targeted by a KZNF is inserted near and regulates the KZNF gene, thereby introducing a negative feedback loop. This phenomenon has been documented in prior studies of KZNF activity in embryogenesis (51) and cancer (115).” [pg. 39-40, L705-711]

Overall, our study presents preliminary evidence that the four EC clusters derived from gene & TE family expression may be distinguished by complex interplay of activators (Figure S8) and repressors (Figure 6) altering the activity of infection-responsive TE families to co-opt specific elements for immune regulatory function. While not yet validated in an experimental setting, we believe these results are of biological significance.

Third, the figures present values that have been very heavily analyzed, and it is difficult to impossible to infer what the underlying data look like. For example, with the exception of a few selected examples in Figs 4 and 5, individual provirus data are lacking. Nor can we tell how consistent the distribution of expression values within a TE group is, whether the TEs included solo LTRs (which constitute the majority of all ERVs), the possible contribution of other TFs to expression (with the exception of a brief mention of STAT1).

Authors: We respectfully disagree that the values presented in our figures are heavily analyzed. As this manuscript represents the first investigation of TEs’ role in HIV-1 elite control, we believe the most reasonable initial approach was to compile and visualize the data at the family level, rather than at the level of individual loci, which is harder to interpret due to mapping issues, commonly low transcription, and often idiosyncratic behavior of individual loci. Nonetheless, we did not limit our analysis to full-length HERVs (proviruses) and thus retain all solo LTR data in our analyses. This was added to the Methods of the revised manuscript.

“To facilitate comprehensive expression quantification, we curated a reference transcriptome by combining gene, TE, and HIV-1 genomic sequences. This was achieved by integrating the locus-level TE classification from RepeatMasker, the hg19 GenCode gene annotation,

and the HXB2 reference HIV-1 annotation. For the TEs, we removed simple repeats, SINE elements, and DNA transposons, retaining LINE and HERV loci, including all solo LTRs. We also removed any loci within gene exons/UTRs. The remaining sequences were appended in fasta format, and all sequences were annotated with their respective gene, TE locus, or HIV subunit and modeled in GTF format.” [pg. 55, L869-878]

For the sake of transparency, all relevant details on sequencing data analysis and the corresponding scripts are available in the Methods and our GitHub Repository.

Additionally, while most of our figures make comparisons at the family level, we do visualize multiple TE loci (Figure 4C) and provide a list of putative locus-level TE-gene pairs from which those shown in Figure 4C were selected (Table S7). In our revisions, we also re-clustered the 128 EC CD4+ T cell RNA-seq samples based only on locus-level TE expression, using the same graph-based k-nearest neighbors method as in Figure 3. The results of this new analysis have been integrated into the revised manuscript as Figure S7.

“To further explore locus-level expression patterns, we re-clustered the same EC samples (n=128) using only locus-level TE expression. This again resolved four EC clusters (Figure S7A), which interestingly appeared even more distinct than those identified by gene and TE family expression (Figure 3A). The TE locus-based clusters (TL-Cs) aligned well with the gene and TE family clusters (GT-Cs), with an average 70% overlap in samples between each GT-C and its corresponding TL-C (Figure S7B), indicating high consistency (Table S8). The remaining 30% of samples that shifted between clusters did so consistently within individuals, not cohorts, maintaining heterogeneous TL-C compositions similar to the GT-Cs (Figures S7C & S5A). An exception to this heterogeneity was TL-C4, comprising 22 samples from GT-C1 that were almost entirely from the CD4+ T cell subsets of only four participants in the Jiang cohort (Figure S7C, Table S8). No other samples from the Jiang cohort shifted to this cluster from other GT-Cs, suggesting that these patterns reflect individual variation rather than cohort bias. Like the GT-Cs, each TL-C included samples from all five CD4+ T cell subsets and was largely heterogeneous (Figure S7C). Notably, TL-C2 mirrored corresponding GT-C3 in its overrepresentation of EM and TM cells, while TL-C1 uniquely showed an overrepresentation of naïve CD4+ T cells. Beyond sample composition, each TL-C was characterized by a unique pattern of expressed TE loci (Figure S7D). These signatures were heterogeneous across families, with subsets of variable loci from one TE family marking separate clusters (Figure S7E), some of which did not reach the threshold of significance in earlier analyses when analyzed at the family-level, like SVA-D. Many families maintained their cluster-specific signatures, like THE1B (a marker of GT-C2), for which the majority of variable loci were found in corresponding TL-C1. However, some TE families, like the L1s that marked GT-C1, showed more heterogeneous signatures with variable loci marking multiple TL-Cs. These findings underscore the need for future locus-level investigations with high-depth sequencing to fully capture the complexity of TE expression.” [pg. 27-28, L462-488]

With this addition, we include significantly more data analyzed at the locus level, which we believe not only validate the distinct clustering observed in Figure 3, but also underscore the potential for locus resolution analysis to reveal additional layers of retrotranscriptomic diversity in EC CD4+ T cells.

Finally, we agree with the reviewer that TFs other than STAT1 may contribute to the observed changes in TE expression. To investigate this, we analyzed several TFs expressed in CD4+ T cells and, for TFs enriched over TEs of interest, subsequently examined the correlation between TF and target TE expression in the deconvoluted EC CD4+ T cell samples used for the clustering. The results of this analysis have been integrated into the revised manuscript at Figure S8.

“In addition to KZNF repressors, transcriptional activators may also drive the differential expression of specific TE families across ECs (83). To investigate this, we focused on transcription factors (TFs) expressed in CD4+ T cells and mined ChIP-seq data from the ENCODE Consortium (84) to identify TFs with binding enrichment to TE families of interest, selected for their elevated, cluster-specific expression in ECs (highlighted in Figures 4, 5, and S4). We then examined the correlation between TF and target TE expression in the deconvoluted CD4+ T cell samples from ECs used for our clustering analysis (Figure 3) (9,37). We observed several significant positive correlations between TF and TE expression across ECs (Figure S8). Thus, differential expression of immune-related TFs may also contribute to the variation in TE expression and cis-regulatory activity across ECs, in tandem with the repressive activities of KZNFs.” [pg. 30, L517-527]

This evidence supports the reviewer’s suggestion that other TFs may be contributing to the unique EC retrotranscriptome we profile in this study. These added analyses, mimicking those conducted for KZNFs in Figure 6B & 6D, demonstrate that transcriptional activators may indeed play a crucial role in shaping the TE landscape in ECs.

Other issues

Figure 1:

A) Log2 fold change of what? TPM values? Needs to be specified.

Authors: Thank you for pointing out this ambiguity. The log2-transformed fold change values plotted in Figure 1A refer to DESeq2-normalized expression. They were extracted from the results of the DESeq2 pipeline, which we applied to the raw count expression matrix (see our Methods for more details). Following your suggestion, we have clarified this point in the figure legend in the revised manuscript.

“Total detected genes and TE loci are plotted by log2-transformed fold change of DESeq2-normalized counts (EC vs. HC).” [pg. 10, L163-164]

We have similarly made these changes to any figure legend which was ambiguous in its description of the expression data.

Why Bonferroni correction? Usually BH q values or other less stringent adjustments are used nowadays.

Authors: In our analysis, we opted for the Bonferroni correction due to its well-established reliability and stringent control of the family-wise error rate when conducting multiple tests. Given the exploratory nature of our investigation and the desire to minimize the risk of false positive findings, we chose to employ this traditional correction method within our analytical pipelines.

B,C): Z-score of what? Scaled, normalized counts? Scaled TPM values?

Authors: Thank you again for highlighting this point of uncertainty. We now clarify this in the figure legend in the revised manuscript.

“Heatmap displaying the expression of the top differentially expressed genes in CD4+ T cells of ECs (n=4; red bar) vs. HCs (n=5; blue bar). Relative expression levels are representative of row-wise scaled, log2-transformed expression in transcripts per million (TPM). Heatmap coloration is based on the z-score distribution from low (gold) to high (purple) expression.” [pg. 11, L167-171]

Figure 2:

B) The blue font color is very difficult to see.

Authors: We have changed the blue font color to make it more easily distinguishable from the black.

C) This heatmap should demarcate or separate genes versus TE clades. If that's not possible, then the two should be shown separately.

Authors: We appreciate your suggestion regarding the heatmap presentation. While we understand the rationale for demarcating genes versus TE clades, we have chosen to retain the original figure layout. In this analysis, TEs were analyzed simultaneously with genes. The order in which they are shown was obtained by default clustering of the expression matrix using the hclust function. We chose to present them together and in this order to provide a comprehensive visualization of the differential expression patterns between the two groups and highlight the homogenous nature of gene and TE expression across VPs.

L191: How many groups (NOT families) and how many total elements were examined?

Authors: We begin with the RepeatMasker annotation of the hg19 assembly and filter out the SINE elements, DNA transposons, simple repeats, and all loci within gene exons/UTRs. These details are provided in the Methods of the revised manuscript, as was quoted above. In total, our analyses examine 1,104,828 loci from 603 TE groups (which we refer to as families). We apologize if this figure is not accurate to a separate classification of TEs into groups, rather than families. Any such method of grouping TEs is unfamiliar to us and outside of the Dfam annotation.

L198: 2B, not C

Authors: Thank you for catching this. The figures labelled were swapped in error and have been changed to reflect in Figure 2 to match the in-text references.

L205: Did the expressed proviruses have STAT1 sites?

Authors: Thank you for your question. The identification of LTR13’s increased expression in ECs compared to VPs was the result of a family level analysis which considered expression additively across the LTR13 loci in our annotation. To answer your question, we analyzed STAT1 ChIP-seq data from the ENCODE Consortium to characterize which LTR13 loci were bound by STAT1 (corroborated by motif prediction calls). We then integrated the EC RNA-seq data and found that the expressed LTR13 proviruses significantly overrepresented those with bound STAT1 sites (Figure R2, available in the Response to Reviewers PDF).

These data suggest that STAT1 binding may play a critical role in the transcriptional regulation of LTR13 in ECs, contributing to their differential expression profile. Further exploration into the contribution of activating, immune-related TFs is explored in Figure S8 in the revised manuscript.

L333: 10 kb is very close. Why was it chosen?

Authors: We chose 10 kb as our cutoff for selection because it allowed for very high confidence in the TE loci’s cis-regulatory capacity over the nearby genes. For transparency, we have made this clearer in the Results text of the revised manuscript.

“These loci and genes were paired based on proximity, with a maximum distance of 10kb between the TE locus and the gene’s transcription start site, to increase the likelihood of a direct cis-regulatory influence of the TE over the nearby gene.” [pg. 21, L360-363]

However, if desired, a less stringent cutoff could also be used with relative confidence (e.g., 50 kb).

L351-352: Again, correlation is not causation. How do the authors know it's not the other way around?

Authors: The candidates that we chose to display in Figure 4 (the figure to which these lines refers) are from MER41, ERV3-16, and LTR12C. Our lab and others have shown that these specific loci or other loci in these TE families are capable of regulating neighboring genes’ expression, with specific evidence in the context of immunity (PMID Smitha, Ed, APOBEC, etc.). Based on this knowledge, we believe that it’s most likely that TE-derived regulatory sequences are the cause of the increased restriction factor expression, rather than TE accessibility being a consequence of the transcriptional activation of the neighboring genes. However, we recognize that these results are correlative, as the reviewer notes, and we emphasize this in the revised manuscript. Most notably:

“We acknowledge that these associations are drawn from correlative patterns and manipulative experiments are needed to infer causality between chromatin changes at these TEs and increased expression of nearby immunity genes.” [pg. 36, L620-623]

Figure 4

B) Need to show a scale of the genome region, the orientation of both the gene and the TE, whether it is a solo LTR

Authors: Thank you for the suggestion. Genomic scale and orientation have been added to Figure 4C. All loci visualized were solo LTRs, save for HCP5, which is a lncRNA derived from a full-length ERV3 element.

Figure 5

A) Would benefit from also showing HCs

Authors: Thank you for the recommendation. The RNA-seq datasets used in this analysis do not include HC samples. Additionally, this analysis is meant to highlight differences in TE expression between the four EC clusters. Thus, we have chosen to keep Figure 5A as it appears in the original manuscript.

C) Would be helped by showing adjusted p-values, and also should show examples of non-correlating relationships between these KZNF genes and other TEs.

Authors: Thank you for the suggestion. All correlation analyses had adjusted p-values below 0.01, derived from corr.test in R. We’ve added this to the figure legends of Figure 6B [pg. 32, L539] and S8B [pg. 53, L835]. However, we have chosen not to integrate non-correlating examples into the revised manuscript for the sake of space.

Figure 6

Title: should start with "proposed model for.." or some such.

Authors: Thank you for the suggestion. The title has been changed to “Proposed model for the interplay of KZNFs and TEs regulating proximal antiviral gene expression in elite controllers of HIV-1” in the revised manuscript [pg. 34, L580-581].

L 537: Again, how do the alleles segregate in the clusters?

Authors: This question has been addressed in response to an earlier comment from Reviewer #3.

Generally, in the correlation analyses, I'd like to see adjusted p-values and examples of non-correlated results.

Authors: Thank you for the suggestion. As mentioned above, all correlation analyses have been annotated with the adjusted p-value threshold. Additionally, below we’ve included examples of non-correlated results from two analyses. First, we show a TE-gene pair whose increased TE accessibility in HCs compared to ECs does not correlate with increased expression of the proximal gene (Figure R3, available in the Response to Reviewers PDF). Notably, this gene does not play a role in HIV-1 infection response. Here, we show that genes with proximal (Second, we show the pairwise correlation and linear regression results of L1PA6 and ZNF2 (Figure R4, available in the Response to Reviewers PDF). ZNF2 is one of the KZNFs highlighted in Figure 6 for its low expression in ECs, anticorrelated to its repressive target LTR12C. On the other hand, L1PA6 is active in ECs, with variably high expression across samples. ZNF2 ChIP-exo revealed that ZNF2 has no capacity to bind to L1PA6 loci (adj. p-value = 1; PMID 37730438). Thus, even though both genes are variable across samples, we observe no significant (anti)correlation between the two variables (rho = 0.051 & p-value = 0.866).

While we have not integrated these results into the revised manuscript for the sake of space, we hope that the provided examples satisfactorily demonstrate the presence of non-correlated results in our analyses, further reinforcing the specificity and robustness of our significant findings.

Significance:

This study presents an in-depth analysis of the reverse transcriptome in Elite controllers. It will be of interest to both HIV researchers and those interested in the regulation of the human retrotranscriptome and its consequences.

Provides an avenue for future explanation into elite controllers and TE involvement in the phenotype.

Does a good job of placing the work in the context of existing lit, synthesizing other papers regarding TEs and immune control.

Potential immune regulatory involvement of specific HERV clades.

Authors: We’d like to thank the reviewer for their encouraging feedback. We’re pleased that they found our analysis of the EC retrotranscriptome to be of broad interest and appreciate their recognition of our efforts to synthesize existing literature, contextualizing our findings within the broader field. We agree that our study opens new avenues for exploring the role of TEs, particularly specific HERV clades, in not only the EC phenotype but immune regulation as a whole.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #3

Evidence, reproducibility and clarity

Summary:

This manuscript presents an analysis of published gene expression (RNA-seq and ATAC-seq) data from a couple of cohorts of HIV-infected elite controllers (EC), as compared to uninfected controls, (HC), virological progressors (VP). The authors report that HIV elite controllers may exhibit 4 distinct patterns of TE (and gene) expression and suggest that TE expression may drive some form of antiviral gene expression. Further, they show that heterogeneous TE expression may be determined by differential KZHF gene activity among the different clusters of elite controllers. These results are very interesting, even though the conclusions are very preliminary. It presents intriguing correlations between expression of certain TE groups of LINES and HERVs, and the clustering into 4 gene expression groups in EC and is a novel finding. That said, correlation is not causation, and the authors need to be more cautious in presenting their highly preliminary model in Figure 6.

Major comments:

Overall, although preliminary, as the authors note, the results are interesting and worthy of follow-up. At this point, however, a number of issues arise that need further clarification and analysis before I would consider this study complete. First, the analyses shown in Figures 3-5 based on data from studies on EC of CD4 cells are apparently motivated by the differential TE expression in total PBMCs shown in Fig 1 and 2. Yet, the TE groups (please don't use taxonomic terms like "subfamily") identified in Fig 2 and Fig 4 are completely different, with no overlap. This discrepancy underscores the possibility that the differential expression observed is, at lest in part, due to the differences among the groups or clusters in cell type composition, as seen in Fig 2D and 3B which, themselves, could be a consequence of HIV infection and elite control (which has been shown to involve ongoing, albeit low-level, virus replication). This issue must be addressed. Similarly, of the 12 DE TE groups in EC in Fig 5A, only 3 overlap with the 16 in EC Fig S1.<br /> Second, The introduction points out the strongly supported, association between elite control and immunogenetic determinants, most notably specific HLA-B types, but also innate immunity factors. This cries out for inclusion of these factors in the analyses of this manuscript, in the format of Figure S4, for example, but none is to be found. The relevant genotypes are likely available in the metadata in the references cited, but, if not, could be inferred from the RNA-seq data. It would also be very useful to present the KZNF data in Figure 5 the same way, since, looking at Fig 5C, the correlation of high and low KZNF expression, while clearly correlated with a that of few groups of elements, with clustering into specific groups does not appear to be well supported. I n general, other than the cell type composition differences, there is no presentation of evidence for any biologically important feature associated with the clusters found.<br /> Third, the figures present values that have been very heavily analyzed, and it is difficult to impossible to infer what the underlying data look like. For example, with the exception of a few selected examples in Figs 4 and 5, individual provirus data are lacking. Nor can we tell how consistent the distribution of expression values within a TE group is, whether the TEs included solo LTRs (which constitute the majority of all ERVs), the possible contribution of other TFs to expression (with the exception of a brief mention of STAT1).

Other issues

Figure 1: A) Log2 fold change of what? TPM values? Needs to be specified.

Why Bonferroni correction? Usually BH q values or other less stringent adjustments are used nowadays. B,C): Z-score of what? Scaled, normalized counts? Scaled TPM values?

Figure 2: B) The blue font color is very difficult to see C) This heatmap should demarcate or separate genes versus TE clades. If that's not possible, then the two should be shown separately.

L191: How many groups (NOT Fam1lies) and how many total elements were examined?

L198: 2B, not C

L205: Did the expressed proviruses have STAT1 sites?

L333: 10 kb is very close. Why was it chosen?

L351-352: Again., correlation is not causation. How do the authors know it's not the other way around?

Figure 4 Title: For "induction" Substitute "correlation"

Panel B: Need to show a sclae of the genome region, the orientation of both the gene and the TE, whether it is a solo LTR 5 Panel A: Would benefit from also showing HCs C: Would be helped by showing adjusted p-values, and also should show examples of non-correlating relationships between these KZNF genes and other TEs. 6 Title: should start with "proposed model for.." or some such. L 537: Again, how do the alleles segregate in the clusters?

General

In the correlation analyses, I'd like to see adjusted p-values and examples of non-correlated results.

Significance

This tudy presents an in depth analysis of the reverse transcriptome in Elite controllers. It will be of interest to both HIV researchers and thos interested in the regulation of the human retrotranscriptome and its consequences

• Describe the nature and significance of the advance (e.g. conceptual, technical, clinical) for the field.

Provides an avenue for future explanation into elite controllers and TE involvement in the phenotype. - Place the work in the context of the existing literature (provide references, where appropriate).

Does a good job of this, synthesizing other papers regarding TEs and immune control. - State what audience might be interested in and influenced by the reported findings.

Potential immune regulatory involvement of specific HERV clades. - Define your field of expertise with a few keywords to help the authors contextualize your point of view.

Indicate if there are any parts of the paper that you do not have sufficient expertise to evaluate.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

The authors have re-analyzed published RNA-Seq data from CD4 T cells isolated from HIV elite controllers and reference cohorts, including HIV negative persons, viremic progressors and ART-treated persons. Their main finding is that in some of their comparisons, EC have higher levels of interferon-stimulated genes (ISG), paired with distinct expression patterns of transposable elements. The authors suggest that expression of transposable elements may induce altered expression of ISG, presumably due to immune recognition of TE. They also suggest that reduced expression of KZNF genes, which encode for transcription factors that can suppress TE, may be responsible for enhanced expression of TE. I have the following comments:

1. All data included in this manuscript derive from previously published data. A new dataset, specifically designed to focus on a high-resolution analysis of TE expression, would be better suited to address the proposed questions.
2. As the authors acknowledge, the described investigations are exploratory, and do not allow to draw firm conclusions. Mechanistic experiments are recommended to address the authors' hypotheses.
3. An important limitation is that virological data of EC are not considered. For example, I believe it is a lot more likely that the upregulation of ISG in EC relates to ongoing low-level viral replication. The authors could analyze cell-associated HIV RNA and DNA levels and determine how they associate with ISG expression.
4. KZNF genes seem downregulated in EC. Can the authors propose a reason/mechanism for that?

Significance

Overall, I think this is an interesting manuscript that proposes a distinct and potentially important mechanisms that may contribute to immune control of HIV. My suggestions to improve the manuscript are complex and cannot be easily addressed through experimental work. I believe a possible option would be to publish the present manuscript without my proposed modifications, but highlight the weaknesses of the current paper more clearly; mechanistic studies could then be deferred to a future study.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

Singh et al. analyze the expression and putative contribution of TEs in CD4+ T cells in HIV elite controllers. Through re-analysis of existing datasets, the authors describe broad differences in expression of TEs in ECs through analysis of RNAseq annd ATACseq data, and come up with convincing examples where differentially-expressed innate immune genes correlate with increased accessibility of proximal TEs. Overall, the authors' conclusions are appropriately measured, though the manuscript text should be re-organized for clarity and a few further analyses are needed to support the main message of the paper.

Major comments: The manuscript would benefit from a re-organization of the figures to focus on TEs - in particular, Fig 1B, Fig 2, and Fig 3 reproduce known transcriptional differences between ECs and HCs and serve as quality controls for the authors' computational analysis. Conversely, Supplementary Fig 6 contains very interesting data on KZNF expression and should be included in the main figures.

It remains unclear whether differences in TE expression described are specific to ECs or to EC-like CD4+ T cell states. As there are plenty of datasets available that compare the transcriptome of naïve, activated, exhausted, and regulatory CD4+ T cells, the authors should compare the TE expression patterns observed in ECs to activated CD4+ T cells, particularly those with a Th1 and cytotoxic phenotype analogous to those observed in ECs, from healthy donors.

In Fig 1, the authors demonstrate differential expression of both innate immune genes and TEs, but the link between the two is unclear. Is there any enrichment in differential expression for TEs located proximal to innate immune genes? This type of analysis should be possible using the authors' own software to map TE expression to specific genomic loci.

Optional: In Fig 3, the authors cluster CD4+ T cells based on transcriptomic profiles. It would be interesting to re-cluster these samples based on TE expression alone, given the differences in TE expression described in Fig 5.

Significance

The manuscript by Singh et al. describes for the first time the role of TEs in HIV elite controllers, suggesting that TEs may be co-opted for cis-regulatory function. This study builds off prior work demonstrating that HIV-infected CD4+ T cells activate LTR elements that may regulate expression of interferon-inducible genes, demonstrating that ECs show further upregulation of innate immune genes. While these findings will need to be experimentally validated, this study constitutes a useful resource and adds to the growing body of evidence implicating TEs in cis-regulatory control of immune genes. This study will be of interest to basic scientists interested in genetic mechanisms of HIV control, and if further developed may comprise a useful source of biomarkers to predict viral kinetics in HIV-infected individuals.

My expertise is in immunology, TE biology, and viral infection

Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

Learn more at Review Commons

---

Reply to the reviewers

Reviewer #1

Evidence, reproducibility and clarity

Sanial et al. carefully analyze the use of in-gel fluorescence as an alternative to immunoblotting. The authors show that simple modifications of common protein extraction protocols can preserve (to varying extents) fluorescent proteins in their native, fluorescent states. This can be exploited in different applications for in-gel fluorescence quantification, bypassing immunoblotting. The experimental results are clear, showcasing the ease and linearity of in-gel fluorescence quantification.

In my opinion, the trick of this approach is also potentially its main drawback, the partial denaturation conditions. I think the manuscript could be strengthened with more extensive benchmarking of the approach and further discussion of potential caveats as detailed below.

Major points:

1. Protein abundance in the original GFP library (and in other FP-tagged libraries constructed in the meanwhile) have been quantified using fluorescence (flow cytometry, microscopy, colony fluorescence) (Ho et al. 2018 10.1016/j.cels.2017.12.004, Weill et al. 2018 10.1038/s41592-018-0044-9, Meurer et al. 2018 10.1038/s41592-018-0045-8). This provides an opportunity to significantly strengthen the manuscript (where most of the test have been done using two abundant cytosolic proteins Bmh1 and Hxk1) if the authors could apply their approach to a representative fraction of the yeast proteome (sampling from such libraries FP-tagged proteins that differ in abundance, localization, membrane vs cytosolic/nuclear, subunits of large stable complexes vs proteins not part of complexes, etc.) and compare their quantification with previous relative abundance estimates. This information would also help future users in case protein-specific issues are identified.

Indeed, Hxk1 and Bmh1 are quite strongly expressed (41,000 and 65,000 copies/cell, according to SGD, ____www.yeastgenome.org____). In the course of our experiments we were able to detect proteins with a much lower expression level (eg. Reg1, 4000 copies/cell). We have selected a number of proteins based on their expression level as detailed in SGD, ranging from 700 to 75000, and plan to detect the signal by IGF and compare it with published data on absolute protein quantifications for ecah protein. However, this will take a bit of time as each gene must be tagged with EGFP – we cannot use the GFP-S65T from the GFP collection which is poorly amenable to IGF because of its sensitivity to denaturation, as we show in our manuscript.

The authors discuss several drawbacks, including the change in apparent molecular weight compared to denatured proteins; differential recognition of folded vs denatured proteins by antibodies.

Other potentials drawbacks should be discussed. For instance, the need of additional steps post-fluorescence imaging for signal normalization against a loading control; the need of antibodies and immunoblotting to decide on the best denaturing temperature for a specific protein or FP tag; complexity of the native protein extraction protocol compared for example to alkaline lysis followed by TCA precipitation (Knop et al. 1999 PMID: 10407276).

• Regarding the the need of additional steps post-fluorescence imaging for signal normalization against a loading control – this doesn’t take extra time for people using gels with protein stain included in the gel (eg. Stain-Free from BioRad). There are other possibilities of total protein fluorescent labeling that will be discussed. We will provide an example of this application.
• On the need of antibodies and immunoblotting to decide on the best denaturing temperature for a specific protein or FP tag – we believe that the system could be set up for a protein of interest for which antibodies are available (as we did for Bmh1 or Hxk1), and once this is done, there is no need to do these controls anymore. We will mention this in the manuscript.
• On the complexity of the native protein extraction protocol compared for example to alkaline lysis followed by TCA precipitation – indeed the protocol is a bit more time-consuming compared to the mentioned method, and we will mention this in the text. However, please note that people studying mammalian cells, for instance, often use this native protocol for total extracts so this is mostly a yeast-model issue. Yet, we will add this comment. Moreover, although the denatured fraction is FP- and temperature-dependent, even under the milder 30{degree sign}C conditions there is a detectable denatured fraction (Fig.s3b). This would seem to preclude the use of this approach for absolute protein quantification.

True, but it depends a lot on the FP used. For instance, sfGFP is not denatured and could potentially be used for absolute quantification. We will comment in the text.

Finally, any evidence that the denatured fraction would depend on the protein tagged with the FP?

We will use several proteins used for point 1 but fused to the most sensitive FP, GFP-S65T, and do a western blot using anti-GFP antibodies to estimate the variation in native vs. denatured forms of the protein.

Minor points: 1. In the experiments designed to test the linearity and sensitivity of the approach, an alternative approach that would not result in dilution of cell extract is to mix wild type cell extract (no GFP fusion) with extract of the GPF-tagged strain in different ratios.

Yes, this was an alternative but it seemed that dilution was easier to control than mixing two extracts.

Define all acronyms at first appearance. For example, DTT and LDS on page 4.

Thank you, we will address all acronyms in the text.

Fig.4D: the colors chosen to represent EGFP and sfGFP data make them hard to tell apart. The same comment to Fig.S6.

Agreed, we will change the figures accordingly.

As the temperature steps are not uniform in Figures 4 and 5, it would be more informative to indicate the exact temperate above each lane (in addition/instead of the ramp cartoon).

Agreed, we will change the figures accordingly.

Regarding linearity, that HRP-based quantification is not linear is expected. A fairer comparison would be to use fluorescently labeled secondary antibodies. It is also puzzling that detection with signal amplification (HRP) is less sensitive than direct quantification of the fluorescence signal from the FP tag.

We will do a sensitivity tets (dilutions) to compare IGF with HRP-based and fluorescent-based antibody-mediated detection.

I appreciate the workflow Figure 10. But in my opinion it is trying to show too much (protocol, troubleshooting, calls to figure panels). Perhaps it could be made clearer by separating the protocol steps/settings from the optimization/troubleshooting tips.

Thank you, we will work on this to make the workflow clearer.

Some of the discussion of different fluorescent proteins, and expression levels of tagged proteins, could be confounded by the different linkers used in the tagging constructs.

Thank you for this remark. Indeed, there are various linkers on these constructs and we don’t know to which extent they contribute to the effect on protein expression level. We will comment his in the text.

Significance

Could be a generally useful and simple approach for in-gel quantification using fluorescent protein tags.

__ ____Thank you for your comments and overall assesment.__

---

Reviewer #2

Evidence, reproducibility and clarity

The present manuscript "Direct observation of fluorescent proteins in gels: a rapid cost-efficient, and quantitative alternative to immunoblotting" describes a method how to visualize bands of fluorescent protein fusions onto a common SDS-PAGE without antibody staining. It is based on ability of GFP-like fluorescent proteins (FPs) to retain their fluorescence under conditions of SDS-PAGE if step of extensive heating (boiling) of protein sample is omitted. This property of FPs is not novel; it was known for more than 20 years (for example, see Fig. 2 in Yanushevich et al. FEBS Lett. 2002 Jan 30, 511:11-4; Supporting Fig. 7 in Campbell et al. Proc Natl Acad Sci USA. 2002 Jun 11, 99:7877-82). However, the authors did perform a very accurate and robust study to quantitatively assess the behavior of several FP fusion protein in SDS-PAGE. A thorough analysis of different conditions for a variety of FPs and target proteins was done; detailed protocols were developed. A surprisingly high sensitivity of FP detection (even superior to that of standard Western blotting) was demonstrated. Considering the simplicity of the proposed approach, it appears to be the method of choice for those working with FP fusion proteins.

Thank you for this comment. Indeed we do not claim to discover that FP remain fluorescent in mild denaturing conditions, as presented in the text. We did our best to include original publications showing precedent for this and we missed Yanushevich et al. FEBS Lett. 2002 that we will add. However the Campbell paper is cited, precisely for the Supplementary figure 7 that the reviewer mentions.

I have only minor, discretionary comments:

1. It is known that under conditions of SDS-PAGE without heating, FPs retain not only fluorescence but also their oligomeric state. The same can be true for proteins of interest (POIs). If so, even for monomeric FPs, the POI-FP band can potentially migrate much slower than expected because of oligomerization of the POI.

__Thank you for this suggestion. Our data in the manuscript already show that Bmh1 and Bmh2, which are tighlty associated 14-3-3 proteins, no longer intereact in these mild denaturation conditions. In the set of proteins that we will use to answer to Reviewer #1 (point 1), we will include proteins in large complexes to assess whether this can happen. __

It might be useful to briefly discuss a possibility to use other types of fluorescent proteins (namely, Flavin-binding FPs, bacteriophytochrome-based FPs, bilirubin-binding FP UnaG) in the same way as proposed here. In particular, biliverdin-binding near-infrared FPs (IFP, iRFP, etc.) can be detected even after fully denaturing SDS-PAGE by zinc-induced orange fluorescence of proteins carrying covalently attached bilin chromophore (Berkelman TR, Lagarias JC. Visualization of bilin-linked peptides and proteins in polyacrylamide gels. Anal Biochem. 1986, 156, 194-201; Stepanenko OV, Kuznetsova IM, Turoverov KK, Stepanenko OV. Impact of Double Covalent Binding of BV in NIR FPs on Their Spectral and Physicochemical Properties. Int J Mol Sci. 2022, 23, 7347).

__Agreed. ____We will extend the discussion to other fluorescent approaches to visualize proteins in gels and compare them. __

---

Significance

A simple method of specific visualization of fluorescent protein fusion bands on SDS-PAGE is proposed.

Thank you for your comments and overall assesment.

Reviewer #3

Evidence, reproducibility and clarity

In this paper, Sanial et al present in-gel fluorescence detection (IGF), a method that allows the direct detection of fluorescent proteins from SDS-PAGE gels with minimal adaptation of existing protocols. The authors test a range of fluorescent proteins routinely used, especially when working with yeast, and describe their behavior in IGF. They identify heat-induced denaturation of fluorescent proteins as the main component influencing their assay and systematically test this on a selection of fluorescent proteins. Next, they compare the detection limit and the linearity of the signal between IGF and chemiluminescence, showing that IGF is not only comparable but also superior to chemiluminescence. This is particularly significant given that chemiluminescence can suffer from issues such as a limited dynamic range and limitations in accurately quantifying very low or high-abundance proteins. The authors further demonstrate the utility of IGF in co-immunoprecipitation experiments and test whether the mild denaturing conditions are compatible with proteins from other organisms. Overall, the study is well-presented and is an asset to the scientific community. I have one major and some minor comments that, in my opinion, would improve this already informative paper: Major comment 1. In all cases where there is signal quantification the authors should perform replicates to account for variability of the signal (in Fig 6, S6 and S7).

__Agreed, we will perform triplicates for the indicated experiments. __

Minor comments 1. The study mainly focuses on soluble protein. While the authors have tested one plasma membrane protein, the study would benefit from including more membrane proteins from different environments (e.g., cell wall, nuclear envelope, mitochondrial). This would help determine if incubation at higher temperatures is necessary to properly solubilize these proteins, in which case the experiment would need adaptation.

Thank you for this suggestion. __In the set of proteins that we will use to answer to Reviewer #1 (point 1), we will include proteins from various subcellular locations. __

The authors show that when fluorescent proteins are partially denatured, their migration behavior changes. One cannot exclude that in some cases, the tagged proteins themselves might also be partially resistant to denaturing at the low temperatures used for IGF. This would lead to more than one fluorescent bands. In such cases one should be careful with interpretation, especially in the context of PTMs or isoforms. Could the authors briefly discuss this?

Thank you for this comment. __We will discuss this in the text. __

Based on Fig 4D and 5D, some fluorescent proteins seem to have a higher signal variability between replicates than others. It would be helpful to add this information next to the behavior of the proteins in different temperatures so it would be easier to choose the fluorescent protein for specific experiments.

__Indeed, there are variations between experiments, but it is not clear whether this inherent to the FP considered or the experiment. We will look back at the data and modify the text accordingly if pertinent. __

The sensitivity experiment (Figure 6) is convincing and important for IP conditions, where the total protein concentration of the sample is radically decreased. Could the authors additionally test if very low abundant proteins can be detected (without any dilution of the total protein content), and compare this to chemiluminescence? This could be done either by tagging some very low abundant proteins (for example a few hundred copies per cell) or diluting the lysate in wild-type lysate to artificially reduce their concentration while maintaining the overall protein load the same.

__We have planned an experiment in which low abundant proteins will be tagged in response to reviewer 1 (point 1) which should address this point. __

It would be useful to address the detection of very high molecular weight proteins - or proteins that are problematic in terms of transfer during western blotting.

Again, in the experiment planned in ____response to reviewer 1 (point 1), proteins or various MW as well as membrane proteins will be studied, which should address this point. __ __

Significance

The authors already discuss the strengths and limitations of their approach. The main strength of IGF is that it does not require transfer of the proteins to a membrane and also does not rely on antibody binding and (potential) chemical reactions. In addition to the fact that this is time, cost, equipment, waste and expertise effective, the sensitivity and signal linearity of IGF seems to not only compare but outperforme western blotting. There are two main limitations. First, IGF relies on the resilience to denaturing of the chosen fluorescent protein that depends, according to the authors, at least on the temperature and overall protein concentration and pH. Second, IGF relied on tagging proteins with fluorescent proteins which might affect the stability or even function of the tagged protein. As the authors mention, these factors do not diminish the value of IGF, they highlight the need for appropriate controls.

A potential development of the technique (not at the present study) could be the compatibility of IGF with different self-labelling proteins (Halo, Snap) and fluorescent dyes.

We have conducted experiments in which we show the applicability of IGF in combination to SNAP-tagging, that we could show if needed.

I think IGF will benefit a rather broad range of scientists. As already mentioned by the authors, there are different applications of IGF. From checking of clones when creating strains, to comparison of protein levels in different conditions and coIP experiments.

Thank you for your comments and overall assesment.

Cross reviews. Reviewer 1: I agree with the assessment by Reviewer #2. Considering the comment about potential oligomerization of a protein of interest, I stand by my point about testing the method with more proteins of interest. How extensive this testing should be or whether additional discussion of possible issues would suffice is a matter of opinion. It is clear from the manuscript in it's current form that the method works and that it has caveats.

We believed that the experiments we have planned will clarify these points.

Reviewer 2: In general, I agree with the points raised by Reviewer #1. However, in my opinion, there is already a large body of reliable experimental results in the manuscript that are worth publishing without a new round of extensive experiments.

Reviewer 1: Fair enough, I don't insist on the experiments in my point 1.

We think that this is an important point that will likely be a common question for readers so we will still do our best to provide data for this point.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #3

Evidence, reproducibility and clarity

In this paper, Sanial et al present in-gel fluorescence detection (IGF), a method that allows the direct detection of fluorescent proteins from SDS-PAGE gels with minimal adaptation of existing protocols. The authors test a range of fluorescent proteins routinely used, especially when working with yeast, and describe their behavior in IGF. They identify heat-induced denaturation of fluorescent proteins as the main component influencing their assay and systematically test this on a selection of fluorescent proteins. Next, they compare the detection limit and the linearity of the signal between IGF and chemiluminescence, showing that IGF is not only comparable but also superior to chemiluminescence. This is particularly significant given that chemiluminescence can suffer from issues such as a limited dynamic range and limitations in accurately quantifying very low or high-abundance proteins. The authors further demonstrate the utility of IGF in co-immunoprecipitation experiments and test whether the mild denaturing conditions are compatible with proteins from other organisms.

Overall, the study is well-presented and is an asset to the scientific community. I have one major and some minor comments that, in my opinion, would improve this already informative paper:

Major comment

1. In all cases where there is signal quantification the authors should perform replicates to account for variability of the signal (in Fig 6, S6 and S7). Minor comments
2. The study mainly focuses on soluble protein. While the authors have tested one plasma membrane protein, the study would benefit from including more membrane proteins from different environments (e.g., cell wall, nuclear envelope, mitochondrial). This would help determine if incubation at higher temperatures is necessary to properly solubilize these proteins, in which case the experiment would need adaptation.
3. The authors show that when fluorescent proteins are partially denatured, their migration behavior changes. One cannot exclude that in some cases, the tagged proteins themselves might also be partially resistant to denaturing at the low temperatures used for IGF. This would lead to more than one fluorescent bands. In such cases one should be careful with interpretation, especially in the context of PTMs or isoforms. Could the authors briefly discuss this?
4. Based on Fig 4D and 5D, some fluorescent proteins seem to have a higher signal variability between replicates than others. It would be helpful to add this information next to the behavior of the proteins in different temperatures so it would be easier to choose the fluorescent protein for specific experiments
5. The sensitivity experiment (Figure 6) is convincing and important for IP conditions, where the total protein concentration of the sample is radically decreased. Could the authors additionally test if very low abundant proteins can be detected (without any dilution of the total protein content), and compare this to chemiluminescence? This could be done either by tagging some very low abundant proteins (for example a few hundred copies per cell) or diluting the lysate in wild-type lysate to artificially reduce their concentration while maintaining the overall protein load the same.
6. It would be useful to address the detection of very high molecular weight proteins - or proteins that are problematic in terms of transfer during western blotting.

Significance

The authors already discuss the strengths and limitations of their approach. The main strength of IGF is that it does not require transfer of the proteins to a membrane and also does not rely on antibody binding and (potential) chemical reactions. In addition to the fact that this is time, cost, equipment, waste and expertise effective, the sensitivity and signal linearity of IGF seems to not only compare but outperforme western blotting. There are two main limitations. First, IGF relies on the resilience to denaturing of the chosen fluorescent protein that depends, according to the authors, at least on the temperature and overall protein concentration and pH. Second, IGF relied on tagging proteins with fluorescent proteins which might affect the stability or even function of the tagged protein. As the authors mention, these factors do not diminish the value of IGF, they highlight the need for appropriate controls.

A potential development of the technique (not at the present study) could be the compatibility of IGF with different self-labelling proteins (Halo, Snap) and fluorescent dyes.

I think IGF will benefit a rather broad range of scientists. As already mentioned by the authors, there are different applications of IGF. From checking of clones when creating strains, to comparison of protein levels in different conditions and coIP experiments.

Keywords/field of expertise: yeast genetics, organelle homeostasis, biochemistry, molecular biology, cell biology, fluorescence microscopy, functional proteomics, gut microbiology

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

The present manuscript "Direct observation of fluorescent proteins in gels: a rapid cost-efficient, and quantitative alternative to immunoblotting" describes a method how to visualize bands of fluorescent protein fusions onto a common SDS-PAGE without antibody staining. It is based on ability of GFP-like fluorescent proteins (FPs) to retain their fluorescence under conditions of SDS-PAGE if step of extensive heating (boiling) of protein sample is omitted. This property of FPs is not novel; it was known for more than 20 years (for example, see Fig. 2 in Yanushevich et al. FEBS Lett. 2002 Jan 30, 511:11-4; Supporting Fig. 7 in Campbell et al. Proc Natl Acad Sci USA. 2002 Jun 11, 99:7877-82). However, the authors did perform a very accurate and robust study to quantitatively assess the behavior of several FP fusion protein in SDS-PAGE. A thorough analysis of different conditions for a variety of FPs and target proteins was done; detailed protocols were developed. A surprisingly high sensitivity of FP detection (even superior to that of standard Western blotting) was demonstrated. Considering the simplicity of the proposed approach, it appears to be the method of choice for those working with FP fusion proteins.

I have only minor, discretionary comments:

1. It is known that under conditions of SDS-PAGE without heating, FPs retain not only fluorescence but also their oligomeric state. The same can be true for proteins of interest (POIs). If so, even for monomeric FPs, the POI-FP band can potentially migrate much slower than expected because of oligomerization of the POI.
2. It might be useful to briefly discuss a possibility to use other types of fluorescent proteins (namely, Flavin-binding FPs, bacteriophytochrome-based FPs, bilirubin-binding FP UnaG) in the same way as proposed here. In particular, biliverdin-binding near-infrared FPs (IFP, iRFP, etc.) can be detected even after fully denaturing SDS-PAGE by zinc-induced orange fluorescence of proteins carrying covalently attached bilin chromophore (Berkelman TR, Lagarias JC. Visualization of bilin-linked peptides and proteins in polyacrylamide gels. Anal Biochem. 1986, 156, 194-201; Stepanenko OV, Kuznetsova IM, Turoverov KK, Stepanenko OV. Impact of Double Covalent Binding of BV in NIR FPs on Their Spectral and Physicochemical Properties. Int J Mol Sci. 2022, 23, 7347).

Referee cross-commenting

In general, I agree with the points raised by Reviewer #1. However, in my opinion, there is already a large body of reliable experimental results in the manuscript that are worth publishing without a new round of extensive experiments.

Significance

A simple method of specific visualization of fluorescent protein fusion bands on SDS-PAGE is proposed.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

Sanial et al. carefully analyze the use of in-gel fluorescence as an alternative to immunoblotting. The authors show that simple modifications of common protein extraction protocols can preserve (to varying extents) fluorescent proteins in their native, fluorescent states. This can be exploited in different applications for in-gel fluorescence quantification, bypassing immunoblotting. The experimental results are clear, showcasing the ease and linearity of in-gel fluorescence quantification.

In my opinion, the trick of this approach is also potentially its main drawback, the partial denaturation conditions. I think the manuscript could be strengthened with more extensive benchmarking of the approach and further discussion of potential caveats as detailed below.

Major points:

1. Protein abundance in the original GFP library (and in other FP-tagged libraries constructed in the meanwhile) have been quantified using fluorescence (flow cytometry, microscopy, colony fluorescence) (Ho et al. 2018 10.1016/j.cels.2017.12.004, Weill et al. 2018 10.1038/s41592-018-0044-9, Meurer et al. 2018 10.1038/s41592-018-0045-8). This provides an opportunity to significantly strengthen the manuscript (where most of the test have been done using two abundant cytosolic proteins Bmh1 and Hxk1) if the authors could apply their approach to a representative fraction of the yeast proteome (sampling from such libraries FP-tagged proteins that differ in abundance, localization, membrane vs cytosolic/nuclear, subunits of large stable complexes vs proteins not part of complexes, etc.) and compare their quantification with previous relative abundance estimates. This information would also help future users in case protein-specific issues are identified.
2. The authors discuss several drawbacks, including the change in apparent molecular weight compared to denatured proteins; differential recognition of folded vs denatured proteins by antibodies.

Other potentials drawbacks should be discussed. For instance, the need of additional steps post-fluorescence imaging for signal normalization against a loading control; the need of antibodies and immunoblotting to decide on the best denaturing temperature for a specific protein or FP tag; complexity of the native protein extraction protocol compared for example to alkaline lysis followed by TCA precipitation (Knop et al. 1999 PMID: 10407276).

Moreover, although the denatured fraction is FP- and temperature-dependent, even under the milder 30{degree sign}C conditions there is a detectable denatured fraction (Fig.s3b). This would seem to preclude the use of this approach for absolute protein quantification.

Finally, any evidence that the denatured fraction would depend on the protein tagged with the FP?

Minor points:

1. In the experiments designed to test the linearity and sensitivity of the approach, an alternative approach that would not result in dilution of cell extract is to mix wild type cell extract (no GFP fusion) with extract of the GPF-tagged strain in different ratios.
2. Define all acronyms at first appearance. For example, DTT and LDS on page 4.
3. Fig.4D: the colors chosen to represent EGFP and sfGFP data make them hard to tell apart. The same comment to Fig.S6.
4. As the temperature steps are not uniform in Figures 4 and 5, it would be more informative to indicate the exact temperate above each lane (in addition/instead of the ramp cartoon).
5. Regarding linearity, that HRP-based quantification is not linear is expected. A fairer comparison would be to use fluorescently labeled secondary antibodies. It is also puzzling that detection with signal amplification (HRP) is less sensitive than direct quantification of the fluorescence signal from the FP tag.
6. I appreciate the workflow Figure 10. But in my opinion it is trying to show too much (protocol, troubleshooting, calls to figure panels). Perhaps it could be made clearer by separating the protocol steps/settings from the optimization/troubleshooting tips.
7. Some of the discussion of different fluorescent proteins, and expression levels of tagged proteins, could be confounded by the different linkers used in the tagging constructs.

Referee Cross-commenting

This session contains comments from all reviewers.

Reviewer 1: I agree with the assessment by Reviewer #2. Considering the comment about potential oligomerization of a protein of interest, I stand by my point about testing the method with more proteins of interest. How extensive this testing should be or whether additional discussion of possible issues would suffice is a matter of opinion. It is clear from the manuscript in it's current form that the method works and that it has caveats.

Reviewer 2: In general, I agree with the points raised by Reviewer #1. However, in my opinion, there is already a large body of reliable experimental results in the manuscript that are worth publishing without a new round of extensive experiments.

Reviewer 1: Fair enough, I don't insist on the experiments in my point 1.

Significance

Could be a generally useful and simple approach for in-gel quantification using fluorescent protein tags.

Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

Learn more at Review Commons

---

Reply to the reviewers

Thank you very much for your editorial handling of our manuscript entitled 'A conserved fungal Knr4/Smi1 protein is vital for maintaining cell wall integrity and host plant pathogenesis'. We have taken on board the reviewers' comments and thank them for their diligence and time in improving our manuscript.

Please find our responses to each of the comments below.

Reviewer(s)' comments

Reviewer #1

---

Major comments:

---

__1.1. As a more critical comment, I find the presentation of the figures somewhat confusing, especially with the mixing of main figures, supplements to the main figures, and actual supplemental data. On top of that, the figures are not called up in the right order (e.g. Figure 4 follows 2D, while 3 comes after 4; Figure 6 comes before 5...), and some are never called up (I think) (e.g. Figure 1B, Figure 2B). __

---

__Response: __The figure order has been revised according to the reviewer's suggestion, while still following eLife's formatting guidelines for naming supplementals. Thank you.

1.2. I agree that there should be more CWI-related genes in the wheat module linked to the FgKnr4 fungal module, or, vice-versa, CW-manipulating genes in the fungal module. It would at least be good if the authors could comment further on if they find such genes, and if not, how this fits their model.

---

Response: Thank you for your insightful suggestion regarding the inclusion of more CWI-related genes in the wheat module linked to the FgKnr4 fungal module F16, or vice versa. We did observe a co-regulated response between the wheat module W05 which is correlated to the FgKnr4 module F16. Namely, we observed an enrichment of oxidative stress genes including respiratory burst oxidases and two catalases (lines 304 - 313) in the correlated wheat module (W05). Early expression of these oxidative stress inducing genes likely induces the CWI pathway in the fungus, which is regulated by FgKnr4. Knr4 functions as both a regulatory protein in the CWI pathway and as a scaffolding protein across multiple pathways in S. cerevisiae (Martin-Yken et al., 2016, https://onlinelibrary.wiley.com/doi/10.1111/cmi.12618 ). Scaffolding protein-encoding genes are typically expressed earlier than the genes they regulate to enable pre-assembly with their interacting partners, ensuring that signaling pathways are ready to activate when needed. In this context, the CWI integrity MAPKs Bck1 and Mkk1 are part of module F05, which includes two chitin synthases and a glucan synthase. This module is highly expressed during the late symptomless phase. The MAPK Mgv1, found in module F13, is expressed consistently throughout the infection process, which aligns with the expectation that MAPKs are mainly post-transcriptionally regulated. Thank you for bringing our attention to this, this is now included in the discussion (lines 427 - 443) along with eigengene expression plots of all modules added to the supplementary (Figure 3 - figure supplement 1).

To explore potential shared functions of FgKnr4 with other genes in its module, we re-analyzed the high module membership genes within module F16, which includes FgKnr4, using Knetminer (Hassani-Pak et al., 2021; https://onlinelibrary.wiley.com/doi/10.1111/pbi.13583 ). This analysis revealed that 8 out of 15 of these genes are associated with cell division and ATP binding. Four of the candidate genes are also part of a predicted protein-protein interaction subnetwork of genes within module F16, which relate to cell cycle and ATP binding. In S. cerevisiae, the absence of Knr4 results in cell division dysfunction (Martin-Yken et al., 2016, https://onlinelibrary.wiley.com/doi/10.1111/cmi.12618 ). Accordingly, we tested sensitivity of ΔFgknr4 to microtubule inhibitor benomyl (a compound commonly used to identify mutants with cell division defects; Hoyt et al., 1991 https://www.cell.com/cell/pdf/0092-8674(81)90014-3.pdf). We found that the ΔFgknr4 mutant was more susceptible to benomyl, both when grown on solid agar and in liquid culture. This data has now been added Figure 7, and referred to in lines 338-348.

__Specific issues: __

---

1.3. In the case of figure 5, I generally find it hard to follow. In the text (line 262/263), the authors state that 5C shows "eye-shaped lesions" caused by ΔFgknr4 and ΔFgtri5, but I can't see neither (5C appears to be a ΔFgknr4 complementation experiment). The figure legend also states nothing in this regard.

__Response: __Thank you for your suggestion. We have amended the manuscript to include an additional panel that shows the dissected spikelet without its outer glumes, making the eye shaped diseased regions more visible in Figure 5.

__1.4. Figure 5D supposedly shows 'visibly reduced fungal burden' in ΔFgknr4-infected plants, but I can't really see the fungal burden in this picture, but the infected section looks a lot thinner and more damaged than the control stem, so in a way more diseased. __

---

Response: __Thank you for your insight. We have revised our conclusions based on this image to state that while ΔFgknr4 can colonise host tissue, it does so less effectively compared to the wild-type strain as we are unable to quantitatively evaluate fungal burden using image-colour thresholding due to the overlapping colours of the fungal cells and wheat tissues. Decreased host colonisation is evidenced by (i) reduced fungal hyphae proliferation, particularly in the thicker adaxial cell layer, (ii) collapsed air spaces in wheat cells, and (iii) increased polymer deposition at the wheat cell walls, indicating an enhanced defence response. __Figure 5 has been amended to include these observations in the corresponding figure legend and the resin images now include insets with detailed annotation.

__1.5. The authors then go on to state (lines 272-273) that they analyzed the amounts of DON mycotoxin in infected tissues, but don't seem to show any data for this experiment. __

Response: __We have amended this to now include the data in __Figure 5 - figure supplement 2B, thank you.

Reviewer #2

---

__Major issues: __

---

2.1 If Knf4 is involved in the CWI pathway, what other genes involved in the CWI pathway are in this fungal module? one of the reasons for developing modules or sub-networks is to assign common function and identify new genes contributing to the function. since FgKnr4 is noted to play a role in the CWI pathways, then genes in that module should have similar functions. If WGCN does not do that, what is the purpose of this exercise?

---

Response: __Thank you for raising this point regarding the role of FgKnr4 in the CWI pathway and the expectations for genes of shared function within the FgKnr4 module F16. We did observe that the module containing FgKnr4 (F16) was also correlated to a wheat module (W05) which was significantly enriched for oxidative stress genes. This pathogen-host correlated pattern led us to study module F16, which otherwise lacks significant gene ontology term enrichment, unique gene set enrichments, and contains few characterised genes. This is now highlighted in __lines 233-246. This underscores the strength of the WGCNA. By using high-resolution RNA-seq data to map modules to specific infection stages, we identified an important gene that would have otherwise been overlooked. This approach contrasts with other network analyses that often rely on the guilt-by-association principle to identify novel virulence-related genes within modules containing known virulence factors, potentially overlooking significant pathways outside the scope of prior studies. Therefore, our analysis has already benefited from several advantages of WGCNA, including the identification of key genes with high module membership that may be critical for biological processes, as well as generating a high-resolution, stage-specific co-expression map of the F. graminearum infection process in wheat. This point is now emphasised in lines 233-252. As discussed in response to reviewer 1, Knr4 functions as both a regulatory protein in the CWI pathway and as a scaffolding protein across multiple pathways in S. cerevisiae (Martin-Yken et al., 2016, https://onlinelibrary.wiley.com/doi/10.1111/cmi.12618 ) which would explain its clustering separate from the CWI pathway genes. The high module membership genes within module F16 containing FgKnr4 were re-analysed using Knetminer (Hassani-Pak et al., 2021; https://onlinelibrary.wiley.com/doi/10.1111/pbi.13583 ), which found that 8/15 of these genes were related to cell division and ATP binding. Four of the candidate genes are also part of a predicted protein-protein interaction subnetwork of genes within module F16, which relate to cell cycle and ATP binding. In S. cerevisiae, the absence Knr4 leads to dysfunction in cell division. Accordingly, we tested sensitivity of ΔFgknr4 to the microtubule inhibitor benomyl (a compound commonly used to identify mutants with cell division defects; Hoyt et al., 1991 https://www.cell.com/cell/pdf/0092-8674(81)90014-3.pdf). We found that the ΔFgknr4 mutant was more susceptible to benomyl, both when grown on solid agar and in liquid culture. This data has now been added as Figure 7 and referred to in lines 338-348.

---

2.2. Due to development defects in the Fgknr1 mutant, I would not equate to as virulence factor or an effector gene.

---

__Response: __We are in complete agreement with the reviewer and are not suggesting that FgKnr4 is an effector or virulence factor, we have been careful with our wording to indicate that FgKnr4 is simply necessary for full virulence and its disruption results in reduced virulence and have outlined how we believe FgKnr4 participates in a fungal signaling pathway required for infection of wheat.

---

2.3. What new information is provided with WGCN modules compared with other GCN network in Fusarium (examples of GCN in Fusarium is below) ____https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5069591/ https://doi.org/10.1186/s12864-020-6596-y____ DOI: 10.1371/journal.pone.0013021. The GCN networks from Fusarium have already identified modules necessary/involved in pathogenesis.

Response: __The 2016 New Phytologist gene regulatory network (GRN) by Guo et al. is large and comprehensive. However, only three of the eleven datasets are in planta, with just one dataset focusing on F. graminearum infection on wheat spikes. The other two in planta datasets involve barley infection and Fusarium crown rot. By combining numerous in planta and in vitro datasets, the previous GRNs lack the fine resolution needed to identify genetic relationships under specific conditions, such as the various stages of symptomatic and symptomless F. graminearum infection of mature flowering wheat plants. This limitation is highlighted in the 2016 paper itself. This network is expanded in the Guo et al., 2020 BMC genomics paper where it includes one additional in planta and nine in vitro datasets. However, the in planta dataset involves juvenile wheat coleoptile infection, which serves as an artificial model for wheat infection but is not on mature flowering wheat plants reminiscent of Fusarium Head Blight of cereals in the field. This model differs significantly in the mode of action of F. graminearum, notably DON mycotoxin is not essential for virulence in this context (Armer et al. 2024, https://pubmed.ncbi.nlm.nih.gov/38877764/ ). The Guo et al., 2020 paper still faces the same issues in terms of resolution and the inability to draw conclusions specific to the different stages of F. graminearum infection. Additionally, these GRNs use Affymetrix data, which miss over 400 genes (~ 3 % of the genome) from newer gene models. In contrast, our study addresses these limitations by analysing a meticulously sampled, stage- and tissue-specific in planta RNA-seq dataset using the latest reference annotation. Our approach provides higher resolution and insights into host transcriptomic responses during the infection process. The importance of our study in the context of these GRNs is now addressed in the introduction (__lines 85-92).

---

2.4. Ideally, the WGCN should have been used identify plant targets of Fusarium pathogenicity genes. This would have provided credibility and usefulness of the WGCN. Many bioinformatic tools are available to identify virulence factors and the utility of WGCN in this regard is not viable. However, if the authors had overlapped the known virulence factors in a fungal module to a particular wheat module, the impact of the WGCN would be great. The module W12 has genes from numerous traits represented and WGCN could have been used to show novel links between Fg and wheat. For example, does tri5 mutant affect genes in other traits?

__Response: __Thank you for your suggestions. In this study we have shown the association between the main fungal virulence factor of F. graminearum, DON mycotoxin, with wheat detoxification responses. Through this we have identified a set of tri5 responsive genes and validated this correlation in two genes belonging to the phenylalanine pathway and one transmembrane detoxification gene. Although we could validate more genes in this tri5 responsive wheat module, our paper aimed to investigate previously unstudied aspects of the F. graminearum infection process and how the fungus responded to changing conditions within the host environment. We accomplished this by characterising a gene within a fungal module that had limited annotation enrichment and few characterised genes. Tri5 on the other hand is the most extensively studied gene in F. graminearum and while the network we generated may offer new insights into tri5 responsive genes, this is beyond the scope of our current study. In addition to the tri5 co-regulated response, we have also demonstrated the coordinated response between the fungal module F16, which contains FgKnr4 that is necessary for tolerance to oxidative stress, and the wheat module W05, which is enriched for oxidative stress genes.

---

While our co-expression network approach can be used to explore and validate other early downstream signaling and defense components in wheat cells, several challenges must be considered: (a) the poor quality of wheat gene calls, (b) genetic redundancy due to both homoeologous genes and large gene families, and (c) the presence of DON, which can inhibit translation and prevent many transcriptional changes from being realised within the host responses. Additionally, most plant host receptors are not transcriptionally upregulated in response to pathogen infection (most R gene studies for the NBS-LRR and exLRR-kinase classes), making their discovery through a transcriptomics approach unlikely. These points will be included in our discussion (lines 408-413), thank you.

Specific issues

• *

2.5. Since tri5 mutant was used a proof of concept to link wheat/Fg modules, it would have been useful to show that TRI14, which is not involved DON biosynthesis, but involved in virulence ( https://doi.org/10.3390/applmicrobiol4020058____) impact the wheat module genes.

---

Response: __Our goal was to show that wheat genes respond to the whole TRI cluster, not just individual TRI genes. Therefore, the tri5 mutant serves as a solid proof-of-concept, because TRI5 is essential for DON biosynthesis, the primary function of the TRI gene cluster, thereby representing the function of the cluster as a whole. This is now clarified in __lines 217-219. Additionally, the uncertainties surrounding other TRI mutants would complicate the question we were addressing-namely, whether a wheat module enriched in detoxification genes is responding to DON mycotoxin, as implied by shared co-expression patterns with the TRI cluster. For instance, the referenced TRI14 paper indicates that DON is produced in the same amount in vitro in a single media. Although the difference is not significant, the average DON produced is lower for the two Δtri14 transformants tested. Therefore, we cannot definitively rule out that TRI14 is involved in DON biosynthesis and extrapolate this to DON production in planta. Despite this, the suggestion is interesting, and would make a nice experiment but we believe it does not contribute to the overall aim of this study.

2.6. Moreover, prior RNAseq studies with tri5 mutant strain on wheat would have revealed the expression of PAL and other phenylpropanoid pathway genes?

__Response: __We agree that this would be an interesting comparison to make but unfortunately no dataset comparing in planta expression of the tri5 mutant within wheat spikes exists.

2.7. Table S1 lists 15 candidate genes of the F16 module; however, supplementary File 1 indicates 74 genes in the same module. The basis of exclusion should be explained. The author has indicated genes with high MM was used as representative of the module. The 59 remaining genes of this module did not meet this criteria? Give examples.

---

Response: __The 15 genes with the highest module membership were selected as initial candidates for further shortlisting from the 74 genes within module F16. In WGCNA, genes with high module membership (MM) (i.e. intramodular connectivity) are predicted to be central to the biological functions of the module (Langfelder and Horvath, 2008; https://bmcbioinformatics.biomedcentral.com/articles/10.1186/1471-2105-9-559 ) and continues to be a metric to identify biologically significant genes within WGCN analyses (https://bmcplantbiol.biomedcentral.com/articles/10.1186/s12870-024-05366-0 Tominello-Ramirez et al., 2024; https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9151341/ ;Zheng et al., 2022; https://www.nature.com/articles/s41598-020-80945-3 Panahi and Hejazi et al 2021). Following methods by Mateus et al. (2019) (https://academic.oup.com/ismej/article/13/5/1226/7475138 ) key genes were defined as those exhibiting elevated MM within the module, which were also strongly correlated (R > |0.70|) with modules of the partner organism (wheat). We have clarified this point in the manuscript. Thank you for the suggestion. (__Lines 253-263).

2.____8. A list from every module that pass this criteria will be useful resource for functional characterization studies.

---

__Response: __A supplementary spreadsheet has been generated which includes full lists of the top 15 genes with the highest module membership within the five fungal modules correlated to wheat modules and a summary of shared attributes among them. Thank you for this suggestion.

2.9. Figure 3 indicates TRI genes in the module F12; your PHI base in Supp File S2 lists only TRI14. Why other TRI genes such as TRI5 not present in this File?

---

Response: For clarity, the TRI genes in module F12 are TRI3, TRI4, TRI11, TRI12, and TRI14 which was stated in Table 1. TRI5 clusters with its neighboring regulatory gene TRI6 in module F11, which exhibits a similar but reduced expression pattern compared to module F12. To improve clarity on this the TRI genes in module F12 are also listed in-text in line 168 and added to Figure 4. The enrichment and correlated relationship of W12 to a cluster's expression still imply a correlated response of the wheat gene to the TRI cluster's biosynthetic product (DON), which is absent in the Δtri5 mutant.

TRI14 and TRI12 are listed in PHI-base. TRI12 was mistakenly excluded due to an unmapped Uniprot ID, which were added separately in the spreadsheet. We will recheck all unmapped ID lists to ensure all PHI-base entries are included in the final output. Thank you for pointing out this error.

---

2.10. What is purpose of listing the same gene multiple times? Example, osp24 (a single gene in Fg) is listed 13 times in F01 module.

---

__Response: __This is a consequence of each entry having a separate PHI ID, which represents different interactions including inoculations on different cultivar. Cultivar and various experimental details were omitted from the spreadsheet to reduce information density, however the multiple PHI base ID's will be kept separate to make the data more user friendly when working with the PHI-base database. An explanation for this is now provided in the file's explanatory worksheet, thank you.

Reviewer #3:

---

3.1. Why only use of high confidence transcripts maize to map the reads and not the full genome like Fusarium graminearum? I have never analyzed plant transcriptome.

---

__Response: __ In the wheat genome, only high-confidence gene calls are used by the global community (Choulet et al., 2023; https://link.springer.com/chapter/10.1007/978-3-031-38294-9_4 ) until a suitable and stable wheat pan-genome becomes available.

3.2. The regular output of DESeq are TPMs, how did the authors obtain the FPKM used in the analysis?

---

Response: FPKM was calculated using the GenomicFeatures package and included on GitHub to enhance accessibility for other users. However, the input for WGCNA and this study as a whole was normalised counts rather than FPKM. The FPKM analysis was done to improve interoperability of the data for future users and made available on Github. To complement this, the information regarding FPKM calculation is now included in the methods section of the revised manuscript (line 491).

3.3. Do the authors have a Southern blot to prove the location of the insertion and number of insertions in Zymoseptoria tritici mutant and complemented strains?

---

__Response: __No, but the phenotype is attributed to the presence or absence of ZtKnr4, as the mutant was successfully complemented in multiple phenotypic aspects. This satisfies Koch's postulates which is the gold standard for reverse genetics experimentation (Falkow 1988; https://www.jstor.org/stable/4454582 ).

__3.4. Boxplots and bar graphs should have the same format. In Figures 5 B and F and supplementary figure 6.3 the authors showed the distribution of samples but it is lacking in figure 3 B and all bar graphs. __

---

__Response: __Graphs have been modified to display the distribution of all samples, thank you.

3.5. Line 247 FGRAMPH1_0T23707 should be FGRAMPH1_01T23707

---

__Response: __Thank you this has now been amended.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #3

Evidence, reproducibility and clarity

The authors of the manuscript entitled "A conserved fungal Knr4/Smi1 protein is vital for maintaining cell wall integrity and host plant pathogenesis" used a weighted gene co-expression network to identify Fusarium graminearum genes highly expressed during early symptomless infection of wheat. Based on its sequence and previous studies, authors selected FgKnr4 from the early symptomless Fusarium modules. The characterization of knockout strains revealed a role in morphogenesis, growth, cell wall stress tolerance, and virulence in F. graminearum and the phylogenetically distant fungus Zymoseptoria tritici.

The methods are properly described and statistical analysis are reasonable so reproducibility is possible. The RNA-seq dataset is already published and the authors provided a repository with the code used to create the co-expression network. However, I have the following questions:

• Why only use of high confidence transcripts maize to map the reads and not the full genome like Fusarium graminearum? I have never analyzed plant transcriptome.
• The regular output of DESeq are TPMs, how did the authors obtain the FPKM used in the analysis?
• Do the authors have a southern blot to prove the location of the insertion and number of insertions in Zymoseptoria tritici mutant and complemented strains?
• Boxplots and bar graphs should have the same format. In Figures 5 B and F and supplementary figure 6.3 the authors showed the distribution of samples but it is lacking in figure 3 B and all bar graphs.
• Line 247 FGRAMPH1_0T23707 should be FGRAMPH1_01T23707

Referees cross-commenting

I agree with reviewer 1, the order in which the figures are called in the text is confusing. Regardless of figures 5C-D I am no expert in the field therefore I can only say they look like they have not been edited.

I agree with reviewer 1, data of DON mycotoxin production in infected issues is need it to support statement in line 272-273.

I agree with Reviewer 2, the criteria to exclude genes from the final selection list should be explained.

Significance

The study showed, once again, that a weighted gene co-expression network is a great method to identify new genes of interest regardless of the organism or condition even if not very popular in the fungal pathogen field yet. The study proved that functions identified in a WGCN module from a pathogen have their opposite in the host module. The authors go beyond the theory and demonstrate the effect of the highest expressed gene during the early symptomless stage of infection in maize and wheat fungal pathogens.

Fungal pathogen, RNA-seq, metabolic models, metabolism, comparative genomics

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

Summary: The authors in this manuscript use "dual weighting" to identify clusters or modules of genes from the fungus F. graminearum (Fg) with coordinated expression patterns with genes in wheat modules - potentially uncover key regulators or pathways linking Fg genes with plant traits, including plant pathogenesis. As proof of concept, the authors use one of the fungal genes FgKnr4 identified in a fungal module that has strong link with the wheat module. They were able to show that this gene is likely involved in CWI pathway and affects virulence properties of the fungus

Major comments:

Does the WGCN provide useful framework to link fungal genes affecting plant traits? If Knf4 is involved in the CWI pathway, what other genes involved in the CWI pathway are in this fungal module? This is not forthcoming. Due to development defects in the Fgknr1 mutant, I would not equate to as virulence factor or an effector gene.

Since tri5 mutant was used a proof of concept to link wheat/Fg modules, it would have been useful to show that TRI14, which is not involved DON biosynthesis, but involved in virulence ( https://doi.org/10.3390/applmicrobiol4020058) impact the wheat module genes. Moreover, prior RNAseq studies with tri5 mutant strain on wheat would have revealed the expression of PAL and other phenylpropanoid pathway genes?

Table S1 lists 15 candidate genes of the F16 module; however, supplementary File 1 indicates 74 genes in the same module.

The basis of exclusion should be explained. The author has indicated genes with high MM was used as representative of the module. The 59 remaining genes of this module did not meet this criteria? Give examples. Did similar exclusion criteria used for other modules and if so, how many genes in each module pass the criteria? For example, Did TRI5 in module F12 pass this criteria. A list from every module that pass this criteria will be useful resource for functional characterization studies.

Minor comments:

Figure 3 indicates TRI genes in the module F12; your PHI base in Supp File S2 lists only TRI14. Why other TRI genes such as TRI5 not present in this File? What is purpose of listing the same gene multiple times? Example, osp24 (a single gene in Fg) is listed 13 times in F01 module.

Referees cross-commenting

agree with both reviewers regarding clarification of Figures.

one of the reasons for developing modules or sub-networks is to assign common function and identify new genes contributing to the function. since FgKnr4 is noted to play a role in the CWI pathways, then genes in that module should have similar functions. If WGCN does not do that, what is the purpose of this exercise?

Significance

What new information is provided with WGCN modules compared with other GCN network in Fusarium (examples of GCN in Fusarium is below)

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5069591/ https://doi.org/10.1186/s12864-020-6596-y DOI: 10.1371/journal.pone.0013021

The GCN networks from Fusarium have already identified modules necessary/involved in pathogenesis. Ideally, the WGCN should have been used identify plant targets of Fusarium pathogenicity genes. This would have provided credibility and usefulness of the WGCN.

Many bioinformatic tools are available to Identify virulence factors and the utility of WGCN in this regard is not viable. However, if the authors had overlapped the known virulence factors in a fungal module to a particular wheat module, the impact of the WGCN would be great. The module W12 has genes from numerous traits represented and WGCN could have been used to show novel links between Fg and wheat. For example, does tri5 mutant affect genes in other traits?

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

Summary:

A public mRNA-seq dataset from Dilks et al. (2019) for wheat spikelets infected by Fusarium graminearum was used to generate a dual weighted gene co-expression network (WGCN). Since colonization of the spike by F. graminearum progresses from spikelet to spikelet, thereby forming an infection-gradient from early to late stages, quasi spatio-temporal resolution for the transcriptomic dataset can be achieved by cutting the spike into equal pieces along this gradient (in this case cuts were done at rachis internodes 1-2, 3-4, 5-6, and 7-8. The authors created co-expression networks for both, fungal and plant genes, and cross-correlated them. They identify several modules specific for each infection stage. For further analysis, the authors focus on two module pairs. (1) the wheat module 12 (W12), which correlates to Fusarium module 12 (F12), and (2) the Fusarium module 16 (F16) and the correlated wheat modules 1 and 5 (W01/W05). The W12/F12 modules were deemed of interest because they were specific to the transition from symptomless to symptomatic infection stage. Here, the authors find genes related to mycotoxin production to be upregulated in the F12 module, while the W12 is enriched in genes involved in detoxification. F16 and W01/W05 are specific to the earliest stages of infection, and thus most likely involved in fungal virulence. Here, one of the key genes identified is FgKnr4, which the authors show to be important for fungal virulence, as gene knockout leads to a premature stop of disease progression. As the authors show that FgKnr4 is involved in activating cell wall-integrity mechanisms, and may function in oxidative stress-resistance, this reduced virulence may be the result a reduced ability of the fungus to withstand plant defense mechanisms. Interestingly, knocking out an orthologue of FgKnr4 in Zymoseptoria tritici led to similarly reduced virulence of this pathogenic fungus on wheat plant.

Comments:

Overall, I find the WGCN analysis to be very interesting and informative, especially because of the different stages of infection. As the dataset is made public (I believe), I think that this will be a really important resource for the community. The exemplary functional analysis of the F16/W01/W05 modules via FgKnr4 is very interesting and demonstrates that novel genes involved in virulence can be identified via this approach. A similar more detailed analysis of the W12/F12 modules with a focus on detoxification mechanisms in the plant (i.e. the W12 module) would be a very interesting bonus, but as much as I would be interested in reading about it, functional gene analyses in wheat are obviously time-consuming, and it is not essential to this manuscript. As a more critical comment, I find the presentation of the figures somewhat confusing, especially with the mixing of main figures, supplements to the main figures, and actual supplemental data. On top of that, the figures are not called up in the right order (e.g. Figure 4 follows 2D, while 3 comes after 4; Figure 6 comes before 5...), and some are never called up (I think) (e.g. Figure 1B, Figure 2B). In the case of figure 5, I generally find it hard to follow. In the text (line 262/263), the authors state that 5C shows "eye-shaped lesions" caused by ΔFgknr4 and ΔFgtri5, but I can't see neither (5C appears to be a ΔFgknr4 complementation experiment). The figure legend also states nothing in this regard. Figure 5D supposedly shows 'visibly reduced fungal burden' in ΔFgknr4-infected plants, but I can't really see the fungal burden in this picture, but the infected section looks a lot thinner and more damaged than the control stem, so in a way more diseased. The authors then go on to state (lines 272-273) that they analyzed the amounts of DON mycotoxin in infected tissues, but don't seem to show any data for this experiment. In contrast to the sometimes confusing data presentation, I find the table of correlated modules (table 1) very helpful, and obviously am happy to see that all data is available in the first author's GitHub account.

Referees cross-commenting

just to clarify in regards to my comment on Figures 5C-D, and Reviewer #3's comment "Regardless of figures 5C-D I am no expert in the field therefore I can only say they look like they have not been edited." - I didn't want to insinuate that the images have been edited. Based on the images provided, I just can't see what the authors state is shown. So this is not about editing/manipulation - just about image quality/choice. The phenotypic descriptions by the authors are quite detailed ("eye-shaped lesions", 'visibly reduced fungal burden'...), but at least for me, the images aren't good enough to illustrate and underpin their statements. Maybe better images are needed, maybe magnifications of the exact regions showing the phenotypes? But this is simply a matter of presentation, not of editing/manipulation.

Second, I agree that there should be more CWI-related genes in the wheat module linked to the FgKnr4 fungal module, or, vice-versa, CW-manipulating genes in the fungal module. It would at least be good if the authors could comment further on if they find such genes, and if not, how this fits their model.

Significance

In summary, I think that the presented WGCN analysis of mRNA-seq data with quasi-spatio-temporal resolution is a very helpful approach to identify novel fungal virulence and plant immunity genes, and with the created datasets made public, this will be an interesting and valuable resource for the community. The identification and functional analysis of FgKnr4 works as proof-of-principle. If the data presentation is improved, I believe that this will be an interesting publication.

Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

Learn more at Review Commons

---

Reply to the reviewers

Manuscript number:

RC-2024-02569

Corresponding author(s): Mary O'Riordan, Teresa O'Meara

1. General Statements

We thank the reviewers for their positive feedback, highlighting the significance and novelty of our work, especially regarding the novel functions of IRE1a in regulating phagosome biology during infection. We also appreciate some overarching themes that were focused on by multiple reviewers, including the role of XBP1S protein and RIDD activity, which we have addressed here. We have also added additional data, made adjustments to data presentation, and added clarifying language to address concerns from Reviewer 3. We appreciate these constructive suggestions and include our planned experiments to address reviewer concerns here. Our specific responses to the reviewer comments are below.

Specific figures used in the response to reviewers are in the attached file as they cannot be pasted here.

2. Description of the planned revisions

Reviewer 1:

1) The demonstration of protein misfolding independent IRE1 activation should also be demonstrated using molecules such as TUDCA or 4PBA that should be innocuous regarding the splicing of XBP1s. It would also be interesting to evaluate the activation of the other arms of the UPR in particular through the phosphorylation of eIF2a, expression of ATF4 and cleavage of ATF6.

We appreciate the suggestion to strengthen our data regarding protein misfolding-independent activation of IRE1 more robust. We note that canonical UPR transcriptional targets are not induced during C. albicans infection (Fig. 2G,H), suggesting that IRE1 is activated in the absence of a standard unfolded protein response. However, we agree that we can use additional chemical chaperones to assay this. To address this point, we will perform the suggested experiments in the presence or absence of TUDCA with C. albicans, LPS, thapsigargin, and tunicamycin. As 4PBA has been shown to inhibit protein synthesis, rather than promoting protein folding or preventing aggregation (PMC9741500), we will avoid using this compound for these assays.

We will also perform western blots for ATF6 cleavage and eIF2a phosphorylation, although we note that eIF2a can be phosphorylated by multiple kinases and can be triggered by nutrient deprivation or changes in intracellular calcium, both of which occur during C. albicans infection (glucose: PMC6709535; calcium: data within this manuscript).

3) The authors use thioflavin to evaluate the extent of protein misfolding. This type of stain can lead to artefactual results and in general it is rather safer to test several stainers (see for instance the work presented in PMC10720158)

We thank the reviewer for this suggestion. We have previously tried Proteostat staining as an additional method to measure protein misfolding, but we found that it bound strongly to the C. albicans cell wall, which would result in a strong false positive signal that is not indicative of host protein misfolding (see below). Congo Red, an additional dye used in the listed reference, is also known to bind to C. albicans and perturbs cell wall synthesis (PMC266468), therefore we have avoided these dyes.

However, to address this point, we will perform experiments utilizing poly-ubiquitin blotting, as in the suggested reference, as an orthogonal readout of protein misfolding during C. albicans infection or treatment with LPS, depleted zymosan, and thapsigargin.

__Figure legend: Proteostat staining with _C. albicans_ infection. __Macrophages were infected with C. albicans, and subsequently stained with Proteostat to measure protein misfolding. Proteostat bound and displayed strong fluorescence on the C. albicans cell wall.

6) The whole study relies on the use of IRE1deltaR to impair IRE1 signaling. The authors should validate their hypothesis with an orthogonal approach, for instance with IRE1 pharmacological inhibitors (eg MKC8866 or KIRA8).

We consider the use of genetic perturbation of IRE1 to be a strength of this manuscript, as IRE1 inhibitors have been shown to cause off-target effects (KIRA8: PMC9600248). However, to address this point, we will attempt to replicate important phenotypes, including the effect of IRE1 on calcium flux and phagolysosome fusion, using MKC8866 and KIRA8 as representative inhibitors.

3. Description of the revisions that have already been incorporated in the transferred manuscript

__Reviewer 1: __

5) The authors focus on the IRE1/XBP1s signaling arm of the UPR but do not explore RIDD activity which has been linked to several infection mechanisms and lysosomal integrity (in particular by regulating the expression of BLOS1 - see PMC9119680 and PMC6446841). The authors should definitely evaluate how RIDD is activated (or not) in their experimental systems.

We thank the reviewer for this suggestion, as we have considered potential effects of RIDD when analyzing our RNA-seq data, and are aware of the potential links between IRE1, BLOS1 (encoded by Bloc1s1) expression, and lysosome perturbations. We now add additional figures to our supplemental data (Fig. S3C-D; also shown below) showing that established RIDD targets, including Bloc1s1 are not depleted during C. albicans infection, and also not increased in IRE1 null macrophages. We add the following text to describe these findings (lines 322-326): "Additionally, we did not observe depletion of published RIDD targets (14, 65, 66) during C. albicans infection in WT macrophages (Fig. S3C; Table S2), nor increased expression of RIDD targets in IRE1ΔR macrophages, compared to IRE1 WT macrophages (Fig. S3D; Table S1.1), suggesting minimal RIDD activity during C. albicans infection." We also note that experiments with LysoSensor (Fig. 3E) suggested lysosome biogenesis is not impaired in IRE1 null macrophages. Therefore, we expect RIDD activity has negligible effects on our reported phenotypes.

Reviewer #1 (Significance (Required)):

The manuscript is interesting and highlights novel aspects towards the interaction between macrophages and a pathogen, candida albicans, involving the likely selective activation of IRE1. The data are novel and experimentally sound. Several controls are however missing.

The strengths of the study are associated with the novelty of the findings, with the links that could potentially derive from this study to connect ER biology, UPR signaling and phagosome maturation

The main weaknesses are associated i) with the fact that the authors did not evaluate RIDD activity which has already been linked with pathogen infection and with lysosome integrity, ii) with methodological aspects, in particular regarding the demonstration of the IRE1 activation independent on protein misfolding and the sole use of a genetic variant of IRE1 to test their hypotheses

We thank Reviewer 1 for their constructive feedback and for noting the novelty of our findings. We believe that the data we have added regarding RIDD activity and our planned experiments to address additional concerns will add additional evidence to support our findings.

Reviewer 2:

1. A point that should be addressed with more detail is the correlation of fungal killing with Ca2+ fluxes and Ire1α activity, given the well-known data regarding the strong ability of the axis dectin/SYK/phospholipase Cγ to induce Ca2+ transients, a response not shared by LPS signaling, and the sequential activation of mitochondrial Ca2+ uniporter (MCU), which is a critical element of fungal killing associated with the citrate-pyruvate shuttle as a NADPH source (Seegren et al., Cell Rep. 33: 108411, 2020). Incidentally, this paper is referred in ref. 46 as a preprint, although it was accessible in Cell Reports in 2020.

This is an excellent suggestion; we have added this topic to our discussion (lines 605-608) and have corrected the citation.

The assay of the expression of V-ATPase complex, mitochondrial calcium uniporter, and mitochondrial uptake 1 and 2 could shed light on the dependence of fungal killing on Ire1α function.

Thank you for this suggestion - below, we plot the transcripts comprising the V-ATPase, as well as Mcu, Micu1, and Micu2. We note that these transcripts are not perturbed in IRE1 null macrophages, suggesting that the basic functions of the V-ATPase complex and mitochondrial calcium uptake are intact in IRE1 null macrophages.

These data are in agreement with our LysoSensor assay (Fig. 3E), which suggested that lysosome biogenesis is not impaired in IRE1 null macrophages.

While we cannot rule out a defect in mitochondrial calcium flux from our RNA-seq data, we have added discussion around this topic to our discussion, as mentioned above.

Expression of V-ATPase subunits and mitochondrial calcium uptake genes in C. albicans-infected IRE1 null macrophages vs C. albicans-infected IRE1 WT macrophages.

Fig. 1A should be explained with more detail to disclose the products of PstI digestion.

Thank you for the suggestion. We have added this information to the Figure 1 legend, "RT-PCR-amplified Xbp1 cDNA was treated with PstI, which recognizes a cleavage site within the 26 base pair intron that is removed by IRE1α activity, resulting in cleavage of the unspliced isoform, specifically."

The anti-XBP1 antibody used to construct the blots in Fig.S1A recognizes epitopes not disclosed by the manufacturers, but they have to pertain to the N-terminal peptide sequence shared by sXBP1 and uXBP1. Showing full lanes encompassing both protein isoforms would allow a better appraisal of protein expression. In connection to point 4, the use of an antibody reactive to the epitopes expressed in sXBP1 in cell lysates or, preferentially in nuclear fractions, could be most valuable to rule out the dependence of the effect of Ire1α on the trans-activating function of sXBP1.

We have un-cropped these westerns and now show spliced and unspliced XBP1 products on a single image in Fig. S1A.

On page 23, the mention to Fig. 5A should be changed to Fig. 5B.

We have fixed this mis-labeling, thank you for calling this to our attention.

Line 209. I understand gene synthesis refers to gene expression.

We have clarified this in the text, thank you for the suggestion.

Line 394. What is the reason to study the cytokine-signature of Candida in LPS-primed cells?

Thank you for the question; we have added the following text (lines 413-414) to clarify that LPS is used for inflammasome priming:

"Therefore, we tested secretion of IL-1β, TNF, and IL-6 from WT and IRE1ΔR macrophages after LPS treatment to transcriptionally prime the NLRP3 inflammasome components, followed by C. albicans infection (Fig. 5D-F)."

Numerous studies have shown that C. albicans can trigger macrophage pyroptosis, resulting in production of pro-inflammatory cytokines like IL-1b, which can also be influenced by phagosome rupture (PMC3910967). However, this requires inflammasome transcriptional priming, and LPS is commonly used to prime macrophages for inflammasome activation in vitro. Therefore, we perform a short pre-treatment with LPS for NLRP3 inflammasome priming to subsequently measure its activation following C. albicans infection, using secreted cytokines as a readout. We also note that macrophages in vivo may not be naive and are often M1-polarized by the microbial or cytokine environment, thus inflammasome priming is likely common during in vivo infection.

Reviewer #2 (Significance (Required)):

This study focuses on an aspect not usually addressed in papers devoted to the UPR.

If more data are shown as suggested, the paper could be of interest for a wider audience

We thank the reviewer for their positive feedback about the novelty of our work and agree that the suggested experiments will bolster our data and story.

---

Reviewer 3:

Fig. 2:

Panel A-B: same question as for Fig. 1. The variation in TG DMSO-induced splicing is huge. The effects of the treatments with CHX or Act D are smaller than the variation between experiments with TG DMSO alone. As long as that variation is not controlled for, it is impossible to draw any conclusion from the inhibitors. In this regard, it is very difficult to interpret data if they are not done in one and the same experiment.

The variability in thapsigargin fold change over mock likely represents differences in basal Xbp1 expression. We consistently see complete Xbp1 splicing in response to thapsigargin treatment (see Fig. 1A). Additionally, we note that thapsigargin treatment is used only as a positive control, not as a physiologically relevant treatment, as it results in unmitigated ER stress that triggers cell death (PMC6986015).

We have removed the following sentence, "Translation inhibition using cycloheximide was sufficient to alleviate Xbp1 splicing specifically in response to thapsigargin, likely by reducing the nascent protein folding burden (Fig. 2B)," since our data are plotted on separate graphs, matched to their respective controls, for appropriate comparisons.

4. Description of analyses that authors prefer not to carry out

---

Reviewer 1:

2) Since the IRE1/XBP1 arm of the UPR is also involved in lipid biosynthesis which might be required for phagosome maturation, the authors should perform XBP1s rescues in IRE1 deficient cells to ensure that their observation is XBP1s dependent or IRE1 dependent.

As we do not see XBP1S protein induced in wild-type macrophages at any timepoint during our C. albicans infection scheme (Fig. S1A-B), we interpret our results as being XBP1S-independent. If we were to add back XBP1S with constitutive expression, we would be overexpressing the protein relative to C. albicans infected wild-type macrophages (in which we do not see measurable XBP1S expression). Therefore, we believe these experiments would not address a physiologically-relevant scenario.

4) The authors should evaluate in what compartment IRE1 is activated upon CA infection, does that happen in the ER or in the ER fraction fused to phagosomes?

This is an interesting question for future exploration. In order to answer this question with existing tools, we would need to perform biochemical fractionation of infected cells to isolate an ER-phagosome contact site fraction, followed by phos-tag gel analysis of IRE1 activation in the ER fraction, compared to the ER-phagosome contact site fraction. However, a biochemical fractionation protocol to distinguish the ER fraction from ER-phagosome contact sites has not yet been developed, to our knowledge, and we believe it is outside the scope of this study to develop such a technique.

We have added additional text regarding this intriguing question to our discussion (lines 549-553).

Reviewer 2:

---

Infection at a MOI 1 of C. albicans is a ratio of infecting agent/susceptible targets not very high for a non-soluble stimulus with limited diffusion in the culture medium. Although I recognize the difficulty of quantitating adhered cell, the mention to 80% confluence makes it more difficult the appraisal of the actual MOI. The delayed time-course of Xbp splicing under these conditions can be explained by the time required for in vitro proliferation, Candida damage, and diffusion of fungal patterns. A study with viable Candida at MOI 5 in human monocyte-derived dendritic cells, which show a robust capacity for non-opsonic phagocytosis associated with C-type lectin receptors only showed initial hypha formation after 2 hours (Rodriguez et al., J. Biol. Chem. 289, P22942-22957, 2014). Consistent with the requirement of a time lag for infecting agent to attain levels of expression consistent with a net response, 16 hours have been considered an appropriate time-course to assay sXBP1 expression following SARS-CoV2 infection (Fernandez et al., Biochim Biophys Acta Mol Basis Dis. 1870(5):167193, 2024). I wonder if a higher MOI could show a similar kinetics.

We use lower MOI in part due to the size and ability of C. albicans to undergo extensive hyphal growth if its numbers greatly exceed the number of host cells. From our microscopy data, we can see that C. albicans spreads well throughout the culture plate (see Fig. 3A, Fig. 4A). We and others have observed considerable death of macrophage cultures after 12 hours with Candida infection, even at low MOI (PMC6709535), therefore we avoid later timepoints in these assays and all other in vitro assays in our manuscript.

As all of our in vitro experiments are performed within an 8 hour window of infection, whether XBP1S is induced at later timepoints by C. albicans or depleted zymosan would not alter the conclusions of the rest of our results.

sXBP1 can be present in nuclear fractions in resting cells, which suggests the involvement of post-translational modifications for the display of transcriptional activity.

As we do not see induction of XBP1S in our lysates after C. albicans infection, it is unlikely that post-translational modification is influencing its function, although we agree post-translational modification is a likely regulatory control over XBP1S during the unfolded protein response.

The independence of sXBP1 transcriptional activity from canonical UPR associated with misfolded protein stress is well known from the seminal paper by Martinon et al., (ref.6). Moreover, the expression of CHOP, the final effector of the PERK route, encoded by DDIT3 gene, has been found to be blunted by Candida (Rodriguez et al., J. Biol. Chem. 289, P22942-22957,2014). This is additional evidence for the recruitment of sXBP1 transcriptional activity in the absence of canonical UPR.

As mentioned, we found that XBP1S protein is not induced during C. albicans infection at any timepoint in our experiments (Fig. S1A-B). Importantly, the work referenced by the reviewer uses RAW267.7 cells, which (as mentioned by the authors) constitutively express CHOP as a result of Abel leukemia virus infection. Based on this specific overexpression, we believe this phenotype is not comparable to our bone marrow-derived macrophages.

Reviewer 3:

Fig. 1:

Panel 1C: please remove outlier in 4h timepoint. This implies that the experiment needs to be redone to reduce variation

We have performed an outlier test on these data, which revealed that this data point is not a statistical outlier, therefore we do not feel that its removal is appropriate (see below).

Panel 1E-H: how is the splicing efficiency determined and normalized? How to explain the big differences in splicing efficiency of Xbp1 upon LPS stimulation (appr. 4 to 6 times in E, G and H versus 30-fold in panel F). Where does this difference come from?

Panel H, outlier needs to be removed.

We do occasionally see differences in magnitude of Xbp1 splicing in different cell lines or experiments, especially with controls, which may be caused by differences in the basal level of Xbp1 expression, especially as Xbp1 levels have been shown to be affected by circadian rhythm in certain cell types (PMCID: PMC11214543; PMCID: PMC6959563).

In panel H, an outlier test reveals that these are not statistical outliers, therefore we feel their removal is inappropriate as we do not wish to mask biological variation. Moreover, this graph includes two cell lines (open and closed circles), showing that our data are robust across multiple independent cell lines and are an appropriate measure of experimental replicates.

Fig. 2:

Panel A-B: same question as for Fig. 1. The variation in TG DMSO-induced splicing is huge. The effects of the treatments with CHX or Act D are smaller than the variation between experiments with TG DMSO alone. As long as that variation is not controlled for, it is impossible to draw any conclusion from the inhibitors. In this regard, it is very difficult to interpret data if they are not done in one and the same experiment.

The variability in thapsigargin fold change over mock likely represents differences in basal Xbp1 expression. We consistently see complete Xbp1 splicing in response to thapsigargin treatment (see Fig. 1A). Additionally, we note that thapsigargin treatment is used only as a positive control, not as a physiologically relevant treatment, as it results in unmitigated ER stress that triggers cell death (PMC6986015).

We have removed the following sentence, "Translation inhibition using cycloheximide was sufficient to alleviate Xbp1 splicing specifically in response to thapsigargin, likely by reducing the nascent protein folding burden (Fig. 2B), since our data are plotted on separate graphs, matched to their respective controls, for appropriate comparisons.

Below, we plot all data together with replicate matching, although our major interpretation of these data is that C. albicans infection can trigger Xbp1 splicing with or without new gene expression, and not about the impact of the inhibitors on the control treatment thapsigargin.

Please provide a scheme of how the experiment was performed, at what time were the inhibitors provided, at what time point the inducers? What are matched mock samples. Which mock samples were chosen since they differ from one experiment to the next? Please plot all the data for one and the same experiment in one graph so that the reader can easily compare the results of DMSO, DMSO + inducer, DMSO + inducer + inhibitor. Indicate whether the points in the graph are technical or experimental repetitions.

-How to explain the increase in XBP1 splicing in combination with ActD? Was this due to differences in Gapdh expression? Where did the authors control for cell death? Please provide the data.

Below is a scheme of the experimental treatments. We have now clarified in the figure legend that inhibitors (ActD and CHX) are added at the same time as experimental treatments (Mock, Ca, TG). All data included in the original submission are biological replicates, as stated in the figure legend. We have now re-written the figure legend to clearly indicate that these are biological replicates.

All data are normalized such that the effects of the drugs are directly compared (for example, the fold change over Mock for Candida is matched to its drug treatment; Mock DMSO vs Ca DMSO and Mock ActD vs Ca ActD, or Mock CHX vs Ca CHX). Actinomycin D does inhibit new transcription, although IRE1 can cleave existing Xbp1 transcript. We now show conditions normalized to DMSO Mock in Supplemental Figure 2, which allows visualization of the effects of ActD and CHX on Xbp1-S abundance in comparison to control DMSO treatment, while also seeing the relative changes in Xbp1 splicing caused by C. albicans or thapsigargin treatment (see below).

-Is RT-qPCR a reliable readout when actinomycinD is used? How can new genes be transcribed.

We interpret RT-qPCR data as a readout of transcript abundance, rather than transcription. Therefore, we are not measuring new gene expression here, but whether the existing Xbp1 transcript can be cleaved by IRE1. Based on the technique, we can still measure changes in Xbp1-S abundance.

Panel D: where is TG at 4h and 6h?

We do not include thapsigargin at later timepoints to avoid autofluorescence from excessive cell death. We include thapsigargin as a positive control at the early 2h timepoint, but note that LPS is sufficient to increase thioflavin T intensity at the 8h timepoint.

Panel G, why was Ddit3 included here as this is not a typical IRE1 dependent gene (rather PERK dependent). What about IRE1 specific genes such as Sec61 or Sec24a?

We have added additional text (lines 235-240; "Finally, we measured induction of UPR-responsive genes by RT-qPCR in response to C. albicans infection, LPS and depleted zymosan treatment, or thapsigargin treatment, to further test whether IRE1α activation occurs without canonical UPR induction (Fig. 2G-H). C. albicans infection and depleted zymosan treatment did not lead to induction of UPR-responsive genes (Ddit3, Grp78, Grp94, and total Xbp1) at 4 or 6 hours.") to clarify that the purpose of this figure is to add evidence that IRE1 activation is independent of the canonical UPR response (indicating that IRE1 is likely specifically activated independently of the other UPR branches) during C. albicans infection. Therefore, the transcripts measured are canonical UPR-responsive transcripts, rather than IRE1/XBP1S targets (although some are overlapping).

Below are RNA-seq data comparing Sec61a1, Sec61a2, and Sec24a in IRE1 null macrophages, compared to IRE1 WT macrophages. While there is less expression of Sec61a1 in IRE1 null macrophages, Sec61a2 and Sec24a are largely unaffected. These data support our finding that XBP1S protein is not induced during C. albicans infection.

Did the authors also check for RIDD activity?

As mentioned above in response to Reviewer 1, we now add additional figures to our supplemental data (Fig. SX; also shown below) showing that established RIDD targets are not depleted during C. albicans infection in WT macrophages, and also not increased in IRE1 null macrophages. Therefore, we expect RIDD activity has negligible effects on our reported phenotypes.

Fig. 3:

Panel C and D look convincing. Lamp1 is a well-known RIDD target gene (see Osorio et al., Nat Imm, 2014). Did the authors check Lamp1 expression in presence and absence of IRE1 and could RIDD explain their phenotype?

As shown above, Lamp1 transcript expression is not strongly perturbed in IRE1 null macrophages. If RIDD activity were depleting Lamp1 transcript abundance, we would expect to see increased Lamp1 expression in IRE1 null macrophages. We also note that our experiments using LysoSensor (Fig. 3E) suggested that lysosome biogenesis is not impaired, but more specifically, lysosome recruitment to the phagosome is impaired in IRE1 null macrophages.

Fig. 4, but especially Fig 5 and Fig 6 suffer from very bad imaging quality. Both Fig 5A and Fig 6A are completely uninterpretable. The SRB staining is all over the cells and it is totally unclear how the authors interpret this as phagosomal leakage or not. Fig. 6A is even worse and appears nothing but vague background. It is difficult to understand how the authors make graphs based on these types of images and dare to draw any conclusions.

In Figure 4, we observe some photobleaching from frequent image acquisition, which is necessary to capture calcium flux dynamics. Image brightness across the timecourse is adjusted in the same way such that we do not attempt to hide the effects of photobleaching. However, our analyses account for photobleaching over time, and the phagosomal calcium flux is clear and quantifiable. `

In Figure 5, the sulforhodamine B pulse-chase assay involves loading of the endosomal system with SRB, thus the cells are expected to ingest a considerable amount of SRB and it will distribute throughout the endosomal network. However, as endosomes fuse, we also observe fusion with the C. albicans-containing phagosome and SRB will surround C. albicans hyphae. Our analysis pipeline first segments C. albicans hyphae (see below) and measures SRB signal in proximity to the phagosome. Thus, we measure loss of phagosome-associated SRB over time, as C. albicans ruptures the phagosome, in hundreds of macrophages. This is a standard assay that has been previously used for this purpose (PMID: 33022213; PMID: 30131363).

For Figure 6, we have added additional wide-field images that we believe will clarify how these images can be readily quantified (Fig. 6A, shown below). The purpose of the previous Fig. 6A (now Fig. 6B) is to demonstrate single cell examples of live and dead C. albicans using the dual-fluorescence assay, although we quantify much wider fields for sufficient numbers. We hope the amended figures provide additional clarity.

Fig. 7 is again an example where differences in expression are mainly due to one or a few complete outliers, and it is hard to understand why the authors did not repeat these experiments to reduce the problems in variation to get proper data sets before submission.

After performing outlier tests, we have found a total of 4 data points that are statistical outliers from all of the panels in Figure 7. These included the highest data point in each genotype in the female IL-1Ra levels (Fig. 7A, second graph), the highest data point among the male IRE1 fl/fl mice IL-1Ra levels (Fig. 7B, second graph), and the highest data point among the male TNF levels in IRE1 fl/fl + LysM-Cre mice (Fig. 7B, third graph). We have removed these data points in our updated graphs and changed the text to only point out differences in serum TNF and IL-6 levels. Moreover, our interpretation includes that serum cytokine levels are not different in male mice. However, no other data points are statistical outliers, therefore we believe their removal is inappropriate.

While the paper started nicely and showed an interesting hypothesis (Fig. 3), the remaining part of the paper was of very poor quality and was not ready for submission.

We thank the reviewer for the constructive feedback and believe that the addition of data and clarifications we have added will demonstrate that our data are of sufficient quality to support our conclusions.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #3

Evidence, reproducibility and clarity

Reviewer comments for "Non-canonical activation of IRE1a by Candida albicans infection promotes macrophage phagosomal calcium flux to enhance fungal killing"

This paper describes a role for IRE1 in controlling Candida albicans (Ca) infection in macrophages. The authors show that Ca infection slightly induces IRE1 activity as monitored by XBP1 splicing, however this does not result in XBP1 protein expression, nor in IRE1-dependent target gene expression. The authors propose that IRE1 controls phagosomal maturation, as measured by defects in LAMP1 recruitment to Ca containing phagosomes. This would be due to a defect in calcium flux at the phagosomal level leading to an increased propensity to rupture and cytosolic escape of the pathogen. While the data are interesting and the defect in LAMP1 recruitment to the phagosome convincing, the majority of the data are difficult to interpret due to the poor quality. This concerns specifically all imaging experiments but also the ELISAs and qPCRs where differences are due to the effect of outliers rather than to the behavior of a complete population. Therefore, most experiments need to be redone and complemented with additional approaches before any firm conclusions can be drawn. Specific details and examples are provided below.

Fig. 1:

Panel 1C: please remove outlier in 4h timepoint. This implies that the experiment needs to be redone to reduce variation

Panel 1E-H: how is the splicing efficiency determined and normalized? How to explain the big differences in splicing efficiency of Xbp1 upon LPS stimulation (appr. 4 to 6 times in E, G and H versus 30-fold in panel F). Where does this difference come from?

Panel H, outlier needs to be removed.

Fig. 2:

Panel A-B: same question as for Fig. 1. The variation in TG DMSO-induced splicing is huge. The effects of the treatments with CHX or Act D are smaller than the variation between experiments with TG DMSO alone. As long as that variation is not controlled for, it is impossible to draw any conclusion from the inhibitors. In this regard, it is very difficult to interpret data if they are not done in one and the same experiment. Please provide a scheme of how the experiment was performed, at what time were the inhibitors provided, at what time point the inducers? What are matched mock samples. Which mock samples were chosen since they differ from one experiment to the next? Please plot all the data for one and the same experiment in one graph so that the reader can easily compare the results of DMSO, DMSO + inducer, DMSO + inducer + inhibitor. Indicate whether the points in the graph are technical or experimental repetitions.

• How to explain the increase in XBP1 splicing in combination with ActD? Was this due to differences in Gapdh expression? Where did the authors control for cell death? Please provide the data.
• Is RT-qPCR a reliable readout when actinomycinD is used? How can new genes be transcribed.

Panel D: where is TG at 4h and 6h?

Panel G, why was Ddit3 included here as this is not a typical IRE1 dependent gene (rather PERK dependent). What about IRE1 specific genes such as Sec61 or Sec24a?

Did the authors also check for RIDD activity?

Fig. 3:

Panel C and D look convincing. Lamp1 is a well-known RIDD target gene (see Osorio et al., Nat Imm, 2014). Did the authors check Lamp1 expression in presence and absence of IRE1 and could RIDD explain their phenotype?

Fig. 4, but especially Fig 5 and Fig 6 suffer from very bad imaging quality. Both Fig 5A and Fig 6A are completely uninterpretable. The SRB staining is all over the cells and it is totally unclear how the authors interpret this as phagosomal leakage or not. Fig. 6A is even worse and appears nothing but vague background. It is difficult to understand how the authors make graphs based on these types of images and dare to draw any conclusions.

Fig. 7 is again an example where differences in expression are mainly due to one or a few complete outliers, and it is hard to understand why the authors did not repeat these experiments to reduce the problems in variation to get proper data sets before submission.

While the paper started nicely and showed an interesting hypothesis (Fig. 3), the remaining part of the paper was of very poor quality and was not ready for submission.

Significance

The study presents interesting hypothesis but unfortunately the data are not of sufficient quality

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

This an interesting report in a widely explored area. This makes it necessary to pigeonhole the new data provided by the study. The paper addresses two issues encompassing a scope distinct from studies focusing on the cytokine-signature and the role of sXbp1. This research singles out fungal killing and Ire1α versus sXbp1 function. More precisely, the reduction of cytokine expression in Ire1fl/fl LysMCre as compared to WT discloses an opposing function of Ire1α and its target sXbp1 in cytokine expression that requires mechanistic explanation.

1. A point that should be addressed with more detail is the correlation of fungal killing with Ca2+ fluxes and Ire1α activity, given the well-known data regarding the strong ability of the axis dectin/SYK/phospholipase Cγ to induce Ca2+ transients, a response not shared by LPS signaling, and the sequential activation of mitochondrial Ca2+ uniporter (MCU), which is a critical element of fungal killing associated with the citrate-pyruvate shuttle as a NADPH source (Seegren et al., Cell Rep. 33: 108411, 2020). Incidentally, this paper is referred in ref. 46 as a preprint, although it was accessible in Cell Reports in 2020.
2. The assay of the expression of V-ATPase complex, mitochondrial calcium uniporter, and mitochondrial uptake 1 and 2 could shed light on the dependence of fungal killing on Ire1α function.
3. Infection at a MOI 1 of C. albicans is a ratio of infecting agent/susceptible targets not very high for a non-soluble stimulus with limited diffusion in the culture medium. Although I recognize the difficulty of quantitating adhered cell, the mention to 80% confluence makes it more difficult the appraisal of the actual MOI. The delayed time-course of Xbp splicing under these conditions can be explained by the time required for in vitro proliferation, Candida damage, and diffusion of fungal patterns. A study with viable Candida at MOI 5 in human monocyte-derived dendritic cells, which show a robust capacity for non-opsonic phagocytosis associated with C-type lectin receptors only showed initial hypha formation after 2 hours (Rodriguez et al., J. Biol. Chem. 289, P22942-22957, 2014). Consistent with the requirement of a time lag for infecting agent to attain levels of expression consistent with a net response, 16 hours have been considered an appropriate time-course to assay sXBP1 expression following SARS-CoV2 infection (Fernandez et al., Biochim Biophys Acta Mol Basis Dis. 1870(5):167193, 2024). I wonder if a higher MOI could show a similar kinetics.
4. Fig. 1A should be explained with more detail to disclose the products of PstI digestion.
5. sXBP1 can be present in nuclear fractions in resting cells, which suggests the involvement of post-translational modifications for the display of transcriptional activity.
6. The independence of sXBP1 transcriptional activity from canonical UPR associated with misfolded protein stress is well known from the seminal paper by Martinon et al., (ref.6). Moreover, the expression of CHOP, the final effector of the PERK route, encoded by DDIT3 gene, has been found to be blunted by Candida (Rodriguez et al., J. Biol. Chem. 289, P22942-22957,2014). This is additional evidence for the recruitment of sXBP1 transcriptional activity in the absence of canonical UPR.
7. The anti-XBP1 antibody used to construct the blots in Fig.S1A recognizes epitopes not disclosed by the manufacturers, but they have to pertain to the N-terminal peptide sequence shared by sXBP1 and uXBP1. Showing full lanes encompassing both protein isoforms would allow a better appraisal of protein expression. In connection to point 4, the use of an antibody reactive to the epitopes expressed in sXBP1 in cell lysates or, preferentially in nuclear fractions, could be most valuable to rule out the dependence of the effect of Ire1α on the trans-activating function of sXBP1.
8. On page 23, the mention to Fig. 5A should be changed to Fig. 5B.
9. Line 209. I understand gene synthesis refers to gene expression.
10. Line 394. What is the reason to study the cytokine-signature of Candida in LPS-primed cells?

Significance

This study focuses on an aspect not usually addressed in papers devoted to the UPR. If more data are shown as suggested, the paper could be of interest for a wider audience

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

The authors show that in macrophages, IRE1 activation (independent of improperly folded proteins) is essential to promote fungicidal activity towards candida albicans (CA) in vitro and in vivo by ensuring phagosome maturation through the preservation of calcium fluxes

Major comments

1. The demonstration of protein misfolding independent IRE1 activation should also be demonstrated using molecules such as TUDCA or 4PBA that should be innocuous regarding the splicing of XBP1s. It would also be interesting to evaluate the activation of the other arms of the UPR in particular through the phosphorylation of eIF2a, expression of ATF4 and cleavage of ATF6.
2. Since the IRE1/XBP1 arm of the UPR is also involved in lipid biosynthesis which might be required for phagosome maturation, the authors should perform XBP1s rescues in IRE1 deficient cells to ensure that their observation is XBP1s dependent or IRE1 dependent.
3. The authors use thioflavin to evaluate the extend of protein misfolding. This type of stain can lead to artefactual results and in general it is rather safer to test several stainers (see for instance the work presented in PMC10720158)
4. The authors should evaluate in what compartment IRE1 is activated upon CA infection, does that happen in the ER or in the ER fraction fused to phagosomes?
5. The authors focus on the IRE1/XBP1s signaling arm of the UPR but do not explore RIDD activity which has been linked to several infection mechanisms and lysosomal integrity (in particular by regulating the expression of BLOS1 - see PMC9119680 and PMC6446841). The authors should definitely evaluate how RIDD is activated (or not) in their experimental systems.
6. The whole study relies on the use of IRE1deltaR to impair IRE1 signaling. The authors should validate their hypothesis with an orthogonal approach, for instance with IRE1 pharmacological inhibitors (eg MKC8866 or KIRA8).

Significance

The manuscript is interesting and highligths novel aspects towards the interaction between marcrophages and a pathogen, candida albicans, involving the likely selective activation of IRE1. The data are novel and experimentally sound. Several controls are however missing.

The strengths of the study are associated with the novelty of the findings, with the links that could potentially derive from this study to connect ER biology, UPR signaling and phagosome maturation

The main weaknesses are associated i) with the fact that the authors did not evaluate RIDD activity which has already been linked with pathogen infection and with lysosome integrity, ii) with methodological aspects, in particular regarding the demonstration of the IRE1 activation independent on protein misfolding and the sole use of a genetic variant of IRE1 to test their hypotheses

Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

Learn more at Review Commons

---

Reply to the reviewers

Manuscript number: RC-2024-02555

__Corresponding author(s): __Maurizio Molinari

[Please use this template only if the submitted manuscript should be considered by the affiliate journal as a full revision in response to the points raised by the reviewers.

• *

If you wish to submit a preliminary revision with a revision plan, please use our "Revision Plan" template. It is important to use the appropriate template to clearly inform the editors of your intentions.]

1. General Statements [optional]

We thank the 3 reviewers for the positive and constructive comments to our manuscript.

Please see below the point-by-point responses to their suggestions.

2. Point-by-point description of the revisions

Reviewer #1 (Evidence, reproducibility and clarity (Required)):

The paper proposes an interesting role for ERp44 in TMX5 retention. The authors identified a list of proposed TMX5 clients which include many Golgi localised proteins but do not discuss the role for TMX for instance in protein folding. In this context it is not absolutely clear whether TMX5 acts as a trafficking chaperone? are clients functionally engaged in the Golgi or ER or both?

This work focuses on the crosstalk between a member of the PDI superfamily lacking a conventional cytosolic ER retention motif (TMX5), and ERp44, a PDI family member previously reported to retrieve in the ER proteins that lack the ER retention motif (ERp44). To support our conclusions on the involvement of ERp44 in control of TMX5’s intracellular distribution, we have added new data obtained by characterization of a cell line lacking ERp44, where more than 50% of TMX5 escapes ER retention (new Fig. 6).

__We agree with the referee that the assessment of the biological role of TMX5 is of interest. We mention in this manuscript that there is a follow-up study (ongoing in the lab) on TMX5 clients and TMX5 function. More specifically, we are monitoring the action of TMX5 on the biogenesis and intracellular trafficking of class I HLA molecules, which are, besides PDI, ERp57 and ERp44, major interactors (clients) of TMX5 (please also refer to the initial and final parts of the new discussion). __

The defining criteria for the client proteins were not included. At last, it might be of interest to evaluate for how long TMX5 clients are retained on the protein, whether it is temporary (as for instance a folding sensor) or more permanent.

The list of interacting proteins is now available (__Data are available via ProteomeXchange with identifier PXD054716."), their selection for presentation in Figure 3B is now explained more clearly (Results, page 6). Also better explained is that we define as “clients of TMX5” those endogenous proteins that associate with TMX5, covalently, via the catalytic Cys220. The mutation of the TMX5 active site cysteine residue does not impact the covalent association of PDI, ERp57 and ERp44 with TMX5. For this reason, we do not consider these PDI family members clients of TMX5. In this submission, we explore the covalent association of ERp44 and its consequences on TMX5 subcellular distribution. Interacting via non-catalytic Cysteine residues 114 and 124 with the catalytic cysteine 29 of ERp44, we identify TMX5 as a client of the latter.__

The preparation of figures could be greatly improved and there is some inconsistency among similar gels.

---

Please refer to the point-by-point answers below.

---

The proposed model of ERp44, ER retention vs ER retrieval, is unclear. Overall, there is more room for improved discussion beyond the conclusions from experiments.

__We thank the referee for these comments. We have improved the description of the results, and we separated the Discussion (written de novo) from the Results section. __

The ERp44 interaction is interesting especially since the protein contains an incomplete thioredoxin domain (such as ERp29, PDIA17 and 18), would the interaction between Erp44 and TMX5 be involved in some holdase/competitor role thereby allowing for client selectivity (or kinetics)? In addition, all the experiments were carried out in Hek293T or MEF cells, would the authors anticipate some interactions of TMX5 with PDIA17/18 in cells where those proteins are highly expressed? Testing whether the observation is a general mechanism occurring between TMX5 and PDI family members with incomplete thioredoxin sites would be an asset.

__We thank the referee for this comment that we implemented in the new discussion. __

Major comments Fig 2 - avoid labels on the blots that might obscure information and impede clarity and interpretation. o The % of resistant protein can be otherwise placed.

---

__This has been modified, thank you. __

• What does the asterix in 2B signify? This should be included in the legend.

We have now specified in the legend of figures that asterisks show cross-reacting polypeptide bands.

• A label for 'deglycosylated' proteins could be included.

__We added a label for de- and for glycosylated proteins in the EndoH essays in Figs. 2A, 2B and 5B. __

• Consider treating with PNGase.

This is now showy in panel 2A, lane 5.

• There is a change in EndoH resistance of about 3-4% among wt, C220A & C114A, is this significant?

We do not consider significant these variations. Our data show that the mutation of TMX5 Cys 114 or of Cys124 to alanine substantially reduce (without abolishing) the co-IP with ERp44. This means that the proteins interact less, or that the interaction is more short living. The EndoH experiment shown in Fig. 2B and the CLSM analyses in Figs. 2D-2O fail to reveal significant differences showing that these reduced or more short living associations with ERp44 are sufficient to control TMX5 distribution.

In the previous submission, the function of ERp44 in retaining TMX5 in the ER was supported by data showing that the co-expression of ERp44 retains TMX5 in the ER, but co-expression of ERp44C29S that cannot bind TMX5 fails to retain TMX5 in the ER. These model is further supported in this new submission by the release of 50% of TMX5 from the ER in cells lacking ERp44, which is substantially inhibited to the levels measured in wild type cells upon back-transfection of ERp44, but not upon the co-transfection of the ERp44C29S mutant (new Figure 6).

---

• Equivalent inputs (cell lysates) for the IPs should be included.

---

__ These have now been added in Figs. 4, 5 6, and 7.__

Fig 3A - Indicate the specific bands that were subject to MS. How did the authors correct for non-specific interactors and false positives? Perhaps a more specifically targeted approach could be utilised.

• How do the authors explain the absence of bands representing the reduced form of interacting TMX5 interactors?

• What was the inclusion and exclusion criteria used to determine which of the proteins listed were clients?

The endogenous proteins present in the entire region of the gel labeled with the red and blue rectangles have been sequenced (see methods section and this is now also better explained in the results section, page 6). Only the proteins that disappear from the corresponding region of the gel when the samples have been reduced are listed in the table. This is also better explained in the text (page 6). These experiments have been repeated few times with a series of controls (e.g., mock-transfected cells and cells transfected with other members of the TMX family (shown to capture and to impact on the fate of other endogenous polypeptides in previous publications from our lab)). An in parallel analysis of mock, TMX3, TMX4 and TMX5 interactors has been published in (Kucinska et al Nature Comm 2023), where we focused on the biological function of TMX4. The references referring to the TMX1 study (Brambilla et al 2015) and the TMX4 study (Kucinska et al) are given in the text.

The Table in Fig. 3B only lists the interacting polypeptides that have a MW __- It might be useful to perform MS on the C220A mutant and compare those results to the WT.

__To validate few interactions with endogenous proteins detected in MS, and to compare the interactions of TMX5 and TMX5C220A, we have used the specifically targeted approach suggested above by the referee (i.e., co-IP validated by WB, Figs. 3C-3F).

__

Fig 4 - Equivalent inputs (cell lysates) for the IPs should be included.

__This is now shown as panel A in Fig. 4.____

__

Fig 5 - Equivalent inputs (cell lysates) for the IPs should be included.

This is now Figure 7, see new panel 7A

Fig 6 - A loading control should be included.

__This is now Fig. 5. Both panels A and B in Fig. 5 show total cell extracts____

__

• Blot using anti-HA to identify ERp44 should be included to substantiate claims.

The ERp44 and TMX5 components of the ERp44-s-s-TMX5 mixed disulfides are detected upon IP:HA followed by WB:V5 (to show the TMX5 component) and upon IP:V5 followed by WB:HA to show the ERp44 component) in Figs. 4B-4E and 7B-7C.

• How do the authors account for the huge difference in TMX5 associated complexes shown in Fig 6A compared to Fig 3A.

Fig. 6 is now Fig. 5. As specified in the legends of the figures, Fig. 3A shows a gel, where the complexes are stained with silver, Fig. 5A is a WB, where the complexes are stained with an antibody. The intensities of the signals cannot be compared.

• Inappropriate marking on the gel area.
• Inconsistencies in protein standard labeling

This has carefully been checked and corrected where needed. Please note that we used two different MW standards for our figures (200, 117, 97, 66, 45, 31 kDa and 270, 175, 130, 95, 66, 53, 37 kDa)

• It might be useful to demonstrate the colocalisation of ERp44 and ERp44C29S with Giantin and with TMX5 considering that ERp44 is known to cycle between the Golgi and ER.

These data are shown in Fig. 5D-5I.

Reviewer #1 (Significance (Required)):

This work provides an additional understanding on how the regulation of Erp44 trafficking might occur (and perhaps additional PDIs), and lead to the characterization of kinetic value that might explain better productive protein folding in the early secretory pathway. This represents a significant advance in the field and may in turn unveil uncharacterized pathophysiological functions in various diseases. This is a serious study well conducted and original by an expert in the field that desserves publication.

Field of expertise: ER homeostasis control

Reviewer #2 (Evidence, reproducibility and clarity (Required)):

The manuscript by Solda et al, investigates TMX5, a poorly understood member of the PDI family that lacks an ER-retrieval motif. They find that it localizes to the ER and the Golgi and that it interacts with ERp44. This interaction requires formation of a mixed disulfide and they identify the cysteine residues in both proteins that mediate this interaction. Overall, this is a well written manuscript that is easy to follow and the story is compact and straight-forward. It provides some new and solid insight into the biology of TMX5 without going into depth of what the cellular role of TMX5 is or might be. I have only very few comments and suggestions:

1- The authors conclude that ERp44 associates only with ER-localized TMX5. I am not sure that this is a valid conclusion based on the data. EndoH sensitivity just means that the protein has not gone to the medial Golgi. The pool of TMX5 could therefore be an ERGIC-based pool, or it could interact with a TMX5 that is recycled directly from the first Golgi cisterna, where complex glycosylation is unlikely to occur. Can this be validated using another type of experiment? Alternatively, the wording could be changed.

We thank the reviewer. We agree with this insightful comment that led us to change the wording used in some part of the text.

2- Is the trafficking of TMX5 dependent on its glycosylation?

This is another insightful comment that we report in the ____new discussion, where we write, page 14 “____It should be noted that in the case of TMX5 the extensive N-glycosylation could engage____ leguminous L-type lectins located in the ER (VIPL), cycling between the ER and the intermediate compartment (ERGIC) (ERGIC-53) or between the ERGIC and the cis-Golgi (VIP36)_33-36_ and have an impact on the subcellular distribution and activity of TMX5.____”

3- Figure 6: The data are not really convincing. Just because the color turns yellow, it does not mean that there is colocalization. The green channel is overexposed in this area of the cell, and anything will produce a yellow color, even if there is no genuine colocalization. Maybe the authors could provide a different example and even better would be a quantification of the colocalization.

__We thank the referee for this comment. We show images of better quality, where the black/white channels clearly show the co-localization (or lack thereof) of TMX5 with the Golgi marker Giantin in cells mock-transfected (co-localization TMX5:Giantin, Fig. 5D), co-transfected with ERp44 (no co-localization TMX5:Giantin, Fig. 5E), or co-transfected with ERp44C29S, co-localization TMX5:Giantin, Fig. 5F). Figs. 5G-5I show the corresponding results for the co-localization or lack thereof between TMX5C220A and Giantin. Importantly, the IF data match the data shown in Fig. 5B, where release from the ER (or arrival in the medial Golgi, see text of the manuscript and comment 1 by the referee) is assessed by monitoring complex glycosylation. __

Reviewer #2 (Significance (Required)):

This is a solid story that will be of interest of scientists working on various aspects of the secretory pathway and protein quality control. The advance is rather incremental, because there are no experiments that provide insight into the cellular roles of TMX5.

Reviewer #3 (Evidence, reproducibility and clarity (Required)):

Solda et al have assembled data on the transmembrane redox enzyme TMX5, on which currently very little information is available. TMX5 does not contain any obvious targeting signal, unlike the other TMX family proteins, which localize to the ER. TMX5 has 5 glycosylation sites, which can be used to determine its intracellular localization biochemically. Indeed, about 20% of TMX5 is found as endoH-resistant, indicating Golgi localization. This is confirmed with beautiful IF imagery and giantin co-localization. The Golgi localization requires two luminal cysteines (C212, 177), which likely form a disulfide bond. TMX5 acts as a natural cysteine-trapping protein, allowing for easy assessment of its interactors. Within its interactome, the authors found multiple members of the thioredoxin family. Many of these interactions occur within the CXXS motif, but notably ERp44 does not require this motif to interact, indicating this and other interactions are not of a catalytic nature. Instead, the authors found this interaction to be essential for ER retention or retrieval and depends on the cysteine within the ERp44 "active" site. The study provides critical first insight about the potential functions and sites of activity of TMX5.

Specific Points: 1. The results are very convincing and of high quality. 2. The cytosolic tail of TMX5 contains an LI motif, which could act as a post-ER localization signal. Since the protein might play a role in ciliogenesis, this motif could be critical. In this context, I am wondering which mutations are known to lead to the disease spectrum.

---

The position of disease-related TMX5 mutations identified so far are given in Xu H, et al (2024) Mol Genet Genomic Med 12: e2340 _https://www.ncbi.nlm.nih.gov/pubmed/38073519_ and in ____Deng T, Xie Y (2024) Mol Genet Genomic Med 12: e2343 https://www.ncbi.nlm.nih.gov/pubmed/38156946____.

They are all distributed in the luminal part of the protein____.

---

Mutation of C114 and C124 abrogates interaction with ERp44. Therefore, I would expect these mutations to increase endoH resistance and Golgi staining. This should be investigated by the authors.

__The mutations C114 and C124 reduce (or make short-living), without abrogating the covalent association between TMX5 and ERp44. The EndoH experiment shown in Fig. 2B and the IF in Figs. 2D-2O fail to reveal significant differences showing that these reduced or more short living associations with ERp44 are sufficient to control TMX5 distribution. To strengthen our conclusion that ERp44 is involved in regulation of the intracellular TMX5 distribution, we have now added data in ERp44 cell (50% of TMX5 displays complex glycans as symptom of traffic to the medial Golgi compartment), back-transfection of ERp44 (but not of the ERp44C29S mutant that does not associate with TMX5) restores the complex glycan fraction to the level measured in wild type cells (Fig. 6). __

Minor Points: 1. The position of the % endoH resistance in Figures 1B and 6B is not ideal, as it obstructs a visual inspection of TMX5 resistance to endoH.

This has been modified, thank you.

Reviewer #3 (Significance (Required)):

Given that no information about TMX5 is currently available, the study provides critical first insight that should allow researchers to tackle the disease relevance of TMX5 in the future.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #3

Evidence, reproducibility and clarity

Solda et al have assembled data on the transmembrane redox enzyme TMX5, on which currently very little information is available. TMX5 does not contain any obvious targeting signal, unlike the other TMX family proteins, which localize to the ER. TMX5 has 5 glycosylation sites, which can be used to determine its intracellular localization biochemically. Indeed, about 20% of TMX5 is found as endoH-resistant, indicating Golgi localization. This is confirmed with beautiful IF imagery and giantin co-localization. The Golgi localization requires two luminal cysteines (C212, 177), which likely form a disulfide bond. TMX5 acts as a natural cysteine-trapping protein, allowing for easy assessment of its interactors. Within its interactome, the authors found multiple members of the thioredoxin family. Many of these interactions occur within the CXXS motif, but notably ERp44 does not require this motif to interact, indicating this and other interactions are not of a catalytic nature. Instead, the authors found this interaction to be essential for ER retention or retrieval and depends on the cysteine within the ERp44 "active" site. The study provides critical first insight about the potential functions and sites of activity of TMX5.

Specific Points:

1. The results are very convincing and of high quality.
2. The cytosolic tail of TMX5 contains an LI motif, which could act as a post-ER localization signal. Since the protein might play a role in ciliogenesis, this motif could be critical. In this context, I am wondering which mutations are known to lead to the disease spectrum.
3. Mutation of C114 and C124 abrogates interaction with ERp44. Therefore, I would expect these mutations to increase endoH resistance and Golgi staining. This should be investigated by the authors.

Minor Points:

1. The position of the % endoH resistance in Figures 1B and 6B is not ideal, as it obstructs a visual inspection of TMX5 resistance to endoH.

Significance

Given that no information about TMX5 is currently available, the study provides critical first insight that should allow researchers to tackle the disease relevance of TMX5 in the future.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

The manuscript by Solda et al, investigates TMX5, a poorly understood member of the PDI family that lacks an ER-retrieval motif. They find that it localizes to the ER and the Golgi and that it interacts with ERp44. This interaction requires formation of a mixed disulfide and they identify the cysteine residues in both proteins that mediate this interaction.

Overall, this is a well written manuscript that is easy to follow and the story is compact and straight-forward. It provides some new and solid insight into the biology of TMX5 without going into depth of what the cellular role of TMX5 is or might be. I have only very few comments and suggestions:

1. The authors conclude that ERp44 associates only with ER-localized TMX5. I am not sure that this is a valid conclusion based on the data. EndoH sensitivity just means that the protein has not gone to the medial Golgi. The pool of TMX5 could therefore be an ERGIC-based pool, or it could interact with a TMX5 that is recycled directly from the first Golgi cisterna, where complex glycosylation is unlikely to occur. Can this be validated using another type of experiment? Alternatively, the wording could be changed.
2. Is the trafficking of TMX5 dependent on its glycosylation?
3. Figure 6: The data are not really convincing. Just because the color turns yellow, it does not mean that there is colocalization. The green channel is overexposed in this area of the cell, and anything will produce a yellow color, even if there is no genuine colocalization. Maybe the authors could provide a different example and even better would be a quantification of the colocalization.

Significance

This is a solid story that will be of interest of scientists working on various aspects of the secretory pathway and protein quality control. The advance is rather incremental, because thereare no experiments that provide insight into the cellular roles of TMX5.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

The paper proposes an interesting role for ERp44 in TMX5 retention. The authors identified a list of proposed TMX5 clients which include many Golgi localised proteins but do not discuss the role for TMX for instance in protein folding. In this context it is not absolutely clear whether TMX5 acts as a trafficking chaperone? are clients functionally engaged in the Golgi or ER or both? The defining criteria for the client proteins were not included. At last, it might be of interest to evaluate for how long TMX5 clients are retained on the protein, whether it is temporary (as for instance a folding sensor) or more permanent.

The preparation of figures could be greatly improved and there is some inconsistency among similar gels. The proposed model of ERp44, ER retention vs ER retrieval, is unclear. Overall, there is more room for improved discussion beyond the conclusions from experiments.

The ERp44 interaction is interesting especially since the protein contains an incomplete thioredoxin domain (such as ERp29, PDIA17 and 18), would the interaction between Erp44 and TMX5 be involved in some holdase/competitor role thereby allowing for client selectivity (or kinetics)? In addition, all the experiments were carried out in Hek293T or MEF cells, would the authors anticipate some interactions of TMX5 with PDIA17/18 in cells where those proteins are highly expressed? Testing whether the observation is a general mechanism occurring between TMX5 and PDI family members with incomplete thioredoxin sites would be an asset.

Major comments

Fig 2

• avoid labels on the blots that might obscure information and impede clarity and interpretation.
• The % of resistant protein can be otherwise placed.
• What does the asterix in 2B signify? This should be included in the legend.
• A label for 'deglycosylated' proteins could be included.
• Consider treating with PNGase to .....
• There is a change in EndoH resistance of about 3-4% among wt, C220A & C114A, is this significant?
• Equivalent inputs (cell lysates) for the IPs should be included.

Fig 3A

• Indicate the specific bands that were subject to MS. How did the authors correct for non-specific interactors and false positives? Perhaps a more specifically targeted approach could be utilised.
• How do the authors explain the absence of bands representing the reduced form of interacting TMX5 interactors?
• What was the inclusion and exclusion criteria used to determine which of the proteins listed were clients?
• It might be useful to perform MS on the C220A mutant and compare those results to the WT.

Fig 4

• Equivalent inputs (cell lysates) for the IPs should be included.

Fig 5

• Equivalent inputs (cell lysates) for the IPs should be included.

Fig 6

• A loading control should be included.
• Blot using anti-HA to identify ERp44 should be included to substantiate claims.
• How do the authors account for the huge difference in TMX5 associated complexes shown in Fig 6A compared to Fig 3A.
• Inappropriate marking on the gel area.
• Inconsistencies in protein standard labeling
• It might be useful to demonstrate the colocalisation of ERp44 and ERp44C29S with Giantin and with TMX5 considering that ERp44 is known to cycle between the Golgi and ER.

Significance

This work provides an additional understanding on how the regulation of Erp44 trafficking might occur (and perhaps additional PDIs), and lead to the characterization of kinetic value that might explain better productive protein folding in the early secretory pathway. This represents a significant advance in the field and may in turn unveil uncharacterized pathophysiological functions in various diseases.

This is a serious study well conducted and original by an expert in the field that desserves publication.

Field of expertise: ER homeostasis control

Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

Learn more at Review Commons

---

Reply to the reviewers

Reviewer #1 (Evidence, reproducibility and clarity (Required)): Summary: To explore the relationship between histone post-translational modifications (H3K4me3 and H3K27me3) and enhancer activation with gene expression during early embryonic development, the authors used a monolayer differentiation approach to convert mouse embryonic stem cells (ESCs) into Anterior Definitive Endoderm (ADE). They monitored differentiation stages using a dual reporter mESC line (B6), which has fluorescent reporters inserted at the Gsc (GFP) and Hhex (Redstar) loci. Their analyses indicate that the differentiating cells advanced through stages similar to those in the embryo, successfully converting into endoderm and ADE with high efficiency. This is elegant and well performed stem cell biology.

Their subsequent genome-wide and nascent transcription analyses confirmed that the in vitro gene expression changes correlated with developmental stages and confirmed that transcriptional activation precedes mRNA accumulation. They then focussed on linking active enhancers and histone modifications (H3K4me3 and H3K27me3) were with gene expression dynamics. Finally, the performed PRC2 inhibition and showed that, while it enhanced differentiation efficiency, it also induced ectopic expression of non-lineage specific genes.

Major comments: In terms of mechanistic advances, they propose that transcriptional up-regulation does not require prior loss of H3K27me3, which they show appears to lag behind gene activation, but critically, on a likely mixed population level. I am sceptical of their interpretation of their data because they are looking at heterogenous populations of cells. To explain, one could imagine a particular H3K27me3 coated gene that gets activated during differentiation. In a population of differentiating cells, while the major sub-population of cells could retain H3K27me3 on this particular gene when it is repressed, a minority sub-population of cells could have no H3K27me3 on the gene when it is actively transcribed. The ChIP and RNA-seq results in this mixed cell scenario would give the wrong impression that the gene is active while retaining H3K27me3, when in reality, it's much more likely that the gene is never expressed when its locus in enriched with the repressive H3K27me3 modification. Therefore, to support their claim, they would have to show that a particular gene is active when its locus is coated with H3K27me3. Personally, I don't feel this approach would be worth pursuing.

They also report that inhibition of PRC2 using EZH2 inhibitor (EPZ6438) enhanced endoderm differentiation efficiency but led to ectopic expression of pluripotency and non-lineage genes. However, this is not surprising considering the established role of Polycomb proteins as repressors of lineage genes.

Reviewer #1 (Significance (Required)): I feel that this is a solid and well conducted study in which the authors model early development in vitro. It should be of interest to researchers with an interest in more sophisticated in vitro differentiation systems, perhaps to knockout their gene of interest and study the consequences. However, I don't see any major mechanistic advances in this work.

*>Author Response *

*We agree with the point regarding the delayed loss of H3K27me3 relative to gene activation, and indeed this same point has been raised by reviewer 3 (see below). Our cell-population based data does not allow us to directly test if gene up-regulation in a small population of cells from TSSs lacking H3K27me3, accounts for the observed result. Furthermore, there are currently no robust methods to determine cell- or allele-specific expression simultaneously with ChIP/Cut and Run for chromatin marks. However, we provide the following additional evidence that strongly supports our conclusions. *

• *

Our FACs isolation strategy used to prepare cell populations for ChIP, microarray expression and 4sU-seq analysis is based on expression (or lack thereof) of a fluorescent GSC-GFP reporter. This means that every cell in the G+ populations express the Gsc fluorescent reporter, at least at the protein level, at the point of isolation. This is despite the presence of appreciable and invariant levels of H3K27me3 at the TSS of the Gsc gene in both G+ and G- populations at day 3 of differentiation. Comparable to our meta-analysis of all upregulated genes shown in the original manuscript (Figure 5 and S5), H3K27me3 levels are then subsequently reduced in the G+ relative to the G- populations at day 4. The transcriptional changes which correspond to the GSG-GFP reporter expression and associated ChIP-seq data are shown in the reviewer figure (Fig R1 A shown in revision plan). To further support our observations, we sought to rule out the possibility that the shift in H3K27me3 and transcription were from mutually exclusive gene sets, from nominal transcription levels or from sites with low level H3K27me3. To do this with a gene set of sufficient size to yield a robust result, we selected upregulated TSSs that had a greater than median value for both transcription (4sU-seq) and H3K27me3 (n=49 of 159 genes; Fig R1 B shown in revision plan). Meta-analysis of these genes showed that, as for all upregulated gene TSS (n=159), transcriptional activation occurred in the presence of substantial and invariant levels of H3K27me3 at day 3 followed by a subsequent reduction by day 4 of differentiation (Fig R1 C shown in revision plan). Importantly, many of these genes yielded high absolute 4sU-seq signal, comparable to that of Gsc, arguing against transcriptional activation being limited to a small subpopulation of cells.

• *

Reviewer #2 (Evidence, reproducibility and clarity (Required)): In this paper the authors profile gene expression, including active transcription, and histone modifications (k4 and k27me3) during a complex differentiation protocol from ES cells, which takes advantage of FACS sorting of appropriate fluorescent reporters. The data is of good quality and the experiments are well performed. The main conclusion, that the analyzed histone marks channel differentiation more than they directly allow/block it, is well supported by the data. The paper is interesting and will represent a good addition to an already extensive literature. I have however a few major concerns, described below:

1/ K4me3 may show more changes than they interpret, at least over the +1 nucl. An alternative quantification to aggregate profiles should be used to more directly address the questions regarding the correlations between histone mods and gene expression.

*>Author Response *

*Whilst we state that H3K4me3 levels are somewhat invariant at differentially expressed genes relative to H3K27me3, quantification of individual TSS (+/- 500 bp) did show a direct correlation with gene expression (Figure 5 and S5). To further explore this in response to the reviewer’s comment we will quantify K4me3 signal at the +1 nucleosome to determine if this yields more substantial differences than that observed more broadly across TSSs. *

2/ Related to the previous point, it appears clear in Fig.4 that the promoters of each gene expression cluster do not belong to a single chromatin configuration. I think it would be important to: 1/ cluster the genes based on promoter histone mods and interrogate gene expression and cluster allocation (basically the reverse to what is presented) 2/ order the genes in the heatmaps identically for K4me3 and K27me3 to more easily understand the respective chromatin composition per cluster

>Author Response

We thank the reviewer for these suggestions and will include these analyses in a revised manuscript.

3/ Also, as it is apparent that not all promoters in every cluster are enriched for the studied marks, could the authors separately analyze these genes? What are they? Do they use alternative promoters?

>Author Response

*Indeed, this is the case. Whilst there is significant enrichment of H3K27me3 at the TSS of developmentally regulated genes, not all genes whose expression changes during the differentiation will be polycomb targets. We will further stratify these clusters as suggested and determine what distinguishes the subsets. If informative, this data will be included in a revised manuscript. *

4/ The use of 4SU-seq to identify active enhancers is welcome; however, I have doubts it is working very efficiently: for instance, in the snapshots shown in Fig.2A, the very active Oct4 enhancers in ES cells are not apparent at all... More validation of the efficiency of the approach seems required.

>Author Response

The 4sU-seq data shown in Figure 2A was generated in samples isolated from day 3 and 4 of the ADE differentiation. It is therefore likely that the enhancers have been partly or wholly decommissioned at this point. Indeed, in a separate study we generated 4sU-seq data using the same protocol and conditions as presented here but in ES cells and differentiated NPCs (day 3 to 7) and indeed see transcription at Oct 4 enhancers in ESCs (arrowed in the screenshot shown in revision plan) which are extinguished upon differentiation to neural progenitor cells (NPCs); data from PMID: 31494034).

5/ The effects of the EZH2 inhibitor are quite minor regarding the efficiency of the differentiation as analyzed by FACS, despite significant gene expression changes. To the knowledge of this referee, this is at odds with results obtained with Ezh2 ko ES cells that display defects in mesoderm and endoderm differentiation. I have issues reconciling these results (uncited PMID: 19026780). Either the authors perform more robust assays (inducible KOs) or they more directly explain the limitations of the study and the controversies with published work.

>Author Response

We agree that this result appears to be at odds with the findings in (PMID: 19026780*). This is likely due to the fact that we are acutely reducing H3K27me3 levels for a short period either during or immediately preceding the differentiation rather than removing PRC2 function genetically. This, likely provides a less pronounced defect on the ability to generate endodermal cells. However, we cannot address this without further experimentation which is beyond the scope of this study. We will more fully discuss the results in the context of this and other studies and discuss the limitations of the study in this regard. *

Minor 1/ please add variance captured to PCA plots 2/ Fig1E add color scales to all heatmaps 3/ Fig4C,D are almost impossible to follow, please find a way to identify better the clusters/samples and make easier to correlate all the variables

• *

>Author Response

*We will address all of these points in a revised manuscript. *

Reviewer #2 (Significance (Required)):

The paper is incremental in knowledge, and not by a big margin, as it is known already that histone mods rather channel than drive differentiation. Though, the authors do not clearly address inconsistencies with published work, especially regarding Ezh2 thought to be important to make endoderm. It is however a good addition to current knowledge, provided a better discussion of differences with published work is provided.

>Author Response

*As outlined above, we will address this with a more complete discussion about the distinction between the studies and what can and can’t be concluded from our approach. *

* *

• *

Reviewer #3 (Evidence, reproducibility and clarity (Required)): This study investigates the role of chromatin-based regulation during cell fate specification. The authors use an ESC model of differentiation into anterior primitive streak and subsequently definitive endoderm, which they traced via a dual-reporter system that combines GSC-GFP and HHEX-RedStar. The authors mapped changes in (nascent) gene expression and histone modifications (H3K4me3/H3K27me3) at key timepoints and within different populations over six days of differentiation. Finally, the authors test the functional implications of H3K27me3 landscapes via PRC2 inhibition.

The majority of data chart the descriptive changes in (epi)genomic and transcriptional dynamics coincident with cell differentiation. The use of nascent transcriptomics improves the temporal resolution of expression dynamics, and is an important strategy. By and large the data reinforce established paradigms. For example, that transcription is the dominant mechanism regulating mRNA levels, or that dynamic chromatin states changes occur and largely corelate of gene activity. They also identify putative enhancers with profiling data, albeit these are not validated, and confirm that PRC2 inhibition impacts cell fate processes - in this case promoting endodermal differentiation efficiency. Overall, the study is relatively well-performed and clearly written, with the omics profiling adding more datasets from in vitro cell types that can be difficult to characterise in vivo. Whilst the majority of the study may be considered incremental, the key finding is the authors conclusion that H3K27me3 is subordinate to gene activity rather than an instructive repressor. If borne out, this would mark an important observation with broad implications. However, in my view this conclusion is subject to many confounders and alternative interpretations, and the authors have not ruled out other explanations. Given the centrality of this to the novelty of the study, I would encourage further analysis/stratification of existing data, and potentially further experiments to provide more confidence in this key conclusion.

Primary issue 1.) The authors show that at the earliest timepoint (d3), nascent gene activation of a handful of genes between G+ and G- populations is not associated with a FC loss of H3K27me3. From this the authors extrapolate their key conclusion that H3K27me3 is subordinate. Causality of chromatin modifications in gene regulation is critical to decipher, and therefore this is an important observation to confirm. Below I go through the possible confounders and issues with the conclusion at this point.

(i) Single-cell penetrance. A possible (likely?) possibility is that gene activation initially occurs in a relatively small subset of cells at d3. Because these genes are expressed lowly prior to this, they will register as a significant upregulation in bulk analysis. However, in this scenario H3K27me3 would only be lost from a small fraction of cells, which would not be detectable against a backdrop of most cells retaining the mark. In short, the authors have not ruled out heterogeneity driving the effect. Given the different dynamic range of mRNA and chromatin marks, and that a small gain from nothing (RNA) is easier to detect than a small loss from a pre-marked state (chromatin), investigating this further is critical to draw the conclusions the authors have.

(ii) Initial H3K27me3 levels. The plots in Fig 5 show the intersect FC of H3K27me3 and gene expression. Genes that activate at d3 show no loss of H3K27me3. However, it is important to characterise (and quantitate) whether these genes are significantly marked by H3K27me3 in the first place, which I could not find in the manuscript. Many/several of the genes may not be polycomb marked or may have low levels to begin with. This would obviously confound the analysis, since an absence/low K27 cannot be significantly lost and is unlikely to be functional. Thus, the DEG geneset should be further stratified into H3K27me3+ and K27me3- promoter groups/bins, with significance and conclusions based on the former only (e.g. boxplot in 5F).

(iii) Sample size. The conclusions are based on a relatively small number of genes that upregulate between G+ and G- (n=55 in figure by my count, text mentions n=52). Irrespective of the other confounders above, this is quite a small subset to make the sweeping general conclusion that "loss of the repressive polycomb mark H3K27me3 is delayed relative to transcriptional activation" in the abstract. Indeed, the small number of DEG suggests the cell types being compared are similar and perhaps therefore have specific genomic features (this could be looked at) that drive .

>Author Response

*These are very good points and are also raised by reviewer 1 (see above). We have one example where we can definitively interrogate single cell protein expression, in our current data. Gsc (as monitored by GSC-GFP FACS and the bulk RNA analysis) meets the criteria of being robustly upregulated in all FACs sorted cells in the presence of high levels of H3K27me3 in the D3G+ population. We believe that the additional analysis (Figure R1A shown in revision plan) and the discussion above addresses the reviewer’s concerns about both the levels of expression and magnitude of H3K27me3. With respect to the third point, the numbers are low (although here I present data from the 4SU analysis with approximately three times more data points) however, the point here is not too say this happens in every instance of gene activation but more that it can happen and not just at a small subset of outlier genes. This is important, as the reviewer notes, in our understanding of how polycomb repression is relieved during development. We will also look to see if there are sequence characteristics/ motifs of these genes. In a revised manuscript we would include this data and further analysis as outlined above. The reviewer points out that the numbers vary a little between analyses. This arises due to the annotation of multiple TSSs per genes in some cases. This will be rectified throughout and made clearer in the legends. *

Other comments: 2.) The authors show that promoter H3K4me3 corelates well with gene expression dynamics in their model. They conclude that "transcription itself is required for H3K4me3 deposition", or in other words is subordinate. This may well be the case but from their correlative data this cannot be inferred. Indeed, several recent and past papers have shown that H3K4me3 itself can directly modulate transcription, for example by triggering RNA II pause-release, by preventing epigenetic silencing and/or by recruiting the PIC. The authors could point out or discuss these alternative possibilities to provide a more balanced discourse.

>Author Response

We agree and this will be discussed more thoroughly and both possibilities put forward in the revised manuscript.

3.) The labelling of some figures is unclear. In Fig 4C and 4D (right) it is impossible to tell what sample each of the lines represents. It is also not clear what the blue zone corresponds to in genome view plots (the whole gene?). Moreover, the replicate numbers are not shown in figure legends.

---

>Author Response

*We agree that the data presented in 4C and D is unclear. We will, as a minimum, collapse profiles into like populations (ESC / G- / G+ / G+H- / G+H+) which makes sense given the similarity of these populations across all analyses (see e.g. PCA analysis in Figure 1). We will also explore alternative ways of presenting the data to better highlight the dynamics and incorporate this with the changes suggested by reviewer 2. The blue shaded area represents the full extent of the key gene being discussed in the screen shot, this is mentioned in the legend but will be made clearer in a revised manuscript. Replication will also be added to the legend throughout (n=2 for ChIP-seq and n=3 for 4sU-seq). *

4.) It would be nice to provide more discussion to reconcile the conclusions that H3K27me3 in endoderm differentiation is subordinate and the final figure showing inhibiting H3K27me3 has a significant effect on differentiation, since the latter is the functional assessment.

>Author Response

*We will build on the points already made that suggests that whilst K27me3 is a passive repressor that serves to act against sub-threshold activating cues, it is nonetheless a critical regulator of developmental fidelity. *

Reviewer #3 (Significance (Required)): Overall, the study's strengths are in that it characterises epigenomic dynamics within a specific and relevant cell fate model. The nascent transcriptomics adds important resolution, and underpins the core conclusions. The weakness is that data is over-interpreted at this point, and other possibilities are not adequately tested. The conclusions should therefore either be scaled back (which reduces novelty) or further analysis and/or experiments should be performed to support the conclusion. If it proves correct, this would be a significant observation for the community,

>Author Response

In a revised manuscript, we will address the reviewer’s concerns with additional data and discussion as indicated above.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #3

Evidence, reproducibility and clarity

This study investigates the role of chromatin-based regulation during cell fate specification. The authors use an ESC model of differentiation into anterior primitive streak and subsequently definitive endoderm, which they traced via a dual-reporter system that combines GSC-GFP and HHEX-RedStar. The authors mapped changes in (nascent) gene expression and histone modifications (H3K4me3/H3K27me3) at key timepoints and within different populations over six days of differentiation. Finally, the authors test the functional implications of H3K27me3 landscapes via PRC2 inhibition.

The majority of data chart the descriptive changes in (epi)genomic and transcriptional dynamics coincident with cell differentiation. The use of nascent transcriptomics improves the temporal resolution of expression dynamics, and is an important strategy. By and large the data reinforce established paradigms. For example, that transcription is the dominant mechanism regulating mRNA levels, or that dynamic chromatin states changes occur and largely corelate of gene activity. They also identify putative enhancers with profiling data, albeit these are not validated, and confirm that PRC2 inhibition impacts cell fate processes - in this case promoting endodermal differentiation efficiency. Overall, the study is relatively well-performed and clearly written, with the omics profiling adding more datasets from in vitro cell types that can be difficult to characterise in vivo. Whilst the majority of the study may be considered incremental, the key finding is the authors conclusion that H3K27me3 is subordinate to gene activity rather than an instructive repressor. If borne out, this would mark an important observation with broad implications. However, in my view this conclusion is subject to many confounders and alternative interpretations, and the authors have not ruled out other explanations. Given the centrality of this to the novelty of the study, I would encourage further analysis/stratification of existing data, and potentially further experiments to provide more confidence in this key conclusion.

Primary issue

1.) The authors show that at the earliest timepoint (d3), nascent gene activation of a handful of genes between G+ and G- populations is not associated with a FC loss of H3K27me3. From this the authors extrapolate their key conclusion that H3K27me3 is subordinate. Causality of chromatin modifications in gene regulation is critical to decipher, and therefore this is an important observation to confirm. Below I go through the possible confounders and issues with the conclusion at this point.

(i) Single-cell penetrance. A possible (likely?) possibility is that gene activation initially occurs in a relatively small subset of cells at d3. Because these genes are expressed lowly prior to this, they will register as a significant upregulation in bulk analysis. However, in this scenario H3K27me3 would only be lost from a small fraction of cells, which would not be detectable against a backdrop of most cells retaining the mark. In short, the authors have not ruled out heterogeneity driving the effect. Given the different dynamic range of mRNA and chromatin marks, and that a small gain from nothing (RNA) is easier to detect than a small loss from a pre-marked state (chromatin), investigating this further is critical to draw the conclusions the authors have.

(ii) Initial H3K27me3 levels. The plots in Fig 5 show the intersect FC of H3K27me3 and gene expression. Genes that activate at d3 show no loss of H3K27me3. However, it is important to characterise (and quantitate) whether these genes are significantly marked by H3K27me3 in the first place, which I could not find in the manuscript. Many/several of the genes may not be polycomb marked or may have low levels to begin with. This would obviously confound the analysis, since an absence/low K27 cannot be significantly lost and is unlikely to be functional. Thus, the DEG geneset should be further stratified into H3K27me3+ and K27me3- promoter groups/bins, with significance and conclusions based on the former only (e.g. boxplot in 5F).

(iii) Sample size. The conclusions are based on a relatively small number of genes that upregulate between G+ and G- (n=55 in figure by my count, text mentions n=52). Irrespective of the other confounders above, this is quite a small subset to make the sweeping general conclusion that "loss of the repressive polycomb mark H3K27me3 is delayed relative to transcriptional activation" in the abstract. Indeed, the small number of DEG suggests the cell types being compared are similar and perhaps therefore have specific genomic features (this could be looked at) that drive .

Other comments:

2.) The authors show that promoter H3K4me3 corelates well with gene expression dynamics in their model. They conclude that "transcription itself is required for H3K4me3 deposition", or in other words is subordinate. This may well be the case but from their correlative data this cannot be inferred. Indeed, several recent and past papers have shown that H3K4me3 itself can directly modulate transcription, for example by triggering RNA II pause-release, by preventing epigenetic silencing and/or by recruiting the PIC. The authors could point out or discuss these alternative possibilities to provide a more balanced discourse.

3.) The labelling of some figures is unclear. In Fig 4C and 4D (right) it is impossible to tell what sample each of the lines represents. It is also not clear what the blue zone corresponds to in genome view plots (the whole gene?). Moreover, the replicate numbers are not shown in figure legends.

4.) It would be nice to provide more discussion to reconcile the conclusions that H3K27me3 in endoderm differentiation is subordinate and the final figure showing inhibiting H3K27me3 has a significant effect on differentiation, since the latter is the functional assessment.

Significance

Overall, the study's strengths are in that it characterises epigenomic dynamics within a specific and relevant cell fate model. The nascent transcriptomics adds important resolution, and underpins the core conclusions. The weakness is that data is over-interpreted at this point, and other possibilities are not adequately tested. The conclusions should therefore either be scaled back (which reduces novelty) or further analysis and/or experiments should be performed to support the conclusion. If it proves correct, this would be a significant observation for the community,

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

In this paper the authors profile gene expression, including active transcription, and histone modifications (k4 and k27me3) during a complex differentiation protocol from ES cells, which takes advantage of FACS sorting of appropriate fluorescent reporters. The data is of good quality and the experiments are well performed. The main conclusion, that the analyzed histone marks channel differentiation more than they directly allow/block it, is well supported by the data. The paper is interesting and will represent a good addition to an already extensive literature. I have however a few major concerns, described below:

1. K4me3 may show more changes than they interpret, at least over the +1 nucl. An alternative quantification to aggregate profiles should be used to more directly address the questions regarding the correlations between histone mods and gene expression.
2. Related to the previous point, it appears clear in Fig.4 that the promoters of each gene expression cluster do not belong to a single chromatin configuration. I think it would be important to:
   • cluster the genes based on promoter histone mods and interrogate gene expression and cluster allocation (basically the reverse to what is presented)
   • order the genes in the heatmaps identically for K4me3 and K27me3 to more easily understand the respective chromatin composition per cluster
3. Also, as it is apparent that not all promoters in every cluster are enriched for the studied marks, could the authors separately analyze these genes? What are they? Do they use alternative promoters?
4. The use of 4SU-seq to identify active enhancers is welcome; however, I have doubts it is working very efficiently: for instance, in the snapshots shown in Fig.2A, the very active Oct4 enhancers in ES cells are not apparent at all... More validation of the efficiency of the approach seems required.
5. The effects of the EZH2 inhibitor are quite minor regarding the efficiency of the differentiation as analyzed by FACS, despite significant gene expression changes. To the knowledge of this referee, this is at odds with results obtained with Ezh2 ko ES cells that display defects in mesoderm and endoderm differentiation. I have issues reconciling these results (uncited PMID: 19026780). Either the authors perform more robust assays (inducible KOs) or they more directly explain the limitations of the study and the controversies with published work.

Minor

1. please add variance captured to PCA plots
2. Fig1E add color scales to all heatmaps
3. Fig4C,D are almost impossible to follow, please find a way to identify better the clusters/samples and make easier to correlate all the variables

Significance

The paper is incremental in knowledge, and not by a big margin, as it is known already that histone mods rather channel than drive differentiation. Though, the authors do not clearly address inconsistencies with published work, especially regarding Ezh2 thought to be important to make endoderm. It is however a good addition to current knowledge, provided a better discussion of differences with published work is provided.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

Summary:

To explore the relationship between histone post-translational modifications (H3K4me3 and H3K27me3) and enhancer activation with gene expression during early embryonic development, the authors used a monolayer differentiation approach to convert mouse embryonic stem cells (ESCs) into Anterior Definitive Endoderm (ADE). They monitored differentiation stages using a dual reporter mESC line (B6), which has fluorescent reporters inserted at the Gsc (GFP) and Hhex (Redstar) loci. Their analyses indicate that the differentiating cells advanced through stages similar to those in the embryo, successfully converting into endoderm and ADE with high efficiency. This is elegant and well performed stem cell biology.

Their subsequent genome-wide and nascent transcription analyses confirmed that the in vitro gene expression changes correlated with developmental stages and confirmed that transcriptional activation precedes mRNA accumulation. They then focussed on linking active enhancers and histone modifications (H3K4me3 and H3K27me3) were with gene expression dynamics. Finally, the performed PRC2 inhibition and showed that, while it enhanced differentiation efficiency, it also induced ectopic expression of non-lineage specific genes.

Major comments:

In terms of mechanistic advances, they propose that transcriptional up-regulation does not require prior loss of H3K27me3, which they show appears to lag behind gene activation, but critically, on a likely mixed population level. I am sceptical of their interpretation of their data because they are looking at heterogenous populations of cells. To explain, one could imagine a particular H3K27me3 coated gene that gets activated during differentiation. In a population of differentiating cells, while the major sub-population of cells could retain H3K27me3 on this particular gene when it is repressed, a minority sub-population of cells could have no H3K27me3 on the gene when it is actively transcribed. The ChIP and RNA-seq results in this mixed cell scenario would give the wrong impression that the gene is active while retaining H3K27me3, when in reality, it's much more likely that the gene is never expressed when its locus in enriched with the repressive H3K27me3 modification. Therefore, to support their claim, they would have to show that a particular gene is active when its locus is coated with H3K27me3. Personally, I don't feel this approach would be worth pursuing.

They also report that inhibition of PRC2 using EZH2 inhibitor (EPZ6438) enhanced endoderm differentiation efficiency but led to ectopic expression of pluripotency and non-lineage genes. However, this is not surprising considering the established role of Polycomb proteins as repressors of lineage genes.

Referee cross-commenting

I see that Reviewer #3 has the same concern with over interpretation of data in places - most notably their (in my view not supported) suggestion that transcriptional up-regulation does not require prior loss of H3K27me3.

Significance

I feel that this is a solid and well conducted study in which the authors model early development in vitro. It should be of interest to researchers with an interest in more sophisticated in vitro differentiation systems, perhaps to knockout their gene of interest and study the consequences. However, I don't see any major mechanistic advances in this work.

Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

Learn more at Review Commons

---

Reply to the reviewers

We thank the reviewers for taking the time to read and comprehensively evaluate our manuscript. We are pleased that, overall, they recognize the quality of our data and that it supports our conclusions. We are grateful for their comments, insights and advice and have revised the manuscript accordingly as described in the point-by-point response below. We believe that the revised manuscript is substantially improved by some experimental additions, additional replicates, improved analysis and increased clarity. Some key enhancements are as follows:

Previously we had found increased expression of the WNT pathway following CHRDL2 treatment, using RNA seq. We have now demonstrated this experimentally using the cellular levels and localisation of β-catenin. Previously we had shown that overexpression of CHRDL2 increased resistance to common chemotherapy treatments, as well as irradiation in colorectal cell lines. We have now shown that cells surviving treatment show a further reduction SMAD1/5/8 phosphorylation indicating a selection for CHRLD2 high cells during the treatment. We have also demonstrated a decrease in chemotherapy sensitivity in intestinal organoids treated with secreted forms of CHRDL2.

1. Point-by-point description of the revisions

This section is mandatory. *Please insert a point-by-point reply describing the revisions that were already carried out and included in the transferred manuscript. *

Reviewer #1

__Evidence, reproducibility and clarity __

Clarkson and Lewis present data suggesting that overexpression of Chordin like 2 (CHRDL2) can affect colorectal cancer cell responses to chemotherapy agents, possibly by modulating stem-cell like pathways. I have the following comments:

1. Fig. 1J-it is standard to show the images of cell migration-this is important here, given the modest effect of CHRDL2 overexpression here.

We have now included 3 replicate control and CHRDL2 overexpressing cell images in this figure panel to support the quantification in the graph.

Fig. 2A-the very small error bars for most of the data on the curves suggests these are n=1 experiments with multiple technical replicates to generate the error bars. Please clarify. The legend says n=3 with ANOVA analysis but no significance detected. Please clarify.

All experiments in this figure were done with 5 technical replicates per experiment, this was replicated at least three times to give n=3 biological replicates. The error bars represent the standard error of the mean of these 3 biological replicates as stated in the legend. Some data points showed very little data variation, hence the small error bars. Raw data is available if requested.

1. Fig. 2B-given the overlapping error bars here, how can there be a pWe have removed this representation of the data as it combined many different experiments with variable cell types and chemotherapeutics and it was difficult to carry out meaningful statistics. An overview of the data can be better seen in table form as shown in the revised figure 2B.

Fig. 2G-did the authors try to estimate the concentration of CHRDL2 in the conditioned medium? Which cell line was used to generate this CM?

Conditioned media was harvested from the matching transgenic cell lines with inducible CHRDL2. eg RKO cells were treated with media collected from doxycycline induced transgenic RKO cells whereas CaCO2 cells were treated with media from CaCO2 cells. The concentration of doxycycline was represented by ++ for 10ug/ml, the same notation we have used for directly induced cells treated with 10ug/ml dox.

We did not try to quantify the absolute concentration of CHRDL2 but we have shown the relative amount on a Western blot normalised with a ponceau stain (quantification now included in supplementary figure 1).

We have clarified our description of this experiment, inserting the following statement, "Conditioned media was harvested from corresponding cell lines with the inducible CHRDL2 transgene and the parental control cells. Induction of CHRDL2 to generate conditioned media was carried out using the same concentration and duration of doxycycline treatment as the cells in figure 2A. "

Fig. 5-what is the potential mechanism for gene expression changes in response to CHRDL2 overexpression? Is it all due to BMP inhibition? More mechanistic detail would be welcome here.

We have suggested other pathways involved in these functional effects based on our RNA seq data but at the moment it is not possible to say whether any changes are independent of BMP signaling. CHRDL2 is relatively understudied and as yet there is not much literature supporting BMP independent actions of CHRDL2. However, we have added some discussion and reference to an article suggesting interactions between CHRLD2 and YAP (Wang et al., 2022) including the following statement on page 17: "While the changes in BMP and WNT signaling shown in our GSEA analysis suggest that the effects of CHRDL2 in our system work directly through inhibition of BMP, it is not possible to rule out that some pathways are affected by BMP independent actions of CHRLD2. Indeed, Wang et al, suggest that CHRDL2 can directly alter phosphorylation and activity of YAP in gastric cancer cell lines, which merits further exploration (Wang et al., 2022)"

Significance

Unclear whether genetically engineered inducible overexpression has any real physiological significance but we all use cell models so this is OK.

Reviewer #2

__Evidence, reproducibility and clarity __

Summary: In the manuscript entitled "BMP antagonist CHRDL2 enhances the cancer stem-cell phenotype and increases chemotherapy resistance in Colorectal Cancer" the authors demonstrated that Chordin-like 2 (CHDRL2), a secreted BMP antagonist, promotes a chemo-resistant colorectal cancer stem cell phenotype through the inhibition of BMP signaling. The authors took advantage of both 2D engineered colorectal cancer (CRC) cells and healthy murine 3D organoid systems. Specifically, the authors showed a decreased proliferation rate and reduced clonogenic capability upon overexpression of CHRDL2 in established human colon cancer cell lines. Subsequently, they identified a chemo-resistant phenotype upon standard therapies (5FU, Oxaliplatin and Irinotecan) in CHDRL2 overexpressing cells by performing MTS assay. The authors showed that this chemo-resistant phenotype is associated with ATM and RAD21 activation, supporting an induction of DNA damage signaling pathway. Of note, the authors assessed that the exposure of 3D murine organoid to CHRDL2 resulted in a stem-like phenotype induction accompanied by a reduction of the differentiated counterpart. From RNA-seq data analysis emerged the upregulation of genes associated to stemness and DNA repair pathways in CHRDL2 overexpressing cells.

Major comments: 1. In the first paragraph of the result section authors assessed that "Colorectal adenocarcinoma cell lines were deliberately chosen to encompass a range of CHRDL2 expression levels and genetic mutations", without showing qRT-PCR or WB data on the differential expression levels of CHRDL2 in a panel of immortalized CRC cell lines. Authors should include these data to better support their choice.

*We have now included some qRT-PCR in supplementary figure 1 alongside a table of some of the key driver mutations in each cell line. Western blotting of these cells shows only a very low concentration of CHRDL2 protein. As shown in figure 1B in the control columns, no significant protein expression is observed in any line. *

In Figure 1F, authors described a reduction of cell proliferation in CRC cell lines expressing high levels of CHRDL2 only under low glucose conditions. Why did the authors perform the assay under these conditions? They should better argue this aspect and validated the role of CHRDL2 in metabolism rewiring by performing additional in vitro assays.

We have removed this aspect of the paper as it does not add significantly to our overall conclusions and we can clearly see the effects of CHRDL2 overexpression under standard growth conditions (Figure 1G).

The authors should evaluate the role of CHRDL2 in promoting a stem-like phenotype in human colon cancer stem cells freshly isolated from patients and characterized.

We would very much like to do experiments such as this but it is beyond the scope of this study and will be included in upcoming grant proposals.

In order to confirm the data obtained on 3D murine organoids system obtained from normal Intestinal Stem Cells, authors should investigate the stemness induction, driven by CHRDL2, also in human intestinal organoids.

Experiments using human intestinal organoids are currently planned and ethical approval applications and grant proposals are underway for future experiments of this nature.

The authors should evaluate the oncogenic role of CHRDL2, through the maintenance of stemness, by performing orthotopic or subcutaneous experiments in vivo model.

Similarly, this is not possible for this manuscript but is planned for the future alongside a transgenic mouse model of inducible CHRDL2 overexpression in the intestine.

BMPs proteins are part of a very broad protein family. In the introduction section, authors should indicate the specific BMP protein on which CHRDL2 exerts its inhibitory function. Moreover, they should have assessed BMP protein levels in CACO2, LS180, COLO320 and RKO cell lines.

We have clarified the interactions between CHRDL2 and specific BMPs in the introduction. We have not specifically assessed the BMP protein levels in our cells however we have now included an analysis of expression data from the Cancer Cell Line Encyclopedia in supplementary figure 1 C.

In first panel, the authors should quantify the secreted levels of CHRDL2 in the media of overexpressing CHRDL2 cell lines.

We have done this using the ponceau staining as a loading control and the results are displayed (supplementary figure 1).

In Figure 2D the authors should use the appropriate controls and describe this with more details in results section.

In this figure we have used Hoechst staining followed by FACs analysis to identify the cell cycle profile of our CHRDL2 treated cells. We have improved the description of this in the methods section. Appropriate controls for staining, both negative and positive, are used when setting up the analysis for this experiment. The cell cycle profile is calculated using the Novocyte in house software. We have now included the histogram plots in the main figure to clarify these data in figure 2D.

In Figure 3A, the authors should have performed the assay by choosing IC50.

*We attempted these experiments with the IC50 levels, however the high amount of cell death and frequency of apoptotic cells meant that clear images were difficult to obtain. We therefore reduced the concentrations and still had very measurable effects. *

In Supplementary Fig. 4A-B. the results are unclear. The control cell lines are already chemo resistant.

Again, we used IC25 levels of the drugs so that our cells were damaged but still live throughout the experiment. This has been explained on page 10.

The authors should review and add statistical analysis in both main and supplementary figures.

*We have now added additional details about statistical analysis throughout the figures, legend and main text, showing all significance levels as well as non-significance for each data set. * Minor comments: 1. The quality of immunofluorescence and WB images should be implemented, and in the immunofluorescence panels scale bars should be added.

We have added or improved scale bars on each immunofluorescence image. Western blot images have been improved.

In the graphical abstract authors reported that CHRDL2 overexpression increase WNT and EMT pathways, without performing any specific assay to demonstrate this. Authors should correct and graphically improve the graphical abstract.

*This is a good point and we have now carried out Beta-catenin immunofluorescence as a measure of WNT signaling on both our cancer cell lines - showing an increase in nuclear beta-catenin (figure 1J and K), and our organoids - showing an increase in overall levels and cytoplasmic staining (Figure 4 F). In terms of EMT markers we have carried out immunofluorescence on IQGAP1 (Figure 1I). IQGAP1 is significantly upregulated in CHRDL2 cells, reflecting its role in reduced cell adhesion and increased migration. This correlates with our data showing increased cellular migration as well as the increase in EMT related transcription in our RNAseq data. *

The term "significantly" in the discussion section is inappropriately referred to data showed in the histogram in Figure 1J. Moreover, in Figure 1Jthe authors should delete from the y-axis the term "corrected".

We have changed significantly to substantially

The term "significant" in discussion is inappropriately referred to BMI1 expression level if compared to the histogram in Figure 4G.

We have changed significantly to "a trend to increase"

In Figure 2C the authors should add the unit of measurement (fold over control) in the table.

We have done this

In Figure 4E the authors should add the figure legend reporting OLFM4 protein.

We have done this

The authors should include few sentences summarizing the findings at the end of each paragraph.

We have added short summaries at the start or end of each section to improve the flow of the results section.

Significance

General assessment: Overall, the work is aimed to elucidate the role of CHRDL2 already considered a poor prognosis biomarker involved in the promotion of CRC (PMID: 28009989), in promoting stem-like properties. The authors elucidated new additional insights into the molecular mechanisms regulating stemness phenotype induced by the BMP antagonist CHRDL2 in CRC. The authors include in the study a large amount of data, which only partially support their hypothesis. However, this manuscript lacks organization and coherence, making it challenging to follow and read. Numerous concerns need to be addressed, along with some sentences to rephrase in the result and discussion sections.

Advance: The manuscript reported some functional insights on the role of CHRDL2 in colorectal cancer, but additional data should be added to support authors 'conclusions.

Audience: The manuscript is suggested for basic research scientists.

__Reviewer #3 (Evidence, reproducibility and clarity (Required)):____ __ Summary BMP antagonist CHRDL2 enhances the cancer stem-cell phenotype and increases chemotherapy resistance in Colorectal Cancer Eloise Clarkson et al. The manuscript explored the function of CHRDL2, a BMP antagonist, on colorectal cancer (CRC). The authors found that CHRDL2 overexpression can enhance the survival of CRC cells during chemotherapy and irradiation treatment with elevated levels of stem-cell markers and reduced differentiation. Further RNA-seq analysis revealed that CHRDL2 increased the expression of stem-cell markers, WNT signaling and other well-established cancer-associated pathways. Overall, the manuscript is well-written and presented. I have some suggestions:

Major points:

1. The authors assert BMP antagonism was demonstrated by assessing levels of phosphorylated SMAD1/5 (Figure 1G). However, the immunoblotting assay only depicted P-SMAD 1/5 levels and B-ACTIN as internal control. It's suggested to include total-SMAD1/5 immunoblotting as an internal control to further support the claim of BMP antagonism.

The reviewer is correct that this is the best control. Western blotting has now been performed with total SMAD1 protein expression used as an internal control and this is shown in Figure 1D and Supplementary figure 1F

The authors argue that CHRDL2 overexpression reduced the proliferation of CRC cell lines, as evidenced by cell proliferation assays. However, from Figure 1E, the reduction in proliferation appears insignificant. It would be beneficial to perform one-way ANOVA tests on each time point for CHRDL2+ and CHRDL2++ with Control in Figure 1E to ascertain significance.

*We now have repeated this experiment to reduce variability and have also provided two-way ANOVA analysis between Control and CHRDL2+ and Control and CHRDL2++. One-way ANOVA at timepoint 96hr also provided with details in the figure legend. *

The findings indicating that overexpressing CHRDL2 can confer resistance to chemotherapy in CRC cells (Figure 2A-C) are noteworthy. To deepen the understanding of BMP signaling in cancer stemness and the molecular underpinning of CHRDL2 antagonism, additional western blot assays on P-SMAD1/5 with CHRDL2 overexpression and drug treatment are recommended.

*Western blotting of P-SMAD1/5 upon cells treated with IC50 5FU has now been performed in figure 2C (in the same experiment as the revised panels in figure 1D). The data suggest that CHRDL2 overexpressing cells able to survive chemotherapy have higher levels of P-SMAD1/5 reduction compared to that of untreated cells, strongly suggesting that chemotherapy treatment acts to select the cells with the highest CHRDL2 expression. We thank reviewer 3 for this suggested experiment and have included further discussion on this on page 8. *

The assertion that extrinsic CHRDL2 addition diminishes differentiation and enhances stem-cell numbers in an intestinal organoid model is intriguing. As BMP signaling inhibition contributes to intestinal cell stemness, incorporating additional layers for BMP antagonism of CHRDL2 on intestinal organoids through immunoblotting or real-time quantitative PCR for treated organoids would augment the conclusions.

As stated in the response to reviewer 2, we have investigated Beta-catenin in our organoids following CHRDL2 treatment using immunofluorescence and find that the levels are increased with the staining shifting from the membrane to the cytoplasm and nucleus (Figure 4F).

The authors claim CHRDL2 overexpression can decrease BMP signaling based on GSEA analysis (Figure 5E). However, the GSEA results did not demonstrate the downregulation of BMP signaling. Reanalysis of this GSEA analysis is warranted.

*We agree with this point and have changed the description of this result since the gene set covers both positive and negative regulators of the BMP pathway. We cannot conclusively say from this RNAseq data set that BMP signaling is "downregulated", however since SMAD phosphorylation is increased and nuclear beta-catenin is increased, overall we suggest that the changes we see are likely to represent the effects of decreased BMP signaling along with increased WNT signaling. *

Minor Points:

6.Provide the threshold/cutoff values chosen for differential expressed genes (DEGs) in CHRDL2+ and CHRDL2++ RNA-seq compared with control cells. Explain the minimal overlap between CHRDL2 LOW and CHRDL2 HIGH DEGs. Consider presenting all DEGs in CHRDL2 LOW and CHRDL2 HIGH compared with control cells in one gene expression heatmap for better visualization.

We have now provided the cutoff values for the DEGs in the legend for figure 5 (PThe minimal overlap of DEGs in the low and high expressing cells is an interesting point. We hypothesize that this may be related to the different effects of intermediate vs high levels of WNT signaling that occurs in colon cancer cells, frequently discussed in the literature as the "Just right hypothesis" (Lamlum et al. 1999, Albuquerque et al., 2002, Lewis et al., 2010). However, we haven't included this in the discussion as it merits further exploration. However, we have mainly focused on specific genes that are modified in both data sets, which are more likely to be the direct result of CHRDL2 modification. *

After DEGs analysis, perform Gene Ontology (GO) analysis on these DEGs to further investigate possible gene functions rather than selectively discussing some genes, enhancing understanding of CHRDL2 functions in CRC cells.

We have carried out this analysis using a variety of tools and have now included a Gene Ontology Panther analysis as supplementary figure 7. We have included a comment on this in the text on page 14 saying "Gene ontology analysis supports these findings with enrichment in biological processes such as cellular adhesion, apoptosis and differentiation. "

Conduct similar experiments in both 2D culture and organoid systems, if feasible, to provide more comprehensive insights into CHRDL2's oncogenic roles in CRC tumor progression.

*We have now performed chemotherapy treatment on our organoid systems, and have found that organoids with extrinsic CHRDL2 addition have a higher survival rate after chemotherapy compared to a control (Figure 4H and I). *

Label significance (*, **, ***, and n.s.) for every CRC cell line treated with CHRDL2 in Figure 2D, 2F, 2J, 4G, 5D, and 5F.

We have done this

Label the antibodies with different colors used for immunofluorescence in the figure text in Figure 4E.

We have done this

* * Include replicate dots for the Control group in the bar plots in Figure 1F and 2B.

We have done this

* * Add scale bars in Figure 3A and correct similar issues in other figures if applicable.

We have done this

* *13.Correct grammar and punctuation mistakes throughout the manuscript. For example:

We have done this and further proofread our revised manuscript

Page 7: "As seen in Figure 1J, CHRDL2 overexpression significantly increased the number of migrated cells (P *We have now added additional details about statistical analysis throughout the figures, legend and main text, showing all significance levels as well as non-significance for each data set. * Reviewer #3 (Significance (Required)):

The current study presents compelling evidence demonstrating that BMP signaling antagonist CHRDL2 enhances colon stem cell survival in colorectal cancer cell lines and organoid models. Further validation through CRC mouse models could offer invaluable insights into the clinical relevance and therapeutic implications of CHRDL2 in colorectal cancer.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #3

Evidence, reproducibility and clarity

Summary

BMP antagonist CHRDL2 enhances the cancer stem-cell phenotype and increases chemotherapy resistance in Colorectal Cancer Eloise Clarkson et al. The manuscript explored the function of CHRDL2, a BMP antagonist, on colorectal cancer (CRC). The authors found that CHRDL2 overexpression can enhance the survival of CRC cells during chemotherapy and irradiation treatment with elevated levels of stem-cell markers and reduced differentiation. Further RNA-seq analysis revealed that CHRDL2 increased the expression of stem-cell markers, WNT signaling and other well-established cancer-associated pathways. Overall, the manuscript is well-written and presented. I have some suggestions:

Major points:

1. The authors assert BMP antagonism was demonstrated by assessing levels of phosphorylated SMAD1/5 (Figure 1G). However, the immunoblotting assay only depicted P-SMAD 1/5 levels and B-ACTIN as internal control. It's suggested to include total-SMAD1/5 immunoblotting as an internal control to further support the claim of BMP antagonism.
2. The authors argue that CHRDL2 overexpression reduced the proliferation of CRC cell lines, as evidenced by cell proliferation assays. However, from Figure 1E, the reduction in proliferation appears insignificant. It would be beneficial to perform one-way ANOVA tests on each time point for CHRDL2+ and CHRDL2++ with Control in Figure 1E to ascertain significance.
3. The findings indicating that overexpressing CHRDL2 can confer resistance to chemotherapy in CRC cells (Figure 2A-C) are noteworthy. To deepen the understanding of BMP signaling in cancer stemness and the molecular underpinning of CHRDL2 antagonism, additional western blot assays on P-SMAD1/5 with CHRDL2 overexpression and drug treatment are recommended.
4. The assertion that extrinsic CHRDL2 addition diminishes differentiation and enhances stem-cell numbers in an intestinal organoid model is intriguing. As BMP signaling inhibition contributes to intestinal cell stemness, incorporating additional layers for BMP antagonism of CHRDL2 on intestinal organoids through immunoblotting or real-time quantitative PCR for treated organoids would augment the conclusions.
5. The authors claim CHRDL2 overexpression can decrease BMP signaling based on GSEA analysis (Figure 5E). However, the GSEA results did not demonstrate the downregulation of BMP signaling. Reanalysis of this GSEA analysis is warranted.

Minor Points:

6.Provide the threshold/cutoff values chosen for differential expressed genes (DEGs) in CHRDL2+ and CHRDL2++ RNA-seq compared with control cells. Explain the minimal overlap between CHRDL2 LOW and CHRDL2 HIGH DEGs. Consider presenting all DEGs in CHRDL2 LOW and CHRDL2 HIGH compared with control cells in one gene expression heatmap for better visualization. 7. After DEGs analysis, perform Gene Ontology (GO) analysis on these DEGs to further investigate possible gene functions rather than selectively discussing some genes, enhancing understanding of CHRDL2 functions in CRC cells. 8. Conduct similar experiments in both 2D culture and organoid systems, if feasible, to provide more comprehensive insights into CHRDL2's oncogenic roles in CRC tumor progression. 9. Label significance (, , **, and n.s.) for every CRC cell line treated with CHRDL2 in Figure 2D, 2F, 2J, 4G, 5D, and 5F. 10. Label the antibodies with different colors used for immunofluorescence in the figure text in Figure 4E. 11. Include replicate dots for the Control group in the bar plots in Figure 1F and 2B. 12. Add scale bars in Figure 3A and correct similar issues in other figures if applicable. 13.Correct grammar and punctuation mistakes throughout the manuscript. For example:

Page 7: "As seen in Figure 1J, CHRDL2 overexpression significantly increased the number of migrated cells (P

Page 7: "As seen in Figure 1J, CHRDL2 overexpression significantly increased the number of migrated cells (P

Page 7: "As seen in Figure 1J, CHRDL2 overexpression significantly increased the number of migrated cells (P < 0.0449)," suggesting increased migratory ability, a hallmark of cancer stem cells."

Page 8: "CHRDL2 overexpression resulted in an approximate twofold increase in IC50 values compared to control cells (P < 0.001)."

Page 10: "As seen in Figure 4B, upon the" should be corrected to "Figure 4B."

1. Specify the statistical methods or estimates used for determining statistical significance.

Significance

The current study presents compelling evidence demonstrating that BMP signaling antagonist CHRDL2 enhances colon stem cell survival in colorectal cancer cell lines and organoid models. Further validation through CRC mouse models could offer invaluable insights into the clinical relevance and therapeutic implications of CHRDL2 in colorectal cancer.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

Summary:

In the manuscript entitled "BMP antagonist CHRDL2 enhances the cancer stem-cell phenotype and increases chemotherapy resistance in Colorectal Cancer" the authors demonstrated that Chordin-like 2 (CHDRL2), a secreted BMP antagonist, promotes a chemo-resistant colorectal cancer stem cell phenotype through the inhibition of BMP signaling. The authors took advantage of both 2D engineered colorectal cancer (CRC) cells and healthy murine 3D organoid systems. Specifically, the authors showed a decreased proliferation rate and reduced clonogenic capability upon overexpression of CHRDL2 in established human colon cancer cell lines. Subsequently, they identified a chemo-resistant phenotype upon standard therapies (5FU, Oxaliplatin and Irinotecan) in CHDRL2 overexpressing cells by performing MTS assay. The authors showed that this chemo-resistant phenotype is associated with ATM and RAD21 activation, supporting an induction of DNA damage signaling pathway. Of note, the authors assessed that the exposure of 3D murine organoid to CHRDL2 resulted in a stem-like phenotype induction accompanied by a reduction of the differentiated counterpart. From RNA-seq data analysis emerged the upregulation of genes associated to stemness and DNA repair pathways in CHRDL2 overexpressing cells.

Major comments:

1. In the first paragraph of the result section authors assessed that "Colorectal adenocarcinoma cell lines were deliberately chosen to encompass a range of CHRDL2 expression levels and genetic mutations", without showing qRT-PCR or WB data on the differential expression levels of CHRDL2 in a panel of immortalized CRC cell lines. Authors should include these data to better support their choice.
2. In Figure 1F, authors described a reduction of cell proliferation in CRC cell lines expressing high levels of CHRDL2 only under low glucose conditions. Why did the authors perform the assay under these conditions? They should better argue this aspect and validated the role of CHRDL2 in metabolism rewiring by performing additional in vitro assays.
3. The authors should evaluate the role of CHRDL2 in promoting a stem-like phenotype in human colon cancer stem cells freshly isolated from patients and characterized.
4. In order to confirm the data obtained on 3D murine organoids system obtained from normal Intestinal Stem Cells, authors should investigate the stemness induction, driven by CHRDL2, also in human intestinal organoids.
5. The authors should evaluate the oncogenic role of CHRDL2, through the maintenance of stemness, by performing orthotopic or subcutaneous experiments in vivo model.
6. BMPs proteins are part of a very broad protein family. In the introduction section, authors should indicate the specific BMP protein on which CHRDL2 exerts its inhibitory function. Moreover, they should have assessed BMP protein levels in CACO2, LS180, COLO320 and RKO cell lines.
7. In first panel, the authors should quantify the secreted levels of CHRDL2 in the media of overexpressing CHRDL2 cell lines.
8. In Figure 2D the authors should use the appropriate controls and describe this with more details in results section.
9. In Figure 3A, the authors should have performed the assay by choosing IC50.
10. In Supplementary Fig. 4A-B. the results are unclear. The control cell lines are already chemoresistant.
11. The authors should review and add statistical analysis in both main and supplementary figures.

Minor comments:

1. The quality of immunofluorescence and WB images should be implemented, and in the immunofluorescence panels scale bars should be added.
2. In the graphical abstract authors reported that CHRDL2 overexpression increase WNT and EMT pathways, without performing any specific assay to demonstrate this. Authors should correct and graphically improve the graphical abstract.
3. The term "significantly" in the discussion section is inappropriately referred to data showed in the histogram in Figure 1J. Moreover, in Figure 1Jthe authors should delete from the y-axis the term "corrected".
4. The term "significant" in discussion is inappropriately referred to BMI1 expression level if compared to the histogram in Figure 4G.
5. In Figure 2C the authors should add the unit of measurement (fold over control) in the table.
6. In Figure 4E the authors should add the figure legend reporting OLFM4 protein.
7. The authors should include few sentences summarizing the findings at the end of each paragraph.

Significance

General assessment:

Overall, the work is aimed to elucidate the role of CHRDL2 already considered a poor prognosis biomarker involved in the promotion of CRC (PMID: 28009989), in promoting stem-like properties. The authors elucidated new additional insights into the molecular mechanisms regulating stemness phenotype induced by the BMP antagonist CHRDL2 in CRC. The authors include in the study a large amount of data, which only partially support their hypothesis. However, this manuscript lacks organization and coherence, making it challenging to follow and read. Numerous concerns need to be addressed, along with some sentences to rephrase in the result and discussion sections.

Advance: The manuscript reported some functional insights on the role of CHRDL2 in colorectal cancer, but additional data should be added to support authors 'conclusions.

Audience: The manuscript is suggested for basic research scientists.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

Clarkson and Lewis present data suggesting that overexpression of Chordin like 2 (CHRDL2) can affect colorectal cancer cell responses to chemotherapy agents, possibly by modulating stem-cell like pathways. I have the following comments:

1. Fig. 1J-it is standard to show the images of cell migration-this is important here, given the modest effect of CHRDL2 overexpression here.
2. Fig. 2A-the very small error bars for most of the data on the curves suggests these are n=1 experiments with multiple technical replicates to generate the error bars. Please clarify. The legend says n=3 with ANOVA analysis but no significance detected. Please clarify.
3. Fig. 2B-given the overlapping error bars here, how can there be a p<0.01 between the groups?
4. Fig. 2G-did the authors try to estimate the concentration of CHRDL2 in the conditioned medium? Which cell line was used to generate this CM?
5. Fig. 5-what is the potential mechanism for gene expression changes in response to CHRDL2 overexpression? Is it all due to BMP inhibition? More mechanistic detail would be welcome here.

Significance

Unclear whether genetically engineered inducible overexpression has any real physiological significance but we all use cell models so this is OK.

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

Learn more at Review Commons

---

Reply to the reviewers

We would like to thank the reviewers for their overall positive assessment of our manuscript. We have used their constructive feedback to substantially improve our manuscript as described below.

Reviewer #1

Evidence, reproducibility and clarity

This study by Reyes at al is a well conducted analysis of memory B cell dynamics of Plasmodium falciparum (Pf) -specific B cell populations over the course of reducing Pf prevalence in ten Ugandan adults. The data is presented well and the authors provide compelling evidence that 1. There is an overall loss of Ag specific B cells with reduction in exposure and 2. Different antigens (MSP1/AMA-1 vs CIDRa-1) generate different flavors of long lived responses. However, additional clarity to the reader should be provided on certain topics (listed below).

Major comments: 1. While the premise of the study (reduced Pf transmission due to the use of indoor residual spraying (IRS)) is an important one, I think the authors must take into consideration that 9/10 subjects had at least one Pf positive episode between Time Points 1 and 2 (Figure 1). Also, it looks from Fig 1 that some samples were collected at a time of Pf positive test (green squares), while in Table S1 none of the subjects have a positive parasite status at TP1.

We recognize that most individuals had detectable parasitemia before and after time point (TP) 1. In our manuscript, we therefore do not report the time between TP1 and TP2, because we agree that the length of this time interval is not relevant in our study methodology. We only mention the time between the last known P. falciparum infection and collection of blood at the second time point. We use the sample collected at TP1 only as a representative sample obtained during a time with high P. falciparum exposure and do not make any claims based on the time between TP1 and TP2. The occurrence of infections after sample collection at TP1 confirms that parasite transmission was still high at this time. We have added a schematic of the relative levels of parasite transmission to Figure 1 to emphasize this.

With respect to infection status, none of the donors were blood smear positive at TP1. However, as mentioned in Table S1, parasites were detected in three individuals using the more sensitive LAMP assay. These three individuals are therefore marked as parasite positive in Figure 1. Table S1 has been modified to highlight the parasite status of these three individuals.

1. Figure S1A: What is trBC? Figure S1B: What is Strep? Are the strep positive cells also CIDR-1 positive and were they gated out? Why is APC used for MZ-1 and one of the MSP1-AMA-1 tetramers? Do these stainings come from multiple panels?

All abbreviations of B cell populations were defined in the figure legend (for example, trBC stands for transitional B cells). To facilitate the interpretation of Figure S1, we have now included the definitions of these abbreviations in the figure.

Strep stands for streptavidin, which has now also been clarified in the figure. In our gating strategy, we used the term “strep” to denote cells that bound to both CIDRa1 and MSP1/AMA1 tetramers, which we interpreted as non-specific binding to streptavidin or other components of the antigen tetramers. Only the “non-strep” cells were used to gate on antigen-specific cells. We have added this clarification to the figure legend.

In panel B, we accidentally used the term MZ (for merozoite) to describe tetramers of the merozoite antigens MSP1 / AMA1. These labels are interchangeable, but to avoid confusion, MZ-1 has been changed to MSP1 / AMA1.

1. Figure 3A: how many cells does the umap plot represent? Were there a total of 3555 Ag specific B cells that were non-naive (Figure 3E)?

It is correct that there were a total of 3,555 antigen-specific B cells used for the clustering shown in panel A. This information has been added to Figure 3A.

1. Could the authors comment on why in Figure 3, Ig isotype expression was not considered for clustering? This would allow for characterization of DN sub populations/ clusters in addition to the CD21-CD27- ABCs? It looks like IgD expression was low across the clusters (Figure 3D). Was this the case for the cells considered in this analysis, or was it excluded? If it was truly low expressed, how were the assessments in Figure 2 made?

From prior experience, we know that Ig isotype information tends to dominate in the clustering, which would result in major clusters based on IgM, IgD, IgG, and IgA expression, not on expression of other markers. This is illustrated in the example below. The UMAP on the left shows clusters in green and red that consist of IgG+ and IgA+ B cells, respectively. The UMAP on the right shows that switched memory (swM) B cells and DN B cells are found in both IgG and IgA clusters. Because we were mainly interested in identifying different subsets of B cells, irrespective of Ig isotype, we did not include Ig isotype in the clustering. We have clarified in the manuscript that Ig isotypes were excluded from the analysis to prevent these from dominating the clustering:

“Unsupervised clustering was then performed based on expression of all markers, except for Ig isotypes to prevent these from dominating the clustering.”

IgD expression among cell clusters shown in Figure 3 was low because only non-naïve B cells were included in the analyis. The majority of non-naïve cells are class-switched memory B cells and DN B cells, which by definition do not express IgD (see gating strategy in Figure S1A). Figure 2 shows all B cell populations, including naïve B cells and non-naïve B cell populations (unswitched memory, switched memory, and DN), that were gated based on IgD and CD27 expression.

5.Are there differences in these designations / phenotypes of DN populations of atBCs vs CD21-CD27- atBCs?

In the malaria field, atypical B cells are typically defined as CD21-CD27-. The definition of DN2 B cells comes from the autoimmunity field and is stricter: IgD-CD27-CD21-CD11c+ B cells. In our manuscript, we define atypical B cells in a stricter way than typically done in the malaria field, following published guidelines for the identification of B cell subsets (https://doi.org/10.3389/fimmu.2019.02458). Using these guidelines, atypical B cells and DN2 B cells are phenotypically identical. We have added a reference to these published guidelines in the Results section:

“Following published guidelines for the identification of B cell populations (21), total CD19+ B cells were divided into naïve B cells (IgD+CD27-), unswitched memory B cells (IgD+CD27+), switched memory B cells (IgD-CD27+), and double negative B cells (IgD-CD27-).”

1. Lines 258-259: In considering only switched MBCs, what clusters from Figure 3a were included? There seem to be 2588 sw MBCs (Table S3, Figure 4). Do the remaining cells (967 cells) come from clusters 2, 5 and 6 (and excludes the atBC clusters)

This analysis did not use the clusters presented in Figure 3, but instead used switched memory B cells gated as shown in Figure S1A. The reason for this is that the clusters in Figure 3 were generated using antigen-specific B cells and cannot be reproduced using non-antigen-specific B cells. Thus, it is not possible to separate all other B cells into the same six clusters. The only way to compare expression of certain markers between antigen-specific and non-antigen specific switched memory B cells is to gate on these populations manually. We have now tried to clarify this in the manuscript as follows:

“we determined the percentages of CD95+ cells and CD11c+ cells among antigen-specific switched memory B cells and the total population of switched memory B cells (gated manually as shown in Figure S1A).”

Minor comments: 1. Line 178- 179: Was there a specific measure of rate of decline made for these cells?

We did not calculate a rate of decline of antigen-specific B cells for several reasons: 1) the time between TP1 and TP2 is not the same for all people in the study, 2) the time between last exposure and TP2 is not the same for all people, and 3) the rate of decline is most likely not linear and cannot accurately be estimated with only two data points. We have changed the wording of this sentence such that we do not use the word “rate”:

“we did not observe a difference in the percentage of B cells with specificity for merozoite antigens or variant surface antigens that were lost.”

In addition, we included the percentage of reduction in size in the paragraph before this section:

“we observed that both populations decreased in size by about 50%, although these differences were not statistically significant.”

Significance

General assessment: Strengths: The authors provide evidence that the dynamics of antigen specific cells in humans can vary with exposure and with the nature of the antigen. They have nicely discussed the potential causes for these differences (Discussion), although they should include the findings of Ambegaonkar et al that ABCs in malaria may be restricted to responding specifically to membrane bound antigens (PMCID: PMC7380957)

As suggested by the reviewer, we have added a paragraph to the Discussion section to discuss the results reported by Ambegaonkar et al. and how the difference between soluble vs. membrane-bound antigens may have an effect on how these antigens are perceived by B cells:

The difference between soluble and membrane-bound antigens may also have a direct effect on how these antigens are perceived by B cells. Atypical B cells have been shown to be restricted to recognition of membrane-bound antigens (41). The interaction of a B cell with membrane-associated antigen allows the formation of an immunological synapse. Inhibitory receptors expressed by atypical B cells are excluded from this synapse, resulting in B cell receptor signaling and differentiation towards antibody-secreting cells (41). This could explain why atypical B cell subset 1 that expresses the highest levels of the inhibitory receptor FcRL5 is enriched for recognition of the CIDRα1 domain of membrane-bound protein PfEMP1. It should however be noted that soluble antigen can also be presented effectively in membrane-context by conventional dendritic cells, follicular dendritic cells, and subcapsular macrophages in secondary lymphoid organs, especially when it is part of an immune complex (reviewed in (42)). This would provide a route for atypical B cells to also respond to soluble merozoite antigens, such as MSP1 and AMA1.

Limitations: 1. Outlined above, and as the authors also mention, a small sample size and homogenous population. 2. The evidence for reduced transmission is not clear, and the negative parasite tests for donors shown in Table S1 do not match with Figure 1 data. 3. Lack of IgD expression across clusters (Figure 3D- the authors will need to clarify this point) would require re-analysis of Figure 2 data

1. We have provided clarification in response to the points raised by the reviewer.

2. We believe there is clear evidence for reduced transmission, from a median of almost 2 infections per person per year prior to the implementation of IRS to a median parasite-free period of 1.7 years prior to sample collection at TP2. To further emphasize this, we have summarized the number of P. falciparum infections among the ten individuals included in this study (now included in Table S3):

year

Pf infections

comment

2012

20

2013

19

TP1

2014

20

TP1

2015

8

Start IRS

2016

0

TP2

This reduced parasite exposure is reflected in a decrease in immune activation as presented in Figure 2. We have clarified that the data in Table S1 did indeed match those shown in Figure 1.

1. We have clarified that IgD expression is low in the clusters presented in Figure 3 because naïve B cells were excluded from this analysis.

Advances: This study highlights the importance of studying antigen specific B cells in humans in the context of natural infection and the use of high-parameter tools such as spectral flow cytometry in assessing a large quantity of data from limited clinical samples. These data are important to inform better vaccine design. Studies in inbred animals can be quite limited or different from human B cell responses.

Audience: This study will be of interest to malariologists and B cell immunologists. Atypical B cells are relevant to many infectious diseases and auto immunity, while the dynamics of memory B cells in malaria all be relevant to those interested in vaccine design against blood stage antigens.

---

Reviewer #2

Evidence, reproducibility and clarity

Summary: In this study, the authors compared long-lived total and antigen (ag)-specific B-cell levels in a cohort of 10 Ugandan malaria patient samples that were collected before and after local reduction of P. falciparum transmission (pre/post-IRS). The focus is on the novel comparison of the two most common malaria antigens: merozoite antigens (MSP1/AMA1) and variant surface antigens (CIDRα1). Using high-parameter spectral flow cytometry, they also characterized the phenotype of the different populations of cells. Their main findings include 1) a decrease in activated but maintenance of resting ag-specific B-cells in the post-IRS sample and 2) CD95 and CD11c, as the only differentially expressed markers between MSP1/AMA1-specific and CIDRα1-specific long-lived memory B cells. Their further phenotypic characterization suggests functional consequences with MSP1/AMA1-specific B-cells being poised for rapid antibody-secreting cell differentiation while CIDRα1-specific B cells were enriched among a subset of atypical B cells that seem poised for antigen presentation (CD86+CD11chi/ AtBC1). Their findings consolidate and further expand our knowledge of long-lived B-cell levels during P. falciparum malaria and report/compare (for the first time to my knowledge) a differential selection of long-lived B-cell levels between these 2 antigen specificities. Overall, the manuscript is straightforward and well-written and the authors did a good job explaining their methodology, findings, and interpretations. I believe the major gap missing in this study is the reconciliation of long-lived antigen-specific B-cell levels with the serum antigen-specific antibody levels of these patients against the same 2 antigens (MSP1/AMA1 and CIDRα1) in the experiments and the discussion. The antibody data would strengthen their main argument and is the main missing piece for characterizing more completely the long-lived antigen-specific humoral responses. Below are my suggestions that would help improve the manuscript:

Major comments: 1. Serum Anti-Pf antibodies: Do the authors have access to the serum/plasma of these patients? It would be important to correlate the total and ag-specific B-cell populations with levels of serum IgG antibodies against those specific Pf antigens (MSP1/AMA1 and CIDRα1) and total IgG levels to strengthen their point about long-lived humoral responses.

To our understanding, the rationale for such an analysis would be that if IgG levels correlated with the size of a certain B cell population, it would suggest that this B cell population is implicated in the production of IgG against a particular antigen. While a correlation between the percentage of memory B cells and IgG titers has been observed for antigens from several viruses and bacteria (1-4), other studies have reported the absence of such a correlation (4-7). Similarly, for P. falciparum antigens, a moderate correlation between memory B cell abundance and IgG titers has been observed for some merozoite antigens, but not for others (8, 9). The lack of a correlation between the magnitude of the memory B cell and the antibody response fits with the prevailing model that memory B cells and plasma cells are two independently controlled arms of the humoral immune system (10, 11). Given the lack of strong evidence that the levels of IgG titers and memory B cells are interconnected, we do not think this analysis will be informative.

An alternative analysis would be to study the contribution of B cell subsets to the production of IgG after re-exposure, similar to a recent study that identified T-bet+ memory B cells as the main contributors to antibody responses following influenza virus vaccination (12). Unfortunately, we are unable to perform this analysis in this study population, because only four of the individuals included in this study (spanning calendar years 2012 – 2016) were recruited into a follow up cohort (calendar years 2017 – 2019), and none of these four people were infected during this later time frame.

We have however added this future direction to the Discussion section:

To determine the contribution of different memory B cell subsets to the recall response against P. falciparum, it would be interesting to analyze IgG responses upon re-infection. However, none of the individuals included in this study experienced a recorded P. falciparum infection post-IRS, preventing us from performing such an analysis.

References

1. Crotty et al., J Immunol (2003), https://doi.org/10.4049/jimmunol.171.10.4969
2. Quinn et al., J Infect Dis (2004), https://doi.org/10.1086/423937
3. Cohen et al., Cell Rep Med (2021), https://doi.org/10.1016/j.xcrm.2021.100354
4. Amanna et al., New England J Med (2007), https://doi.org/10.1056/nejmoa066092
5. Leyendeckers et al., Eur J Immunol (1999), https://doi.org/10.1002/(sici)1521-4141(199904)29:04%3C1406::aid-immu1406%3E3.0.co;2-p
6. Nanan et al., Vaccine (2001), https://doi.org/10.1016/s0264-410x(01)00328-0
7. Goel et al., Science Immunol (2021), https://doi.org/10.1126/sciimmunol.abi6950
8. Rivera-Correa et al., eLife (2019), https://doi.org/10.7554/elife.48309
9. Jahnmatz et al., Front Immunol (2021), https://doi.org/10.3389/fimmu.2020.619398
10. Weisel et al., Immunity (2016), https://doi.org/10.1016/j.immuni.2015.12.004
11. Shinnakasu et al., Nat Immunol (2016), https://doi.org/10.1038/ni.3460
12. Nellore et al., Immunity (2023), https://doi.org/10.1016/j.immuni.2023.03.00

    1. Correlation between populations and initial parasite load: Are the levels between any of the populations at any time point correlated significantly in any way? If the statistical power/N allows it, please perform a correlation array between all populations using all samples both total and ag-specific and initial parasite load.

We agree that this analysis could be very interesting. However, in most recorded infection cases, parasitemia was submicroscopic and parasite load was not reported. Information about parasite density in the blood prior to TP1 is available for only half of the individuals in this study. In these people, the last known parasite density was recorded between three months to two years prior to TP1. Given the small number of individuals for whom these data are available and the large variation in time between parasitemia and sampling, we do not have sufficient data to perform this analysis.

1. Figure 2: Why were total and ag-specific plasmablasts/plasma cells not included in this figure? Please include to compare levels in these two time points.

We did not include the levels of total and antigen-specific plasmablasts (PBs) in Figure 2 because the percentages of PBs are relatively low, and very few antigen-specific PBs were detected. We have now included the levels of total PBs in Figure 2A and the percentages of antigen-specific PBs in Supplementary Figure 2. The percentage of PBs among total B cells decreased by about 50% between TP1 and TP2, in line with a decrease in immune activation.

1. Healthy baseline: The study is missing "healthy" controls as a reference. I presume this is because each patient is its uninfected control in the post-IRS sample. In methods, they mentioned they used two naïve-USA B-cells as technical controls. It would be important to include and maybe expand (to match age and gender)on that specific data from those controls as supplementary figures to support their findings:
2. Show negative Tetramer staining for these samples (to understand the background).
3. Levels of all the USA controls total B cell populations and compared to the pre/post-IRS samples to understand "baseline" or "non-endemic" control levels.

1. We have included flow cytometry plots of tetramer staining for the non-P. falciparum exposed donors (pooled B cells from two US donors) to show the level of background for these probes. These plots are shown in Figure S1B.

2. We have used data from P. falciparum-naive US donors (n = 7) that we generated for a prior study to show the average level of total B cell populations in Figure 2, and the percentage of switched memory B cells that express CD95, CD11c, T-bet, and FcRL5 in Figure 4.

Minor comments: 1. In the gating strategy (S1), please include the percentage of each population of that representative example.

We have added the percentages for all gated populations to Figure S1.

1. For Figure 2, since not every panel has the same N, please include the N for each panel in the figure or a supplementary table.

All panels in Figure 2 show data for all 10 individuals. However, since some data points are overlapping, it may appear that some panels show data from fewer individuals. Specifically, no antigen-specific DN1 cells were detected pre- and post-IRS for four individuals. These data points therefore overlap and are not visible. To avoid confusion, we had mentioned this in the legend to Figure 2 (see text in orange). We have tried to further clarify this by emphasizing in the figure legend that data from all 10 individuals are shown (see text in red):

“Figure 2: Abundance of total and antigen-specific B cell subsets in the circulation during high parasite transmission and in the absence of P. falciparum exposure. The percentage of B cell subsets among circulating B cells is shown for total B cells (A), MSP1/AMA1-specific B cells (B), and CIDRα1-specific B cells (C). For MSP1/AMA1-specific B cells and CIDRα1-specific B cells, the total percentage among all circulating B cells is also shown (right most graphs in each panel). All panels show data for all 10 individuals. In panels B and C, no antigen-specific DN1 cells were detected pre- and post-IRS for four individuals. These data points therefore overlap and are not clearly visible. Differences between groups were evaluated using a Wilcoxon matched-pairs signed-rank test. P values

1. Please mention the history of past and chronic co-infections of these 10 patients. Particularly if they had any other active or recent infection when the sample was taken.

Four individuals had active or recent infections in the three months prior to sample collection, with upper respiratory tract infections being the most common. This information has been included in Table S3, with a reference to these data in the Methods section. We have also included a link to ClinEpiDB where additional information about the cohort participants, including medical history, can be found.

1. Discussion: further discussion with relevant literature on the following points is needed to consolidate cellular and antibody studies: a. Whether the presence of long-lived ag-specific B-cell responses correlates with sustained levels of IgG against Pf antigens. b. The different types of antibodies (protective/pathogenic) that these different B-cell populations have been reported to produce during malaria.

a. We have added the following paragraph to the Discussion section:

To determine how these different long-lived B cell subsets contribute to protection against P. falciparum infection, it would be important to analyze the connection between the cellular repertoire and plasma IgG. For P. falciparum antigens, a moderate correlation between memory B cell abundance and IgG titers has been observed for some merozoite antigens, but not for others (28, 44). This is in line with studies for other pathogens, that showed a correlation between the percentage of memory B cells and IgG titers for antigens from several viruses and bacteria (48-51), while other studies have reported the absence of such a correlation (51-54). The lack of a correlation between the magnitude of the memory B cell and the antibody response fits with the prevailing model that memory B cells and plasma cells are two independently controlled arms of the humoral immune system (55, 56). To determine the contribution of different memory B cell subsets to the recall response against P. falciparum, it would be interesting to analyze IgG responses upon re-infection. However, none of the individuals included in this study experienced a recorded P. falciparum infection post-IRS, preventing us from performing such an analysis.

b. We have added additional discussion about the types of antigens recognized by atypical B cells to the Discussion section:

Prior studies have shown that while atypical B cells harbor reactivity against P. falciparum antigens (9,18), they are also enriched for autoreactivity (43). Specifically, atypical B cells produce antibodies against the membrane lipid phosphatidylserine, which can induce the destruction of uninfected erythrocytes and contribute to anemia (44).

Significance

General assessment:

Strengths: - Novelty in contrasting two different types of P. falciparum antigen responses at the B-cell level. - The use of tetramers is a cutting-edge technique to assess this question. - Analyses were thorough and found contrasting differences in antigen-specific B-cell populations (atypical vs classical) between these 2 antigens for the first time (to my knowledge). - Well-written manuscript with clear data, methodology, and conclusions

Limitations: - Missing serum/plasma antibody data to support their claim about long-lived humoral responses and reconciliation of ag-specific B-cell levels and ag-specific antibody levels in experiments and discussion. - Limited N of 10 patients of the same gender (female), some population analyses had even fewer samples. - Missing baseline levels for non-endemic uninfected control for B-cell populations for comparison.

• We have included a discussion about the correlation between plasma antibody and memory B cell responses in the Discussion section.

• We have clarified that some data points overlap in Figure 2, giving the impression that data from fewer than 10 individuals were shown.

• We have included baseline levels of 1) tetramer reactivity (Figure S1), 2) the size of B cell populations (Figure 2), and 3) expression of select markers (Figure 4).

Advance: The study consolidates antigen-specific responses with the discovery of recently characterized populations (ex. atypical) and finds novel differences between two types of malaria antigen responses at the B-cell level and between specific populations responding differentially to these antigens. The findings are incremental (role of B-cell population in malaria-specific responses), conceptual (contrasting two types of B-cell antigen responses in the same infection), and clinical (finding significant differences in patients).

Audience: This study will attract basic B-cell immunology scientists, infectious disease clinicians/scientists, vaccinologists, and both basic malaria immunology and clinical audiences.

Reviewer expertise: Malaria, immunology, antibodies.

__Reviewer #3 __

Evidence, reproducibility and clarity: The authors analysed the antigen specificity and phenotypes of B cells during high P falciparum transmission and after a period of successful malaria control with IRS in Uganda. The gap between the two sampling time points is close to two years.

They use antigen probes for MSP1/AMA1 and CIDRalpha1, two antigens expressed at different stages of P. falciparum life cycle-merozoites and infected red cells, respectively. While MSP1/AMA1 are involved in the parasite's invasion of red blood cells, CIDRalpha1 is a domain of PFEMP1, a large family of antigenically variant proteins that mediates the sequestration of infected red cells in small blood vessels.

They found that the percentage of activated antigen-specific memory B cells declined with malaria control. However, detectable frequencies of antigen-specific memory B cells were retained after malaria control, which confirms earlier reports.

However, they also demonstrate that the phenotypic characteristics of memory B cells are associated with antigen specificity. The retained MSP1/AMA1-specific B cells were mostly CD95+CD11c+ memory B cells and FcRL5-Tbet- atypical B cells. In contrast, the retained CIDRalpha1-specific B cells were enriched among a subpopulation of atypical B cells.

These findings suggest differences exist in how the MSA1/AMA1 and CIDRalpha1 y are recognised and processed by the human immune system and how the immune response responds to them upon re-infection with P falciparum.

Major issues affecting the conclusion: The findings and conclusions of this study, whilst positively exciting and informative, are based on the analyses of very few cells (at times). Even the authors themselves acknowledge this. I expect the authors to address this issue by toning down their reporting and conclusions (where appropriate). Ultimately, we need to have the confidence that these results are reproducible.

We appreciate the reviewer’s concern about the numbers of antigen-specific cells included in our analyses, which is an inherent limitation of this approach. However, we would like to point out that most analyses included a substantial number of antigen-specific B cells:

Figure 3D: 158 to 2,038 cells per group

Figure 4: an average of 26 to 184 cells per donor

Figure 5B: 55 to 508 cells per group

Figure 5C: 10 to 334 cells per group*

* The group with 10 cells is an outlier here. All other groups contain at least 104 cells. Because this one condition had such a small number of cells, we specifically mentioned this number in the text.

The numbers of cells for analyses shown in Figures 3D and 5B were already included in the figures. All the other numbers were mentioned in Table S3. To further clarify the number of cells included in each analysis, we have added the number of cells to Figures 4 and 5C.

To tone down our reporting, we have rephrased some of our conclusions, and now present our section headers in past tense to make these statements reflect our observation instead of a definitive conclusion. For example:

Conclusion: “The loss of MSP1/AMA1-specific and CIDRα1-specific B cells in the circulation was similar, but the phenotype of long-lived MSP1/AMA1-specific and CIDRα1-specific B cells appeared to differ.”

Section header: “Long-lived MSP1/AMA1-specific and CIDRα1-specific B cells differed in phenotype”

Finally, in the Discussion section, we have added a statement to our paragraph describing the limitations of our study to stress the importance of reproducing our findings:

All in all, it will be important to perform additional studies of the phenotype and functionality of long-lived B cells with specificity for P. falciparum antigens to reproduce and extend our findings.

Minor comments: Figure 2D-I found this figure, and its presentation is unclear. Notably, using contour plots doesn't allow the reader to appreciate the density of the cells being presented.

To facilitate the interpretation of this figure, we have changed the plot type to a contour plot with density color gradient, and added the number of cells shown in each plot. (Please note that this panel has been renumbered to C.)

Figure 4 - label the y-axis.

The y-axis was labeled with “%”, which we have expanded to “% of B cells expressing marker of interest”.

__Significance: __The study design-as outlined-allowed for the analyses of the specificity and phenotypic characteristics of residual P falciparum-specific memory B cells after 1.7 years of little to no P falciparum exposure. The cell phenotyping methods presented are also appropriate. However, antigen-specific cells are rare in blood circulation, and as the authors themselves acknowledge in the discussion, some of the results are based on very few cells. This means we cannot be sure all the results presented are reproducible.

Previous studies demonstrated that P falciparum memory B cells are maintained long after cessation of antigen exposure. However, few (if any) detailed antigen-specific and phenotypic analyses of the characteristics of P falciparum-specific memory B cells following a long period of no exposure exist. Thus, this study presents an incremental advance in our knowledge. In addition, the association of antigen specificity with cell phenotypes is a new concept in malaria immunology. The research presented will greatly interest infectious disease immunologists and vaccinologists.

I am an infectious disease immunologist with substantial experience in malaria immunology.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #3

Evidence, reproducibility and clarity

The authors analysed the antigen specificity and phenotypes of B cells during high P falciparum transmission and after a period of successful malaria control with IRS in Uganda. The gap between the two sampling time points is close to two years.

They use antigen probes for MSP1/AMA1 and CIDRalpha1, two antigens expressed at different stages of P. falciparum life cycle-merozoites and infected red cells, respectively. While MSP1/AMA1 are involved in the parasite's invasion of red blood cells, CIDRalpha1 is a domain of PFEMP1, a large family of antigenically variant proteins that mediates the sequestration of infected red cells in small blood vessels.

They found that the percentage of activated antigen-specific memory B cells declined with malaria control. However, detectable frequencies of antigen-specific memory B cells were retained after malaria control, which confirms earlier reports.

However, they also demonstrate that the phenotypic characteristics of memory B cells are associated with antigen specificity. The retained MSP1/AMA1-specific B cells were mostly CD95+CD11c+ memory B cells and FcRL5-Tbet- atypical B cells. In contrast, the retained CIDRalpha1-specific B cells were enriched among a subpopulation of atypical B cells.

These findings suggest differences exist in how the MSA1/AMA1 and CIDRalpha1 y are recognised and processed by the human immune system and how the immune response responds to them upon re-infection with P falciparum.

Major issues affecting the conclusion:

The findings and conclusions of this study, whilst positively exciting and informative, are based on the analyses of very few cells (at times). Even the authors themselves acknowledge this. I expect the authors to address this issue by toning down their reporting and conclusions (where appropriate). Ultimately, we need to have the confidence that these results are reproducible.

Minor comments:

Figure 2D-I found this figure, and its presentation is unclear. Notably, using contour plots doesn't allow the reader to appreciate the density of the cells being presented.

Figure 4 - label the y-axis.

Significance

The study design-as outlined-allowed for the analyses of the specificity and phenotypic characteristics of residual P falciparum-specific memory B cells after 1.7 years of little to no P falciparum exposure. The cell phenotyping methods presented are also appropriate. However, antigen-specific cells are rare in blood circulation, and as the authors themselves acknowledge in the discussion, some of the results are based on very few cells. This means we cannot be sure all the results presented are reproducible.

Previous studies demonstrated that P falciparum memory B cells are maintained long after cessation of antigen exposure. However, few (if any) detailed antigen-specific and phenotypic analyses of the characteristics of P falciparum-specific memory B cells following a long period of no exposure exist. Thus, this study presents an incremental advance in our knowledge. In addition, the association of antigen specificity with cell phenotypes is a new concept in malaria immunology. The research presented will greatly interest infectious disease immunologists and vaccinologists.

I am an infectious disease immunologist with substantial experience in malaria immunology.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

Summary:

In this study, the authors compared long-lived total and antigen (ag)-specific B-cell levels in a cohort of 10 Ugandan malaria patient samples that were collected before and after local reduction of P. falciparum transmission (pre/post-IRS). The focus is on the novel comparison of the two most common malaria antigens: merozoite antigens (MSP1/AMA1) and variant surface antigens (CIDRα1). Using high-parameter spectral flow cytometry, they also characterized the phenotype of the different populations of cells. Their main findings include 1) a decrease in activated but maintenance of resting ag-specific B-cells in the post-IRS sample and 2) CD95 and CD11c, as the only differentially expressed markers between MSP1/AMA1-specific and CIDRα1-specific long-lived memory B cells. Their further phenotypic characterization suggests functional consequences with MSP1/AMA1-specific B-cells being poised for rapid antibody-secreting cell differentiation while CIDRα1-specific B cells were enriched among a subset of atypical B cells that seem poised for antigen presentation (CD86+CD11chi/ AtBC1). Their findings consolidate and further expand our knowledge of long-lived B-cell levels during P. falciparum malaria and report/compare (for the first time to my knowledge) a differential selection of long-lived B-cell levels between these 2 antigen specificities. Overall, the manuscript is straightforward and well-written and the authors did a good job explaining their methodology, findings, and interpretations. I believe the major gap missing in this study is the reconciliation of long-lived antigen-specific B-cell levels with the serum antigen-specific antibody levels of these patients against the same 2 antigens (MSP1/AMA1 and CIDRα1) in the experiments and the discussion. The antibody data would strengthen their main argument and is the main missing piece for characterizing more completely the long-lived antigen-specific humoral responses. Below are my suggestions that would help improve the manuscript:

Major comments:

1. Serum Anti-Pf antibodies: Do the authors have access to the serum/plasma of these patients? It would be important to correlate the total and ag-specific B-cell populations with levels of serum IgG antibodies against those specific Pf antigens (MSP1/AMA1 and CIDRα1) and total IgG levels to strengthen their point about long-lived humoral responses.
2. Correlation between populations and initial parasite load: Are the levels between any of the populations at any time point correlated significantly in any way? If the statistical power/N allows it, please perform a correlation array between all populations using all samples both total and ag-specific and initial parasite load.
3. Figure 2: Why were total and ag-specific plasmablasts/plasma cells not included in this figure? Please include to compare levels in these two time points.
4. Healthy baseline: The study is missing "healthy" controls as a reference. I presume this is because each patient is its uninfected control in the post-IRS sample. In methods, they mentioned they used two naïve-USA B-cells as technical controls. It would be important to include and maybe expand (to match age and gender)on that specific data from those controls as supplementary figures to support their findings:
5. Show negative Tetramer staining for these samples (to understand the background).
6. Levels of all the USA controls total B cell populations and compared to the pre/post-IRS samples to understand "baseline" or "non-endemic" control levels.

Minor comments:

1. In the gating strategy (S1), please include the percentage of each population of that representative example.
2. For Figure 2, since not every panel has the same N, please include the N for each panel in the figure or a supplementary table.
3. Please mention the history of past and chronic co-infections of these 10 patients. Particularly if they had any other active or recent infection when the sample was taken.
4. Discussion: further discussion with relevant literature on the following points is needed to consolidate cellular and antibody studies: a. Whether the presence of long-lived ag-specific B-cell responses correlates with sustained levels of IgG against Pf antigens. b. The different types of antibodies (protective/pathogenic) that these different B-cell populations have been reported to produce during malaria.

Significance

General assessment:

Strengths:

• Novelty in contrasting two different types of P. falciparum antigen responses at the B-cell level.
• The use of tetramers is a cutting-edge technique to assess this question.
• Analyses were thorough and found contrasting differences in antigen-specific B-cell populations (atypical vs classical) between these 2 antigens for the first time (to my knowledge).
• Well-written manuscript with clear data, methodology, and conclusions

Limitations:

• Missing serum/plasma antibody data to support their claim about long-lived humoral responses and reconciliation of ag-specific B-cell levels and ag-specific antibody levels in experiments and discussion.
• Limited N of 10 patients of the same gender (female), some population analyses had even fewer samples.
• Missing baseline levels for non-endemic uninfected control for B-cell populations for comparison.

Advance:

The study consolidates antigen-specific responses with the discovery of recently characterized populations (ex. atypical) and finds novel differences between two types of malaria antigen responses at the B-cell level and between specific populations responding differentially to these antigens. The findings are incremental (role of B-cell population in malaria-specific responses), conceptual (contrasting two types of B-cell antigen responses in the same infection), and clinical (finding significant differences in patients).

Audience:

This study will attract basic B-cell immunology scientists, infectious disease clinicians/scientists, vaccinologists, and both basic malaria immunology and clinical audiences.

Reviewer expertise:

Malaria, immunology, antibodies.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

This study by Reyes at al is a well conducted analysis of memory B cell dynamics of Plasmodium falciparum (Pf) -specific B cell populations over the course of reducing Pf prevalence in ten Ugandan adults. The data is presented well and the authors provide compelling evidence that 1. There is an overall loss of Ag specific B cells with reduction in exposure and 2. Different antigens (MSP1/AMA-1 vs CIDRa-1) generate different flavors of long lived responses. However, additional clarity to the reader should be provided on certain topics (listed below).

Major comments:

1. While the premise of the study (reduced Pf transmission due to the use of indoor residual spraying (IRS)) is an important one, I think the authors must take into consideration that 9/10 subjects had at least one Pf positive episode between Time Points 1 and 2 (Figure 1). Also, it looks from Fig 1 that some samples were collected at a time of Pf positive test (green squares), while in Table S1 none of the subjects have a positive parasite status at TP1.
2. Figure S1A: What is trBC? Figure S1B: What is Strep? Are the strep positive cells also CIDR-1 positive and were they gated out? Why is APC used for MZ-1 and one of the MSP1-AMA-1 tetramers? Do these stainings come from multiple panels?
3. Figure 3A: how many cells does the umap plot represent? Were there a total of 3555 Ag specific B cells that were non-naive (Figure 3E)?
4. Could the authors comment on why in Figure 3, Ig isotype expression was not considered for clustering? This would allow for characterization of DN sub populations/ clusters in addition to the CD21-CD27- ABCs? It looks like IgD expression was low across the clusters (Figure 3D). Was this the case for the cells considered in this analysis, or was it excluded? If it was truly low expressed, how were the assessments in Figure 2 made?
5. Are there differences in these designations / phenotypes of DN populations of atBCs vs CD21-CD27- atBCs?
6. Lines 258-259: In considering only switched MBCs, what clusters from Figure 3a were included? There seem to be 2588 sw MBCs (Table S3, Figure 4). Do the remaining cells (967 cells) come from clusters 2, 5 and 6 (and excludes the atBC clusters)

Minor comments:

1. Line 178- 179: Was there a specfic measure of rate of decline made for these cells?

Significance

General assessment:

Strengths: The authors provide evidence that the dynamics of antigen specific cells in humans can vary with exposure and with the nature of the antigen. They have nicely discussed the potential causes for these differences (Discussion), although they should include the findings of Ambegaonkar et al that ABCs in malaria may be restricted to responding specifically to membrane bound antigens (PMCID: PMC7380957)

Limitations:

1. Outlined above, and as the authors also mention, a small sample size and homogenous population.
2. The evidence for reduced transmission is not clear, and the negative parasite tests for donors shown in Table S1 do not match with Figure 1 data.
3. Lack of IgD expression across clusters (Figure 3D- the authors will need to clarify this point) would require re-analysis of Figure 2 data

Advances: This study highlights the importance of studying antigen specific B cells in humans in the context of natural infection and the use of high-parameter tools such as spectral flow cytometry in assessing a large quantity of data from limited clinical samples. These data are important to inform better vaccine design. Studies in inbred animals can be quite limited or different from human B cell responses.

Audience: This study will be of interest to malariologists and B cell immunologists. Atypical B cells are relevant to many infectious diseases and auto immunity, while the dynamics of memory B cells in malaria all be relevant to those interested in vaccine design against blood stage antigens.

Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

Learn more at Review Commons

---

Reply to the reviewers

The authors do not wish to provide a response at this time.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #3

Evidence, reproducibility and clarity

Summary

The manuscript presents IGNITE (Inference of Gene Networks using Inverse kinetic Theory and Experiments), an unsupervised machine learning framework for constructing gene regulatory networks from single-cell RNA sequencing (scRNA-seq) data. IGNITE utilizes a kinetic inverse Ising model to infer gene interactions from binarized expression data and can predict genetic perturbation effects, such as those from knockout experiments. Although the application of inverse Ising models to network reconstruction is not entirely novel, IGNITE's specific implementation and its application to single-cell RNA sequencing data represent a new development. The method is tested on the transition from naive to formative states in murine pluripotent stem cells, a system the authors are highly knowledgeable about, and its performance is compared to state-of-the-art alternative methods.

Major concerns

My concern regards the generality of the method, particularly the entire pipeline presented, and the fairness of the performance comparison. These concerns can be easily addressed by the authors by better explaining their choices and their general applicability, and by toning down the conclusions about the comparison with existing inference methods.

The pre-processing steps are extensive, and their rationale is not always clear, though the results heavily depend on this analysis. Several steps appear to involve arbitrary choices optimized for specific outcomes, potentially introducing biases. The authors should better explain the rationale behind their choices to mitigate these concerns.

Specifically, part of the pipeline seems to be built to reproduce a specific expression pattern of 24 genes that some of the authors discovered in a previous paper. Although this prior knowledge could be useful and relevant in this specific system, it could limit the generality of the method. For example, the authors selected approximately 2000 genes based on prior knowledge and used a combination of t-SNE and UMAP for dimensionality reduction (although the two techniques have a similar goal). This specific combination seems to reproduce the pseudotime alignment the authors were expecting to find, but such prior information might not be available in general. Therefore, feature selection and the methods used to project data need more justification, especially if the goal is to create a general tool applicable across different biological systems.

Analogously, the clustering seems manually adjusted to match known expression patterns of 24 relevant genes, rather than being the result of an optimized clustering method. Additionally, the clusters overlap with different time points, raising concerns about potential batch effects. These issues should be addressed to strengthen the validity of the method.

The claims about the comparison with existing methods should be toned down. While the comparisons are useful and interesting, they might be biased due to the method's fine-tuning for the specific system studied. The claim that the model requires only scRNA-seq data is misleading, as strong prior biological knowledge was used to select, for example, the genes analyzed.

Significance

The manuscript is scientifically sound, clearly written, and deserves publication. The proposed method is quantitative, novel, theoretically grounded, and was tested in detail with appropriate null models and statistical methods. Moreover, IGNITE can be applied to various biological systems as the availability of scRNA-seq datasets is continuously growing. The paper will be of interest to a broad community of computational biologists and biology labs interested in gene regulation using scRNA-seq data.

The limitation, in my opinion, is the method's (particularly the pre-processing pipeline) fine-tuning for the specific biological system tested. Testing IGNITE on another biological system without pre-selected pre-processing steps or detailed biological priors would be more convincing and make the paper's conclusions much stronger. The comparison with other methods also may be slightly biased due to this fine-tuning.

My background is in statistical physics, with expertise in biological physics, specifically in mathematical modeling and data analysis in molecular biology.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

Corridori et al introduce IGNITE, a computational framework to infer gene regulatory networks (GRNs) from scRNA-seq data leveraging the kinetic Ising model, which can be used to simulate synthetic gene expression and perform in-silico knockout experiments. Other similar frameworks exist, but none combine these three aspects together. The authors have generated a scRNA-seq of murine ESCs differentiation which they use to compare their method with others. Specifically they show that they can infer known regulatory interactions, that they can generate similar data than the original and that it can potentially predict gene expression changes in transcription factor knock-out perturbations.

Major comments:

• Many of the authors' claims are backed by qualitative results and not properly quantified. In Fig2, authors qualitatively compare intra gene correlations between genes for the original data and their prediction. Instead of just visualizing they should compute and report the Spearman correlation between the original expression and the predicted one. The Fraction of Agreement is not a good metric to compare knockout predictions since it is completely dependent on the class imbalance of signs, for example if the selected genes are 75% positive and 25% negative, a naive predictor that only outputs positive predictions will still have a high score. Instead, the authors should quantify this with Spearman correlation or RMSE and compare across methods. In FigS4a-b the authors qualitatively claim that other methods could not predict the expected cell composition, which they should quantify and report the values across methods. When comparing against the ground truth network, the fraction of correctly inferred interactions is technically the same as precision but is ignoring recall. I suggest the authors compute precision, recall and a combined F1 score to compare the evaluated methods. Authors claim that the method is scalable to a larger number of genes but no data is provided, they should show how their method compares to others when using a different number of cells and number of genes at memory usage and running time.
• The authors need to better describe which tests were performed when talking about significance, which thresholds and which corrections, if any, were employed.
• To reduce the number of dimensions of scRNA-seq data the authors use t-SNE and then from the obtained result UMAP to project the data into a lower dimensional space. This is fundamentally wrong since distances are not well preserved in t-SNE. Instead the authors should first employ PCA and then UMAP. Additionally, the authors use UMAP distances in the Slingshot pseudotime calculation. Similar to t-SNE, UMAP distances have no real meaning and should only be used for visualization purposes. Instead, the authors should provide Slingshot the obtained PCA embeddings.
• Dictys (PMID: 37537351) is a known GRN inference method that also can simulate gene expression but is missing in the benchmark, the authors should add it to the method comparison.
• The current manuscript is not reproducible since it is missing the method's code, the code to reproduce the figures and the generated scRNA-seq data.
• Authors claim that the method is scalable to a larger number of genes but no data is provided to back this claim. They should show how their method compares to others when using a different number of cells and number of genes.

Minor points:

• In the introduction, authors mention multimodal GRN inference methods but do not provide any references.
• In Table 1, CellOracle is annotated as not being able to do multiple KO which is wrong. Additionally, the authors mention that IGNITE uses no prior knowledge which is not really true since it requires pseudotime ordering. The authors should add a column to Table 1 whether methods require pseudotime.
• It is unclear what the dashed arrow of Fig1b means. Moreover, plotting gene expression values on top of UMAPs can be misleading, instead authors should plot the gene expression distributions binned by pseudotime.
• The authors report a p-value of 1.04x10-171 which is below detection limit (see PMID: 30921532). Authors should change it to an interval such as p < 2.2×10-16.
• To make CellOracle results easier to interpret and more comparable, authors should run it at the atlas level instead of at the cell type level, this way generating only one GRN. This can be achieved by assigning the same cluster label to all cells.
• Experimental values in FigS3b seem to have been repeated and do not match the previous ones for IGNITE and SCODE.
• It is unclear what the different circles mean in Fig5b.

Significance

This manuscript is an incremental and methodological work for specialized audiences. Its strengths are that the authors employ kinetic Ising model for GRN inference and that they provide a single framework capable of inferring, simulating and perturbing gene expression. The main limitations are that the claims should be better quantified and that the code and data need to be made accessible.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

Summary

Corridori and colleagues propose IGNITE, a novel method to recover Gene Regulatory Networks (GRN) from single cell RNA-sequencing (scRNA-seq) data. Their method solves the inverse Ising problem generating a cohort of candidate GRN optimising it to minimise the difference to the input expression matrix. Authors report the IGNITE is able to predict wild type data and simulate both single and multiple gene knockouts. Authors benchmark this method on a in-house data set of differentiating pluripotent stem cells (PSC). They focus on a small set of genes known to be involved in PSC differentiation into formative cells. Authors benchmark IGNITE against state of the art tools (SCODE, MaxEnt and CELLORACLE). They evaluate IGNITE ability to predict wild type gene expression by comparing their data with experimental data and with SCODE. They conclude the tool has generative capacity comparable with SCODE. They also evaluate IGNITE ability to recover known interactions with respect to other tools without finding it to significantly outperform them.

Major comments

• Are the key conclusions convincing?

Conclusions appear convincing although model generalizability could be shown in a more thorough manner. For instance, analysing some other publicly available dataset could help demonstrate hyperparameters effects on GRN predictions and their robustness across different experiments. - Should the authors qualify some of their claims as preliminary or speculative, or remove them altogether?

Claims are well supported by data. - Would additional experiments be essential to support the claims of the paper? Request additional experiments only where necessary for the paper as it is, and do not ask authors to open new lines of experimentation.

I think the work would benefit from an additional benchmark on a different cellular system. This experiment would show how hyperparameters generalise across datasets and would provide potential users insights how to tweak them.

Also, how does the model scale with the number of genes? A benchmark on computation time and resources required to infer GRN of growing size would be valuable in the adoption of this tool.

In addition, I think the GRN comparison benchmark presented in section (3.4) would benefit from a quantitative discussion. Authors show inferred GRNs in Figure 4 and S5. For instance, measuring matrix similarity (when appropriate) would help understanding how predicted GRN compare. I understand authors attempt to do so by focusing on validated interactions and computing the fraction of correctly inferred interactions (FCI) but I think a measurement of the overall similarity (eg. Pearson correlation) would add on this.

Another comment regards the dependency between Correlation Matrices Distance (CMD) and FCI, shown in Figure 5. I understand that IGNITE GRN that maximise FCI are not the same that minimise CMD. However, it looks like GRN that maximise FCI have higher value in terms of biological information. I wonder whether optimization for one or the other metric could be left to the end user as a tunable parameter.

Authors should discuss why the expression of some genes does not follow the expected trends (Fig 1C vs Fig S1A). Out of the 24 genes they select for their analysis, at least four do not follow the expected trends: Sox2, according to literature, is a Naive gene, however, in Figure 1C its gene expression pattern is more similar to Formative late genes. Other genes with similar "unexpected" patterns are Zic3, Etv4 and Sall4.

Are the suggested experiments realistic in terms of time and resources? It would help if you could add an estimated cost and time investment for substantial experiments.

I think suggested experiments are doable as long as authors get publicly available data, i.e. the in-house dataset they generated for this study is enough to show applicability. For example datasets analysed in SCODE paper (https://doi.org/10.1093/bioinformatics/btx194) could be used as second benchmark. The point of applying the tool to another dataset is to show how it generalises across different biological systems, experiments and, potentially, sequencing technologies. - Are the data and the methods presented in such a way that they can be reproduced?

The methods section is really clear. To enable reproducibility both raw scRNA-seq data, the IGNITE source code and code written to benchmark it should be released in the public domain in appropriate repositories (eg. ENA, GitHub, Binder etc). - Are the experiments adequately replicated and statistical analysis adequate?

Yes.

Minor comments

• Specific experimental issues that are easily addressable.

Related to the Sox2 expression pattern is the binarization shown in Figure 2D. How is it possible that Sox2 is always marked as active? Could the authors clarify how these outlier behaviours emerge and propose mitigation strategies, if any?

In section 5.11.2 it is unclear if xi are in log scale or not. Since the model starts from binarized, log transformed expression values, should not generated ones be in the same scale as the input? - Are prior studies referenced appropriately?

Yes, referencing is clear. - Are the text and figures clear and accurate?

Yes, figures appear to be clear, readable and well documented both in captions and main text. - Do you have suggestions that would help the authors improve the presentation of their data and conclusions?

Section 3.3 could be improved by better describing experimental datasets. Only in the methods section it is clearly stated that experimental data for single KO experiments were retrieved from the literature.

Check typesetting:

• parenthesis missing in Eq. 1
• Leftover $ in section 3.1
• Parenthesis missing in Section 3.3
• Misplaced comma in section 5.2.1

Significance

• Describe the nature and significance of the advance (e.g. conceptual, technical, clinical) for the field.

The paper presents a method to infer GRN from scRNA-seq data alone. Applications include GRN prediction and their perturbations. This paper represents a technical advance in the field as it is the first application of the inverse Ising problem GRN inference. - Place the work in the context of the existing literature (provide references, where appropriate).

The paper itself presents the landscape of GRN inference tools using scRNA-seq data: SCODE, MaxEnt and CELLORACLE. More tools exist, for instance SCENIC (https://doi.org/10.1038/nmeth.4463) mainly relies on co-expression matrices. Other tools exist but require additional data types e.g. GRaNIE and GRaNPA (https://doi.org/10.15252/msb.202311627) leverage on physical interaction data (ATAC-seq, ChIP-seq). Similarly DeepFlyBrain uses deep neural networks to infer eGRN in Drosophila (https://doi.org/10.1038/s41586-021-04262-z). The value of tools like IGNITE and its competitors is that they do not require additional data types, which, in turn, helps in controlling experimental costs. - State what audience might be interested in and influenced by the reported findings.

The paper might be of interest to biologists interested in regulation of gene expression. The tool might turn out to be useful in planning experimental work by guiding the choice of perturbations to introduce in experimental systems. - Define your field of expertise with a few keywords to help the authors contextualize your point of view. Indicate if there are any parts of the paper that you do not have sufficient expertise to evaluate.

I am a computational biologist.

I have no sufficient expertise to evaluate the mathematical details of the method.

Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

Learn more at Review Commons

---

Reply to the reviewers

REVIEWER 1

Reviewer #1 Evidence, reproducibility and clarity: Nornes et al. have generated a cohort of arterial enhancers based on in silico analysis and validation with transgenic lines in both zebrafish and mice. They utilized publicly available datasets for chromatin marks, including ATAC-seq on endothelial cells either from cell culture or isolated from mice, as well as EP300 binding, H3K27Ac, and H3K4Me1. Focusing on eight arterial-expressed genes, they identified a putative enhancer region marked by at least one enhancer feature. After validating the activity of these enhancers in zebrafish and mice, the authors assessed the regulatory pathways upstream of these genes. Using ChIP-seq and Cut&Run for key endothelial transcription factors, they discovered that binding sites for SoXF and ETS factors are shared in arterial enhancers, whereas binding sites for Notch, MEF2, and Fox are present only in the subset of identified enhancers. Together this study provides an arterial enhancer atlas that allows further characterisation of regulatory network behind endothelial cell identity.

__Reviewer #1 Major Comment 1: __The authors have assessed 15 enhancers for arterial-venous specificity, by assessing the expression in DA, ISV, cardinal and ventral veins at 2 dpf. Interestingly there is a clear difference in the expression patterns of these enhancers in the zebrafish axial vasculature, especially seen at the level of ISV. The co-localization of the enhancer expression in the endothelium was done using endothelial marks expressed in both venous and arterial EC (kdrl). To fully distinguish if the expression is venous or arterial endothelial compartment colocalization with Tg expressed in arterial (flt1) or venous (lyve1) EC would be informative.

RESPONSE: We agree with the reviewer that a more detailed description of arterial-venous specificity of each enhancer could be included. In the original manuscript, the expression pattern of each enhancer within the vasculature was primarily assessed at 2 days post fertilization (dpf) in Fig 1-2. This identified arteries using direction of blood flow and available descriptive information, as arterial development in 2pdf zebrafish is very stereotypical and already well characterized.

__REVISION (PLANNED): __The original Figure 3A includes a more detailed assessment of arterial-venous specificity at 3dpf for four arterial enhancers (Cxcr4+135, Cxcr4+151, Gja5-78 and Gja5-7, chosen as enhancers representing the four types of expression patterns seen). We will now extend this more detailed analysis to all arterial enhancer:GFP lines. This analysis uses kdrl-mCherry to mark the entire vasculature, comparative to the expression of the arterial enhancers (GFP). This allows us to clearly identify the intersegmental arteries (as opposed to intersegmental veins) by looking for direct connection to the dorsal aorta, and by assessing the direction of blood flow within these vessels. This analysis is done at 3dpf to give time for the intersegmental arteries to acquire identity and connect definitively with the dorsal aorta, and for the diminishment of any GFP expression originating from the initial sprouting from the dorsal aorta. By extending this analysis to the other arterial enhancer zebrafish lines shown in Figure 2, we will be able to more clearly classify the activity of each enhancer within different vascular beds. This information will also be recorded in a new Table better detailing the timing and specificity of activity of each enhancer.

We chose not to use arterial or venous "marker lines" (e.g. Flt1:reporter or Lyve1:reporter) for the simple reason that these are also enhancer:GFP transgenes, and therefore are not necessarily definitive of the arterial or venous lineage per se (e.g. Flt1:GFP expression is controlled by the transcription factors binding the Flt1 enhancer in the same way that Cxcr4+135 and the others are, with the added caveat that the transcriptional regulation of the Flt1 and Lyve enhancers are not well defined). We felt that morphological determination based on direct connections and blood flow direction was therefore more accurate.

__Reviewer #1 Major Comment 2: __In addition, it is striking that cxc4+135 drives the expression in nearly every ISV as cxcl12+269 only every other. Similarly, not all the enhancers are enriched in the DA to the same level. Is there biological significance to this? could authors discuss these results further? The pattern of expression of the unc5b-identified enhancer is also striking, does this reflect the known roles of unc5b in the vascular formation?

__RESPONSE: __We agree, the diversity of enhancer expression patterns within the arterial compartment is notable, and really very interesting. The variations in enhancer expression pattern must be largely influenced by the transcription factor motifs within each enhancer, as these patterns were seen in both transient and stable transgenic zebrafish and therefore largely independent of chromatin integration location.

__REVISION (PLANNED): __The extension of Figure 3A to all enhancer lines (see previous comment) will permit us to more clearly classify the activity of each arterial enhancer within different beds and at different time points. Currently there were no clear links between a particular transcription factor motif/binding and expression pattern, something that is discussed briefly in the original Results and Discussion sections. However, the expansion of Figure 3A to all enhancers, and the creation of a Table summarizing this more systematically will make the link (or lack of one) between expression patterns within the arterial tree and TF motifs easier to appreciate and discuss.

__Reviewer #1 Major Comment 3: __The final part of the paper focuses on defining the presence of "deeply conserved" transcription factor binding sites (TFSB), defined as TFBS that are as conserved as the enhancer sequence surrounding them. In literature, the term 'deep conservation' refers to evolutionary conservation (genomic sequence preservation) in a wide range of species. Therefore, the additional classification presented by the authors based on the surrounding sequence is not clear. As, the KLF motifs in the Ece1in1, which is conserved between mouse and human, are defined as "deeply conserved". However, the FLK motif in the following enhancer, Flk1in10 (one line below), gets classified as non-deeply conserved, despite also being conserved between mouse and human. Thus, in the current form, there is a contradiction in the way the authors use the term 'deeply conserved' and the accepted meaning of this term. To avoid confusion, it would be important to revise this nomenclature.

RESPONSE: We agree that this nomenclature should be revised. Our aim was to develop a standard approach to transcription factor motif analysis that could be applied to enhancers regardless of conservation levels and size, and easily replicated by others. Because not all functional transcription factor motifs within enhancers are necessarily conserved between species, we were careful to label both conserved and non-conserved motifs for each TF examined. Nonetheless, extra emphasis was placed on motifs with confirmatory TF binding evidence (e.g. ChIP-seq/CUT&RUN), and those conserved at the same depth as the surrounding sequence. This was because our previous work on endothelial enhancers clearly indicates that these motifs are far more likely to play a key role in regulation. However, the reviewer is correct to note that referring to such motifs as "deeply" conserved could be misinterpreted.

REVISION (COMPLETED): We have altered our nomenclature. This is explained in the relevant Results sections: "Because the level of conservation of motifs can often be an indication of their importance to enhancer activity, we classified each motif into three categories: strongly conserved (motif conserved to the same depth of the surrounding sequence), weakly conserved (motif conserved in orthologous human enhancer but not to the same depth as the surrounding sequence) and not conserved (motif is not conserved within the orthologous human sequence)".

Two enhancers (Unc5b-57 and Cdh1-1) were only conserved human-mouse, therefore each TF motif within these enhancers could be annotated as both weakly and strongly conserved. As the reviewer noted, this does create confusion. We have now adjusted Figure 5 to use a distinct shape for motifs for which no distinction between weak and strong motif can be made. This does not cover Ece1in1, which is conserved human-mouse-tenrec but was erroneously originally labelled human-mouse only. This error has been corrected.

__Reviewer #1 Minor Comment 1: __Details on how the corresponding non-coding regions between mice and humans were established are missing, what alignment tool was used?

RESPONSE AND REVISION (COMPLETED): This information has now been included in the relevant Results section: "Orthologous human enhancer sequences were identified for every enhancer using the Vertebrate Multiz Alignment & Conservation Track on the UCSC genome browser"

__Reviewer #1 Minor Comment 2: __Not sufficient details are provided for the re-analysis of siRNA data. E.g., which clustering method was used? How the clusters were assigned to cell identities?

RESPONSE AND REVISION (COMPLETED): The details regarding the re-analysis of scRNA data has been expanded in the Methods sections: "Publicly available E12 and E17.5 scRNA-seq data from EC isolated from BmxCreERT2;RosatdTomato lineage traced murine hearts54 was obtained from GEO (GSE214942) prior to processing FASTQ files with the 10X Genomics CellRanger pipeline (V7.0.0). RNA-seq reads were aligned to the mm10 genome reference downloaded from 10X Genomics with the addition of the TdTomato-WPRE sequence. Exclusion of low quality cells with either a UMI count >100,000, total gene count 10%) was performed using Scater55. Data normalisation was performed using the MultiBatchNormalisation method prior to merging of TdTomato positive and negative datasets from individual timepoints. The top 2000 most highly variable genes (excluding mitochondrial and ribosomal genes) in the merged datasets were identified using the Seurat FindVariableFeatures method and utilised to calculate principal component analysis (PCA). Normalised data was scaled using the ScaleData function. Cell clustering was performed using the standard unsupervised graph-based clustering method implemented within Seurat (V4)56. Clusters were visualised in two dimensions using UMAP based non-linear dimensional reduction following the standard Seurat (V4) workflow49. Identified clusters were assigned identities based on marker genes shown to be differentially expressed between populations previously identified in the original study47. Key markers include Npr3 (endocardial), Fabp4 (coronary vascular endothelial), and Nfatc1 (valvular endothelial). The E12.5 sinus venosus EC cluster was assigned based in Aplnr as previously described54. Arterial and venous EC clusters in the E17.5 datasets were annotated based on their enriched expression of Gja5 and Nr2f2, respectively."

__Reviewer #1 Minor Comment 3: __Details about the first HOMER analysis (in the assessment of transcription factor motifs and binding patterns at arterial enhancers) seem to be missing from the methods section.

RESPONSE AND REVISION (COMPLETED): This has been included in the methods: "Analysis of overrepresented motifs within our validated arterial enhancer cohort was performed with HOMER's findMotifsGenome tool using the full validated region of the arterial enhancers. The analysis used the hg38 masked genome and otherwise default settings for all other parameters including randomly selected background regions".

__Reviewer #1 Minor Comment 4: __Pg 12: "For ETS, 23/23 arterial enhancers contained at least one conserved motif (all "deeply" conserved to the same depth as the surrounding enhancer, see S7)". Is it S8, where conservation is indicated?

__ ____RESPONSE AND REVISION (COMPLETED):__ We have corrected this error in the text - no figure actually needed to be referenced here as the previous sentence contained the full list of relevant figures to this statement (Table 2 and Figures 5 and S9, previously called S8, are the places to see this information).

__Reviewer #1 Minor Comment 5: __Figure 1 and 2 for non-zebrafish readers it would be useful to indicate in Figures 1 and 2 the non EC expression that can be observed in the embryos.

RESPONSE AND REVISION (COMPLETED): In addition to arterial expression, a number of the enhancer:GFP transgenes also showed GFP expression within the neural tube. In addition, some transient transgenic embryos also showed ectopic expression in muscle fibres. These have now been indicated on the images in Figure 1 and 2.

__Reviewer #1 Minor Comment 6: __Table S1: Please, indicate in the legend what the asterisk in the H DNAseI column stands for

RESPONSE AND REVISION (COMPLETED): The asterisk indicates where DNaseI hypersensitivity is also seen in multiple non-EC lines. This explanation has been added to the legend.

__Reviewer #1 Minor Comment 7: __Figure S8: The phrasing "conserved to animal" in Figure S8 is misleading. There is no difference in something being conserved to tenrec or manatee, as both are Afrotherians. Hence, the data show that both Efnb2-141 and Ephb4-2 were present in the common ancestor of Afrotherians and humans, namely the ancestor of all placentals. Instead, it would be good to indicate the phylogenetic group for which the presence of the enhancer can be inferred (in this case, Placentalia).

__RESPONSE: __Whilst I appreciate the point, it is the exact sequence that is important here - obviously tenrec and manatee are similar species but still contain differences in nucleotide sequences. The information about conservation leads the reader to the exact species with which the comparison is being made. We tried to restrict this to just one species per phylogenetic group (e.g. tenrec, opossum, chicken, zebrafish) but occasionally this was not possible.

Reviewer #1 Significance

To date, a systematic approach to identifying the regulatory networks driving endothelial cell identity is missing. This study provides important datasets and validation of enhancers involved in arterial gene expression and the associated transcription factors. Although this is only the tip of the iceberg, this work represents a significant milestone in the systematic understanding of how arterial gene expression is regulated. Overall, this study offers a powerful resource for understanding arterial gene regulation and conducting genome-wide studies of arterial enhancers.

__RESPONSE: __We thank the reviewer for these kind words. Whilst we agree this is only a very small snapshot of all the arterial enhancers involved in gene regulation, we would like to stress that not only is this a massive increase to what has been known previously, but is also deliberately focused on the genes used to define arterial identity during development and in the adult, therefore these enhancers by themselves form an extremely valuable dataset with which to study the key factors driving arterial differentiation and identity.

__ __

---

REVIEWER 2

__Reviewer #2 Evidence, reproducibility and clarity: __In this work, Nornes and collaborators have described a cohort of arterial enhancers that drive gene expression in arteries and not in veins. The paper is very well written and it is very informative. The authors have used in silico models to identified the potential artery enhancers and then used different developmental in vivo systems, zebrafish and mice, to validate their findings. Finally, the authors have explored what transcription factors may be binding the identified enhancer sequences and thus, drive arterial gene expression. I would like to congratulate the authors for this work that it has been a pleasure to read and review.

Reviewer #2 Major Comment 1: In their identification of enhancers, the authors consider a candidate every enhancer that has a putative mark in both mouse and human. Nevertheless, all the human data comes from in vitro analysis. Considering how much cell culture affects endothelial cell identity, inducing effects like EndoMT, would this have any effect on the enhancer selection? Would it be possible to search any human in vivo data? Would this allow for even stronger and more relevant sequences?

__RESPONSE: __We agree that the use of human endothelial cells in culture raises some potential issues. However, we stress that the mouse EC enhancer marks, which played a key role in defining putative enhancers, come from in vivo analysis (E11 embryos, P6 retina and adult aorta), limiting the potential for significant impact from cell culture-induced issues. Whilst we would have enthusiastically incorporated human in vivo data had it been available, our approach was still indisputably successful at identifying arterial enriched/specific enhancers.

We consider it unlikely that culture/identity-related problems with human cultured ECs led to a significant undercount of enhancers, in part because comparatively few regions with enhancer marks in mouse in vivo ECs were excluded due to the absence of human enhancer marks. In fact, Cxcr4, Cxcl12, and Gja5 were poorly transcribed in the human cell lines studied here and consequently only enhancer marks in mouse were used to define putative enhancers for these three genes (this is clearly stated in the Results section). If a similar rational had applied to the remaining five genes, only an additional six putative enhancers would have been tested (one for Gja4, two for Nrp1 and three for Unc5b). However, we felt it made sense to include analysis of human enhancer marks for these five genes, as all were expressed in the human ECs used (as indicated by H3K1Me3 and DNaseI hypersensitivity at promoter regions) and orthologous human enhancers were identified for all. Additionally, our retrospective analysis of previously described mammalian in vivo-validated EC enhancers (Table S1 in the original manuscription, including eight arterial enhancers) found that all 32 were marked by at least one enhancer mark in human samples (1/32 did not contain mouse enhancer marks). We also tested eleven regions that did not reach our putative enhancer threshold, including five with only mouse marks. None of these directed expression in transgenic analysis.

Reviewer #2 Major Comment 2: The human data comes from vein endothelial or microvasculature endothelial cells. Specially because some of the enhancers identified by the authors drive also vein expression, could the authors discriminate whether this is due to the identification coming from vein cells. Is there available data from HAECs? Would this not be conceptually more correct that using vein endothelial cells data? This should be at least discussed in the paper.

__RESPONSE AND REVISION (COMPLETED): __We have now included a comparison with enhancer marks from HAECs, telo-HAECs and HUAECs as a new Figure S5. The enhancer marks seen in these cells were very similar to those in the HUVEC and microvascular cells already surveyed. Had enhancer marks within HAECs/telo-HAEC/HUAECs been included as a human enhancer mark in our initial survey, it would have been unlikely to have altered our analysis, although we agree it would have made it more conceptually correct. We chose not to go back and engineer this into our original enhancer selection rational however as we felt it would be intellectually dishonest. A paragraph has been added to the Results section about this analysis.

Reviewer #2 Major Comment 3: Although the authors use the mouse embryo to further validate their finding beyond the zebrafish, the expression are a bit different. While on the fish the enhancers label smaller vessels of arterial identity, in the mouse, only bigger arteries are marked. Is this defined by the time of the analysis?

__RESPONSE: __This experiment was conducted to demonstrate that these enhancers were arterial enriched in both zebrafish and mouse transgenesis, and feel this is clearly shown by the current data. Whilst I do not really agree that the expression pattern is different (for example, the Gja5 enhancers are more restricted to the major arteries in both zebrafish and mouse, compared to the more widely expressed Efnb2-333), this is challenging to ascertain at a single time-point in a transient transgenic mouse assay. Whilst it would be potentially interesting to better assess the activity of these enhancers over time in mice, we consider this a lengthy experiment (multiple stable lines would need to be established and characterized for each enhancer) which would not add particular benefit to this paper.

Reviewer #2 Major Comment 4: The analysis of the enhancers is only done during development. Is the activity of these enhancers maintained through live or only important for artery vs vein determination? Is the expression of the different enhancer reporters maintained into adulthood?

RESPONSE AND REVISION (PLANNED): We agree this would be interesting to ascertain. We plan to examine the activity of enhancer:GFP activity in adult fish fins (which are accessible even without crossing into a casper background, which is beyond the timescale of this project) in the fully revised version of this paper. We have already conducted a feasibility study on four arterial enhancers:GFP lines (Gja5-7:GFP, Gja5-78:GFP, Gja4+40:GFP and Efnb2-333:GFP), which found that all four were still active, and arterial-specific, in the adult.

Reviewer #2 Significance

This is a very well done study with potential interest for vascular biologists, in particular to those interested in the determination between artery and veins in a context of development. It advances our knowledge on the field of vascular biology as it not only proposes potential enhancers but also goes on to validation of the enhancers. Nevertheless, it is important to note that some of this enhancers have been identified from in vitro human data. In vitro culture of endothelial cells affects their cellular identity and thus, this study may have underscored many potential enhancers.

REVIEWER 3

__Reviewer #3: Evidence, reproducibility and clarity: __This manuscript by Nornes et al analyzed multiple published databases and identified a group of putative enhancers for 8 selected non-Notch arterial genes in mouse and human ECs. These enhancers were cloned and screened in fish embryos to test their effect in driving GFP reporter expression, which narrowed down a cohort of enhancers for further testing of expression activities in mouse embryonic arteries. The authors then analyzed the sequences of these enhancers, and identified binding motifs of ETS, SOX-F, FOX and MEF2 family TFs and Notch transcription regulator RBPJ commonly present in closed proximity in these arterial enhancers, suggesting interaction between these TFs in determination of arterial identity.

Reviewer #3 Major Comment : This study provides an enormous amount of bioinformatic data analysis and screening results in transgenic fish and mouse models, which led to the discovery of a group of arterial enhancers and TFs binding motifs essential in regulating arterial identity.

Reviewer #3 Other Comments ____1: Choice of arterial genes is slightly biased. Acvrl1/Alk1 is not enriched in arterial ECs. Sema3G, which is highly expressed in arterial ECs, is missing. UNC5B is enriched in arterial ECs but also expressed by sprouting ECs (PMID: 38866944).

__RESPONSE: __When we started this project, scRNA-seq datasets in the developing vasculature were less available. Consequently, we initially based our choice of genes on data from Raftrey et al., Circ Res 2021 (available earlier on bioRxiv), which was focused on mouse coronary arterial ECs at the timepoints that arteries differentiate. This found Acvrl1 to be arterial enriched (not a novel observation, many publications treat Acvrl1 as arterial specific or arterial-enriched) and did not list Sema3g. We also considered a wider dataset from mouse and human mid-gestation embryos when available (Hou et al., Cell Research 2022). However, it is important to note that we did not aim to investigate every arterial-enriched gene, rather to use these datasets to help identify loci associated with gene expression patterns which indicated a high likelihood of containing arterial enhancers active during arterial differentiation.

Sc-RNAseq data from both Raftery et al., and Hou et al., indicated that arterial ECs are subdivided into two groups, reflecting maturity but also potentially slightly different developmental trajectories. The genes studied here were therefore selected to evenly cover both subgroups, with Acvrl1, Cxcl12, Gja5 and Nrp1 primarily restricted to the mature arterial EC subgroup, while Cxcr4, Efnb2, Gja4 and Unc5b were also expressed in the less mature/arterial plexus/pre-arterial EC subgroup. It is notable that genes within the latter subgroup are also associated with angiogenic/sprouting ECs (Dll4 also belongs to this subgroup), which likely indicates biological links between angiogenesis and arterial identity rather than a problem in gene choice and specificity.

__REVISION (COMPLETED): __This is already discussed in the Results section (angiogenic expression of arterial genes is discussed within the MEF2 and RBPJ sections) and in the Discussion (paragraph 2, referring to different expression patterns within arterial ECs). However, we have now edited the relevant Results section to better explain gene selection: "It is therefore clear that a better understanding of the regulatory pathways directing arterial differentiation requires the identification and characterization of a larger number of arterial enhancers directing the expression of key arterial identity genes. To identify a cohort of such enhancers, we looked in the loci of eight non-Notch genes: Acvrl1(ALK1) Cxcr4, Cxcl12, Efnb2, Gja4(CX37), Gja5 (CX40), Nrp1 and Unc5b. Although not a definitive list of arterial identity genes, single cell transcriptomic analysis indicates these genes are all significantly enriched in arterial ECs4,20, and are commonly used to define arterial EC populations in mouse and human scRNAseq analysis4,5,20,54. Additionally, single-cell transcriptomic data indicates that arterial ECs can be divided into two subgroups4,20. The genes selected here are equally split between subgroups (Acvrl1, Cxcl12, Gja5 and Nrp1 from the mature arterial EC subgroup, Cxcr4, Efnb2, Gja4 and Unc5b from the less mature/arterial plexus/pre-arterial EC subgroup)4,20. We did not exclude genes also implicated in angiogenesis/expressed in sprouting ECs, as these genes formed that vast majority of those associated with the less mature EC subgroup".

Reviewer #3 Other Comments ____2: Exclusion of Notch genes. Although the reason for choosing non-notch genes and excluding notch genes for screening is addressed in this paper, it would be interesting to examine how the arterial enhancers identified in this study are present in the Notch genes, especially Dll4 (enriched in arterial and sprouting ECs) and Jag1 (enriched in arterial ECs).

__RESPONSE: __Previous work from our lab and others has already examined arterial enhancers for Notch pathway genes. We already included these enhancers in all our later analysis (Figure 5-6 and relevant supplemental figures), including analysis of TF motifs.

Reviewer #3 Other Comments ____3: SoxF family TFs. Among the 3 members of SoxF TFs, only Sox17 and Sox7 were assessed. Though not specific, Sox18 is highly expressed in the arteries. On the contrary, Sox7 is highly expressed in the vein and shows weak expression in arterial ECs (PMID: 26630461).

__RESPONSE AND REVISION (PLANNED): __We agree. We will include assessment of SOX18 binding in our final revised manuscript. An antibody for this analysis has been identified already.

Reviewer #3 Other Comments ____4: Minor inaccuracy in Intro/paragraph 3: though sox17 is reported as indispensable for arterial specification (PMID: 24153254), losing a single SoxF factor does not seem to completely compromise the arterial program (PMID: 24153254, PMID: 26630461). A combined loss of Sox17/18, or Sox 7/17/18, seems to do the job (PMID: 26630461).

__RESPONSE: __We have altered this section: "The evidence linking SOXF transcription factors to arterial differentiation is more extensive, with loss of either SOX17 (the SOXF factor most specific to arterial ECs) or SOX7 resulting in arterial defects21-24. Whilst losing a single SOXF factor does not entirely compromise the arterial program, arterial differentiation appears absent after compound Sox17;Sox18 and Sox7;Sox17;Sox18 deletion, although this occurs alongside significantly impaired angiogenesis and severe vascular hyperplasia21-24. PMID 24153254 is reference 23, PMID 26630461 is reference 24.

Reviewer #3 Other Comments ____5: Fig.4 e14.5 mouse embryos. If the observation aims to assess the dorsal aorta, it would be better to use mouse embryos at mid-gestation (e9.5-10.5), when the paired DAs are formed with arterial identity but haven't been remodelled and fused as one single aorta. The morphological data in this figure would be better to show the colocalization of LacZ expression and an arterial marker (e.g. Sox17) using immulfluorescence staining instead of purely lacZ.

RESPONSE: This experiment was primarily conducted to demonstrate that our enhancers were arterial enriched in both zebrafish and mouse transgenesis, and feel this is clearly shown with the e14.5 transgenic embryos originally shown. We chose e14.5 because it matched the timepoints used for the single cell transcriptomics first used to select the target arterial identity genes, and feel it is a good match to 2-3 dpf zebrafish in terms of arterial differentiation mechanisms. We agree that E9-10 would have also been an additional useful timepoint, but we do not have the resources to generate this data nor consider it essential for the conclusions of our work here.

__REVISION (PLANNED): __We are unable to perform immunofluorescence in the e14.5 transgenic embryos due to the fixation and staining solutions used for X-gal staining (which was done by an external company and could not be altered), but agree additional information is needed to demonstrate arterial endothelial specificity. We will therefore expand the analysis of sectioned embryos (currently restricted to just the Efnb2-333:LacZ transgene) to all enhancers shown in Figure 4. This analysis has some limitations due to infiltration of the X-gal solution to deeper tissues, but is anticipated it will clearly show enhancer activity in arterial endothelial cells rather than venous ECs or smooth muscle cells.

__Reviewer #3 (Significance (Required)): __This novel work establishes an important foundation for future understanding of how TFs may interact to determine arterial specification.

Other revisions

In addition to changes suggested by the reviewers, we also made one additional adjustment to the paper to include analysis of two additional putative enhancers (Efnb2-159 and Cxcr4+119). These were initially omitted in error yet both regions reach the standard of testable putative enhancers (noted in small changes to Figure S1 and Table S2). When tested in zebrafish transient transgenic embryos, Cxcr4+119 was inactive whilst Efnb2-159 was active in arterial endothelial cells. The relevant tables and figures have been adjusted to reflect these changes, the most significant of which are the inclusion of Efnb2-159 positive zebrafish in Figure 1 (and the necessity to create an additional supplemental Figure (S3) to accommodate the increased number of images), and analysis of Efnb2-159 transcription factor motifs/binding as part of Figure 5 and 6. No conclusions were altered by the inclusion of this additional data.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #3

Evidence, reproducibility and clarity

This manuscript by Nornes et al analyzed multiple published databases and identified a group of putative enhancers for 8 selected non-Notch arterial genes in mouse and human ECs. These enhancers were cloned and screened in fish embryos to test their effect in driving GFP reporter expression, which narrowed down a cohort of enhancers for further testing of expression activities in mouse embryonic arteries. The authors then analyzed the sequences of these enhancers, and identified binding motifs of ETS, SOX-F, FOX and MEF2 family TFs and Notch transcription regulator RBPJ commonly present in closed proximity in these arterial enhancers, suggesting interaction between these TFs in determination of arterial identity.

Major comments:

This study provides an enormous amount of bioinformatic data analysis and screening results in transgenic fish and mouse models, which led to the discovery of a group of arterial enhancers and TFs binding motifs essential in regulating arterial identity.

Other comments:

1. Choice of arterial genes is slightly biased. Acvrl1/Alk1 is not enriched in arterial ECs. Sema3G, which is highly expressed in arterial ECs, is missing. UNC5B is enriched in arterial ECs but also expressed by sprouting ECs (PMID: 38866944).
2. Exclusion of Notch genes. Although the reason for choosing non-notch genes and excluding notch genes for screening is addressed in this paper, it would be interesting to examine how the arterial enhancers identified in this study are present in the Notch genes, especially Dll4 (enriched in arterial and sprouting ECs) and Jag1 (enriched in arterial ECs).
3. SoxF family TFs. Among the 3 members of SoxF TFs, only Sox17 and Sox7 were assessed. Though not specific, Sox18 is highly expressed in the arteries. On the contrary, Sox7 is highly expressed in the vein and shows weak expression in arterial ECs (PMID: 26630461). Minor inaccuracy in Intro/paragraph 3: though sox17 is reported as indispensable for arterial specification (PMID: 24153254), losing a single SoxF factor does not seem to completely compromise the arterial program (PMID: 24153254, PMID: 26630461). A combined loss of Sox17/18, or Sox 7/17/18, seems to do the job (PMID: 26630461).
4. Fig.4 e14.5 mouse embryos. If the observation aims to assess the dorsal aorta, it would be better to use mouse embryos at mid-gestation (e9.5-10.5), when the paired DAs are formed with arterial identity but haven't been remodeled and fused as one single aorta. The morphological data in this figure would be better to show the colocalization of LacZ expression and an arterial marker (e.g. Sox17) using immulfluorescence staining instead of purely lacZ.

Significance

This novel work establishes an important foundation for future understanding of how TFs may interact to determine arterial specification.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

In this work, Nornes and collaborators have described a cohort of arterial enhancers that drive gene expression in arteries and not in veins. The paper is very well written and it is very informative. The authors have used in silico models to identified the potential artery enhancers and then used different developmental in vivo systems, zebrafish and mice, to validate their findings. Finally, the authors have explored what transcription factors may be binding the identified enhancer sequences and thus, drive arterial gene expression. I would like to congratulate the authors for this work that it has been a pleasure to read and review.

Major comments:

1. In their identification of enhancers, the authors consider a candidate every enhancer that has a putative mark in both mouse and human. Nevertheless, all the human data comes from in vitro analysis. Considering how much cell culture affects endothelial cell identity, inducing effects like EndoMT, would this have any effect on the enhancer selection? Would it be possible to search any human in vivo data? Would this allow for even stronger and more relevant sequences?
2. The human data comes from vein endothelial or microvasculature endothelial cells. Specially because some of the enhancers identified by the authors drive also vein expression, could the authors discriminate whether this is due to the identification coming from vein cells. Is there available data from HAECs? Would this not be conceptually more correct that using vein endothelial cells data? This should be at least discussed in the paper.
3. Although the authors use the mouse embryo to further validate their finding beyond the zebrafish, the expression are a bit different. While on the fish the enhancers label smaller vessels of arterial identity, in the mouse, only bigger arteries are marked. Is this defined by the time of the analysis?
4. The analysis of the enhancers is only done during development. Is the activity of these enhancers maintained through live or only important for artery vs vein determination? Is the expression of the different enhancer reporters maintained into adulthood?

Significance

This is a very well done study with potential interest for vascular biologists, in particular to those interested in the determination between artery and veins in a context of development. It advances our knowledge on the field of vascular biology as it not only proposes potential enhancers but also goes on to validation of the enhancers. Nevertheless, it is important to note that some of this enhancers have been identified from in vitro human data. In vitro culture of endothelial cells affects their cellular identity and thus, this study may have underscored many potential enhancers.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

Summary:

Nornes et al. have generated a cohort of arterial enhancers based on in silico analysis and validation with transgenic lines in both zebrafish and mice. They utilized publicly available datasets for chromatin marks, including ATAC-seq on endothelial cells either from cell culture or isolated from mice, as well as EP300 binding, H3K27Ac, and H3K4Me1. Focusing on eight arterial-expressed genes, they identified a putative enhancer region marked by at least one enhancer feature. After validating the activity of these enhancers in zebrafish and mice, the authors assessed the regulatory pathways upstream of these genes. Using ChIP-seq and Cut&Run for key endothelial transcription factors, they discovered that binding sites for SoXF and ETS factors are shared in arterial enhancers, whereas binding sites for Notch, MEF2, and Fox are present only in the subset of identified enhancers. Together this study provides an arterial enhancer atlas that allows further characterisation of regulatory network behind endothelial cell identity.

Major comments:

The authors have assessed 15 enhancers for arterial-venous specificity, by assessing the expression in DA, ISV, cardinal and ventral veins at 2 dpf. Interestingly there is a clear difference in the expression patterns of these enhancers in the zebrafish axial vasculature, especially seen at the level of ISV. The co-localization of the enhancer expression in the endothelium was done using endothelial marks expressed in both venous and arterial EC (kdrl). To fully distinguish if the expression is venous or arterial endothelial compartment colocalization with Tg expressed in arterial (flt1) or venous (lyve1) EC would be informative. In addition, it is striking that cxc4+135 drives the expression in nearly every ISV as cxcl12+269 only every other. Similarly, not all the enhancers are enriched in the DA to the same level. Is there biological significance to this? could authors discuss these results further? The pattern of expression of the unc5b-identified enhancer is also striking, does this reflect the known roles of unc5b in the vascular formation? The final part of the paper focuses on defining the presence of "deeply conserved" transcription factor binding sites (TFSB), defined as TFBS that are as conserved as the enhancer sequence surrounding them. In literature, the term 'deep conservation' refers to evolutionary conservation (genomic sequence preservation) in a wide range of species. Therefore, the additional classification presented by the authors based on the surrounding sequence is not clear. As, the KLF motifs in the Ece1in1, which is conserved between mouse and human, are defined as "deeply conserved". However, the FLK motif in the following enhancer, Flk1in10 (one line below), gets classified as non-deeply conserved, despite also being conserved between mouse and human. Thus, in the current form, there is a contradiction in the way the authors use the term 'deeply conserved' and the accepted meaning of this term. To avoid confusion, it would be important to revise this nomenclature.

Minor:

Details on how the corresponding non-coding regions between mice and humans were established are missing, what alignment tool was used?

Not sufficient details are provided for the re-analysis of siRNA data. E.g., which clustering method was used? How the clusters were assigned to cell identities?

Details about the first HOMER analysis (in the assessment of transcription factor motifs and binding patterns at arterial enhancers) seem to be missing from the methods section.

Pg 12: "For ETS, 23/23 arterial enhancers contained at least one conserved motif (all "deeply" conserved to the same depth as the surrounding enhancer, see S7)". Is it S8, where conservation is indicated?

Figure 1 and 2 for non-zebrafish readers it would be useful to indicate in Figures 1 and 2 the non EC expression that can be observed in the embryos.

Table S1: Please, indicate in the legend what the asterisk in the H DNAseI column stands for

Figure S8: The phrasing "conserved to animal" in Figure S8 is misleading. There is no difference in something being conserved to tenrec or manatee, as both are Afrotherians. Hence, the data show that both Efnb2-141 and Ephb4-2 were present in the common ancestor of Afrotherians and humans, namely the ancestor of all placentals. Instead, it would be good to indicate the phylogenetic group for which the presence of the enhancer can be inferred (in this case, Placentalia).

Significance

To date, a systematic approach to identifying the regulatory networks driving endothelial cell identity is missing. This study provides important datasets and validation of enhancers involved in arterial gene expression and the associated transcription factors. Although this is only the tip of the iceberg, this work represents a significant milestone in the systematic understanding of how arterial gene expression is regulated. Overall, this study offers a powerful resource for understanding arterial gene regulation and conducting genome-wide studies of arterial enhancers.

Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

Learn more at Review Commons

---

Reply to the reviewers

REVIEWER 1

Reviewer #1 Evidence, reproducibility and clarity: Nornes et al. have generated a cohort of arterial enhancers based on in silico analysis and validation with transgenic lines in both zebrafish and mice. They utilized publicly available datasets for chromatin marks, including ATAC-seq on endothelial cells either from cell culture or isolated from mice, as well as EP300 binding, H3K27Ac, and H3K4Me1. Focusing on eight arterial-expressed genes, they identified a putative enhancer region marked by at least one enhancer feature. After validating the activity of these enhancers in zebrafish and mice, the authors assessed the regulatory pathways upstream of these genes. Using ChIP-seq and Cut&Run for key endothelial transcription factors, they discovered that binding sites for SoXF and ETS factors are shared in arterial enhancers, whereas binding sites for Notch, MEF2, and Fox are present only in the subset of identified enhancers. Together this study provides an arterial enhancer atlas that allows further characterisation of regulatory network behind endothelial cell identity.

__Reviewer #1 Major Comment 1: __The authors have assessed 15 enhancers for arterial-venous specificity, by assessing the expression in DA, ISV, cardinal and ventral veins at 2 dpf. Interestingly there is a clear difference in the expression patterns of these enhancers in the zebrafish axial vasculature, especially seen at the level of ISV. The co-localization of the enhancer expression in the endothelium was done using endothelial marks expressed in both venous and arterial EC (kdrl). To fully distinguish if the expression is venous or arterial endothelial compartment colocalization with Tg expressed in arterial (flt1) or venous (lyve1) EC would be informative.

RESPONSE: We agree with the reviewer that a more detailed description of arterial-venous specificity of each enhancer could be included. In the original manuscript, the expression pattern of each enhancer within the vasculature was primarily assessed at 2 days post fertilization (dpf) in Fig 1-2. This identified arteries using direction of blood flow and available descriptive information, as arterial development in 2pdf zebrafish is very stereotypical and already well characterized.

__REVISION (PLANNED): __The original Figure 3A includes a more detailed assessment of arterial-venous specificity at 3dpf for four arterial enhancers (Cxcr4+135, Cxcr4+151, Gja5-78 and Gja5-7, chosen as enhancers representing the four types of expression patterns seen). We will now extend this more detailed analysis to all arterial enhancer:GFP lines. This analysis uses kdrl-mCherry to mark the entire vasculature, comparative to the expression of the arterial enhancers (GFP). This allows us to clearly identify the intersegmental arteries (as opposed to intersegmental veins) by looking for direct connection to the dorsal aorta, and by assessing the direction of blood flow within these vessels. This analysis is done at 3dpf to give time for the intersegmental arteries to acquire identity and connect definitively with the dorsal aorta, and for the diminishment of any GFP expression originating from the initial sprouting from the dorsal aorta. By extending this analysis to the other arterial enhancer zebrafish lines shown in Figure 2, we will be able to more clearly classify the activity of each enhancer within different vascular beds. This information will also be recorded in a new Table better detailing the timing and specificity of activity of each enhancer.

We chose not to use arterial or venous “marker lines” (e.g. Flt1:reporter or Lyve1:reporter) for the simple reason that these are also enhancer:GFP transgenes, and therefore are not necessarily definitive of the arterial or venous lineage per se (e.g. Flt1:GFP expression is controlled by the transcription factors binding the Flt1 enhancer in the same way that Cxcr4+135 and the others are, with the added caveat that the transcriptional regulation of the Flt1 and Lyve enhancers are not well defined). We felt that morphological determination based on direct connections and blood flow direction was therefore more accurate.

__Reviewer #1 Major Comment 2: __In addition, it is striking that cxc4+135 drives the expression in nearly every ISV as cxcl12+269 only every other. Similarly, not all the enhancers are enriched in the DA to the same level. Is there biological significance to this? could authors discuss these results further? The pattern of expression of the unc5b-identified enhancer is also striking, does this reflect the known roles of unc5b in the vascular formation?

__RESPONSE: __We agree, the diversity of enhancer expression patterns within the arterial compartment is notable, and really very interesting. The variations in enhancer expression pattern must be largely influenced by the transcription factor motifs within each enhancer, as these patterns were seen in both transient and stable transgenic zebrafish and therefore largely independent of chromatin integration location.

__REVISION (PLANNED): __The extension of Figure 3A to all enhancer lines (see previous comment) will permit us to more clearly classify the activity of each arterial enhancer within different beds and at different time points. Currently there were no clear links between a particular transcription factor motif/binding and expression pattern, something that is discussed briefly in the original Results and Discussion sections. However, the expansion of Figure 3A to all enhancers, and the creation of a Table summarizing this more systematically will make the link (or lack of one) between expression patterns within the arterial tree and TF motifs easier to appreciate and discuss.

__Reviewer #1 Major Comment 3: __The final part of the paper focuses on defining the presence of "deeply conserved" transcription factor binding sites (TFSB), defined as TFBS that are as conserved as the enhancer sequence surrounding them. In literature, the term 'deep conservation' refers to evolutionary conservation (genomic sequence preservation) in a wide range of species. Therefore, the additional classification presented by the authors based on the surrounding sequence is not clear. As, the KLF motifs in the Ece1in1, which is conserved between mouse and human, are defined as "deeply conserved". However, the FLK motif in the following enhancer, Flk1in10 (one line below), gets classified as non-deeply conserved, despite also being conserved between mouse and human. Thus, in the current form, there is a contradiction in the way the authors use the term 'deeply conserved' and the accepted meaning of this term. To avoid confusion, it would be important to revise this nomenclature.

RESPONSE: We agree that this nomenclature should be revised. Our aim was to develop a standard approach to transcription factor motif analysis that could be applied to enhancers regardless of conservation levels and size, and easily replicated by others. Because not all functional transcription factor motifs within enhancers are necessarily conserved between species, we were careful to label both conserved and non-conserved motifs for each TF examined. Nonetheless, extra emphasis was placed on motifs with confirmatory TF binding evidence (e.g. ChIP-seq/CUT&RUN), and those conserved at the same depth as the surrounding sequence. This was because our previous work on endothelial enhancers clearly indicates that these motifs are far more likely to play a key role in regulation. However, the reviewer is correct to note that referring to such motifs as “deeply” conserved could be misinterpreted.

REVISION (COMPLETED): We have altered our nomenclature. This is explained in the relevant Results sections: “Because the level of conservation of motifs can often be an indication of their importance to enhancer activity, we classified each motif into three categories: strongly conserved (motif conserved to the same depth of the surrounding sequence), weakly conserved (motif conserved in orthologous human enhancer but not to the same depth as the surrounding sequence) and not conserved (motif is not conserved within the orthologous human sequence)”.

Two enhancers (Unc5b-57 and Cdh1-1) were only conserved human-mouse, therefore each TF motif within these enhancers could be annotated as both weakly and strongly conserved. As the reviewer noted, this does create confusion. We have now adjusted Figure 5 to use a distinct shape for motifs for which no distinction between weak and strong motif can be made. This does not cover Ece1in1, which is conserved human-mouse-tenrec but was erroneously originally labelled human-mouse only. This error has been corrected.

__Reviewer #1 Minor Comment 1: __Details on how the corresponding non-coding regions between mice and humans were established are missing, what alignment tool was used?

RESPONSE AND REVISION (COMPLETED): This information has now been included in the relevant Results section: “Orthologous human enhancer sequences were identified for every enhancer using the Vertebrate Multiz Alignment & Conservation Track on the UCSC genome browser”

__Reviewer #1 Minor Comment 2: __Not sufficient details are provided for the re-analysis of siRNA data. E.g., which clustering method was used? How the clusters were assigned to cell identities?

RESPONSE AND REVISION (COMPLETED): The details regarding the re-analysis of scRNA data has been expanded in the Methods sections: “Publicly available E12 and E17.5 scRNA-seq data from EC isolated from BmxCreERT2;RosatdTomato lineage traced murine hearts54 was obtained from GEO (GSE214942) prior to processing FASTQ files with the 10X Genomics CellRanger pipeline (V7.0.0). RNA-seq reads were aligned to the mm10 genome reference downloaded from 10X Genomics with the addition of the TdTomato-WPRE sequence. Exclusion of low quality cells with either a UMI count >100,000, total gene count 10%) was performed using Scater55. Data normalisation was performed using the MultiBatchNormalisation method prior to merging of TdTomato positive and negative datasets from individual timepoints. The top 2000 most highly variable genes (excluding mitochondrial and ribosomal genes) in the merged datasets were identified using the Seurat FindVariableFeatures method and utilised to calculate principal component analysis (PCA). Normalised data was scaled using the ScaleData function. Cell clustering was performed using the standard unsupervised graph-based clustering method implemented within Seurat (V4)56. Clusters were visualised in two dimensions using UMAP based non-linear dimensional reduction following the standard Seurat (V4) workflow49. Identified clusters were assigned identities based on marker genes shown to be differentially expressed between populations previously identified in the original study47. Key markers include Npr3 (endocardial), Fabp4 (coronary vascular endothelial), and Nfatc1 (valvular endothelial). The E12.5 sinus venosus EC cluster was assigned based in Aplnr as previously described54. Arterial and venous EC clusters in the E17.5 datasets were annotated based on their enriched expression of Gja5 and Nr2f2, respectively.”

__Reviewer #1 Minor Comment 3: __Details about the first HOMER analysis (in the assessment of transcription factor motifs and binding patterns at arterial enhancers) seem to be missing from the methods section.

RESPONSE AND REVISION (COMPLETED): This has been included in the methods: “Analysis of overrepresented motifs within our validated arterial enhancer cohort was performed with HOMER’s findMotifsGenome tool using the full validated region of the arterial enhancers. The analysis used the hg38 masked genome and otherwise default settings for all other parameters including randomly selected background regions”.

__Reviewer #1 Minor Comment 4: __Pg 12: "For ETS, 23/23 arterial enhancers contained at least one conserved motif (all "deeply" conserved to the same depth as the surrounding enhancer, see S7)". Is it S8, where conservation is indicated?

__ ____RESPONSE AND REVISION (COMPLETED):__ We have corrected this error in the text – no figure actually needed to be referenced here as the previous sentence contained the full list of relevant figures to this statement (Table 2 and Figures 5 and S9, previously called S8, are the places to see this information).

__Reviewer #1 Minor Comment 5: __Figure 1 and 2 for non-zebrafish readers it would be useful to indicate in Figures 1 and 2 the non EC expression that can be observed in the embryos.

RESPONSE AND REVISION (COMPLETED): In addition to arterial expression, a number of the enhancer:GFP transgenes also showed GFP expression within the neural tube. In addition, some transient transgenic embryos also showed ectopic expression in muscle fibres. These have now been indicated on the images in Figure 1 and 2.

__Reviewer #1 Minor Comment 6: __Table S1: Please, indicate in the legend what the asterisk in the H DNAseI column stands for

RESPONSE AND REVISION (COMPLETED): The asterisk indicates where DNaseI hypersensitivity is also seen in multiple non-EC lines. This explanation has been added to the legend.

__Reviewer #1 Minor Comment 7: __Figure S8: The phrasing "conserved to animal" in Figure S8 is misleading. There is no difference in something being conserved to tenrec or manatee, as both are Afrotherians. Hence, the data show that both Efnb2-141 and Ephb4-2 were present in the common ancestor of Afrotherians and humans, namely the ancestor of all placentals. Instead, it would be good to indicate the phylogenetic group for which the presence of the enhancer can be inferred (in this case, Placentalia).

__RESPONSE: __Whilst I appreciate the point, it is the exact sequence that is important here – obviously tenrec and manatee are similar species but still contain differences in nucleotide sequences. The information about conservation leads the reader to the exact species with which the comparison is being made. We tried to restrict this to just one species per phylogenetic group (e.g. tenrec, opossum, chicken, zebrafish) but occasionally this was not possible.

Reviewer #1 Significance

To date, a systematic approach to identifying the regulatory networks driving endothelial cell identity is missing. This study provides important datasets and validation of enhancers involved in arterial gene expression and the associated transcription factors. Although this is only the tip of the iceberg, this work represents a significant milestone in the systematic understanding of how arterial gene expression is regulated. Overall, this study offers a powerful resource for understanding arterial gene regulation and conducting genome-wide studies of arterial enhancers.

__RESPONSE: __We thank the reviewer for these kind words. Whilst we agree this is only a very small snapshot of all the arterial enhancers involved in gene regulation, we would like to stress that not only is this a massive increase to what has been known previously, but is also deliberately focused on the genes used to define arterial identity during development and in the adult, therefore these enhancers by themselves form an extremely valuable dataset with which to study the key factors driving arterial differentiation and identity.

__ __

---

REVIEWER 2

__Reviewer #2 Evidence, reproducibility and clarity: __In this work, Nornes and collaborators have described a cohort of arterial enhancers that drive gene expression in arteries and not in veins. The paper is very well written and it is very informative. The authors have used in silico models to identified the potential artery enhancers and then used different developmental in vivo systems, zebrafish and mice, to validate their findings. Finally, the authors have explored what transcription factors may be binding the identified enhancer sequences and thus, drive arterial gene expression. I would like to congratulate the authors for this work that it has been a pleasure to read and review.

Reviewer #2 Major Comment 1: In their identification of enhancers, the authors consider a candidate every enhancer that has a putative mark in both mouse and human. Nevertheless, all the human data comes from in vitro analysis. Considering how much cell culture affects endothelial cell identity, inducing effects like EndoMT, would this have any effect on the enhancer selection? Would it be possible to search any human in vivo data? Would this allow for even stronger and more relevant sequences?

__RESPONSE: __We agree that the use of human endothelial cells in culture raises some potential issues. However, we stress that the mouse EC enhancer marks, which played a key role in defining putative enhancers, come from in vivo analysis (E11 embryos, P6 retina and adult aorta), limiting the potential for significant impact from cell culture-induced issues. Whilst we would have enthusiastically incorporated human in vivo data had it been available, our approach was still indisputably successful at identifying arterial enriched/specific enhancers.

We consider it unlikely that culture/identity-related problems with human cultured ECs led to a significant undercount of enhancers, in part because comparatively few regions with enhancer marks in mouse in vivo ECs were excluded due to the absence of human enhancer marks. In fact, Cxcr4, Cxcl12, and Gja5 were poorly transcribed in the human cell lines studied here and consequently only enhancer marks in mouse were used to define putative enhancers for these three genes (this is clearly stated in the Results section). If a similar rational had applied to the remaining five genes, only an additional six putative enhancers would have been tested (one for Gja4, two for Nrp1 and three for Unc5b). However, we felt it made sense to include analysis of human enhancer marks for these five genes, as all were expressed in the human ECs used (as indicated by H3K1Me3 and DNaseI hypersensitivity at promoter regions) and orthologous human enhancers were identified for all. Additionally, our retrospective analysis of previously described mammalian in vivo-validated EC enhancers (Table S1 in the original manuscription, including eight arterial enhancers) found that all 32 were marked by at least one enhancer mark in human samples (1/32 did not contain mouse enhancer marks). We also tested eleven regions that did not reach our putative enhancer threshold, including five with only mouse marks. None of these directed expression in transgenic analysis.

Reviewer #2 Major Comment 2: The human data comes from vein endothelial or microvasculature endothelial cells. Specially because some of the enhancers identified by the authors drive also vein expression, could the authors discriminate whether this is due to the identification coming from vein cells. Is there available data from HAECs? Would this not be conceptually more correct that using vein endothelial cells data? This should be at least discussed in the paper.

__RESPONSE AND REVISION (COMPLETED): __We have now included a comparison with enhancer marks from HAECs, telo-HAECs and HUAECs as a new Figure S5. The enhancer marks seen in these cells were very similar to those in the HUVEC and microvascular cells already surveyed. Had enhancer marks within HAECs/telo-HAEC/HUAECs been included as a human enhancer mark in our initial survey, it would have been unlikely to have altered our analysis, although we agree it would have made it more conceptually correct. We chose not to go back and engineer this into our original enhancer selection rational however as we felt it would be intellectually dishonest. A paragraph has been added to the Results section about this analysis.

Reviewer #2 Major Comment 3: Although the authors use the mouse embryo to further validate their finding beyond the zebrafish, the expression are a bit different. While on the fish the enhancers label smaller vessels of arterial identity, in the mouse, only bigger arteries are marked. Is this defined by the time of the analysis?

__RESPONSE: __This experiment was conducted to demonstrate that these enhancers were arterial enriched in both zebrafish and mouse transgenesis, and feel this is clearly shown by the current data. Whilst I do not really agree that the expression pattern is different (for example, the Gja5 enhancers are more restricted to the major arteries in both zebrafish and mouse, compared to the more widely expressed Efnb2-333), this is challenging to ascertain at a single time-point in a transient transgenic mouse assay. Whilst it would be potentially interesting to better assess the activity of these enhancers over time in mice, we consider this a lengthy experiment (multiple stable lines would need to be established and characterized for each enhancer) which would not add particular benefit to this paper.

Reviewer #2 Major Comment 4: The analysis of the enhancers is only done during development. Is the activity of these enhancers maintained through live or only important for artery vs vein determination? Is the expression of the different enhancer reporters maintained into adulthood?

RESPONSE AND REVISION (PLANNED): We agree this would be interesting to ascertain. We plan to examine the activity of enhancer:GFP activity in adult fish fins (which are accessible even without crossing into a casper background, which is beyond the timescale of this project) in the fully revised version of this paper. We have already conducted a feasibility study on four arterial enhancers:GFP lines (Gja5-7:GFP, Gja5-78:GFP, Gja4+40:GFP and Efnb2-333:GFP), which found that all four were still active, and arterial-specific, in the adult.

Reviewer #2 Significance

This is a very well done study with potential interest for vascular biologists, in particular to those interested in the determination between artery and veins in a context of development. It advances our knowledge on the field of vascular biology as it not only proposes potential enhancers but also goes on to validation of the enhancers. Nevertheless, it is important to note that some of this enhancers have been identified from in vitro human data. In vitro culture of endothelial cells affects their cellular identity and thus, this study may have underscored many potential enhancers.

REVIEWER 3

__Reviewer #3: Evidence, reproducibility and clarity: __This manuscript by Nornes et al analyzed multiple published databases and identified a group of putative enhancers for 8 selected non-Notch arterial genes in mouse and human ECs. These enhancers were cloned and screened in fish embryos to test their effect in driving GFP reporter expression, which narrowed down a cohort of enhancers for further testing of expression activities in mouse embryonic arteries. The authors then analyzed the sequences of these enhancers, and identified binding motifs of ETS, SOX-F, FOX and MEF2 family TFs and Notch transcription regulator RBPJ commonly present in closed proximity in these arterial enhancers, suggesting interaction between these TFs in determination of arterial identity.

Reviewer #3 Major Comment : This study provides an enormous amount of bioinformatic data analysis and screening results in transgenic fish and mouse models, which led to the discovery of a group of arterial enhancers and TFs binding motifs essential in regulating arterial identity.

Reviewer #3 Other Comments ____1: Choice of arterial genes is slightly biased. Acvrl1/Alk1 is not enriched in arterial ECs. Sema3G, which is highly expressed in arterial ECs, is missing. UNC5B is enriched in arterial ECs but also expressed by sprouting ECs (PMID: 38866944).

__RESPONSE: __When we started this project, scRNA-seq datasets in the developing vasculature were less available. Consequently, we initially based our choice of genes on data from Raftrey et al., Circ Res 2021 (available earlier on bioRxiv), which was focused on mouse coronary arterial ECs at the timepoints that arteries differentiate. This found Acvrl1 to be arterial enriched (not a novel observation, many publications treat Acvrl1 as arterial specific or arterial-enriched) and did not list Sema3g. We also considered a wider dataset from mouse and human mid-gestation embryos when available (Hou et al., Cell Research 2022). However, it is important to note that we did not aim to investigate every arterial-enriched gene, rather to use these datasets to help identify loci associated with gene expression patterns which indicated a high likelihood of containing arterial enhancers active during arterial differentiation.

Sc-RNAseq data from both Raftery et al., and Hou et al., indicated that arterial ECs are subdivided into two groups, reflecting maturity but also potentially slightly different developmental trajectories. The genes studied here were therefore selected to evenly cover both subgroups, with Acvrl1, Cxcl12, Gja5 and Nrp1 primarily restricted to the mature arterial EC subgroup, while Cxcr4, Efnb2, Gja4 and Unc5b were also expressed in the less mature/arterial plexus/pre-arterial EC subgroup. It is notable that genes within the latter subgroup are also associated with angiogenic/sprouting ECs (Dll4 also belongs to this subgroup), which likely indicates biological links between angiogenesis and arterial identity rather than a problem in gene choice and specificity.

__REVISION (COMPLETED): __This is already discussed in the Results section (angiogenic expression of arterial genes is discussed within the MEF2 and RBPJ sections) and in the Discussion (paragraph 2, referring to different expression patterns within arterial ECs). However, we have now edited the relevant Results section to better explain gene selection: “It is therefore clear that a better understanding of the regulatory pathways directing arterial differentiation requires the identification and characterization of a larger number of arterial enhancers directing the expression of key arterial identity genes. To identify a cohort of such enhancers, we looked in the loci of eight non-Notch genes: Acvrl1(ALK1) Cxcr4, Cxcl12, Efnb2, Gja4(CX37), Gja5 (CX40), Nrp1 and Unc5b. Although not a definitive list of arterial identity genes, single cell transcriptomic analysis indicates these genes are all significantly enriched in arterial ECs4,20, and are commonly used to define arterial EC populations in mouse and human scRNAseq analysis4,5,20,54. Additionally, single-cell transcriptomic data indicates that arterial ECs can be divided into two subgroups4,20. The genes selected here are equally split between subgroups (Acvrl1, Cxcl12, Gja5 and Nrp1 from the mature arterial EC subgroup, Cxcr4, Efnb2, Gja4 and Unc5b from the less mature/arterial plexus/pre-arterial EC subgroup)4,20. We did not exclude genes also implicated in angiogenesis/expressed in sprouting ECs, as these genes formed that vast majority of those associated with the less mature EC subgroup”.

Reviewer #3 Other Comments ____2: Exclusion of Notch genes. Although the reason for choosing non-notch genes and excluding notch genes for screening is addressed in this paper, it would be interesting to examine how the arterial enhancers identified in this study are present in the Notch genes, especially Dll4 (enriched in arterial and sprouting ECs) and Jag1 (enriched in arterial ECs).

__RESPONSE: __Previous work from our lab and others has already examined arterial enhancers for Notch pathway genes. We already included these enhancers in all our later analysis (Figure 5-6 and relevant supplemental figures), including analysis of TF motifs.

Reviewer #3 Other Comments ____3: SoxF family TFs. Among the 3 members of SoxF TFs, only Sox17 and Sox7 were assessed. Though not specific, Sox18 is highly expressed in the arteries. On the contrary, Sox7 is highly expressed in the vein and shows weak expression in arterial ECs (PMID: 26630461).

__RESPONSE AND REVISION (PLANNED): __We agree. We will include assessment of SOX18 binding in our final revised manuscript. An antibody for this analysis has been identified already.

Reviewer #3 Other Comments ____4: Minor inaccuracy in Intro/paragraph 3: though sox17 is reported as indispensable for arterial specification (PMID: 24153254), losing a single SoxF factor does not seem to completely compromise the arterial program (PMID: 24153254, PMID: 26630461). A combined loss of Sox17/18, or Sox 7/17/18, seems to do the job (PMID: 26630461).

__RESPONSE: __We have altered this section: “The evidence linking SOXF transcription factors to arterial differentiation is more extensive, with loss of either SOX17 (the SOXF factor most specific to arterial ECs) or SOX7 resulting in arterial defects21–24. Whilst losing a single SOXF factor does not entirely compromise the arterial program, arterial differentiation appears absent after compound Sox17;Sox18 and Sox7;Sox17;Sox18 deletion, although this occurs alongside significantly impaired angiogenesis and severe vascular hyperplasia21–24. PMID 24153254 is reference 23, PMID 26630461 is reference 24.

Reviewer #3 Other Comments ____5: Fig.4 e14.5 mouse embryos. If the observation aims to assess the dorsal aorta, it would be better to use mouse embryos at mid-gestation (e9.5-10.5), when the paired DAs are formed with arterial identity but haven't been remodelled and fused as one single aorta. The morphological data in this figure would be better to show the colocalization of LacZ expression and an arterial marker (e.g. Sox17) using immulfluorescence staining instead of purely lacZ.

RESPONSE: This experiment was primarily conducted to demonstrate that our enhancers were arterial enriched in both zebrafish and mouse transgenesis, and feel this is clearly shown with the e14.5 transgenic embryos originally shown. We chose e14.5 because it matched the timepoints used for the single cell transcriptomics first used to select the target arterial identity genes, and feel it is a good match to 2-3 dpf zebrafish in terms of arterial differentiation mechanisms. We agree that E9-10 would have also been an additional useful timepoint, but we do not have the resources to generate this data nor consider it essential for the conclusions of our work here.

__REVISION (PLANNED): __We are unable to perform immunofluorescence in the e14.5 transgenic embryos due to the fixation and staining solutions used for X-gal staining (which was done by an external company and could not be altered), but agree additional information is needed to demonstrate arterial endothelial specificity. We will therefore expand the analysis of sectioned embryos (currently restricted to just the Efnb2-333:LacZ transgene) to all enhancers shown in Figure 4. This analysis has some limitations due to infiltration of the X-gal solution to deeper tissues, but is anticipated it will clearly show enhancer activity in arterial endothelial cells rather than venous ECs or smooth muscle cells.

__Reviewer #3 (Significance (Required)): __This novel work establishes an important foundation for future understanding of how TFs may interact to determine arterial specification.

Other revisions

In addition to changes suggested by the reviewers, we also made one additional adjustment to the paper to include analysis of two additional putative enhancers (Efnb2-159 and Cxcr4+119). These were initially omitted in error yet both regions reach the standard of testable putative enhancers (noted in small changes to Figure S1 and Table S2). When tested in zebrafish transient transgenic embryos, Cxcr4+119 was inactive whilst Efnb2-159 was active in arterial endothelial cells. The relevant tables and figures have been adjusted to reflect these changes, the most significant of which are the inclusion of Efnb2-159 positive zebrafish in Figure 1 (and the necessity to create an additional supplemental Figure (S3) to accommodate the increased number of images), and analysis of Efnb2-159 transcription factor motifs/binding as part of Figure 5 and 6. No conclusions were altered by the inclusion of this additional data.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #3

Evidence, reproducibility and clarity

This manuscript by Nornes et al analyzed multiple published databases and identified a group of putative enhancers for 8 selected non-Notch arterial genes in mouse and human ECs. These enhancers were cloned and screened in fish embryos to test their effect in driving GFP reporter expression, which narrowed down a cohort of enhancers for further testing of expression activities in mouse embryonic arteries. The authors then analyzed the sequences of these enhancers, and identified binding motifs of ETS, SOX-F, FOX and MEF2 family TFs and Notch transcription regulator RBPJ commonly present in closed proximity in these arterial enhancers, suggesting interaction between these TFs in determination of arterial identity.

Major comments:

This study provides an enormous amount of bioinformatic data analysis and screening results in transgenic fish and mouse models, which led to the discovery of a group of arterial enhancers and TFs binding motifs essential in regulating arterial identity.

Other comments:

1. Choice of arterial genes is slightly biased. Acvrl1/Alk1 is not enriched in arterial ECs. Sema3G, which is highly expressed in arterial ECs, is missing. UNC5B is enriched in arterial ECs but also expressed by sprouting ECs (PMID: 38866944).
2. Exclusion of Notch genes. Although the reason for choosing non-notch genes and excluding notch genes for screening is addressed in this paper, it would be interesting to examine how the arterial enhancers identified in this study are present in the Notch genes, especially Dll4 (enriched in arterial and sprouting ECs) and Jag1 (enriched in arterial ECs).
3. SoxF family TFs. Among the 3 members of SoxF TFs, only Sox17 and Sox7 were assessed. Though not specific, Sox18 is highly expressed in the arteries. On the contrary, Sox7 is highly expressed in the vein and shows weak expression in arterial ECs (PMID: 26630461). Minor inaccuracy in Intro/paragraph 3: though sox17 is reported as indispensable for arterial specification (PMID: 24153254), losing a single SoxF factor does not seem to completely compromise the arterial program (PMID: 24153254, PMID: 26630461). A combined loss of Sox17/18, or Sox 7/17/18, seems to do the job (PMID: 26630461).
4. Fig.4 e14.5 mouse embryos. If the observation aims to assess the dorsal aorta, it would be better to use mouse embryos at mid-gestation (e9.5-10.5), when the paired DAs are formed with arterial identity but haven't been remodeled and fused as one single aorta. The morphological data in this figure would be better to show the colocalization of LacZ expression and an arterial marker (e.g. Sox17) using immulfluorescence staining instead of purely lacZ.

Significance

This novel work establishes an important foundation for future understanding of how TFs may interact to determine arterial specification.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

In this work, Nornes and collaborators have described a cohort of arterial enhancers that drive gene expression in arteries and not in veins. The paper is very well written and it is very informative. The authors have used in silico models to identified the potential artery enhancers and then used different developmental in vivo systems, zebrafish and mice, to validate their findings. Finally, the authors have explored what transcription factors may be binding the identified enhancer sequences and thus, drive arterial gene expression. I would like to congratulate the authors for this work that it has been a pleasure to read and review.

Major comments:

1. In their identification of enhancers, the authors consider a candidate every enhancer that has a putative mark in both mouse and human. Nevertheless, all the human data comes from in vitro analysis. Considering how much cell culture affects endothelial cell identity, inducing effects like EndoMT, would this have any effect on the enhancer selection? Would it be possible to search any human in vivo data? Would this allow for even stronger and more relevant sequences?
2. The human data comes from vein endothelial or microvasculature endothelial cells. Specially because some of the enhancers identified by the authors drive also vein expression, could the authors discriminate whether this is due to the identification coming from vein cells. Is there available data from HAECs? Would this not be conceptually more correct that using vein endothelial cells data? This should be at least discussed in the paper.
3. Although the authors use the mouse embryo to further validate their finding beyond the zebrafish, the expression are a bit different. While on the fish the enhancers label smaller vessels of arterial identity, in the mouse, only bigger arteries are marked. Is this defined by the time of the analysis?
4. The analysis of the enhancers is only done during development. Is the activity of these enhancers maintained through live or only important for artery vs vein determination? Is the expression of the different enhancer reporters maintained into adulthood?

Significance

This is a very well done study with potential interest for vascular biologists, in particular to those interested in the determination between artery and veins in a context of development. It advances our knowledge on the field of vascular biology as it not only proposes potential enhancers but also goes on to validation of the enhancers. Nevertheless, it is important to note that some of this enhancers have been identified from in vitro human data. In vitro culture of endothelial cells affects their cellular identity and thus, this study may have underscored many potential enhancers.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

Summary:

Nornes et al. have generated a cohort of arterial enhancers based on in silico analysis and validation with transgenic lines in both zebrafish and mice. They utilized publicly available datasets for chromatin marks, including ATAC-seq on endothelial cells either from cell culture or isolated from mice, as well as EP300 binding, H3K27Ac, and H3K4Me1. Focusing on eight arterial-expressed genes, they identified a putative enhancer region marked by at least one enhancer feature. After validating the activity of these enhancers in zebrafish and mice, the authors assessed the regulatory pathways upstream of these genes. Using ChIP-seq and Cut&Run for key endothelial transcription factors, they discovered that binding sites for SoXF and ETS factors are shared in arterial enhancers, whereas binding sites for Notch, MEF2, and Fox are present only in the subset of identified enhancers. Together this study provides an arterial enhancer atlas that allows further characterisation of regulatory network behind endothelial cell identity.

Major comments:

The authors have assessed 15 enhancers for arterial-venous specificity, by assessing the expression in DA, ISV, cardinal and ventral veins at 2 dpf. Interestingly there is a clear difference in the expression patterns of these enhancers in the zebrafish axial vasculature, especially seen at the level of ISV. The co-localization of the enhancer expression in the endothelium was done using endothelial marks expressed in both venous and arterial EC (kdrl). To fully distinguish if the expression is venous or arterial endothelial compartment colocalization with Tg expressed in arterial (flt1) or venous (lyve1) EC would be informative. In addition, it is striking that cxc4+135 drives the expression in nearly every ISV as cxcl12+269 only every other. Similarly, not all the enhancers are enriched in the DA to the same level. Is there biological significance to this? could authors discuss these results further? The pattern of expression of the unc5b-identified enhancer is also striking, does this reflect the known roles of unc5b in the vascular formation? The final part of the paper focuses on defining the presence of "deeply conserved" transcription factor binding sites (TFSB), defined as TFBS that are as conserved as the enhancer sequence surrounding them. In literature, the term 'deep conservation' refers to evolutionary conservation (genomic sequence preservation) in a wide range of species. Therefore, the additional classification presented by the authors based on the surrounding sequence is not clear. As, the KLF motifs in the Ece1in1, which is conserved between mouse and human, are defined as "deeply conserved". However, the FLK motif in the following enhancer, Flk1in10 (one line below), gets classified as non-deeply conserved, despite also being conserved between mouse and human. Thus, in the current form, there is a contradiction in the way the authors use the term 'deeply conserved' and the accepted meaning of this term. To avoid confusion, it would be important to revise this nomenclature.

Minor:

Details on how the corresponding non-coding regions between mice and humans were established are missing, what alignment tool was used?

Not sufficient details are provided for the re-analysis of siRNA data. E.g., which clustering method was used? How the clusters were assigned to cell identities?

Details about the first HOMER analysis (in the assessment of transcription factor motifs and binding patterns at arterial enhancers) seem to be missing from the methods section.

Pg 12: "For ETS, 23/23 arterial enhancers contained at least one conserved motif (all "deeply" conserved to the same depth as the surrounding enhancer, see S7)". Is it S8, where conservation is indicated?

Figure 1 and 2 for non-zebrafish readers it would be useful to indicate in Figures 1 and 2 the non EC expression that can be observed in the embryos.

Table S1: Please, indicate in the legend what the asterisk in the H DNAseI column stands for

Figure S8: The phrasing "conserved to animal" in Figure S8 is misleading. There is no difference in something being conserved to tenrec or manatee, as both are Afrotherians. Hence, the data show that both Efnb2-141 and Ephb4-2 were present in the common ancestor of Afrotherians and humans, namely the ancestor of all placentals. Instead, it would be good to indicate the phylogenetic group for which the presence of the enhancer can be inferred (in this case, Placentalia).

Significance

To date, a systematic approach to identifying the regulatory networks driving endothelial cell identity is missing. This study provides important datasets and validation of enhancers involved in arterial gene expression and the associated transcription factors. Although this is only the tip of the iceberg, this work represents a significant milestone in the systematic understanding of how arterial gene expression is regulated. Overall, this study offers a powerful resource for understanding arterial gene regulation and conducting genome-wide studies of arterial enhancers.

Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

Learn more at Review Commons

---

Reply to the reviewers

We thank the reviewers for going through our manuscript and providing valuable feedback. We are grateful to all 3 reviewers for describing our findings as important and valuable, well-designed and robust, and of value to the Parkinson's and Crohn's disease communities studying LRRK2. Below we detail a point-by-point response to the reviewers.

__Reviewer #1 (Evidence, reproducibility and clarity (Required)): __

The paper by Dikovskaya and collaborators investigated the activitiy and expression of LRRK2 in different subtypes of splenic and intestinal immune cells, taking advantage of a novel GFP-Lrrk2 knockin mouse. Interestingly, they found that T-cell-released IL-4 stimulates Lrrk2 expression in B cells. I have a few comments and suggestions for the authors. 1) Figure 1C. LRRK2 KO cells display residual Rab10 phosphorylation. Do the authors have any idea of which kinase other than LRRK2 could be involved in this phosphorylation?

As far as we are aware no other kinase is known to phosphorylate Rab10 at T73 in vivo. In vitro, recombinant Rab10 can be phosphorylated by MST3 at this site (Knebel A. et al, protocols.io https://dx.doi.org/10.17504/protocols.io.bvjxn4pn), but its relevance in vivo or in cells has not been shown. It is possible that the residual band recognised by anti-pT73 Rab10 ab in splenocytes is unspecific background, as it is mainly seen in LRRK2 KO spleen cells and not in other tissues. But to be certain that our assay assesses LRRK2-dependent Rab10 phosphorylation, we have always compared with the MLi-2 control.

2) Since there are no good antibodies for IF/IHC as pointed by the authors, the GFP-Lrrk2 mouse gives the opportunity to check endogenous LRRK2 localization, i.e. in cells untreated or treated with IL-4 or other cytokines. Also, does endogenous GFP-LRRK2 accumulate into filaments/puncta upon MLi2 inhibition? The relocalization into filaments of inhibited LRRK2 has been observed in overexpression but not under endogenous expression. This analysis would be interesting also in light of the observed side effect of type-I inhibitors.

We thank the reviewer for this suggestion. We will attempt a super-resolution microscopy using Airyscan with isolated B-cells treated with cytokine and/or LRRK2 inhibitor to address this question.

3) Figure 5. The authors need to label more clearly the graphs referring to wt mice versus GFP-Lrrk2 KI mice.

We have now labelled the panels referring to the WT mice only with "WT mice", to distinguish them from the other panels that incorporate data from both EGFP-Lrrk2 mice and their WT littermates used as a background.

They should also replace GFP-LRRK2 with GFP-Lrrk2 since they edited the endogenous murine gene.

Thank you, we have corrected it, and also the other mouse genotypes.

4) In the material and methods MLi-2 administration in mice is indicated at 60 mg/kg for 2 hr whereas in suppl. figure 5 the indicated dose is 30 mg/kg. Please correct with the actual dose used.

Thank you, we have corrected the mistake.

5) The discovery of IL-4 as a Lrrk2 activator in B cells is a very interesting and novel finding. The authors could take advantage of the GFP tag to investigate LRRK2 interactome upon IL-4 stimulation (optional). Also, is the signaling downstream of IL-4 attenuated in Lrrk2 KO cells?

We thank the reviewer for these interesting suggestions. The role of LRRK2 in IL-4 activated B-cells is currently under active research in the lab.

Reviewer #1 (Significance (Required)):

The manuscript is well designed and organized, and the experimental approaches are robust. These results are significant for the field as they add additional layers in the complex regulation and regulatory roles of LRRK2 in immunity, with implication for inflammatory disorders and Parkinson's disease.

We thank the reviewer for their positive comments and for recognising our efforts to provide some clarity to a complex field.

__Reviewer #2 (Evidence, reproducibility and clarity (Required)): __

The authors present a flow cytometry methodology to assess LRRK2 expression and pathway markers in mouse models and explore LRRK2 in splenic and intestinal immune cells. This is a highly valuable study given the emerging understanding that LRRK2 pathway activity in peripheral tissues may be of crucial importance to Parkinson's disease and Crohn's disease. P8 : the authors state that their results indicate 'that the effects of LRRK2-R1441C mutation and inflammation on LRRK2 activity represent two different parallel pathways'. This seems like an overinterpretation as pathway suggests the presence of additional partners in the pathway while R1441C is a LRRK2 intrinsic modification. The results can equally be explained by synergistic effects between both activation mechanisms (mutant and inflammation).

We agree with the reviewer, and have added this into the text. The sentence now reads "suggesting that the LRRK2-R1441C mutation and inflammation have different impacts on LRRK2 activity, either in parallel or in synergy."

Methods and experiment descriptions in results : the authors appear to use the terms anti-CD3 stimulation and CD3 stimulation interchangeably, although it is not always clear in the text that these are synonymous. This should be clarified.

We thank reviewer for pointing out this error on our part. We have made the necessary changes to always refer to the stimulation as anti-CD3.

One major observation in this paper is that LRRK2 is not detected in gut epithelial cells as previously has been reported. It would be useful to comment on any differences between the presented protocol and the previous reports, in particular relating to the antigen retrieval step. In order to reinforce the finding, it would be useful to include in situ hybridization data that could further strengthen the observations of which cellular subtypes express LRRK2 and which do not. Indeed, while the KO control shows that there is an unacceptable high non-specific staining, it does not prove absence of expression. Also, can any conclusions be made about expression of LRRK2 in neural cells of the gut? This important information on LRRK2 detection in gut should be mentioned in the abstract and highlighted in the discussion.

We thank the reviewer for pointing this out. In fact, we think the observation that LRRK2 is not detected in epithelial cells is so important that we have a separate manuscript exploring this point. Please see 1. Tasegian, A. et al.https://doi.org/10.1101/2024.03.07.582590 (2024). In this manuscript we have explored the expression of LRRK2 in human and murine intestinal epithelial cells using qPCR. Although we do not have in situ hybridization data, we believe that using both the EGFP-LRRK2 and the pRab10 flow cytometry, as well as qPCR and proteomics on selected cell types, corroborates our findings on the cell types that express LRRK2. We did not analyse LRRK2 expression in the neural cells of the gut, as the focus was on the immune cells, however we hope that others will use the tools developed here to explore this further.

The authors mention in the discussion that they 'show for the first time that eosinophils also express active LRRK2 at levels comparable to B-cells and DCs.' The relevance of this finding should be further developed. Why is this important?

We thank the reviewer for this point. We don't know how LRRK2 is important in these cells. However, as the role of LRRK2 in eosinophils and neutrophils has not yet been explored and both cell types play important roles in IBD, we think it is important to point out. We have now added a sentence to the discussion highlighting the importance of eosinophils in IBD. "Since eosinophils have recently been implicated as key player in intestinal defense and colitis(Gurtner et al, 2022), it will be interesting to evaluate LRRK2 functions in these cells."

In the isolation of lamina propria cells, what efforts were made to characterize the degree of purification of the lamina propria cells compared to cells of other gut wall layers such as epithelium, muscularis mucosa, or deeper layers? Please specify.

Isolation of lamina propria cells is a very well-established process (LeFrancois and Lycke, 'Isolation of Mouse Small Intestinal Intraepithelial Lymphocytes, Peyer's Patch, and Lamina Propria Cells.' Curr. Protocols in Immunology 2001), where we extensively wash off the epithelial layer before digesting the tissue for the LP. After the digestion the muscle and wall of the gut are still intact, so we do not get any contamination with other deeper layers. The subsets of cells we find in the LP are in line with isolations from other labs.

Minor comments Figure 5G, for the graphs indicating LRRK2 activity and LRRK2 phosphorylation, the specific measures should be specified in the graph titles to avoid any ambiguity (pT73-Rab10, pS935-LRRK2).

We have added the specifications to the new version of the figure.

Suppl figure 1 : please specify the figure label and abbreviation AF568 in the legend. Suppl figure 2 : please specify the figure label and abbreviation anti-rb in the legend

Thank you, we added the abbreviations to the legends. The Figure labels for both figures have been already included at the top of figure legends.

Reviewer #2 (Significance (Required)):

The authors present a flow cytometry methodology to assess LRRK2 expression and pathway markers in mouse models and explore LRRK2 in splenic and intestinal immune cells. This is a highly valuable study given the emerging understanding that LRRK2 pathway activity in peripheral tissues may be of crucial importance to Parkinson's disease and Crohn's disease.

We thank the reviewer for recognising the value of this study.

Reviewer #3

Evidence, reproducibility and clarity

The paper describes a set of experiments to analyse LRRK2 activity in tissues and despite it has very important findings and technical developments is largely descriptive. It does look like a collection of experiments more than a defined hypothesis and experiments to address that.

We thank the reviewer for recognising the importance of our findings and the technical developments. We agree that the paper's focus is to describe where LRRK2 is expressed in immune cells, and in which cells is it active or activated after inflammation in a hypothesis-free unbiased manner. We believe this is important data to share as a resource for the wider LRRK2 community and we will submit the manuscript as a Resource.

The flow cytometry assay of the first part is a great technical challenge and represents the establishment of a potentially very useful tool for the field. It would have been important to test other organs, either as controls or for example because of their relevance e.g. lungs. This first part is disconnected from the second part below.

We thank the reviewer for pointing out that the pRab10 assay would be useful to apply to other organs too. Since we are interested in the role of LRRK2 in IBD, we had focused on applying the pRab10 assay on intestinal tissue, with spleens also analysed as major lymphoid organ and a source of immune cells that can translocate to the gut in inflammation. We hope that the publication of this method would allow other researchers to analyse other tissues in the future.

The authors generated a new mouse KI mouse expressing EGFP-LRRK2 and show data the levels of LRRK2 expression are reduced in tissues at different degrees and established a flow cytometry assay to measure LRRK2 expression by monitoring the GFP signal. Interestingly they found that expression does not correlate with activity (as measured by phospho-Rabs). I suggest taking this part out as it breaks the flow of the paper. If data using this mouse is included, then microscopy should be included to complement the flow cytometry data. I understand the mice were used later with the anti-CD3 treatment, but it is very confusing that some experiments are done with EGFP-LRRK2 mice and others not. It does look in general like the mice do not behave as wild types and this is an important caveat. Without microscopy of the tissues or even cells (Figure 4) is hard to conclude much about these experiments.

We thank the reviewer for this point and would like to explain. It is true that in Suppl Figure 5, we show reduction of LRRK2 signal in the EGFP-Lrrk2-KI mice. However, based on immunoblotting, a significant reduction in EGFP-LRRK2 expression levels was seen only in the brain, but not in the tissues we analysed, that is the spleen and the intestine. Further, we have shown clearly using proteomics (Fig. 3D and 5E), that the GFP signal in immune cells correlates very well with the WT LRRK2 expression. Therefore, we think that the GFP signal in these mice reflects WT LRRK2 expression pattern. Further, despite the limitations of reduced kinase activity that we thoroughly describe, we think this model is very useful since no antibodies work to stain for LRRK2 in mice. We therefore respectfully disagree with this reviewer that the EGFP-LRRK2 data should be taken out, as it has proven to be an invaluable tool to measure and track changes in endogenous LRRK2 expression. Moreover, we think the fact that LRRK2 expression does not correlate with levels of activity, that is, LRRK2 is more active in some immune cells than in others, is a very important finding that evidences the cell-specific regulation of LRRK2 activity beyond its expression level.

We tried but failed to visualize the EGFP-LRRK2 signal using fluorescence microscopy in the tissue. This is most likely due to the low expression of LRRK2 (proteomics data suggests that even neutrophils express less than 9000 copies), confounded further by the high background autofluorescence in tissues, especially in the gut. We now explain the lack of tissue images from the EGFP-LRRK2 mice in the text. However, we can visualize the EGFP-LRRK2 in B cells, and we will provide these images in a revised version of the manuscript.

We have also added the following paragraph to the discussion:

"We complemented the pRab10 assay with the development of the EGFP-Lrrk2-KI reporter mouse. Although the reporter was initially designed as a fluorescent tracker for imaging LRRK2 localisation in cells and tissues, the low expression of LRRK2, combined with high and variable autofluorescence in tissues precluded its use for microscopy. Even in neutrophils, which express highest level of LRRK2 among immune cells, there are less than 9000 copies of LRRK2 per cell (Sollberger et al, 2024), making it difficult to identify localization. However, the EGFP signal was sufficient for flow cytometry-based measurements, where background autofluorescence of each cell type was taken into account and subtracted."

Then the authors show that LRRK2 expression and activity is different in different cell types and depends on inflammation. The anti-CD3 strategy to induce inflammation is very different from physiological inflammation such as sepsis and LPS stimulation, so experiments with other stimuli could be important here to contribute to the message of inflammatory trigger of LRRK2 activation and decoupling of cell type.

We thank the reviewer for this suggestion. We used the anti-CD3 model as it also causes intestinal inflammation, and mimics T-cell cytokine storms that happens in many diseases. However, for the revisions we will also test another model of inflammation as suggested, such as LPS stimulation, to measure how inflammation affects LRRK2 expression and activity.

The IL-4 data is intriguing but too preliminary. The lack of strong effect of IFN-gamma is expected as the promoter of LRRK2 in mice and humans is different and human cells responds much better with regards to LRRK2 expression after IFN-gamma stimulation.

We are confused by what the reviewer means by saying the IL-4 data is preliminary. We have shown by flow cytometry, immunoblotting, qPCR and proteomics that IL-4 induced LRRK2 expression in B-cells. So we are uncertain as to how else this can be shown. As to the effect of IFNγ on LRRK2 expression, it may indeed be that human cells respond better than murine cells. Importantly, the IL-4 ability to induce LRRK2 in B-cells is a novel and important finding, regardless of the effects of IFNγ.

Reviewer #3 (Significance (Required))

The paper describes a set of experiments to analyse LRRK2 activity in tissues and despite it has very important findings and technical developments is largely descriptive. It does look like a collection of experiments more than a defined hypothesis and experiments to address that.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #3

Evidence, reproducibility and clarity

The authors present a flow cytometry methodology to asses LRRK2 epxression and pathway markers in mouse models and explore LRRK2 in splenic and intestinal immune cells. This is a highly valuable study given the emerging understanding that LRRK2 pathway activity in peripheral tissues may be of crucial importance to Parkinson's disease and Crohn's disease.

P8 : the authors state that their results indicate 'that the effects of LRRK2-R1441C mutation and inflammation on LRRK2 activity represent two different parallel pathways'. This seems like an overinterpretation as pathway suggests the presence of additional partners in the pathway while R1441C is a LRRK2 intrinsic modification. The results can equally be explained by synergistic effects between both activation mechanisms (mutant and inflammation).

Methods and experiment descriptions in results : the authors appear to use the terms anti-CD3 stimulation and CD3 stimulation interchangeably, although it is not always clear in the text that these are synonymous. This should be clarified.

One major observation in this paper is that LRRK2 is not detected in gut epithelial cells as previously has been reported. It would be useful to comment on any differences between the presented protocol and the previous reports, in particular relating to the antigen retrieval step. In order to reinforce the finding, it would be useful to include in situ hybridization data that could further strengthen the observations of which cellular subtypes express LRRK2 and which do not. Indeed, while the KO control shows that there is an unacceptable high non-specific staining, it does not prove absence of expression. Also, can any conclusions be made about expression of LRRK2 in neural cells of the gut? This important information on LRRK2 detection in gut should be mentioned in the abstract and highlighted in the discussion. The authors mention in the discussion that they 'show for the first time that eosinophils also express active LRRK2 at levels comparable to B-cells and DCs.' The relevance of this finding should be further developed. Why is this important ?

In the isolation of lamina propria cells, what efforts were made to characterize the degree of purification of the lamina propria cells compared to cells of other gut wall layers such as epithelium, muscularis mucosa, or deeper layers? Please specify.

Minor comments

Figure 5G, for the graphs indicating LRRK2 activity and LRRK2 phosphorylation, the specific measures should be specified in the graph titles to avoid any ambiguity (pT73-Rab10, pS935-LRRK2).

Suppl figure 1 : please specify the figure label and abbreviation AF568 in the legend.

Suppl figure 2 : please specify the figure label and abbreviation anti-rb in the legend

Significance

The authors present a flow cytometry methodology to asses LRRK2 epxression and pathway markers in mouse models and explore LRRK2 in splenic and intestinal immune cells. This is a highly valuable study given the emerging understanding that LRRK2 pathway activity in peripheral tissues may be of crucial importance to Parkinson's disease and Crohn's disease.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

The paper by Dikovskaya and collaborators investigated the activitiy and expression of LRRK2 in different subtypes of splenic and intestinal immune cells, taking advantage of a novel GFP-Lrrk2 knockin mouse. Interestingly, they found that T-cell-released IL-4 stimulates Lrrk2 expression in B cells.

I have a few comments and suggestions for the authors.

1. Figure 1C. LRRK2 KO cells display residual Rab10 phosphorylation. Do the authors have any idea of which kinase other than LRRK2 could be involved in this phosphorylation?
2. Since there are no good antibodies for IF/IHC as pointed by the authors, the GFP-Lrrk2 mouse gives the opportunity to check endogenous LRRK2 localization, i.e. in cells untreated or treated with IL-4 or other cytokines. Also, does endogenous GFP-LRRK2 accumulate into filaments/puncta upon MLi2 inhibition? The relocalizaiton into filaments of inhibited LRRK2 has been observed in overexpression but not under endogenous expression. This analysis would be interesting also in light of the observed side effect of type-I inhibitors.
3. Figure 5. The authors need to label more clearly the graphs referring to wt mice versus GFP-Lrrk2 KI mice. They should also replace GFP-LRRK2 with GFP-Lrrk2 since they edited the endogenous murine gene.
4. In the material and methods MLi-2 administration in mice is indicated at 60 mg/kg for 2 hr whereas in suppl. figure 5 the indicated dose is 30 mg/kg. Please correct with the actual dose used.
5. The discovery of IL-4 as a Lrrk2 activator in B cells is a very interesting and novel finding. The authors could take advantage of the GFP tag to investigate LRRK2 interactome upon IL-4 stimulation (optional). Also, is the signaling downstream of IL-4 attenuated in Lrrk2 KO cells?

Significance

The manuscript is well designed and organized, and the experimental approaches are robust. These results are significant for the field as they add additional layers in the complex regulation and regulatory roles of LRRK2 in immunity, with implication for inflammatory disorders and Parkinson's disease.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

The paper describes a set of experiments to analyse LRRK2 activity in tissues and despite it has very important findings and technical developments is largely descriptive. It does look like a collection of experiments more than a defined hypothesis and experiments to address that.

The flow cytometry assay of the first part is a great technical challenge and represents the establishment of a potentially very useful tool for the field. It would have been important to test other organs, either as controls or for example because of their relevance e.g. lungs. This first part is disconnected from the second part below.

The authors generated a new mouse KI mouse expressing EGFP-LRRK2 and show data the levels of LRRK2 expression are reduced in tissues at different degrees and established a flow cytometry assay to measure LRRK2 expression by monitoring the GFP sugnal. Interestengly they found that expression does not correlate with activity (as measured by phospho-Rabs). I suggest taking this part out as it breaks the flow of the paper. If data using this mouse is included, then microscopy should be included to complement the flow cytometry data. I understand the mice were used later with the anti-CD3 treatment, but it is very confusing that some experiments are done with EGFP-LRRK2 mice and others not. It does look in general like the mice do not behave as wild types and this is an important caveat. Without microscopy of the tissues or even cells (Figure 4) is hard to conclude much about these experiments.

Then the authors show that LRRK2 expression and activity is different in different cell types and depends on inflammation. The anti-CD3 strategy to induce inflammation is very different from physiological inflammation such as sepsis and LPS stimulation, so experiments with other stimuli could be important here to contribute to the message of inflammatory trigger of LRRK2 activation and decoupling of cell type.

The IL-4 data is intriguing but too preliminary. The lack of strong effect of IFN-gamma is expected as the promoter of LRRK2 in mice and humans is different and human cells responds much better with regards to LRRK2 expression after IFN-gamma stimulation.

Significance

The paper describes a set of experiments to analyse LRRK2 activity in tissues and despite it has very important findings and technical developments is largely descriptive. It does look like a collection of experiments more than a defined hypothesis and experiments to address that.

Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

Learn more at Review Commons

---

Reply to the reviewers

Reviewer #1

Evidence, reproducibility and clarity:

In this work, Anandi et al. propose an ex vivo model that can be used to recapitulate the in vivo structure of the tumor microenvironment, which allows the observation of morphological and functional changes in tumor cells in a 3D context. Due to the ability of cancer cells to induce hypoxic condition within the TME, authors propose this model to tackle the study of metastasis initiation in vitro. The proposed system successfully displays an ischemic gradient with cells accessing nutrients at different rates, similarly to what happens in solid tumors in vivo. Moreover, in line with the literature, tumor cell migration and invasiveness were promoted by hypoxic conditions. Authors also show that the system could be used to study cell-cell interaction, as co-cultures of macrophages and cancer cells were successfully cultured in the system and studied in the context of tumor hypoxia.

The study proposed is interesting and timely, as cancer cell invasion remains an important area of tumor biology that needs further exploration. The methodology is well explained and proposed in a linear flow. However, the work could benefit from some improvement and changes, as well as from additional experiments. On an important note, authors do not properly refer to the current literature, as several studies on 3D culture systems/chambers have already been studied and developed to investigate the tumor microenvironment, but they are not cited nor referred to in the manuscript. Authors should refer to such literature and explain how this system is different and adds to it.

Major comments:

1. Authors propose this method to study the TME in 3D. When culturing cells with different ECM (Collagen vs. matrigel+collagen I) authors should take into consideration the effect of these materials on different cell types. It is known how collagen and matrigel can differently influence the polarization and phenotype of stromal cells (particularly in regards of fibroblasts - major components of solid cancers - e.g., PMID 21029367), therefore these points should be addressed at least in the discussion.

We completely agree with the reviewer so we added this point (and reference) to our manuscript's introduction (lines 45-46) and discussion (lines 442-445).

1. In addition to the previous comment, matrigel and collagen are also known to alter cancer cell phenotype (e.g., PMID 21029367) and this point should be taken into account.

We completely agree with the reviewer so we added this point (and reference) to our discussion in the main text (lines 442-445).

1. The need for novel 3D systems to study different aspects of the TME in vitro/ex vivo are certainly needed, however they are not inexistent. Authors should address this in the text, as the current literature already started to propose 3D models (including models involving matrigel/collagen in combination with other materials). 3D chambers (of different materials, and with different aims) are being used and designed and can be found in the literature. These works are not cited in the current study at all. For instance, Anguiano 2017; Cavo et al. 2018; Anguiano et al., 2020; Sodek et al. 2008, etc.

We agree so we have now added those references to the main text (line 56-57).

1. Even though the focus is on hypoxia and the achievement of an ischemic gradient in the chamber to allow resemblance of an in vivo tumor, the authors write in line 123 (and also in other parts of the text) that: "these results show that consumer cells in the 3MIC form ischemic gradients that can influence the local metabolic microenvironment experienced by neighboring tumor spheroids". The addition of the use of the PMDS membrane partly supports the claim, however it would be interesting to check whether this is indeed true, by measuring for example the levels of certain metabolites (e.g., glucose, glutamine, glutamate, lactate, aspartate) reached with the system, or pH levels, etc., in presence or absence of the hypoxic gradient/consumer cells.

This is an insightful question and defining the exact composition of this complex ischemic microenvironment is a major ambition of our lab, so we completely agree with the reviewer's comment. However, as the 3MIC was designed specifically for microscopy, measuring specific metabolites it is unfortunately outside its capabilities.

Having said that, and following the spirit of the reviewer's comments, we used microscopy to measure additional signs of metabolic stress. Specifically, we used fluorescent probes to detect changes in intracellular pH (pHrodo, Molecular Probes) and in Redox status (CellROX, Molecular Probes) and glucose (2-NBDG - a fluorescent D-glucose analog). As we explain below, we found exciting results from our pH measurements which led us to additional functional experiments. We are very excited about these new results, and we thank the reviewer for encouraging these experiments. These new results also provide evidence that other parameters in ischemia - and not just hypoxia - change along the 3MIC and can have an impact on tumor cells.

1. When looking at the references presented in the manuscript, authors quote too many review articles, rather than scientific articles. Given the extremely wide literature on cancer metastasis, more of these works should be quoted in this context. For example: in the introduction - text lines 27-38 - only 4 references are research articles, out of 14 references presented in that paragraph.

The reviewer is correct in pointing this out. Our intention was to use reviews on topics that are well established where citing primary research could be unfair to other contributions. But again, we agree with the reviewer, so we replaced reviews with primary research articles in multiple locations along the manuscript.

1. As authors showed successfully how macrophages and cancer cells can interact in the chamber, recapitulating cell interactions in an in vivo context, it would be very interesting to see whether different consumer cells would induce similar or different changes to the spheroids and the ischemic gradient (for instance using stromal cells or non-tumor cell lines as consumers, instead of cancer cells only), as we know how tumors are a multitude of cell subsets, each contributing to nutrient production, oxygen consumption, etc.

This is a great point. We thought about that very same point and conducted several experiments to test the combinatorial effects of different consumer cells. In broad terms, we did not observe major differences when using different consumer cells. However, we agree that this system may provide compelling opportunities to test the effect of different cell types on each other. Still, for consistency and ease, we conducted most of our experiments using the same cells in both consumers and in spheroids.

In the resubmitted version, we added an experiment where we looked at the sprouting of SVEC endothelial cells using the same cells or Lung KPs as consumers (Fig. S6A).

Minor comments:

1. Studying the early metastatic development/seeding remains a timely quest, however authors should refer to several new studies in which various mouse models are used to study metastasis from different points of view (e.g., PMID 25822788; PMID 36991128; PMID 25171411; PMID 25633981; PMID 34632412; PMID 35921474; etc). Or line 41, three reviews are quoted (refs 27-29), whilst there are several works that could be quoted on metabolism in solid tumors also in the context of metastasis (e.g., PMID 36522548; PMID: 26719539, PMID 34303764). This comment applies to the rest of the text.

We thank the reviewer for their help in processing this vast literature. We were aware of most of those works but some were new to us so thanks again! We have now added these references.

1. The order of the references is not properly presented. In the introduction, the first reference is n. 4 (text line 22), instead of it being reference 1. Moreover, the subsequent literature ref. is number 12 and not number 2. Please revise the order of the references, and position them within the bibliography from first cited to last cited in the text.

We apologize for this confusion. We have now revised all the references and we hope they are correctly formatted and numbered. The origin of this confusion may have been that we had references in the abstract thus their numbering started there rather than from the introduction. To avoid further confusions, we removed all references from the abstract.

1. Lines 98-104. It would be helpful to the reader to define here what these consumer cells are. Even though it is explained in the methods that the consumer cells are cancer cells, it is important to make it clear in the text, as it could be misleading at times.

We agree with the reviewer although we did not mean to be misleading. As mentioned above, we chose to use the same cells for both: consumers and spheroids and we have now added a new figure to illustrate this point (Fig S6A). Following the advice, we are also including additional text to make the message clearer (lines 107-109).

1. The English grammar and spelling should be revised in some parts, as well as typos and missing words throughout the text (e.g., Line 38, the word "interraction" is misspelled and should be corrected with "interaction". Line 49, the first sentence seems incomplete. Lines 68-69 should be revised as the sentences do not flow well together, probably due to a missing word. In line 77 it should be "presents". Line 341 should be "cannot be explained").

We apologize for these typos and mistakes. We have tried our best to avoid these type of errors in the new manuscript version.

Referees cross-commenting

I find the comments from the other reviewers to be in line with one another as well as with my general assessment. The major and comments of all reviewers should be addressed. The minor comments should be taken into account as well, as they would render the text and the figures more precise. I suggest that 3-6 months to complete the revision process is an appropriate time frame for the authors.

Finally, I strongly encourage the authors to add in the discussion the points and questions raised by all reviewers, as well as to improve the bibliography in terms of organisation, linearity, and state of the art.

Significance:

General assessment:

The work by Anandi et al. offers an additional tool to tackle the issue of studying the tumor microenvironment, in a 3D culture system.

The authors show a model that can be used to study tumor hypoxia in 3D, offering the possibility to study the TME in a more in vivo-like manner without turning to mice models. The development of new tools to study the TME avoiding the excessive use of animals is definitely a timely quest. In addition, the system has the potential to be applied to tackle different biological questions, as the methodology is well explained and could be suitable to many other fields of cancer biology (e.g., drug resistance or uptake). The work is overall presented in a clear way and the methodology is explained thoroughly and it has the potential to be a useful tool for the study of cancer hypoxia.

However, authors should address how their method could differently impact other cells when applied to other systems. As one major claim is the potential use of this methodology to study the TME, it should be taken into consideration how stromal cells are strongly affected by the ECM, and how certain settings or features of the system may impact such cell populations. In addition, the work does not properly refer to the current state of the art. As other studies started to propose 3D systems for the study of TME and cell-cell interactions - besides organoids - the authors should cite these works and frame their own study in a more appropriate context, pointing out differences with the current 3D chambers available, the advantages of one vs the other, and so on.

Advance: the study adds to the current literature as the study of tumor hypoxia in 3D remains a complicated issue. The interesting co-culture settings with macrophages suggests potential uses of this model to study cell-cell interactions.

Audience: the study is very methodological and offers a tool that could be used by cancer biologists - and maybe by other biology fields.

Reviewer #2

Evidence, reproducibility and clarity:

Summary

Anandi and colleagues present a manuscript describing a nice assay for exploring the progressive effect of metabolic depletion of the nutrients and oxygen on the invasion of cancer cells. This builds upon and extends a device that they previously described - MEMIC - and now enables 3D analysis of small numbers of cells. The key to their method is the inclusion of a layer of consumer cells that deplete oxygen and nutrients. Using this tool, they demonstrate that depleted environments promote invasive behavior and lower cell-cell adhesion. This is related to the nutrient-deprived and hypoxic environments found in the center of many tumors. Cellular Potts Modelling is used to explore ideas around the cooperation between reduced cell-cell adhesion and increase ECM adhesion in promoting invasion. Overall, this is a well-constructed manuscript that will be of interest to cell biologists and cancer biologists.

Major comments

I realize this work is submitted to review commons and this complicates the recommendation regarding publication. My view is that the 'more prestigious' journals would require greater mechanistic insight, but that the work could find a suitable place in other members of the review commons stable. My comments are divided into those essential for any journal and those that might be journal dependent.

We hope that the mechanistic experiments added to our new manuscript version will appeal the reviewer and merit publication in any of the review commons journals.

Essential regardless of journal

1. Many of the figures lack information about the number of spheroids analyzed and from how many biological repeats they are derived.

We have now added this information to all our experiments. This information can be found in the figures and on the figure legends.

1. The authors need to provide citations for their assertion that only gases can cross the PDMS, but not other small metabolites. They should also comment on whether the build-up of CO2 might be relevant.

We have now added the original reference where they describe PDMS's properties (Cox and Dunn, 1986).

The point raised about CO2 is very interesting, but we do not expect a buildup of this gas. When using PDMS, CO2 would not accumulate as PDMS membranes are permeable to gases - including CO2. When using glass covers, the lack of oxygen should minimize CO2 production as hypoxic cells will not be able to conduct oxidative phosphorylation and produce lactic acid instead.

1. The data on the directionality of migration when consumers are present are not significant and doesn't warrant the speculation in lines 186-189.

Following the reviewer's advice we have removed this speculation.

1. The ECM degradation in Figure 3 should be quantified.

We agree. We added additional quantifications for the gelatin degradation assay. We also highlight the quantification we already had of the ECM degradation assessed via DQ collagen. Those data can be found in the new figures 4 and S4, respectively.

1. Do the authors have evidence that the hypoxia-exposed cells are more adhesive to ECM. This is central to their Potts model and I could not locate the supporting experimental data. If not, then the Potts model should include matrix proteolysis, which they do have data about.

Again, this is a very insightful observation, and we completely understand this confusion. We think that this may part of the inherent challenge of trying to condense biological problems into analogies or "metaphors" when using physical/mathematical models.

The algorithm in a Cellular Potts model (CPM) tries to minimize the energy of the system (the entire group of cells/ECM that we are modelling). This global energy reduction is achieved by minimizing local energies in the cell-cell and cell-ECM interactions. The way the algorithm executes this minimization, is by always (probability p=1) accepting a configuration that decrease the energy while restricting the configurations that lead to higher energies (with a probability of p = e-DHT) where DH is the difference between the current and previous energy.

So, the only thing the model is really doing is to increase the likelihood that cells are in a more "comfortable" environment - i.e. that the energy from the interactions with their neighboring cells and ECM is as low as possible. For example, if cell 1 and cell 2 adhere strongly but not to cell 3, in a CPM this is modelled as a low DH between cell 1 and 2 and a higher DH with cell 3. Conversely, when people model cells better at "invading" into a new "territory" they choose a lower energy between that cell type and that type of substratum.

In other words, our CPM does not "care" whether ischemic cells invade the ECM because they create space through increased proteolysis or because they are more adherent to the ECM. These two scenarios are the same in a CPM and it is consistent with previous CPM models of similar scenarios (e.g.: PMID: 18835895, 33933478, 26436883, 23596570).

We have now reworded the description of the model on the main text, and we added an illustration hoping to make this aspect of the model clearer (Fig. S4F).

1. Is the down-regulation of E-cadherin transcriptional - i.e. is the mRNA level reduced?

This is a great question. After the reviewer posed this question, we looked at out data and we concluded E-cad's downregulation is transcriptional. Assessing local mRNA levels in the 3MIC is challenging. However, our E-Cad reporter (pHAGE-E-cadherin-RFP, addgene #79603) is a red fluorescent protein driven by the CDH1 (E-Cad) reporter. RFP levels decrease with ischemia indicating that this regulation occurs at the promoter/transcriptional level. We now added this point to the revised manuscript (lines 259-261). We thank the reviewer for this insight!

1. The title of figure 6 is misleading. The authors do not demonstrate chemoresistance in terms of cell survival or cell proliferation, which is how the term is normally used. The authors should measure cell number, proliferation, and cell viability. The data presented in the Supplementary Figure are inadequate with no quantification. The FUCCI reporter cells would be a good tool for this. Also, why use 150nM paclitaxel when the IC50 is 817nM? This seems bizarre. Lastly, there is a typo in the figure that suggest 150mM drug was used.

We apologize if these experiments caused confusion. Our intention was to look at the anti-migratory effects of Taxol-related drugs. As such, we first determined the concentrations at which the drug was lethal to our cells (this is the LD50 of ~800nM). Then, we tested if lower concentrations - which we knew where not lethal - would affect cell migration, protrusions, etc. Hence the 30-150nM range we used in our experiments.

We have now completely rewritten this section hoping that our approach is now clearer. We have also changed the title of the section and the figure legend to clarify that we are studying the effects of Taxol as anti-motility drug rather than its effects on cell survival and proliferation (now Fig. 7). Finally, we have now fixed the 150mM/150nM typo in the figure legend.

Journal dependent

1. The authors have not excluded that either changes in nutrients, or even a pro-invasive factor, produced by consumer cells are necessary for the increased invasion. They have only shown that they are not sufficient. The authors should perform a series of experiments comparing hypoxic conditions with normal media and normoxic conditions with nutrient depleted/condition media by prior culturing of KP cancer cells.

This is a great point. We actually do not want or propose to exclude this possibility. So, we have now added text to clarify this issue (lines 431-435).

In fact, we would be thrilled if there is a pro-invasive factor. If that would be the case, our results indicate this factor is only effective under ischemia. Because the same consumer cells do not have an effect on the same type of tumor spheroids under well-nurtured environments. In addition, our new pH measurements and perturbations experiments agree with this reviewer's intuition about additional factors being key in the increased invasion (see new Figure 2). We are very excited about these new results, and we hope this reviewer will be excited too.

1. What is the oxygen sensor for increased invasion? PHD1-3 would be a good place to start looking. Is the PHD2-HIF axis important? Do VHL mutant cells still show responses to the consumer cells?

Following the reviewer's feedback, we generated isogenic HIF1A KO cell lines to study whether HIF1A was directly needed in the invasion of tumor spheroids within the 3MIC. We complemented these loss-of-function experiments with For HIF1A gain-of-function using pharmacological interventions that stabilize HIF1A under normal oxygen levels (CoCl2 and DMOG).

As shown in the new figure 2, these experiments mirrored our hypoxia experiments: HIF1A activity was not sufficient but it was required to drive the invasion of ischemic spheroids. We think that these new results are particularly interesting when taken together with our new pH-perturbation experiments. Briefly, our new experiments results show that in addition to the requirement of hypoxia/HIF1A, media acidity also has a strong effect on spheroid invasion. More excitingly, a drop in pH is sufficient to dramatically increase invasion - even in control well-nurtured spheroids. We think that the effects of pH and hypoxia are linked. HIF1A activation and hypoxia the increase glycolysis and thus lactic acid secretion. We speculate that this glycolytic switch is where hypoxia is important, but it is not sufficient because under well-perfused conditions (e.g. healthy tissue or large culture media volume) lactic acid levels may not buildup enough to significantly lower the extracellular pH. In contrast, under poor perfused conditions (3MIC and solid tumors) or if we flood cell cultures with lactic acid, the media's pH drops dramatically (Fig. 2).

1. If they include both spheroids of endothelial cells and cancer cells, will the resulting protrusions in hypoxia grow towards each other? Would macrophages enhance this process?

We agree with the reviewer this is an interesting question and we have anecdotally observed this effect. In the manuscript, we used these chimeric endothelial/tumor spheroids rather than separate ones (Fig. 6E). We do not find strong evidence that their protrusions grew towards each other, but this is something that we would like to explore in the future with more detail.

Significance:

The main advance is technical, as many previous studies have related hypoxia to increased cancer cell invasion, which the authors correctly acknowledge and cite. It is scholarly study, which will be of interest to many readers, and the method reported is likely to be adopted by several groups.

Reviewer #3

Evidence, reproducibility and clarity:

In this work, Anandi et al., developed a cell culture system to live image the initial transformation of cells in deprivation of oxygen and nutrients in a 3D context. Using this system, 3MIC, they were able to create oxygen and nutrient gradients to simulate ischemic conditions that arise deep within tumors and that typically precede metastasis. With the 3MIC system they validated that ischemia triggers cell migration and invasion of tumor cells. In addition, 3MIC also allowed them to study the interaction of tumor spheroids with stromal cells such as macrophages and endothelial cells. Interestingly, the authors showed that co-culturing tumor spheroids with stromal cells increased the pro-metastatic features induced by ischemia conditions. Lastly, using 3MIC allowed the authors to discern that a poor paclitaxel response in ischemic-like cells is driven by intrinsic cellular resistance rather than due to lower drug concentration.

Overall, the work is very well-written, and the results are consisting, convincing and support the conclusions. The methods are clear and complete and allow the reproducibility of the experiments. The experiments are adequately replicated and statistical analyses are well described. However, I have few suggestions to improve the impact of the manuscript:

1. The authors conclude that 3MIC results in the accumulation of lactic acid and nutrient deprivation in an increasing manner when moving far from the opening site. Is there a way to actually show this? So far, the authors employ a hypoxia sensor only. A sensor for internal pH or other method for nutrient deprivation would help to support the conclusion and further validate the model.

This is an excellent point. Following the reviewer's feedback, we tested additional sensors including for extra- and intra-cellular pH. As mentioned above, we observed dramatic changes in extracellular pH levels. We followed up these observations with a series of experiments that showed a key functional role for media acidification in driving invasion (Figure 2).

1. According to figure S3E, the main cell line used by the authors is already quite mesenchymal. It would be good to know if the results showed here are consistent in cells with a more basal epithelial phenotype. Do epithelial cells need stronger ischemic conditions to undergo phenotypic changes?

This is a great catch. To explore this further, we run a Western Blot analysis to compare epithelial and mesenchymal markers expressed by the main cells we used here (Lung KPs) and to compare them to levels in a stereotypical epithelial (MCF-7) and a mesenchymal (MDA-MB-231) cell line (new Fig. S4D). As the reviewer correctly points out, we do see that E-Cad and Vimentin are co-expressed in KP cells.

So far, our observations in a range of cell lines are a consistent decrease in E-Cad levels with no significant effects in vimentin levels - regardless of the basal levels of this protein.

Interestingly, a recent study[1] demonstrated in triple-negative breast cancer models, that an EMT hybrid phenotype - including the presence of Vimentin - is required for metastasis. A compelling hypothesis then is that ischemia in the tumor microenvironment may favor these hybrid phenotypes. We briefly discuss this topic in the revised version of this manuscript.

1. The number of replicates should be included in each figure legend and not only in the methods section. From data presented it is not clearly stated what do points mean in boxplots (e.g, Fig1H, 2B,G...). How many cells/spheroids did the authors count in each experiment?

We have now added this information to all our experiments. This information can be found in the figures and on the figure legends.

1. Figure 3B is not mentioned in the main text.

We apologize for this error, and we thank the reviewer for catching this issue, which have now corrected.

1. Line 295: "In the absence of macrophages, clusters of endothelial cells remained mostly rounded, even in the presence of consumer cells and regardless of their location along the ischemic gradient (Fig. 5A; Video S6)." However, in Video S6, both images show endothelial cells co-cultured with macrophages. I consider that Video S6 should be not referenced here.

The reviewer is correct so have removed that reference.

1. References style should be homogeneous (e.g, in Ref 13 appears "Nature Reviews Cancer" whereas in Ref 14 "Nat Rev Cancer"). Also, in Ref 25, the journal is missing.

We apologize for this oversight, and we have not tried to be more consistent in our references.

1. In plots where distance to open chamber site is not especify (e.g. 6B), at what distance were the data recorded? Please, indicate in the figure legend.

We have now added this information to our figures.

1. In the experiment showed in Fig 4, the sorting strategy would include stromal cells such as fibroblasts and endothelial cells in the GFP- population (as only CD45+ cells are removed). These cells will likely also grow in the 3MIC system and have an effect in migration. Can the authors rule out this confounding effect?

The reviewer is correct. We still think that the possibility of fibroblast contamination is low. First, the fluorescence of HRE-GFP cells under normoxic, is still higher than the autofluorescence of cells not expressing this constructs (such as fibroblasts). This is quite normal as most sensors/reporter have some leakage and thus there is a small amount of transcription. Second, intradermal and subcutaneous tumors are quite poor in fibroblasts. In fact, to study the role of fibroblasts in these tumors, they are usually co-injected with tumor cells (PMID: 20138012). Third, in the process of tumor dissociation and in vitroestablishment, non-transformed cells tend to die more. Since these are more technical points, we moved the cell sorting details to the material and methods section.

1. In Fig 5C the panel of proximal + macrophages is missing

We apologize for this mistake, and we have corrected in the new version of the manuscript.

1. In Fig. 5, Linifanib is used to study the effect of blocking VEGF. Linifanib can also interact with RTKs and PDGF. This fact should be acknowledged.

We agree with this point. Following the reviewer's advice, we now acknowledged the potential off-target effects of these inhibitors (lines 354-355).

Significance

This is a very interesting work with the development of a simple and cost-effective system that allows to continuously monitor biological processes in 3D cultures under nutrient-modified conditions. In general, these data would be broadly interesting to cancer community in general, as 3MIC is a very versatile system, where several aspects can be studied and precisely discerned.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #3

Evidence, reproducibility and clarity

In this work, Anandi et al., developed a cell culture system to live image the initial transformation of cells in deprivation of oxygen and nutrients in a 3D context. Using this system, 3MIC, they were able to create oxygen and nutrient gradients to simulate ischemic conditions that arise deep within tumors and that typically precede metastasis. With the 3MIC system they validated that ischemia triggers cell migration and invasion of tumor cells. In addition, 3MIC also allowed them to study the interaction of tumor spheroids with stromal cells such as macrophages and endothelial cells. Interestingly, the authors showed that co-culturing tumor spheroids with stromal cells increased the pro-metastatic features induced by ischemia conditions. Lastly, using 3MIC allowed the authors to discern that a poor paclitaxel response in ischemic-like cells is driven by intrinsic cellular resistance rather than due to lower drug concentration.

Overall, the work is very well-written, and the results are consisting, convincing and support the conclusions. The methods are clear and complete and allow the reproducibility of the experiments. The experiments are adequately replicated and statistical analyses are well described. However, I have few suggestions to improve the impact of the manuscript:

1. The authors conclude that 3MIC results in the accumulation of lactic acid and nutrient deprivation in an increasing manner when moving far from the opening site. Is there a way to actually show this? So far, the authors employ a hypoxia sensor only. A sensor for internal pH or other method for nutrient deprivation would help to support the conclusion and further validate the model.
2. According to figure S3E, the main cell line used by the authors is already quite mesenchymal. It would be good to know if the results showed here are consistent in cells with a more basal epithelial phenotype. Do epithelial cells need stronger ischemic conditions to undergo phenotypic changes?
3. The number of replicates should be included in each figure legend and not only in the methods section. From data presented it is not clearly stated what do points mean in boxplots (e.g, Fig1H, 2B,G...). How many cells/spheroids did the authors count in each experiment?
4. Figure 3B is not mentioned in the main text.
5. Line 295: "In the absence of macrophages, clusters of endothelial cells remained mostly rounded, even in the presence of consumer cells and regardless of their location along the ischemic gradient (Fig. 5A; Video S6)." However, in Video S6, both images show endothelial cells co-cultured with macrophages. I consider that Video S6 should be not referenced here.
6. References style should be homogeneous (e.g, in Ref 13 appears "Nature Reviews Cancer" whereas in Ref 14 "Nat Rev Cancer"). Also, in Ref 25, the journal is missing.
7. In plots where distance to open chamber site is not especify (e.g. 6B), at what distance were the data recorded? Please, indicate in the figure legend.
8. In the experiment showed in Fig 4, the sorting strategy would include stromal cells such as fibroblasts and endothelial cells in the GFP- population (as only CD45+ cells are removed). These cells will likely also grow in the 3MIC system and have an effect in migration. Can the authors rule out this confounding effect?
9. In Fig 5C the panel of proximal + macrophages is missing
10. In Fig. 5, Linifanib is used to study the effect of blocking VEGF. Linifanib can also interact with RTKs and PDGF. This fact should be acknowledged.

Significance

This is a very interesting work with the development of a simple and cost-effective system that allows to continuously monitor biological processes in 3D cultures under nutrient-modified conditions. In general, these data would be broadly interesting to cancer community in general, as 3MIC is a very versatile system, where several aspects can be studied and precisely discerned.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

Summary

Anandi and colleagues present a manuscript describing a nice assay for exploring the progressive effect of metabolic depletion of the nutrients and oxygen on the invasion of cancer cells. This builds upon and extends a device that they previously described - MEMIC - and now enables 3D analysis of small numbers of cells. The key to their method is the inclusion of a layer of consumer cells that deplete oxygen and nutrients. Using this tool, they demonstrate that depleted environments promote invasive behavior and lower cell-cell adhesion. This is related to the nutrient-deprived and hypoxic environments found in the center of many tumors. Cellular Potts Modelling is used to explore ideas around the cooperation between reduced cell-cell adhesion and increase ECM adhesion in promoting invasion. Overall, this is a well-constructed manuscript that will be of interest to cell biologists and cancer biologists.

Major comments

I realize this work is submitted to review commons and this complicates the recommendation regarding publication. My view is that the 'more prestigious' journals would require greater mechanistic insight, but that the work could find a suitable place in other members of the review commons stable. My comments are divided into those essential for any journal and those that might be journal dependent. Essential regardless of journal

• Many of the figures lack information about the number of spheroids analyzed and from how many biological repeats they are derived.
• The authors need to provide citations for their assertion that only gases can cross the PDMS, but not other small metabolites. They should also comment on whether the build-up of CO2 might be relevant.
• The data on the directionality of migration when consumers are present are not significant and doesn't warrant the speculation in lines 186-189.
• The ECM degradation in Figure 3 should be quantified.
• Do the authors have evidence that the hypoxia-exposed cells are more adhesive to ECM. This is central to their Potts model and I could not locate the supporting experimental data. If not, then the Potts model should include matrix proteolysis, which they do have data about.
• Is the down-regulation of E-cadherin transcriptional - i.e. is the mRNA level reduced?
• The title of figure 6 is misleading. The authors do not demonstrate chemoresistance in terms of cell survival or cell proliferation, which is how the term is normally used. The authors should measure cell number, proliferation, and cell viability. The data presented in the Supplementary Figure are inadequate with no quantification. The FUCCI reporter cells would be a good tool for this. Also, why use 150nM paclitaxel when the IC50 is 817nM? This seems bizarre. Lastly, there is a typo in the figure that suggest 150mM drug was used.

Journal dependent

• The authors have not excluded that either changes in nutrients, or even a pro-invasive factor, produced by consumer cells are necessary for the increased invasion. They have only shown that they are not sufficient. The authors should perform a series of experiments comparing hypoxic conditions with normal media and normoxic conditions with nutrient depleted/condition media by prior culturing of KP cancer cells.
• What is the oxygen sensor for increased invasion? PHD1-3 would be a good place to start looking. Is the PHD2-HIF axis important? Do VHL mutant cells still show responses to the consumer cells?
• If they include both spheroids of endothelial cells and cancer cells, will the resulting protrusions in hypoxia grow towards each other? Would macrophages enhance this process?

Significance

The main advance is technical, as many previous studies have related hypoxia to increased cancer cell invasion, which the authors correctly acknowledge and cite. It is scholarly study, which will be of interest to many readers, and the method reported is likely to be adopted by several groups.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

Summary

In this work, Anandi et al. propose an ex vivo model that can be used to recapitulate the in vivo structure of the tumor microenvironment, which allows the observation of morphological and functional changes in tumor cells in a 3D context. Due to the ability of cancer cells to induce hypoxic condition within the TME, authors propose this model to tackle the study of metastasis initiation in vitro. The proposed system successfully displays an ischemic gradient with cells accessing nutrients at different rates, similarly to what happens in solid tumors in vivo. Moreover, in line with the literature, tumor cell migration and invasiveness were promoted by hypoxic conditions. Authors also show that the system could be used to study cell-cell interaction, as co-cultures of macrophages and cancer cells were successfully cultured in the system and studied in the context of tumor hypoxia. The study proposed is interesting and timely, as cancer cell invasion remains an important area of tumor biology that needs further exploration. The methodology is well explained and proposed in a linear flow. However, the work could benefit from some improvement and changes, as well as from additional experiments. On an important note, authors do not properly refer to the current literature, as several studies on 3D culture systems/chambers have already been studied and developed to investigate the tumor microenvironment, but they are not cited nor referred to in the manuscript. Authors should refer to such literature and explain how this system is different and adds to it.

Major comments:

• Authors propose this method to study the TME in 3D. When culturing cells with different ECM (Collagen vs. matrigel+collagenI) authors should take into consideration the effect of these materials on different cell types. It is known how collagen and matrigel can differently influence the polarization and phenotype of stromal cells (particularly in regards of fibroblasts - major components of solid cancers - e.g., PMID 21029367), therefore these points should be addressed at least in the discussion.
• In addition to the previous comment, matrigel and collagen are also known to alter cancer cell phenotype (e.g., PMID 21029367) and this point should be taken into account.
• The need for novel 3D systems to study different aspects of the TME in vitro/ex vivo are certainly needed, however they are not inexistent. Authors should address this in the text, as the current literature already started to propose 3D models (including models involving matrigel/collagen in combination with other materials). 3D chambers (of different materials, and with different aims) are being used and designed and can be found in the literature. These works are not cited in the current study at all. For instance, Anguiano 2017; Cavo et al. 2018; Anguiano et al., 2020; Sodek et al. 2008, etc.
• Even though the focus is on hypoxia and the achievement of an ischemic gradient in the chamber to allow resemblance of an in vivo tumor, the authors write in line 123 (and also in other parts of the text) that: "these results show that consumer cells in the 3MIC form ischemic gradients that can influence the local metabolic microenvironment experienced by neighboring tumor spheroids". The addition of the use of the PMDS membrane partly supports the claim, however it would be interesting to check whether this is indeed true, by measuring for example the levels of certain metabolites (e.g., glucose, glutamine, glutamate, lactate, aspartate) reached with the system, or pH levels, etc., in presence or absence of the hypoxic gradient/consumer cells.
• When looking at the references presented in the manuscript, authors quote too many review articles, rather than scientific articles. Given the extremely wide literature on cancer metastasis, more of these works should be quoted in this context. For example: in the introduction - text lines 27-38 - only 4 references are research articles, out of 14 references presented in that paragraph.
• As authors showed successfully how macrophages and cancer cells can interact in the chamber, recapitulating cell interactions in an in vivo context, it would be very interesting to see whether different consumer cells would induce similar or different changes to the spheroids and the ischemic gradient (for instance using stromal cells or non-tumor cell lines as consumers, instead of cancer cells only), as we know how tumors are a multitude of cell subsets, each contributing to nutrient production, oxygen consumption, etc.

Minor comments:

• Studying the early metastatic development/seeding remains a timely quest, however authors should refer to several new studies in which various mouse models are used to study metastasis from different points of view (e.g., PMID 25822788; PMID 36991128; PMID 25171411; PMID 25633981; PMID 34632412; PMID 35921474; etc). Or line 41, three reviews are quoted (refs 27-29), whilst there are several works that could be quoted on metabolism in solid tumors also in the context of metastasis (e.g., PMID 36522548; PMID: 26719539, PMID 34303764). This comment applies to the rest of the text.
• The order of the references is not properly presented. In the introduction, the first reference is n. 4 (text line 22), instead of it being reference 1. Moreover, the subsequent literature ref. is number 12 and not number 2. Please revise the order of the references, and position them within the bibliography from first cited to last cited in the text.
• Lines 98-104. It would be helpful to the reader to define here what these consumer cells are. Even though it is explained in the methods that the consumer cells are cancer cells, it is important to make it clear in the text, as it could be misleading at times.
• The English grammar and spelling should be revised in some parts, as well as typos and missing words throughout the text (e.g., Line 38, the word "interraction" is misspelled and should be corrected with "interaction". Line 49, the first sentence seems incomplete. Lines 68-69 should be revised as the sentences do not flow well together, probably due to a missing word. In line 77 it should be "presents". Line 341 should be "cannot be explained").

Referees cross-commenting

I find the comments from the other reviewers to be in line with one another as well as with my general assessment. The major and comments of all reviewers should be addressed. The minor comments should be taken into account as well, as they would render the text and the figures more precise. I suggest that 3-6 months to complete the revision process is an appropriate time frame for the authors. Finally, I strongly encourage the authors to add in the discussion the points and questions raised by all reviewers, as well as to improve the bibliography in terms of organisation, linearity, and state of the art.

Significance

General assessment:

The work by Anandi et al. offers an additional tool to tackle the issue of studying the tumor microenvironment, in a 3D culture system. The authors show a model that can be used to study tumor hypoxia in 3D, offering the possibility to study the TME in a more in vivo-like manner without turning to mice models. The development of new tools to study the TME avoiding the excessive use of animals is definitely a timely quest. In addition, the system has the potential to be applied to tackle different biological questions, as the methodology is well explained and could be suitable to many other fields of cancer biology (e.g., drug resistance or uptake). The work is overall presented in a clear way and the methodology is explained thoroughly and it has the potential to be a useful tool for the study of cancer hypoxia.

However, authors should address how their method could differently impact other cells when applied to other systems. As one major claim is the potential use of this methodology to study the TME, it should be taken into consideration how stromal cells are strongly affected by the ECM, and how certain settings or features of the system may impact such cell populations. In addition, the work does not properly refer to the current state of the art. As other studies started to propose 3D systems for the study of TME and cell-cell interactions - besides organoids - the authors should cite these works and frame their own study in a more appropriate context, pointing out differences with the current 3D chambers available, the advantages of one vs the other, and so on.

Advance: the study adds to the current literature as the study of tumor hypoxia in 3D remains a complicated issue. The interesting co-culture settings with macrophages suggests potential uses of this model to study cell-cell interactions.

Audience: the study is very methodological and offers a tool that could be used by cancer biologists - and maybe by other biology fields.

Note: This response was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Reply to the reviewers

We would first like to thank the reviewers for their careful reading and thoughtful feedback.

We have substantially revised the manuscript and included additional experimental evidence on O-GlcNAc and OGT/OGA protein levels in the placenta of embryos bearing the OGT-Y851A hypomorphic mutation.

Overall, we believe our improved manuscript provides compelling evidence that the glycosyltransferase activity of OGT, and thus the O-GlcNAc modification itself, plays a sexually dimorphic function in placental development and the developmental repression of retrotransposons in the developing embryo.

We have addressed each of the reviewers' comments below. The original comments (C) are in italic, our responses (R) in Roman font.

Reviewer #1

Evidence, reproducibility and clarity

C1: Formichetti at el. developed mice with OGT catalytic dead mutations and then studied their function during early embryogenesis. Not surprisingly, dramatic reduction in OGT activity failed to produce embryos; however, mild reduction in OGT did produce animals. The authors then use the T931 animals that have a mild reduction in activity to further characterize the function in the early embryo. Not surprisingly, male mice showed changes in gene expression, implantation sub-lethality, and an uptick in loss of retrotransposon silencing. The authors also show that an even milder reduction in OGT activity (Y851A) effects male placenta function and chromatin remodeling. Finally, the authors make a less stable OGT transgene within the mouse and again found embryogenesis issues in the males and alterations in numerous gene families including mTOR signaling and p53 function. All in all, this is an interesting study that track functions of OGT in early embryonic development. The studies are well-controlled and rigorous.

R1: We thank the reviewer for their clear understanding and their appreciation of the rigor and impact of this work.

Significance

C2: This is a good study and novel. Not only is it of interest to reproductive biologist, but it echos themes found in O-GlcNAc biology.

R1: We are pleased that the reviewer underlined the novelty of the study and its impact across fields.

Reviewer #2

Evidence, reproducibility and clarity

Comments to authors

C3: To investigate the function of OGT at specific developmental stages, the authors perturbed OGT's function in vivo by creating a murine allelic series featuring four single amino acid substitutions that variably reduced OGT's catalytic activity. The goal was to identify the direct effect of O-GlcNAcylation, using a sophisticated collection of genetic mutants to evaluate in vivo the role of this modification at early stages of development. Overall, the severity of embryonic lethality correlated with the extent of catalytic impairment of OGT, demonstrating that the O-GlcNAc modification is essential for early development.The study represents a substantial advance in our understanding of OGT and O-GlcNAcylation in mammalian development. The creation of novel murine models and inducible systems is an important contribution, providing powerful tools for future research in this field. The insights into the role of OGT's catalytic activity and its involvement in epigenetic regulation during embryonic development are noteworthy, opening new avenues for research.

R3: We thank the reviewer for their insightful comments. We are grateful for the supporting statements. Please find below detailed response to all your comments.

However, there are a few considerations and concerns:

Major:

C4: 1. An assumption of the study is that different mutations cause different levels of O-GlcNAcylation rather than alterations in substrate specificity. It might be important to test, at least in cultured cells, that the different mutations do not change the preference of OGT to modify certain proteins rather than others, which can provide alternative explanations for their findings.

R4: Thanks for asking this question, it helped us to better explain the rationale behind the choice of the Ogt amino-acid substitutions.

This is a critical point that we carefully considered in the design of the single amino-acid substitutions. Two lines of evidence support that the precise mutations created impact the catalytic rate without modifying the substrate specificity:

First, as explained in the text, the choice of the single amino-acid substitutions was driven by previous structural and enzymology knowledge. The impact of the four point mutations selected on OGT protein stability and on the Michaelis-Menten kinetic values had previously been determined experimentally (Fig. 1A legend and Martinez-Fleites, C. et al. Nature Structure Molecular Biology 2008; https://doi.org/10.1038/nsmb.1443).

There is a second important rationale that we added in the revised manuscript: the four point mutations selected are all located in the catalytic domain (specifically, H568A in the N-Cat domain and Y851A, T931A and Q849A in the C-Cat domain), while the substrate recognition is operated via two other domains namely the intervening domain (Int-D) https://doi.org/10.1038/s41589-023-01422-2) and the tetratricopeptide Repeat (TPR) superhelix (10.1021/jacs.7b13546; https://doi.org/10.1073/pnas.2303690120). Therefore, for both these reasons, it is extremely unlikely that these mutations could influence the substrate specificity.

C5.1: 2. In Fig 1D and 1H, the thresholds to define a gene or TE as differentially expressed are not strong. According to the figure legends, "any" change in terms of log2Fc was considered as DE and colored. I think the figures should illustrate better that the changes are subtle, by for example adding a dotted line (at least) in the value 0.5 of the y-axis. These subtle transcriptional changes should be reflected better in certain paragraphs where the expression of TEs are presented/and discussed as a hallmark of the absence of O-GlcNAcylation in the OGT-mutants. The same happens with Suppl Fig 3C (changes are very minor). {. Applying a stronger threshold, among the upregulated genes, only Xist will be significantly overexpressed. If a gentle threshold needs to be applied to this data, authors should at least justify the reasons behind doing so. Same for Fig2D.

R5.1: The reviewer means Figure 2D for MA plot of gene expression and Figure 2H for retrotransposons expression. These figures now include a dash line to indicate Log2FC = 0.5 (as all MA plots).

The text is explicit on the subtle changes in transcription, it reads "with 2/3 of the genes downregulated and 90% of the significant changes below 1 log__2__FC"; "most of the Ogt__T931del/Y embryos showed a low magnitude upregulation of retrotransposons".

The revised text states "Notably, most of the OgtT931__del/Y embryos showed a low magnitude (log2FC < 1) upregulation of retrotransposons".

We expand on this topic in the next response (R5.2) noting that changes in gene expression upon O-GlcNAc perturbation in different systems were previously characterized as subtle and widespread. We suggest that this phenotype may arise from the scarcely understood pleiotropic function of O-GlcNAc in fine-tuning gene expression; this phenotype could have a biological significance.

C5.2: If a gentle threshold needs to be applied to this data, authors should at least justify the reasons behind doing so. Same for Fig2D.

R5.2: Previous studies in different systems reported that O-GlcNAc perturbation causes a widespread change in gene expression of low magnitude (https://doi.org/10.1101/2024.01.22.576677, https://www.pnas.org/doi/10.1073/pnas.2218332120). We use the same thresholds as a recent functional Ogt study in ES cells to call differentially expressed genes, specifically: p<0.05 (Wald test), any FC (Li et al. PNAS 2023, https://www.pnas.org/doi/10.1073/pnas.2218332120). The p value threshold is standard; the absence of FC threshold is dictated by the insufficient knowledge of the significance of the low magnitude changes observed across many transcripts.

C6: 3. In Figure 2B, the T931del allele was recovered in the blastocyst population with a very high frequency, even higher than the male WT group (T931del: 10; WT: 3). This observation suggests that the T931del allele did not significantly affect blastocyst survival. Further clarification or additional experiments might be necessary to understand the implications of this finding on early developmental stages.

R6: This is only a hint as the numbers of blastocysts recovered were too small to perform statistics on Mendelian distribution. Thus, more experiments are needed to perform these statistical tests. These experiments are onerous because the low frequency of germline transmission is incompatible with maintaining this mutation by breeding heterozygous animals. Because of this, a new mouse line needs to be created by CRISPR-HDR targeting in the zygote in order to compute statistics on Mandelian ratios. Importantly, this question - does T931del affect blastocyst survival? - is peripheral, and the results of these experiments would not affect our conclusions in any way.

C7: 4. Similarly, in Figure 2G, there is an apparent higher expression of TE expression in the T931A/Y embryos group than in the T931del/Y group, which combined with the higher frequency of blastocyst generated in this latest group it may indicate a deeper molecular consequence after the deletion of the T931. A comparison of the transcriptome between these two cell lines help to address this possibility. Also, the authors should compare the O-GlcNAc levels of WT, T931A, and T931del mutant blastocysts by immunostaining, similar to what was done in Figure S5F.

R7: We agree that a direct comparison between the two mutations of the T931 residue would be interesting; however, this comment is very difficult to address experimentally for the reasons outlined below:

Firstly, it is not possible to perform a statistical comparison of the transcriptome T931A/Y VS. T931del/Y with the data generated because the number of hemizygous T931A/Y (n=2) is too small. Hence, it cannot be ruled out that the seemingly milder retrotransposon reactivation in one of the T931A/Y embryos could have occurred by chance.

Secondly, considering the low magnitude effect on gene expression changes upon O-GlcNAc genetic perturbation, to statistically assess the penetrance of the molecular phenotype and perform the differential expression analysis, numerous (>>3) hemizygous blastocysts of each genotype would be needed. Because females heterozygous for the T931 mutations transmit the mutant allele at very low frequency, these experiments require numerous de novo CRISPR injection sessions.

Thirdly, for the immunostaining of O-GlcNAc to be semi-quantitative, a large number of hemizygous blastocysts for each genotype would be required (note that in Figure S5F, 29 morulae per condition were imaged), thus requiring numerous CRISPR injection experiments as discussed above. Moreover, O-GlcNAc changes could be subtler than what expected based on the strong reduction of OGT activity, since as a compensatory mechanism Ogt expression is upregulated in the Ogt__T931A/del blastocysts (Fig. S2D), making a quantification even more challenging despite a high number of stained embryos.

In sum, these in vivo experiments are difficult and require sacrificing many animals (about 20 females per CRISPR injection experiment). Because the results would bring refinement to the study but would not change our conclusions, we suggest that the cost/benefit is too high.

C8: 5. In Boulard et al. 2019 O-GlcNAcylation was shown to be sufficient to modulate expression of DNA methylation-dependent TEs. It would be interesting to know (or at least discuss) if the changes in TE expression observed in OGT-mutant embryos in this study involve changes in DNA methylation. Ideally, some DNA methylation measurement optimized for low input numbers of cells would be useful.

R8: Thank you for making the link with our previous study. In the PNAS paper, we report that targeted removal of O-GlcNAc at proteins bound to specific TEs (e.g. IAPez) causes their full-blown reactivation without detectable changes in DNA methylation, thus suggesting a role of the O-GlcNAc modification for the silencing of methylated TEs downstream or independent of DNA methylation. We agree that it would be informative to quantify DNA methylation in the T931-mutant blastocysts to test if the in vitro result is the same in vivo, but this would require performing onerous microinjection sessions as explained above.

C9: 6. The data related with the OGT-degron system in MEs seem disconnected with the rest of the manuscript. While the developmental models (blastocyst, etc) elegantly assess the contribution of O-GlcNAcylation to the control of cell survival and gene expression through the use of different OGT mutants, the degron system is a system of graded depletion that unfortunately was only possible to be used in MEFs (instead of embryos). Thus, the results obtained with the degron system in MEFs are difficult to intersect with the data from the use of OGT-mutants in embryos. Even though there are obvious interesting questions that one may want to know about this OGT degron MEF system, none of them would demonstrate a direct role for O-GlcNAcylation in cellular function, the major point addressed in the developmental system. Using the degron system in embryonic stem cells might have provided a more parallel comparison. The authors should discuss this point in more detail and either use ESC instead of MEFs or provide a stronger justification for the use of MEFs over ESC.

R9: We thank the reviewer for their clear understanding of the system. The choice of primary MEF as an in vitro model was imposed by technical limitations we encountered during the study. We fully agree that ES cells is the model of choice for preimplantation embryos; thus we initially derived ES cells and obtained only one male clone bearing the AID degron system. Upon auxin addition to the culture media, OGT's level remained unchanged in ES cells. Thus, the ES cells model was not usable. To test the AID degron in a different cell type, we then derived MEFs and showed its effectiveness (Figures 4C and S4C-E), which also allowed to collect functional data on OGT's cellular function (Figures 4D-F). We took the comment on board and clarified the rationale of studying MEFs in the revised manuscript. We agree that it remains to be verified that the OGT-dependent pathways uncovered in MEFs are relevant in the preimplantation embryo. Despite this caveat, we feel the mouse model for endogenous OGT-degron, as well as the negative results in vivo and conclusions in MEFs should be shared with the community, which could take advantage of our results to refine the system.

Minor:C10: 7. In Fig 2C the color and shape codes are confusing to understand - there are some colors/shapes that are not represented in the PCA plot. The same in Fig 3H, where in the PCA plot there are pink triangles that do not match with the code legends.

R10: We apologize for the confusion with the legends of Figures 2C and 3H, that we have made unambiguous in the revised version (as well as Figures S2B,C and S3C).

C11: 8. In the figure legends of Figures 2D, 2E, 2F, and 2H, the notation should be corrected from "OgtT931A/Y" to "OgtT931del/Y".

R11: This has been corrected; many thanks for bringing it to our attention.

Significance

C12: To investigate the function of OGT at specific developmental stages, the authors perturbed OGT's function in vivo by creating a murine allelic series featuring four single amino acid substitutions that variably reduced OGT's catalytic activity. The goal was to identify the direct effect of O-GlcNAcylation, using a sophisticated collection of genetic mutants to evaluate in vivo the role of this modification at early stages of development. Overall, the severity of embryonic lethality correlated with the extent of catalytic impairment of OGT, demonstrating that the O-GlcNAc modification is essential for early development.

R12: We thank the reviewer for their clear understanding of our work and their appreciation of the biological importance of the findings.

Reviewer #3

Evidence, reproducibility and clarity

C13: This is a conceptually interesting paper that attempts to leverage the knowledge of OGT catalysis to begin to dissect OGT function. The evidence is presented I a straightforward fashion and is in general well documented. The breeding strategies are well informed and the paper draws heavily on previous work carried out in the mouse.

R13: We greatly appreciate the overall supporting review. However, we fail to understand what they mean with "the paper draws heavily on previous work carried out in the mouse". This comment may stem from a misunderstanding because this work is not based on any previously published study. Specifically, neither the seven murine alleles presented and analyzed nor the single embryo-transcriptomic data sets on which our conclusions are based have been published elsewhere.

To put this work into context, before our study there were two seminal studies published two decades ago that reported the essential role of Ogt for mouse development, but no molecular profiling was performed (10.1073/pnas.100471497, 10.1128/mcb.24.4.1680-1690.2004). The two Ogt loss-of-function alleles studied in these papers were deemed as not suitable for interrogating molecular phenotypes because they caused cell death that confounds molecular profiling and embryonic lethality at implantation, thus preventing study of the sexually-dimorphic role of Ogt placenta. To overcome this long-standing problem, we created new seven murine alleles, which allowed us to tease apart molecular phenotypes at key stages of mouse embryonic development, focusing on the blastocyst and the placenta.

Significance

C14: The paper describes tools which will help dissect the many potential roles of O-GlcNAc addition in early development. As it stands, this is a descriptive manuscript that will lead to hypothesis generation and testing and this should not be undervalued. The biological reagents produced and characterized will be of general interest to the field. Most of the findings presented represented a verification of existing ideas in the field but this is not meant as a criticism since part of the motivation for the approach was to generate a reproducible system for analyzing the biological phenomena.

R14: We thank the reviewer for their appreciation of the importance of experimentally testing ideas shared in the field without direct evidence.

However, we must respectfully disagree with the qualification of "descriptive manuscript". This qualification may stem from the particularly difficult challenge to accessing the molecular details on how the O-GlcNAc modification exerts the biological functions we report. We are fully cognizant of the limitations of the study that we discussed in the discussion section and in R20.2. However, we feel that the adjective "descriptive" is not a fair qualification because we provide numerous novel functional evidence. Specifically, we introduce two novel orthogonal in vivo perturbations for endogenous Ogt that allowed us to interrogate for the first time its function in the developing mouse embryo. These perturbations allow us to draw causative conclusions (not descriptive) on the essential role of the O-GlcNAc modification itself for preimplantation development, its sexually-dimorphic role in the placenta and its requirement in vivo for the stable repression of retrotransposons.

C15: There are perhaps some bioinformatic shortcuts taken that may need to be corrected upon thorough review. These do not lessen the overall impact of the contribution.

R15: All the code written for the bioinformatic analyses performed in this study is publicly available: https://github.com/boulardlab/Ogt_mouse_models_Formichetti2024. The reviewer needs to specify which bioinformatic analysis they suggest could be improved.

Reviewer #4 (Evidence, reproducibility and clarity (Required)):

Summary

C16: O-GlcNAcylation is the fundamental post-translational modification of numerous nuclear and cytosolic proteins. OGT is the sole enzyme catalyzing O-GlcNAc addition onto the proteins. The essentiality of OGT for early development and cellular viability has been established by using OGT-KO mice and cell lines. However, it remains to be elucidated whether the catalytic activity of OGT is required for the early development, and if the catalytic activity of OGT is required what are the functions of OGT or O-GlcNAcylation in early development due to a lack of appropriate mouse models. To overcome the technical difficulty of manipulating the levels of O-GlcNAcylation in early embryos, Formichetti et al. created the series of four mouse models (OgtY851A, OgtT931A, OgtQ849N, and OgtH568A) with different OGT activity by introducing single amino acid substitution in the catalytic domain. By analyzing the inheritance of the hypomorphic OGT alleles and the lethality of mouse embryos, they discovered OGT activity is a critical factor for early development. Subsequently, RNA-seq analyses with two mouse models showing the maternal inheritance of the hypomorphic OGT alleles indicated that sever hypo-OGT activity altered transcription and silencing of retrotransposon in preimplantation development while mild reduction of OGT's activity affected placental development in a sexually dimorphic manner rather than preimplantation development. Furthermore, to study the function of OGT at specific developmental stages, they developed a mouse model bearing endogenously AID-tagged OGT for acute degradation of OGT. Although the degron system wasn't efficient in preimplantation embryos, they discovered quick transcriptional changes upon OGT deletion in MEFs. The quality of the manuscript is good because the question to be solved was appropriately set, the approach was well designed, and their findings were interesting, although their writing was sometimes hard to understand as I raised in my following comments. Nevertheless, there are several points to be fixed before being published.

R16: We thank the reviewer for their clear understanding of our work and their appreciation of the biological importance of the findings. Your comprehensive review of the manuscript and the questions you raised were extremely helpful in improving the manuscript and fully addressing its limitations. Below, we respond to comments in full, have revised the manuscript to improve clarity and have included novel results.

Major Comments

C17: 1. Although the authors showed in vitro activity of each mutant of OGT used in this manuscript by referencing the previous literature, they never showed the levels of global O-GlcNAcylation (and OGT itself) in their established mouse embryos. Although it could be impossible to determine O-GlcNAc levels in OgtQ849N and OgtH568A embryos because of the lack of germline transmission and founder line, respectively, they could do that in OgtY851A and OgtT931A embryos. Given that Y851A and T931A mutants had similar VMAX/KM with different VMAX, it is possible that their activity is comparable or Y851A has even lower activity in vivo depending on the concentration of UDP-GlcNAc in embryos. Therefore, it is critical to assess whether in vivo OGT activity is correlated with that in vitro as expected to conclude that severity of sub-Mendelian inheritance is proportional to the reduction of activity of OGT in vivo. Moreover, since the authors developed the elegant system to deplete OGT, the activity of Q849N and H568A mutant OGT can be examined at least in cells by expressing them in MEFs with OGT-degron system. Thus, I propose determination of global O-GlcNAc levels compensated by OGT levels by western blotting in OgtY851A, and OgtT931A embryos or MEFs with the OGT degron system re-expressing the individual four mutant OGTs. If the protein amount is insufficient for western blotting in the embryos because of the sizes of the earlier stages of embryos, I believe the author could address this by utilizing immunofluorescence as shown in Figure S5.

R17: We fully agree that this is an important point that requires revision. The only mutation for which the level of O-GlcNAc and OGT can be assessed by western blot in vivo is Y851A, the other mutations resulting in embryonic lethality before the blastocyst stage.

We have included in the revised manuscript western blot analyses of protein expression for OGT, OGA and O-GlcNAc levels in the placenta of the OgtY851A mutants (new Figures 3C,D). The new data show that OGT is upregulated at the protein level in homozygous females, in good agreement with our transcriptomic analysis. Furthermore, O-GlcNAc levels were slightly reduced in homozygous and hemizygous placentae thus showing the impact of the point mutation on global O-GlcNAc levels in the placentae. Moreover, the analysis of OGA protein level unexpectedly revealed the enrichment of a previously uncharacterized OGA fast migrating isoform in hemizygous and homozygous placentae.

We agree that it would be informative to compare O-GlcNAc levels in OgtT931A versus OgtY851A embryos. A comparison implies performing the experiment at the same developmental stage, which has to be the blastocyst stage or prior because T931A/Y embryos die around implantation. The blastocyst being made of approximately 140 cells, it would require to pool many single blastocysts to obtain the necessary protein input for western blot. We are not aware of another study performing western blot with pooled blastocysts. An additional great challenge for this experiment is the necessity to genotype and sex the blastocysts before pooling. Thus, the feasibility of this experiment is uncertain.

As an alternative, the reviewer suggests measuring O-GlcNAc levels in the degron MEFs after introduction of OGT transgenes bearing the mutation studied. This experiment would not be conclusive because of residual O-GlcNAc after OGT degradation (Figure S4E). Furthermore, the O-GlcNAc proteome is dynamic during development (as shown in the developing brain by Liu et al. https://doi.org/10.1371/journal.pone.0043724), therefore the MEFs results would have limited value to explain our results in the early embryo.

In sum, available technologies to quantify O-GlcNAc (e.g. western bot, mass spectrometry) are inadequate for low input samples as the early embryo. However, our series of hypomorphic alleles backed up with in vitro enzymology measurements brings indirect evidence to this question. Specifically, the qualitative correlation between the measured OGT activity in vitro and the developmental phenotype indicates that the resulting relative levels of O-GlcNAc are consistent with in vitro measurements.

C18.1 : 2. I didn't understand why the authors couldn't find any founder lines of the OgtH568A mutant. Was that because mosaic mice with OgtH568A mutation are lethal?

R18.1: To answer to this question, it is important to recall two key features of the biological system:

1) The mutation H568A was reported to disrupt the glycosyltransferase activity completely (10.1038/nsmb.1443). Hence, OGT-H588A is catalytic dead.

2) We performed the CRISPR-HDR targeting in the 1-cell embryo.

Based on these premises, the absence of F0 with the OgtH568A mutation (0/31) suggests that introducing this mutation causes embryonic lethality in both males and females. This hypothesis is consistent with the previously reported lethality around implementation of Ogt-null alleles (10.1128/mcb.24.4.1680-1690.2004). It is possible that the sgRNA is very efficient and results in homozygous mutations in all female zygotes injected (as we have not obtained heterozygous females bearing these mutations). High efficiency of the targeted mutagenesis in the zygote results in mutants where all or the majority of cells bear the mutation (no or low mosaicism). The high number of microinjections performed (416 embryos over the 3 injection sessions) allows us to make these claims.

C18.2 : Also, I believe there was no explanation why the OgtQ849N allele showed no maternal inheritance. Was that because Q849N possesses enough activity for sustaining mosaic embryos, but not oocytes? The authors should better explain these points in the manuscript text.

R18.2: Thanks for this comment, we agree that this maternal effect phenotype demands further explanation.

The phenotype observed suggests two possibilities: either that the oocyte cannot maturate or that the cleavage-stage embryo cannot develop with the resulting lower levels of O-GlcNAc. The cleavage-stage embryo does not transcribe a catalytically active OGT before the 8-cell stage and thus relies on the OGT protein inherited from the oocyte until this stage (https://doi.org/10.1101/2024.01.22.576677).

Thank you for this comment, we added this interpretation of the result in the text:<br /> "The lack of maternal transmission of the Q849N allele from seemingly mosaic founder females is likely explained by the reliance of the cleavage stage embryo onto the oocyte payload of OGT and O-GlcNAc modified proteins. Specifically, Ogt's exons encoding for the catalytic domains are not detectable before the 8-cell stage, while OGT full-length protein is present and thus maternally inherited (Formichetti et al, 2024)."

C19: 3. The authors serendipitously found a T931del-allele in the "WT" allele of the OgtT931A line, and suggested that T931del had milder activity loss, although the lethality of embryos was greatly mitigated. Nevertheless, transcriptome analyses in male blastocysts revealed that 120 genes' expression was changed in T931del/Y males. This raised the question about which mutant OGT has higher activity, Y851A or T931del. I think comparing the activity of Y851A and T931del mutants in MEFs with OGT-degron system is important to confirm the proportional relationship between activity and phenotypic severity.

R19: We agree that it is a limitation that the effect of the T931del mutation on OGT activity has not been biochemically characterized. However, the important point here is that our assessment of phenotypic severity based on maternal inheritance of the mutant allele and embryonic lethality is based on the point mutations for which the catalytic activity has been determined, namely Y851A, T931A, Q849N and H568A, but not T931del.

We studied the serendipitously discovered T931del mutation to obtain transcriptional insights in the blastocyst. Because the deleted residue T931 is key for the binding to the donor substrate, we can reasonably assume that this mutation affects the catalytic activity, albeit to an undetermined level.

Hence, our conclusions regarding the requirement of O-GlcNAcylation for development are unaffected by the lack of biochemical knowledge on T931del.

C20.1: 4. Regarding transcriptomes of T931del/Y, the authors found the upregulation of proteasomal activity and stress granules along with the downregulation of amino acid metabolism, mitochondrial respiration, and so on. To validate the results, the authors should perform qPCR on several up- or down-regulated genes.

R20.1 : We agree that, in principle, qPCR validation is suitable. However, this validation experiment is particularly expensive in this case because of the requirement of numerous CRISPR zygote pronuclear injection sessions.

The conclusions of the RNA-seq analysis are strongly supported by a high number of biological replicates (n=10). This high number of biological replicates was essential to obtain sufficient statistical power to quantify with a high level of confidence transcriptional changes of low magnitudes (below 2-fold change, see R5.1 and R5.2).

Therefore, the qPCR validation experiment would require to repeat the CRISPR zygote pronuclear injection sessions with the same high number of animals. This represents a major investment in experimental work and the sacrificing of about 40 animals. Importantly, the RNA-seq results presented are authoritative because of a high number of biological replicates and high number of sequencing reads per sample. Thus, we argue that qPCR validation is not essential and thus the high cost of this experiment is difficult to justify.

C20.2: In addition, according to Figure S2E, the authors pointed out that at least for genes upregulated in OgtT931A embryos, the changes were not explained by a developmentally delayed transcriptome, suggesting that upregulation of these genes was the cause of developmental delay. Therefore, I strongly encourage them to discuss in the manuscript text how up-regulated genes could contribute to developmental delay.

R20.2: Throughout the manuscript, we have been cautious to avoid establishing causal relationships between the differentially expressed genes uncovered and the developmental phenotypes (e.g. delayed development). There are two main obstacles which we believe prevent us from establishing causality with the data available. Firstly, it is not possible to disentangle differentially expressed genes and developmental delay (in other words, we have no way to tell which is the cause and which is the consequence). Secondly, O-GlcNAc modifies over 5000 proteins and the developing embryo is a particularly dynamic system; thus we cannot know whether the differentially expressed promoters are direct targets of O-GlcNAc modified proteins (or alternatively secondary effect of another molecular alteration, for example of the proteome). We discuss this limitation of the study in the discussion section.

C21: 5. Regarding the transcriptome in OgtY851A mice, Y851A/Y male mice had huge transcriptomic differences, while Y851A/Y851A female mice barely had any. Although it seems to agree with the number of Ogt alleles, I wonder whether other X-linked genes expressed higher in female placenta as shown in Figure 3C could attenuate the effects of decreased OGT activity. I don't think this possibility can be excluded, unless the authors further decrease OGT activity in Y851A/Y851A female placenta and obtain the similar results as for male placenta. Or if they compared the levels of global O-GlcNAcylation between Y851A/Y and Y851A/Y851A mouse placentas and discovered they had similar levels of O-GlcNAcylation, then the authors could conclude that the number of Ogt alleles was not the reason of sexual-dimorphism. The authors should determine the levels of O-GlcNAcylation in Y851A/Y and Y851A/Y851A mouse placentas and/or at least discuss the above possibilities in the manuscript text.

R21: Thank you for the thoughtful feedback. We agree that the most likely explanation for the higher sensitivity of males placenta as compared to females to OGT reduced activity is the difference in Ogt copy number, especially because Ogt escapes X-chromosome inactivation in the placenta (new Figure S3A).

Western blot quantification of global O-GlcNAc levels was now performed (new Figures 3C,D). We measured similar level of O-GlcNAc in Y851A/Y and Y851A/Y851A placentas (lowered than WT males in both cases), but we cannot exclude that the WB does not have the dynamic range required to detect a subtle difference. In fact, female homozygous were expected to have an intermediate level between WT males and hemizygous males, and the difference between the two male genotypes (also considering sample-to-sample variability) is already small when quantified from the blot (new Figure 3D). It is possible that a X-linked modifier attenuates the impact of hypo-O_GlcNAcylation in female mutant placenta in the case of identical O-GlcNAc levels in homozygous females and hemizygous males. Thank you for the idea that we included in the revised manuscript:

"Of note, the lower sensitivity of the homozygous females' transcriptome to Ogt disruption (Fig. 3F,I and S3B) seems difficult to reconcile with their lower O-GlcNAc level comparable (lower) O-GlcNAc level to the hemizygous males (Fig. 3C). It is possible that the western blot technique is not sensitive enough to detect subtle differences in O-GlcNAcylation. An alternative hypothesis, if O-GlcNAc levels were truly identical between Y851A/Y and Y851A/Y851A, could be the existence of a modifier in female that could be a XCI-escapee."

C22: 6. In terms of the transcriptome in OgtY851A mice, similar to comment 4, the authors should confirm their transcriptomics data shown as Figure 3D by qPCR. In addition, the authors should describe the potential mechanisms by which the differentiation of precursor cells of LaTPs and JZPs were disrupted. Were master regulators of the differentiation known to be O-GlcNAcylated and loss of O-GlcNAcylation perturbed the function?

R22: As for the whole embryo discussed in R20.2, we also interpret cautiously the gene expression phenotype observed in the placenta. Specifically, we state in the manuscript that it could either be caused by an impact of lower O-GlcNAcylation on placental differentiation or by a general delay in placentation or in the development of the embryo as a whole. The hypothesis of a general delay (of the whole embryo and/or of placental formation specifically) is supported by the downregulation of essentially all markers of more differentiated cell types and the upregulation of the precursor marker. We favor this hypothesis because it is consistent with what observed with the T931 mutants and also with the enzymatic removal of O-GlcNAc in the zygote (Formichetti et al., 2024 BioRxiv). Because of the thousands of O-GlcNAcylated proteins present in the cell, it is impossible to know which is the responsible molecular mechanism, which could even start at much earlier stages.

Minor Comments

C23: 1. Regarding DFP461-463 mutant, I couldn't understand the point of this figure because the results had no difference, and the meaning of the mutation was quite different from the others. Thus, the figure was awkward and a little confusing to me. If the authors still want to include the figures, I would suggest that they should reorganize the position of the figure (maybe after figure 3 is better to show you had tried to investigate the effects of nuclear localization of OGT on the changes of transcriptomes) and add some results. Since WT OGT seems to be localized mainly in the cytosol at steady state (Figure S1B and S1C), the effect of mutation on its nuclear localization should not be obvious. Therefore, it is difficult to conclude the mutation had no effect on the nuclear localization unless the ratio of nuclear and cytosol localization is quantified. Also, I wonder whether the O-GlcNAc levels of nuclear and cytosolic proteins in the mutant cells were comparable to those in WT cells. If this is the case, the results would also support the authors' conclusion.

R23: We took the comments on board and made it clearer that the rationale for the DFP461-463 mutant was an attempt to separate OGT's nuclear and cytosolic functions. We fully agree that these results are peripheral, and thus we presented these results in Supplementary Figure 1 (not in the main figure).

The biochemical evidence presented in Fig S1C shows that the genetic substitution of DFP to AAA on endogenous OGT has no detectable impact on its nuclear localization in primary MEFs. This result is far more authoritative than the evidence provided by Seo et al. 2016 (doi: 10.1038/srep34614), which is based on the overexpression of OGT transgenes in HeLa cells. Importantly, Seo et al. 2016 did not assess the impact of their mutations on endogenous OGT.

We believe that the negative results we obtained with the DFP461-463 mouse model shall be extremely valuable for the field. Firstly, science can move forward only if both negative and positive results are shared. In this specific case, we found that mutation of endogenous OGT in MEFs yielded to a different result than previously reported overexpression of the same mutant construct in HeLa cells. Secondly, we want to make the Ogt-NLS- mouse model available for further investigations.

C24: 2. Since OGT or O-GlcNAcylation regulates chromatin status, the authors analyzed the gene expression profiles of retrotransposons in T931del/Y or T931A/Y mice. Is it possible to investigate if the release of gene silencing is also seen in non-retrotransposon genes? I assumed retrotransposons might be a well-established system to analyze gene silencing status, however, if the authors could find similar effects on genes other than retrotransposons, that would be highly valuable.

R24: This is an interesting idea. This notion refers to the activation of promoters that are normally epigenetically repressed (e.g. silent despite the presence of all trans-active factors required for their expression). Epigenetically repressed promoters include retrotransposons, imprinted genes and germline specific genes that are normally expressed in germ cells and maintained in a repressed state in somatic cells (10.1038/s41580-019-0159-6). Testing of mono-allelic expression of imprinted genes required F1-hybrid. Thus, we assessed whether well-studied germline specific genes could be realized from silencing in T931del/Y or T931A/Y blastocyst and found no evidence for it (see dot plot below). The unbiased transcriptomic analysis presented in the manuscript shows that the product of upregulated genes are enriched in mRNA processing (Figure 2E), but these genes are not normally epigenetically repressed. Thus, contrary to retrotransposons, the role of O-GlcNAc at cellular gene promoters appears not to be linked to epigenetic silencing. This could be explained by the many different protein substrates for O-GlcNAc.

C25: 3. OgtY851A mice with milder OGT activity loss didn't exhibit impaired preimplantation development, but did display postimplantation development such as placental development, suggesting that O-GlcNAcylation of proteins required for preimplantation and postimplantation development relies on different degrees of OGT activity. I wonder whether global O-GlcNAc levels in embryos in preimplantation and postimplantation developmental stages are different or not. This might include both the pattern of blotting and intensities. The results would give the authors an explanation why the dependency on OGT activity was different in two developmental stages. Can the authors provide data? If not, then the authors should at least describe hypotheses in the manuscript to address these questions.

R25: We recently reported that the subcellular patterns of O-GlcNAc are highly dynamic during preimplantation development (Formichetti et al. 2024, BioRxiv). The most striking O-GlcNAc remodeling we observed is the enrichment of nuclear O-GlcNAc as compared to cytoplasmic O-GlcNAc that is concomitant to embryonic genome activation (Formichetti et al. 2024, BioRxiv). We quantified the ratio of the nuclear/cytoplasmic signal by immunofluorescence, but absolute quantification is not possible with this method. Due to the limited number of cells of the preimplantation embryo, this analysis cannot be performed by western blot. Hence, there is no appropriate method to quantitatively compare O-GlcNAc levels between preimplantation and postimplantation embryos.

C26: 4. The authors' AID-degron system elegantly worked in MEFs but was inefficient in preimplantation embryos. I wonder if this was because of the high expression of the shorter isoform of OGT detected as OGTp78 in the author's western blot. Is it possible to examine this possibility in the embryos? Either way, the authors should describe a potential explanation for why the efficiency in the embryos was low. In addition, the authors should describe why they inserted the AID tag only into the longest OGT isoform.

R26: This is a good point. The smallest isoform OGTp78 bears the catalytic domain and thus can partially compensate for the degradation of OGTp110. Note that the level of OGTp78 is low and does not increase upon OGTp110 degradation; thus a compensation can only be partial (Figures S4A and S4D). Alternative hypotheses for the ineffectiveness of the degron system in ex vivo grown embryos include: i) the expression level of OsTIR that may be too low in the early embryo (Rosa26 promoter not being activated at EGA), ii) a possible steric hindrance of the N-ter AID tag in these cells, iii) the lower concentration of Auxin imposed by toxicity on the embryo is likely suboptimal. Testing these possibilities is very difficult in preimplantation embryos.

It is unclear how the OGTp78 isoform is produced; it was hypothesized to originate from an alternative transcription start site (https://doi.org/10.1007/s00335-001-2108-9). We initially attempted to target both isoforms by inserting the AID tag at the C-terminus, but we were unsuccessful in producing this mouse model. It is possible that the C-terminus that is near the catalytic site cannot tolerate the AID knock-in.

C27: 5. In Figure S1C, is the band detected right below OGTp78 in nuclei fractions non-specific or do both bands correspond to OGTp78 ?

R27: To answer this question, a knockout control would be needed. OGTp78 being not targeted by our AID-degron, we cannot test the specificity of these bands using our perturbation tool kit.

C28: 6. Figure 1D top row third column: hemizgous -> hemizygous

R28: Many thanks; the embarrassing typo has been corrected.

C29: 7. Figure 1D second row third column: hemyzygous -> hemizygous

R29: Thanks for bringing this other typo to our attention, it is now corrected.

Reviewer #4 (Significance (Required)):

General assessment: strengths and limitations

C30: Strength: This manuscript elegantly revealed the requirement of OGT in mammalian development by taking advantage knock-in mouse models with different OGT activity. In addition, the manuscript provided the interesting and important transcriptomics data in both pre- and post-implantation embryos of OGT mutant mice. These data sets could explain detailed mechanisms how OGT or O-GlcNAcylation regulates mammalian development in the future. Furthermore, development of AID-tagged OGT system would be a useful tool for other researchers studying OGT function.

Limitation: Although they found interesting changes in terms transcriptomes in developing mice with different OGT activity, they lack the data showing how these changes caused the observed phenotypes. In other words, there are less mechanistic insights behind the developmental problems seen in mice with different OGT activity.

In addition, although I agree the question about whether OGT activity itself is crucial for the early development of mammals has not been completely solved for a long time, I assume people thought OGT activity is actually important for the mammalian development thorough the observation of OGT-linked congenital disorders of glycosylation.

Therefore, I would say the novelty of the manuscript is a little less impactful. Furthermore, although AID-tagged OGT system revealed fundamental questions regarding the transcriptional changes upon acute depletion of OGT in cellular levels, the system was inefficient in mouse embryos. So, they showed nothing about developmental-stage specific requirements of OGT.

Advance: The manuscript can fill a current gap regarding requirement of OGT in mammalian development. Also, the manuscript developed a series of mutant mice with different OGT activity and an AID-tagged OGT mouse line. These mice provide technical advances.

Audience: The manuscript will be interested in researchers in specific fields such as glycobiology, developmental biology, and clinical fields.

Describe your expertise: Biochemistry, Glycobiology, Cell biology

R30: We are thankful for the constructive and supportive review.

We fully agree with the limitations of the study and discussed them in the manuscript. Our in vivo approach revealed the most phenotypically relevant transcriptional phenotypes resulting from OGT catalytic impairment during embryonic development. We make the mouse models created for this study available to the community to facilitate follow-up studies aiming at exploring the underlying molecular details.

As pointed out in the comments, the requirement of OGT glycosyltransferase activity for mammalian development was widely assumed by the field, but this belief was without direct experimental evidence. This study provides the first in vivo evidence for this important conclusion.

Conclusion: The reviewers' comments were tremendously useful to improving the clarity of the manuscript and adding important new in vivo evidence. We note that none of the reviewers provided any reason to doubt our important conclusions:

• The demonstration that the enzymatic activity of Ogt, thus the O-GlcNAc modification itself, is essential for preimplantation development.
• The finding that a mild reduction of OGT's activity is sufficient to perturb the silencing of multiple families of retrotransposons in the growing embryo.

• The indication, from transcriptomes of hypo-O-GlcNAcylated embryos, of a developmental retardation upon a mild O-GlcNAc perturbation.

• The discovery that OGT's rapid depletion in vitro downregulates basal cellular function, including translation. This result provides mechanistic support to the embryonic growth delay resulting from decreasing O-GlcNAc in vivo.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #4

Evidence, reproducibility and clarity

Summary

O-GlcNAcylation is the fundamental post-translational modification of numerous nuclear and cytosolic proteins. OGT is the sole enzyme catalyzing O-GlcNAc addition onto the proteins. The essentiality of OGT for early development and cellular viability has been established by using OGT-KO mice and cell lines. However, it remains to be elucidated whether the catalytic activity of OGT is required for the early development, and if the catalytic activity of OGT is required what are the functions of OGT or O-GlcNAcylation in early development due to a lack of appropriate mouse models. To overcome the technical difficulty of manipulating the levels of O-GlcNAcylation in early embryos, Formichetti et al. created the series of four mouse models (OgtY851A, OgtT931A, OgtQ849N, and OgtH568A) with different OGT activity by introducing single amino acid substitution in the catalytic domain. By analyzing the inheritance of the hypomorphic OGT alleles and the lethality of mouse embryos, they discovered OGT activity is a critical factor for early development. Subsequently, RNA-seq analyses with two mouse models showing the maternal inheritance of the hypomorphic OGT alleles indicated that sever hypo-OGT activity altered transcription and silencing of retrotransposon in preimplantation development while mild reduction of OGT's activity affected placental development in a sexually dimorphic manner rather than preimplantation development. Furthermore, to study the function of OGT at specific developmental stages, they developed a mouse model bearing endogenously AID-tagged OGT for acute degradation of OGT. Although the degron system wasn't efficient in preimplantation embryos, they discovered quick transcriptional changes upon OGT deletion in MEFs. The quality of the manuscript is good because the question to be solved was appropriately set, the approach was well designed, and their findings were interesting, although their writing was sometimes hard to understand as I raised in my following comments. Nevertheless, there are several points to be fixed before being published.

Major Comments

1. Although the authors showed in vitro activity of each mutant of OGT used in this manuscript by referencing the previous literature, they never showed the levels of global O-GlcNAcylation (and OGT itself) in their established mouse embryos. Although it could be impossible to determine O-GlcNAc levels in OgtQ849N and OgtH568A embryos because of the lack of germline transmission and founder line, respectively, they could do that in OgtY851A and OgtT931A embryos. Given that Y851A and T931A mutants had similar VMAX/KM with different VMAX, it is possible that their activity is comparable or Y851A has even lower activity in vivo depending on the concentration of UDP-GlcNAc in embryos. Therefore, it is critical to assess whether in vivo OGT activity is correlated with that in vitro as expected to conclude that severity of sub-Mendelian inheritance is proportional to the reduction of activity of OGT in vivo. Moreover, since the authors developed the elegant system to deplete OGT, the activity of Q849N and H568A mutant OGT can be examined at least in cells by expressing them in MEFs with OGT-degron system. Thus, I propose determination of global O-GlcNAc levels compensated by OGT levels by western blotting in OgtY851A, and OgtT931A embryos or MEFs with the OGT degron system re-expressing the individual four mutant OGTs. If the protein amount is insufficient for western blotting in the embryos because of the sizes of the earlier stages of embryos, I believe the author could address this by utilizing immunofluorescence as shown in Figure S5.
2. I didn't understand why the authors couldn't find any founder lines of the OgtH568A mutant. Was that because mosaic mice with OgtH568A mutation are lethal? Also, I believe there was no explanation why the OgtQ849N allele showed no maternal inheritance. Was that because Q849N possesses enough activity for sustaining mosaic embryos, but not oocytes? The authors should better explain these points in the manuscript text.
3. The authors serendipitously found a T931del-allele in the "WT" allele of the OgtT931A line, and suggested that T931del had milder activity loss, although the lethality of embryos was greatly mitigated. Nevertheless, transcriptome analyses in male blastocysts revealed that 120 genes' expression was changed in T931del/Y males. This raised the question about which mutant OGT has higher activity, Y851A or T931del. I think comparing the activity of Y851A and T931del mutants in MEFs with OGT-degron system is important to confirm the proportional relationship between activity and phenotypic severity.
4. Regarding transcriptomes of T931del/Y, the authors found the upregulation of proteasomal activity and stress granules along with the downregulation of amino acid metabolism, mitochondrial respiration, and so on. To validate the results, the authors should perform qPCR on several up- or down-regulated genes. In addition, according to Figure S2E, the authors pointed out that at least for genes upregulated in OgtT931A embryos, the changes were not explained by a developmentally delayed transcriptome, suggesting that upregulation of these genes was the cause of developmental delay. Therefore, I strongly encourage them to discuss in the manuscript text how up-regulated genes could contribute to developmental delay.
5. Regarding the transcriptome in OgtY851A mice, Y851A/Y male mice had huge transcriptomic differences, while Y851A/Y851A female mice barely had any. Although it seems to agree with the number of Ogt alleles, I wonder whether other X-linked genes expressed higher in female placenta as shown in Figure 3C could attenuate the effects of decreased OGT activity. I don't think this possibility can be excluded, unless the authors further decrease OGT activity in Y851A/Y851A female placenta and obtain the similar results as for male placenta. Or if they compared the levels of global O-GlcNAcylation between Y851A/Y and Y851A/Y851A mouse placentas and discovered they had similar levels of O-GlcNAcylation, then the authors could conclude that the number of Ogt alleles was not the reason of sexual-dimorphism. The authors should determine the levels of O-GlcNAcylation in Y851A/Y and Y851A/Y851A mouse placentas and/or at least discuss the above possibilities in the manuscript text.
6. In terms of the transcriptome in OgtY851A mice, similar to comment 4, the authors should confirm their transcriptomics data shown as Figure 3D by qPCR. In addition, the authors should describe the potential mechanisms by which the differentiation of precursor cells of LaTPs and JZPs were disrupted. Were master regulators of the differentiation known to be O-GlcNAcylated and loss of O-GlcNAcylation perturbed the function?

Minor Comments

1. Regarding DFP461-463 mutant, I couldn't understand the point of this figure because the results had no difference, and the meaning of the mutation was quite different from the others. Thus, the figure was awkward and a little confusing to me. If the authors still want to include the figures, I would suggest that they should reorganize the position of the figure (maybe after figure 3 is better to show you had tried to investigate the effects of nuclear localization of OGT on the changes of transcriptomes) and add some results. Since WT OGT seems to be localized mainly in the cytosol at steady state (Figure S1B and S1C), the effect of mutation on its nuclear localization should not be obvious. Therefore, it is difficult to conclude the mutation had no effect on the nuclear localization unless the ratio of nuclear and cytosol localization is quantified. Also, I wonder whether the O-GlcNAc levels of nuclear and cytosolic proteins in the mutant cells were comparable to those in WT cells. If this is the case, the results would also support the authors' conclusion.
2. Since OGT or O-GlcNAcylation regulates chromatin status, the authors analyzed the gene expression profiles of retrotransposons in T931del/Y or T931A/Y mice. Is it possible to investigate if the release of gene silencing is also seen in non-retrotransposon genes? I assumed retrotransposons might be a well-established system to analyze gene silencing status, however, if the authors could find similar effects on genes other than retrotransposons, that would be highly valuable.
3. OgtY851A mice with milder OGT activity loss didn't exhibit impaired preimplantation development, but did display postimplantation development such as placental development, suggesting that O-GlcNAcylation of proteins required for preimplantation and postimplantation development relies on different degrees of OGT activity. I wonder whether global O-GlcNAc levels in embryos in preimplantation and postimplantation developmental stages are different or not. This might include both the pattern of blotting and intensities. The results would give the authors an explanation why the dependency on OGT activity was different in two developmental stages. Can the authors provide data? If not, then the authors should at least describe hypotheses in the manuscript to address these questions.
4. The authors' AID-degron system elegantly worked in MEFs but was inefficient in preimplantation embryos. I wonder if this was because of the high expression of the shorter isoform of OGT detected as OGTp78 in the author's western blot. Is it possible to examine this possibility in the embryos? Either way, the authors should describe a potential explanation for why the efficiency in the embryos was low. In addition, the authors should describe why they inserted the AID tag only into the longest OGT isoform.
5. In Figure S1C, is the band detected right below OGTp78 in nuclei fractions non-specific or do both bands correspond to OGTp78 ?
6. Figure 1D top row third column: hemizgous -> hemizygous
7. Figure 1D second row third column: hemyzygous -> hemizygous

Significance

General assessment: strengths and limitations

Strength: This manuscript elegantly revealed the requirement of OGT in mammalian development by taking advantage knock-in mouse models with different OGT activity. In addition, the manuscript provided the interesting and important transcriptomics data in both pre- and post-implantation embryos of OGT mutant mice. These data sets could explain detailed mechanisms how OGT or O-GlcNAcylation regulates mammalian development in the future. Furthermore, development of AID-tagged OGT system would be a useful tool for other researchers studying OGT function.

Limitation: Although they found interesting changes in terms transcriptomes in developing mice with different OGT activity, they lack the data showing how these changes caused the observed phenotypes. In other words, there are less mechanistic insights behind the developmental problems seen in mice with different OGT activity. In addition, although I agree the question about whether OGT activity itself is crucial for the early development of mammals has not been completely solved for a long time, I assume people thought OGT activity is actually important for the mammalian development thorough the observation of OGT-linked congenital disorders of glycosylation. Therefore, I would say the novelty of the manuscript is a little less impactful. Furthermore, although AID-tagged OGT system revealed fundamental questions regarding the transcriptional changes upon acute depletion of OGT in cellular levels, the system was inefficient in mouse embryos. So, they showed nothing about developmental-stage specific requirements of OGT.

Advance: The manuscript can fill a current gap regarding requirement of OGT in mammalian development. Also, the manuscript developed a series of mutant mice with different OGT activity and an AID-tagged OGT mouse line. These mice provide technical advances.

Audience: The manuscript will be interested in researchers in specific fields such as glycobiology, developmental biology, and clinical fields.

Describe your expertise: Biochemistry, Glycobiology, Cell biology

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #3

Evidence, reproducibility and clarity

This is a conceptually interesting paper that attempts to leverage the knowledge of OGT catalysis to begin to dissect OGT function. The evidence is presented I a straightforward fashion and is in general well documented. The breeding strategies are well informed and the paper draws heavily on previous work carried out in the mouse.

Significance

The paper describes tools which will help dissect the many potential roles of O-GlcNAc addition in early development. As it stands, this is a descriptive manuscript that will lead to hypothesis generation and testing and this should not be undervalued. The biological reagents produced and characterized will be of general interest to the field. Most of the findings presented represented a verification of existing ideas in the field but this is not meant as a criticism since part of the motivation for the approach was to generate a reproducible system for analyzing the biological phenomena.

There are perhaps some bioinformatic shortcuts taken that may need to be corrected upon thorough review. These do not lessen the overall impact of the contribution.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

Comments to authors

To investigate the function of OGT at specific developmental stages, the authors perturbed OGT's function in vivo by creating a murine allelic series featuring four single amino acid substitutions that variably reduced OGT's catalytic activity. The goal was to identify the direct effect of O-GlcNAcylation, using a sophisticated collection of genetic mutants to evaluate in vivo the role of this modification at early stages of development. Overall, the severity of embryonic lethality correlated with the extent of catalytic impairment of OGT, demonstrating that the O-GlcNAc modification is essential for early development.<br /> The study represents a substantial advance in our understanding of OGT and O-GlcNAcylation in mammalian development. The creation of novel murine models and inducible systems is an important contribution, providing powerful tools for future research in this field. The insights into the role of OGT's catalytic activity and its involvement in epigenetic regulation during embryonic development are noteworthy, opening new avenues for research. However, there are a few considerations and concerns:

Major:

1. An assumption of the study is that different mutations cause different levels of O-GlcNAcylation rather than alterations in substrate specificity. It might be important to test, at least in cultured cells, that the different mutations do not change the preference of OGT to modify certain proteins rather than others, which can provide alternative explanations for their findings.
2. In Fig 1D and 1H, the thresholds to define a gene or TE as differentially expressed are not strong. According to the figure legends, "any" change in terms of log2Fc was considered as DE and colored. I think the figures should illustrate better that the changes are subtle, by for example adding a dotted line (at least) in the value 0.5 of the y-axis. These subtle transcriptional changes should be reflected better in certain paragraphs where the expression of TEs are presented/and discussed as a hallmark of the absence of O-GlcNAcylation in the OGT-mutants. The same happens with Suppl Fig 3C (changes are very minor). Similarly, in Fig2C, the changes in gene expression are lower than log2FC 1 (which represent the double in absolute expression). Applying a stronger threshold, among the upregulated genes, only Xist will be significantly overexpressed. If a gentle threshold needs to be applied to this data, authors should at least justify the reasons behind doing so. Same for Fig2D.
3. In Figure 2B, the T931del allele was recovered in the blastocyst population with a very high frequency, even higher than the male WT group (T931del: 10; WT: 3). This observation suggests that the T931del allele did not significantly affect blastocyst survival. Further clarification or additional experiments might be necessary to understand the implications of this finding on early developmental stages.
4. Similarly, in Figure 2G, there is an apparent higher expression of TE expression in the T931A/Y embryos group than in the T931del/Y group, which combined with the higher frequency of blastocyst generated in this latest group it may indicate a deeper molecular consequence after the deletion of the T931. A comparison of the transcriptome between these two cell lines help to address this possibility. Also, the authors should compare the O-GlcNAc levels of WT, T931A, and T931del mutant blastocysts by immunostaining, similar to what was done in Figure S5F.
5. In Boulard et al. 2019 O-GlcNAcylation was shown to be sufficient to modulate expression of DNA methylation-dependent TEs. It would be interesting to know (or at least discuss) if the changes in TE expression observed in OGT-mutant embryos in this study involve changes in DNA methylation. Ideally, some DNA methylation measurement optimized for low input numbers of cells would be useful.
6. The data related with the OGT-degron system in MEs seem disconnected with the rest of the manuscript. While the developmental models (blastocyst, etc) elegantly assess the contribution of O-GlcNAcylation to the control of cell survival and gene expression through the use of different OGT mutants, the degron system is a system of graded depletion that unfortunately was only possible to be used in MEFs (instead of embryos). Thus, the results obtained with the degron system in MEFs are difficult to intersect with the data from the use of OGT-mutants in embryos. Even though there are obvious interesting questions that one may want to know about this OGT degron MEF system, none of them would demonstrate a direct role for O-GlcNAcylation in cellular function, the major point addressed in the developmental system. Using the degron system in embryonic stem cells might have provided a more parallel comparison. The authors should discuss this point in more detail and either use ESC instead of MEFs or provide a stronger justification for the use of MEFs over ESC.

Minor:

1. In Fig 2C the color and shape codes are confusing to understand - there are some colors/shapes that are not represented in the PCA plot. The same in Fig 3H, where in the PCA plot there are pink triangles that do not match with the code legends.
2. In the figure legends of Figures 2D, 2E, 2F, and 2H, the notation should be corrected from "OgtT931A/Y" to "OgtT931del/Y".

Significance

To investigate the function of OGT at specific developmental stages, the authors perturbed OGT's function in vivo by creating a murine allelic series featuring four single amino acid substitutions that variably reduced OGT's catalytic activity. The goal was to identify the direct effect of O-GlcNAcylation, using a sophisticated collection of genetic mutants to evaluate in vivo the role of this modification at early stages of development. Overall, the severity of embryonic lethality correlated with the extent of catalytic impairment of OGT, demonstrating that the O-GlcNAc modification is essential for early development.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

Formichetti at el. developed mice with OGT catalytic dead mutations and then studied their function during early embryogenesis. Not surprisingly, dramatic reduction in OGT activity failed to produce embryos; however, mild reduction in OGT did produce animals. The authors then use the T931 animals that have a mild reduction in activity to further characterize the function in the early embryo. Not surprisingly, male mice showed changes in gene expression, implantation sub-lethality, and an uptick in loss of retrotransposon silencing. The authors also show that an even milder reduction in OGT activity (Y851A) effects male placenta function and chromatin remodeling. Finally, the authors make a less stable OGT transgene within the mouse and again found embryogenesis issues in the males and alterations in numerous gene families including mTOR signaling and p53 function. All in all, this is an interesting study that track functions of OGT in early embryonic development. The studies are well-controlled and rigorous.

Significance

This is a good study and novel. Not only is it of interest to reproductive biologist, but it echos themes found in O-GlcNAc biology.

Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

Learn more at Review Commons

---

Reply to the reviewers

Point-to-point answer to reviewers comments

Reviewer #1

Evidence, reproducibility and clarity

*Summary: *

*The study by Cottignies-Calamarte et al. describes that AMP-activated protein kinase (AMPK) regulates cell energy balance by suppressing energy-consuming pathways like lipid and protein synthesis and promoting nutrient availability through autophagy. These pathways contribute to SARS-CoV-2 infection by hijacking autophagy and accumulating lipid droplets for viral replication. The antiviral activity of MK-8722, a direct pan-AMPK allosteric activator, was evaluated in vitro. MK-8722 effectively inhibited Alpha and Omicron SARS-CoV-2 variants in Vero76 and human bronchial epithelial Calu-3 cells at micromolar concentrations. This inhibition restored autophagic flux, degrading newly synthesized viral proteins, and reduced lipid metabolism, affecting viral factories. Additionally, MK-8722 treatment increased the type I interferon (IFN-I) response. Post-infection treatment with MK-8722 efficiently suppressed viral replication and restored the IFN-I response without altering the SARS-CoV-2-specific CD8+ T cell response elicited by Spike vaccination. The authors concluded that, MK-8722 acts as an effective antiviral against SARS-CoV-2 infection, even when applied post-exposure, suggesting potential for preclinical tests to inhibit viral replication and alleviate patient symptoms. *

__Major comments: __

• • Are the key conclusions convincing?** Partially. See comments below! *

- Should the authors qualify some of their claims as preliminary or speculative, or remove them altogether?

From my perspective, the title "Direct pharmacological AMPK activation inhibits mucosal SARS-CoV-2 infection by reducing lipid metabolism, restoring autophagy flux and the type I IFN response" is a clear overstatement. In no way can the authors make statements about the autophagic flux, as it simply was not measured. The study would greatly benefit from conducting an autophagy/autophagic flux assay. See specific comments below!

Answer: As suggested by the author, we have now investigated the autophagic flux by staining the cells for LC3b expression and colocalization with the lysosomal marker LAMP1. Results are shown in the new Fig4.D and E and detailed in the results section to read lines 468-475 page 23-24:

“To assess directly the impact of MK-8722 on the autophagic flux, Calu3 cells were infected by SARS-CoV-2 without and with MK-8722 (5uM), double labelled with LC3b and LAMP1, and co-localisation of the two marker quantified (Fig.4D). MK-8722 treatment, compared with no treatment, increased LC3b colocalization in the LAMP1 compartment as shown by the increase in MOCs in treated versus non treated infected cells (LAMP1 signal in LC3b signal 0.078±0.014 vs 0.01±0.006, Mann-Whitney pand lines 480-481 page 24:

“This result indicates that MK-8722 restores the autophagic flux to address viral components to the lysosome, where they are degraded.”

- Would additional experiments be essential to support the claims of the paper? Request additional experiments only where necessary for the paper as it is, and do not ask authors to open new lines of experimentation.

See below specific comments regarding cell line consistency and autophagy measurements.

- Are the suggested experiments realistic in terms of time and resources? It would help if you could add an estimated cost and time investment for substantial experiments.

This question depends on various factors such as access to relevant biosafety labs, availability of required reagents, etc. In my estimation, experiments involving WT viruses and autophagy measurements could be conducted within 3-4 months. The proposed experiments with the delta-N-SARS-CoV-2 mutants, of course, depend on access to such viruses. Overall, I believe all experiments could be completed within 6 months. The costs of those assays are not very high.

- Are the data and the methods presented in such a way that they can be reproduced?

In the present study, a total of 3 cell lines and human PBMCs were utilized for various experiments. Please indicate why each cellular model was chosen and highlight the differences between these models considering what is known for SARS-CoV-2 infection and autophagy! Furthermore, the study would greatly benefit if the key findings were consistently demonstrated in a single cell line.

Answer: We agree with the reviewer that three cell lines and PBMC from patients were used but it was necessary for the experimental design of our experiments as detailed below.

In Figures 1 and 2, experiments use both Vero76 and Calu3 cells for the following reasons. We first used the simian Vero76 cells in order to validate the inhibitory effect of MK-8722 in widely a used cell line in virology, but which lacks the interferon response. This lack of the IFN response renders Vero cells a poor model for the pathophysiology of SARS-CoV-2. We therefore then used the IFN-competent human lung Calu-3 cell line as a more relevant cell model.

In Figures 3 both cells were again and western blots show a similar pattern of activation downstream AMPK after activation by MK-8722 and similar antiviral activity of MK-8722 in Vero76 and Calu3 cells. In Figure.4, we then mostly focused on deciphering the antiviral mechanism of MK-8722 and therefore focused on Calu3, which is more relevant from a pathophysiological point of view.

In Figure 5, we then investigated the T-cell response and therefore used primary human, namely PBMC/purified CD8 T-cells from healthy donors. As the reviewer knows, T-cell activation is restricted by HLA-TCR interaction, or in other words, only matched HLA cells can activate CD8 T-cells. As HLA-A2 is the main HLA expressed in human and only a HLA-A2 antibody is available on the market, for these experiments we used only CD8 T cells from HLA-A2 patients and had to find an additional HLA-A2-expressing epithelial cell susceptible to SARS-CoV-2 infection. It is the case of Caco2 cells, which are HLA-A2+ susceptible to infection and competent for IFN responses. We could not use simian Vero cells that are IFN-deficient and do not express human MHC, nor Calu3 cells being HLA-A2 negative. Altogether, in experiments in Figure 5 addressing HLA-A2 restricted antigen presentation, the use of Caco-2 cells was appropriate in contrast to that of Vero or Calu3.

Furthermore, for sake of clarification, we have now added the following p. 19 lines 390-392:

“The trends in modification of all markers by MK-8722 treatment by being conserved between cell lines indicates a common antiviral mechanism. We therefore focused our study on Calu-3 cells since they are more relevant to SARS-CoV-2 infection.”

The authors conclude that selective activation of AMPK has a pro-autophagic effect which in turn leads to a reduced SARS-CoV-2 replication. I would generally agree with this statement, but throughout the entire manuscript, no real autophagy assays are shown. This should definitely be rectified. It is important to demonstrate that (1) MK-8722 is capable of increasing autophagy and particularly autophagic flux in the cell models used, (2) that in the cell models employed, SARS-CoV-2 infection leads to modulation of autophagy, and (3) that SARS-CoV-2 infected cells, when co-treated with MK-8722, lead to a re-established autophagy. The autophagy assays should be performed according to the expert-curated guidelines by Klionsky et al. This is extremely important so that the results can be compared with the now vast number of existing autophagy-SARS-CoV-2 studies.

We fully agree with the reviewer that it was important for our study to formally address the impact of MK-8722 on the autophagic flux. Following reviewer recommendation and reading of the guidelines for autophagy study, we therefore have evaluated whether MK-8722 affected localization of LC3b, which is essential for autophagosome biogenesis/maturation and also functions as an adaptor protein for selective autophagy, in lysosomal compartment (labelled, by LAMP1). Therefore we labelled cells infected in the presence or absence of MK-8722 for LC3b expression and colocalization with the lysosomal marker LAMP1. Results are shown in the new Fig4.D, E. Please see our above answer (p1) for results description. These results indicate that MK-8722 restored the autophagic flux that had been interrupted by SARS-CoV-2 infection in Calu3 cells.

- Are the experiments adequately replicated and statistical analysis adequate?

Minor comments:

- Specific experimental issues that are easily addressable.

Typo: Line 288: It should be "MK-8722" instead of "MK-7288".

In general, a space between value and unit is not consistently used.

Please indicate always the phosphosite of substrate proteins when phosphorylation is described. E.g. line 288 and throughout the manuscript: regarding ACC phosphorylation.

Answer: We apologize for these typos, which have now been corrected in the revised MS.

• Are prior studies referenced appropriately?

See comment below. Fundmental work that describes the virus-autophagy relationship, such as the work by the Beth Levine lab would be important to add. Also the work bei Konstantin Sparrer and colleagues is important and leads to the current work presented here.

Answer: Konstantin Sparrer’s work is already cited as by ref 17. However, as suggested by the reviewer, the work by the Beth Livine lab has now been added in the revised MS in p29-30 lines 610-613, to read:

.“Convergence of Beclin-Atg14 and P62 activation could stimulated the selective clearance of viral components, in a process called virophagy64–66. We thus propose that AMPK pharmacological activation induce virophagy and is responsible, at least partially, for its antiviral effect.”

- Are the text and figures clear and accurate?

Introduction: In general, for some of the statements claimed in the introduction, which is in sum nicely written, informative and well structured, citations are required. For example - line 59 "...autophagy is sequentially activated and inhibited."

Answer: Citation has been added, namely: Koepke 2021 Autophagy ref 17

Lines 59-65. In the introduction, the interaction between autophagy and SARS-CoV-2 is primarily described in a one-sided manner. There are now several studies demonstrating that both inhibition of autophagy and also induction of autophagy in context of coronaviral infection. Both aspects should be illuminated and introduced here.

Answer: The reviewer points towards a crucial point of autophagy during ß-coronaviruses infection. Indeed, we fully agree that the autophagy is both inhibited and activated, at different levels, by different proteins as exemplified by, Sparrer’s group (ref 17). As we have mentioned in the initial version of our MS: “Throughout the SARS-CoV-2 viral cycle, autophagy is sequentially activated and inhibited 17–20”. As it may have not been clear enough, we have reformulated this paragraph as follows to read line 68-75 p5:

“Throughout the SARS-CoV-2 viral cycle, autophagy is sequentially activated and inhibited 17–20. Indeed, autophagy is initiated by the early expressed nsp6, resulting in the formation of autophagosomes that are essential for the establishment of viral factories 17,21 and subsequent viral proteins expression 17,21. In turn, viral proteins OFR3a and ORF7a expressed at a latter time post-infection, prevent the fusion between autophagosomes and lysosomes, thereby blocking completion of autophagy, as evidenced by increased LC3-B expression, activation of the ULK1 kinase and increase in the autophagy cargo receptor sequestosome-1/p62. Overall, this disruption of autophagy protects newly formed virus from degradation in the LAMP1+ lysosome 19,22–25”

Discussion: The selectivity of the compound should be discussed.

Answer: Manuscript have been revised as follows to discuss both specificity in regards of AMPK and tissue accessibility:

• lines 557-560, p27: “We show here that the blockade of AMPK activation upon infection can be reversed by MK-8722, the pharmacological allosteric pan-activator of AMPK, which blocks infection at a µM concentration, in agreement with the predicted role of AMPK activity on SARS-CoV-2 infection48 and with MK-8722 action on infection by other viruses49”

• Line 576-581, p28: “In contrast, MK-8722, as systemic drug, may reach these tissues with minimal side effects, as a daily treatment in diabetic Non-human primates (NHPs) with MK-8722 (10 mg/kg) for a month induced only a limited and reversible cardiac hypertrophy 36”

• Do you have suggestions that would help the authors improve the presentation of their data and conclusions?

The presentation of the data is understandable. For reader with a less mechanistic background it would be helpful to present a schematic figure like a graphical abstract.

Answer: As suggested by the reviewer, we have now introduced a graphical abstract in the revised MS.

SECTION B – Significance

======================== - Describe the nature and significance of the advance (e.g. conceptual, technical, clinical) for the field.

• Even though there is now a plethora of studies on autophagy and SARS-CoV-2, this study is important and of great interest to a broad readership. Not only virologists, immunologists, and autophagy researchers will eagerly anticipate this study, but especially researchers focusing on pharmacology around AMPK and autophagy will recognize the importance of the data presented here.*

Answer: We thank the reviewer for this positive appreciation of our work.

- Place the work in the context of the existing literature (provide references, where appropriate).

The work builds upon a variety of studies on the interplay between coronaviruses and the mechanism of autophagy. Foundational contributions to this research stem from groundbreaking preliminary work conducted in the laboratories of Gassen and Müller (Gassen et al. 2019 and Gassen et al. 2021), as well as general virus-autophagy studies from the laboratory of Beth Levine. Additionally, recent work by Konstantin Sparrer and colleagues would be important to cite, as it underscores the insights gained in this manuscript.

Answer: As suggested and previously mentioned, these references have now been added to the discussion p27 lines 557-560: “We show here that the blockade of AMPK activation upon infection can be reversed by MK-8722, the pharmacological allosteric pan-activator of AMPK, which blocks infection at a µM concentration, in agreement with the predicted role of AMPK activity on SARS-CoV-2 infection48 and with MK-8722 action on infection by other viruses49”.

In addition, Gassen et al. 2019 is already cited as ref 18

- State what audience might be interested in and influenced by the reported findings. See comment above!__ __

Reviewer #2

Evidence, reproducibility and clarity

In the current manuscript, Cottignes-Calamarte et al. have shown tha pharmacological activation of AMPK can be a strategy for overcoming SARS-CoV-2 infection induced reprogramming of host degradation pathways and innate immune response, without hindering the efficacy of spike expression from vaccine agents. Though the suggestion of selective activity of MK-8722 in degrading viral proteins in infection but not the ectopic spike peptides expression is interesting, the evaluation of the mechanism and providing therapeutic index for the drug will overall improve the study.

• Majorly, I have these suggestions;*

• • The manuscript did not clarify the mechanism clearly but correlated the reduction in viral proteins and their association to LAMP1 as the mechanisim of activity of MK-8722. In this way, the authors did not separate whether the reduction in SARS-CoV-2 infection could lead to the potentiation of these effects. There is mentioning of potential mechanism of the drug by inhibiting the activity of SARS-CoV-2 proteins that reduce the autophagic flux without directly showing this. SARS-CoV-2 Orf3a is a known inhibitor of autophagosome maturation and hence the authors should directly probe the activity of MK-8722 in overcoming the suppression of autophagic flux in cells by ectopically expressing Orf3a.* Answer: We fully agree with the reviewer that our work correlated the reduction in viral proteins and their association to LAMP1 to explain the mechanism of MK-8722 activity and did not focus on particular viral protein(s) that could induced autophagic flux.

Indeed, our approach has been to describe the MK-8722 antiviral activity against full primary SARS-CoV-2 viruses (not ectopically expressed viral proteins or recombinant viruses) in a pathologically relevant cell model including primary CD8+ T cells to mimic at best the initial steps of SARS-CoV-2 infection. Although the ectopical expression of Orf3a is recognized as a potent tool to study the principle of autophagy in the guidelines of autophagy measurement methods, when applied to infection, this ectopical expression of Orf3a will modify the endogenous level of Orf3a (as compared to infection level) and probably affect by itself the autophagy kinetics. To address the question of autophagic flux as suggested by the reviewer, we therefore preferred to use another recommended technique from the guidelines of autophagy measurement methods directly applicable to infected cells. We now evaluated the colocalization of LC3b with LAMP1 during treatment and infection of Calu-3 cells by primary viruses as now described in figure 4.D and E and discussed lines 468-475 pages 23-24 :

“To assess directly the impact of MK-8722 on the autophagic flux, Calu3 cells were infected by SARS-CoV-2 without and with MK-8722 (5uM), double labelled with LC3b and LAMP1, and co-localisation of the two marker quantified (Fig.4D, E). MK-8722 treatment, compared with no treatment, increased LC3b colocalization in the LAMP1 compartment as shown by the increase in MOCs in treated versus non treated infected cells (LAMP1 signal in LC3b signal 0.078±0.014 vs 0.01±0.006, Mann-Whitney p * As the authors propose MK-8722 as a preclinical candidate, they should present therapeutic measures and indexes. All across the data presented, there was no mentioning or measurement of drug toxicity for extended uses up to 36h pi.*

Answer: Our study aimed to describe the antiviral activity of MK-8722 in a cellular model mimicking only the initial steps of infection up to 3 dpi. Complementary experiments were conducted as required and show that treatment with MK-8722 up to 10mM did not result in an increase in cell death at any time post treatment as shown in the new Fig S2I and corresponding description in line 328-332 p17, which read__: __

__“__Furthermore, we also investigated MK-8722 toxicity at 4h, 24h and 96h (Fig. S2I) and found that the drug was not toxic up to 10mM. The 50% toxicity dose was calculated to be 57mM at 96h in Calu3 cells. This allows us to determine a therapeutic index of 76 against the SARS-CoV-2 Alpha variant and 36 against the Omicron variant, thus, placing MK-8722 as an attractive antiviral candidate.”

• Also, it was not clarified if the drug reduced the replication or exclusively worked by restoration of autophagic flux. For the earlier, the authors can consider two assays, i.e., a direct assay of looking into the replicon of SARS-CoV-2 (Bigotti et al 2024; PMID 38387750), or looking at earlier time points of infection (up to 8h pi; Twu et al. 2021; PMID 34788596) with intracellular sgRNA specific RNA probes.In this regards, 24h or 32h (Fig 1F-G) is too long to only measure single-round of infection.*

Answer: We thank the reviewer for this interesting question. The drug likely acts on both steps. Concerning replication, as shown in figure 1 GH at 24 and 32hrs, there is a block in replication but not in virus entry into the cell, since there is no difference in viral N cellular content after 1hr chase. Concerning the autophagic flux, the new Fig4 D and E shows that MK-8722 restores the autophagic flux in infected cells, which had been interrupted by the infection in the absence of the drug.

The viral cycle of ß-coronaviruses is highly dependent on hijacking cellular autophagy and lipid biosynthesis, which are both necessary for the assembly of DMV, essential for viral replication. Since MK-8722 treatment reduces cellular lipid content and increases autophagic flux, the use of viral replicons deficient in structural proteins or early post-infection timepoints will most likely show a decrease in viral genome replication under the effect of MK-8722 due to the inability of the replicons to establish such favourable environments (indicated in the corresponding Fig3A, B and C). However, the effect of MK-8722 on the replicon system will not distinguish between whether MK-8722 affects the viral replication complex and whether it is unable to induce viral factory formation.

• Are the viral components degraded by restored lysosomal activity include replicase or replication organelle components? This needs to be shown with intracellular nsp3/nsp4 levels in infection or ectopic expression.*

Answer: The reviewer raises a question highly relevant to the biology of coronaviruses, which we have addressed in Figure 4. Indeed, the Replication-Transcription Complex (RTC) is a multifactorial complex, in which N protein bound to the viral RNA is believed to help recruiting RdRp (Cong 2020 J Virol 10.1128/jvi.01925-19, Scherer 2022 Sci Adv DOI: 10.1126/sciadv.abl4895). The increased colocalisation of in N staining in LAMP-1+ compartments observed after MK-8722 treatment indicates that MK-8722 addresses RTC to the lysosomes and therefore our results indicate that replication-transcription organelles could be degraded in lysosomes.

• Previous studies suggested the decrease in AMPK phospharhorylation in SARS-CoV-2 infection What is inconsistent to previous studies should be commented by the authors. Eg, why is AMPK suppression not see in SARS-CoV- 2 inefction as reported before (Parthasarathy et al 2023; PMID 36417940)*

Answer: We agree with the reviewer and had already mentioned in the discussion that SARS-CoV-2 infection can have different outcome on AMPK activation as reported in ref 45 and 46 in the original version. Indeed, Gassen and colleagues reports that activation of AMPK and resulting phosphorylation are decreased in cells infected by SARS-CoV-2, as shown in (Ref 18), whereas Parthasarathy reports an increase in AMPK phosphorylation which is clear only at 96 h p.i. (Ref 47). Interestingly in this later study at earlier time point (24hr p.i.), authors report a tendency of AMPK phosphorylation to decrease although not in a statistically significant manner but for only n=3. In our present study, no statistically significant differences were found in AMPK phosphorylation although a small increase tendency is observed. Furthermore, cells used in all these studies differ: Gassen et al. used Vero-FM cells whereas Parthasarathy et al. used intestinal Caco2 cells, and in our study Vero 76 and lung Calu3 cells. Thus the differences reported on the effect of SARS-CoV-2 infection on AMPK activation are likely due to a question of kinetics and/or cellular background. Of note, the level of AMPK phosphorylation observed after SARS-CoV-2 infection is not to compare with the AMPK phosphoryalton induced by drug activation such as MK-8722.

This heterogenity in AMPK activation during SARS-CoV-2 infection might contribute to the various effect of AMPK activator as antiviral reported in the literature. Indeed, as mentioned p27 lines 560-563 “metformin, a drug approved by FDA since 1994, and the adenosine analogue AICAR that activates AMPK indirectly can inhibit replication of SARS-CoV-2 as well as Flaviviruses, but at a concentrations of 10 and 1 mM, respectively 35,47 ”. The reported discrepancy on the antiviral effect of AMPK activation might rely then on an activation threshold, as 10mM metformin and 1mM AICAR blocks infection, while 25__m__M AICAR does not (ref 18, 47). This later concentration was not evaluated in Gassen et al. on AMPK activation. Furthermore; the time frame evaluated in each study is different and could also be a confusing factor. Finally, as shown by Myers and colleagues (Myers 2017, Science ref 36), MK-8722 activates AMPK more efficiently than AICAR, most probably due to its direct activation of AMPK. We believe that direct AMPK activation by MK-8722 and the higher activation level of AMPK is responsible for the antiviral effect reported in our study.

As suggested by the reviewer, we have now clarified this section in the revised MS to read p27-28, lines 557-568:

“We show here that the blockade of AMPK activation upon infection can be reversed by MK-8722, the pharmacological allosteric pan-activator of AMPK, which, at a µM concentration, blocks infection, in agreement with the predicted role of AMPK activity on SARS-CoV-2 infection48 and with MK-8722 action on infection by other viruses49” .In line, metformin, a drug approved by FDA since 1994, and the adenosine analogue AICAR that activates AMPK indirectly can inhibit replication of SARS-CoV-2 as well as Flaviviruses, but at concentrations of 10 and 1 mM, respectively 35,47. These AMPK-sensitive viruses replicate all in viral factories, disturbing lipid synthesis and escaping autophagy 50,51. AMPK activation by metformin at high concentration (10mM) inhibits SARS-CoV-2 replication in vitro 47. However, AMPK activation with 5-amino-imidazolecarboxamide riboside (AICAR), a non-metabolised analogue of AMP able to activate AMPK, used at 25mM is unable to inhibit SARS-CoV-2 infection 18 while at 1mM has been proven to reduce by 10-fold viral production 47, suggesting that AMPK needs to reach an activation threshold to inhibit viral replication. “

• *

• Although AMPK activation is shown to be antiviral for SARS-CoV-2 before, e.g., with Metformin (Parthasarathy et al 2023; PMID 36417940), the role of AMPK-related kinases are shown to be pro-viral (e.g., NUAK2; Prasad et al. 2023; PMID 37421942). The authors should discuss these points to provide the reader a context in this sub-field.*

Answer: As suggested by the reviewer, we have now included in the discussion the following p28 lines 568-571:

“Conversely, NUAK2, an AMPK-related kinase, was reported to stimulate viral replication in A549 and Calu3 cells50. Altogether our results, in agreement with the literature, indicate that AMPK-dependent antiviral activity is restricted to AMPK-members only, confirming their distance with AMPK-related kinases such as NUAK2 53”

Minor

1. • Was there an increase of lipid droplets in SARS-CoV-2 infected cells to non-infected condition? It is not mentioned here.* Answer: Yes indeed, lipid droplets content was increased in Infected non treated cells as indicated in Figure Suppl. 4B (p21 lines 409-411) and as stated in the Result section as follows:

” Infection increased the overall Nile Red staining by 40±10% compared to non-infected cells (Fig. S4B, ANOVA: p * There are several typos and grammatical errors.*

Answer: We have now carefully checked the text for typographical and grammatical errors, which have been corrected.

There is no Fig S2I but is mentioend in the legend.

Answer: We apologize for this mistake in labelling figure S2 panel and have corrected their labelling as follows:

“____G-H: __ACE2 expression and viability were evaluated in Vero76 cells after 24h or in Calu-3 cells after 4days of MK-8722 continuous treatment (1 µM or 5 µM respectively) by flow cytometry. ACE2 expression is expressed as MFI (__G) and viability as frequency of cells non-stained by the amine-reactive dye Viobility (H). n=3 independent experiments. “

• Fig 4A - degradation is not shown, only association.*

Answer: We agree with the reviewer and have modified the text accordingly to read p37 lines 873-874:

Figure 4: MK-8722 treatment increases the ____autophagic flux directing viral components ____to ____lysosomes and restores type I interferon response.”

---

• Line 333 - there is no 3d pi post exposure treatment with MK-8722 in Fig 1l.*

Answer: We meant that treatment initiated at 1dpi lasted until harvest at 3dpi. This is clarified now to read, p18 line 351

“Post-exposure treatment at 1dpi and until harvest at 3dpi, …”

Significance

General assessment*: *

Strengths and limitations: The study provides evidence supporting previous results that the activation AMPK can be harnessed as a strategy to combat SARS-CoV2- infection. Whereas the results suggesting that this is targeted in infection by restoration of autopahgic flux and hence is selective is interesting, but the authors did not directly investigated the SARS-CoV-2 protein that is implicated in this effect. The study also does not discuss the context of AMPK and related kinases in discussion.*

*

Advance -* The study makes a pre-clinical advance, and has potential to make it fundamentally or conceptually sound. *

Audience - Virologists and Clinicians.

Expertise - SARS-CoV-2 infection biology, Cell biology an Molecular virology__ __

__Reviewer #3 __

Evidence, reproducibility and clarity (Required):

• Summary.** The authors address a topic of great importance to the development of antivirals, particularly in light of the possibility of future pandemics due to hitherto understudied viruses: the evaluation of interfering with host pathways (host-directed antivirals), potentially leading to compounds that can be used against a broad spectrum of viruses which require the same host cell functions for their life cycle. The authors studied the effects of activating AMPK with a small molecule (MK-8722) on various aspects of infectivity of SARS-CoV-2. They find that the compound indeed markedly reduced viral infectivity in two cell lines (Vero and Calu3), which correlated with the ability of the compound to activate signaling downstream of AMPK and was associated with increased lysosomal genesis/function and reduced density of "lipid viral factories". *

• *

General assessment. The study provides important proof of concept in reductionist cellular models, which may lead to pharmacologically more conclusive in vivo studies later on. It is limited by the use of cell lines, instead of primary human cells, for viral infectivity. The caption "MK-8722 restores the IFN I pathway" is an overstatement, as the observed increase in ISG expression is quite modest.

Answer: We agree with the reviewer that we did not use primary cells in our study, which was designed to evaluate the antiviral activity of MK-8722 in a simple epithelial cell model still relevant for the pathophysiology. However, we believe that MK-8722 antiviral effect we observed in the present study will be conserved using primary cells given the breadth of cell lines we used such as Vero, pulmonary Calu-3 and intestinal Caco2 cells. Furthermore, concerning the magnitude of the IFN I response we report, we remind the reviewer that Calu-3 cells are infected with an MOI of 0.05, which does not result in 100% cells infected after 24h. Consequently, the IFN-I response as limited to the sole infected cells in a limited amount in this setup and thus, bulk RT-qPCR of Type I IFN and targeted ISG mRNA is expected to remain modest.

In future studies, as suggested by the reviewer, we plan to include few ALI culture treated primary cells. We have clarified this limitation of our study in the Discussion p. 31 line 636-638 to read:

“The antiviral effect of AMPK activation in primary lung reconstruction and preclinical models such as hamster remains to be tested.”

__ Significance (Required): __

Taken as an early proof-of-concept study, the findings are quite important to investigators aiming to develop host-directed antivirals. However, the overall interest and impact of the manuscript would greatly benefit from verifying key findings in a primary cell model. Also, checking at least one other viral species (the authors mention flaviviruses) would be helpful to test how broadly applicable the current compound (or others to be developed in the future) would be as host-directed antivirals.

Answer: We thank the reviewer for the positive evaluation of our work. We fully agree with the reviewer in that preparedness is the key to fighting future pandemics. However to our knowledge, the literature about Flaviviruses is already stating that AMPK activation by metformin is an antiviral strategy with regards to its lipid metabolism normalization and autophagy activation (Farfan-Morales 2021 Sci Rep doi: 10.1038/s41598-021-87707-9. Ref 50). Furthermore in a previous review, we discussed the role of AMPK activation on viral infection that could be either pro- or antiviral depending on the viral family and even within a viral family such as HSV-1 which first inhibits AMPK in early infection while activating AMPK in the latter stage (Moreira, D. et al. Curr Drug Targets 17, 942–953 (2016), ref 34). Hence, adding more viruses to our study will not add novelty to the study, even though the molecule is much stronger compared to metformin.

• *

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #3

Evidence, reproducibility and clarity

Summary. The authors address a topic of great importance to the development of antivirals, particularly in light of the possibility of future pandemics due to hitherto understudied viruses: the evaluation of interfering with host pathways (host-directed antivirals), potentially leading to compounds that can be used against a broad spectrum of viruses which require the same host cell functions for their life cycle. The authors studied the effects of activating AMPK with a small molecule (MK-8722) on various aspects of infectivity of SARS-CoV-2. They find that the compound indeed markedly reduced viral infectivity in two cell lines (Vero and Calu3), which correlated with the ability of the compound to activate signaling downstream of AMPK and was associated with increased lysosomal genesis/function and reduced density of "lipid viral factories".

General assessment. The study provides important proof of concept in reductionist cellular models, which may lead to pharmacologically more conclusive in vivo studies later on. It is limited by the use of cell lines, instead of primary human cells, for viral infectivity. The caption "MK-8722 restores the IFN I pathway" is an overstatement, as the observed increase in ISG expression is quite modest.

Significance

Taken as an early proof-of-concept study, the findings are quite important to investigators aiming to develop host-directed antivirals. However, the overall interest and impact of the manuscript would greatly benefit from verifying key findings in a primary cell model. Also, checking at least one other viral species (the authors mention flaviviruses) would be helpful to test how broadly applicable the current compound (or others to be developed in the future) would be as host-directed antivirals.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

In the current manuscript, Cottignes-Calamarte et al. have shown tha pharmacological activation of AMPK can be a strategy for overcoming SARS-CoV-2 infection induced reprogramming of host degradation pathways and innate immune response, without hindering the efficacy of spike expression from vaccine agents. Though the suggestion of selective activity of MK-8722 in degrading viral proteins in infection but not the ectopic spike peptides expression is interesting, the evaluation of the mechanism and providing therapeutic index for the drug will overall improve the study.

Majorly, I have these suggestions;

1. The manuscript did not clarify the mechanism clearly but correlated the reduction in viral proteins and their association to LAMP1 as the mechanisim of activity of MK-8722. In this way, the authors did not separate whether the reduction in SARS-CoV-2 infection could lead to the potentiation of these effects. There is mentioning of potential mechanism of the drug by inhibiting the activity of SARS-CoV-2 proteins that reduce the autophagic flux without directly showing this. SARS-CoV-2 Orf3a is a known inhibitor of autophagosome maturation and hence the authors should directly probe the activity of MK-8722 in overcoming the suppression of autophagic flux in cells by ectopically expressing Orf3a.
2. As the authors propose MK-8722 as a preclinical candidate, they should present therapeutic measures and indexes. All across the data presented, there was no mentioning or measurement of drug toxicity for extended uses up to 36h pi.
3. Also, it was not clarified if the drug reduced the replication or exclusively worked by restoration of autophagic flux. For the earlier, the authors can consider two assays, i.e., a direct assay of looking into the replicon of SARS-CoV-2 (Bigotti et al 2024; PMID 38387750), or looking at earlier time points of infection (up to 8h pi; Twu et al. 2021; PMID 34788596) with intracellular sgRNA specific RNA probes.In this regards, 24h or 32h (Fig 1F-G) is too long to only measure single-round of infection.
4. Are the viral components degraded by restored lysosomal activity include replicase or replication organelle components? This needs to be shown with intracellular nsp3/nsp4 levels in infection or ectopic expression.
5. Previous studies suggested the decrease in AMPK phospharhorylation in SARS-CoV-2 infection What is inconsistent to previous studies should be commented by the authors. Eg, why is AMPK suppression not see in SARS-CoV- 2 inefction as reported before (Parthasarathy et al 2023; PMID 36417940)
6. Although AMPK activation is shown to be antiviral for SARS-CoV-2 before, e.g., with Metformin (Parthasarathy et al 2023; PMID 36417940), the role of AMPK-related kinases are shown to be pro-viral (e.g., NUAK2; Prasad et al. 2023; PMID 37421942). The authors should discuss these points to provide the reader a context in this sub-field.

Minor

1. Was there an increase of lipid droplets in SARS-CoV-2 infected cells to non-infected condition? It is not mentioned here.
2. There are several typos and grammatical errors.
3. There is no Fig S2I but is mentioend in the legend.
4. Fig 4A - degradation is not shown, only association.
5. Line 333 - there is no 3d pi post exposure treatment with MK-8722 in Fig 1l.

Significance

General assessment

Strengths and limitations: The study provides evidence supporting previous results that the activation AMPK can be harnessed as a strategy to combat SARS-CoV2- infection. Whereas the results suggesting that this is targeted in infection by restoration of autopahgic flux and hence is selective is interesting, but the authors did not directly investigated the SARS-CoV-2 protein that is implicated in this effect. The study also does not discuss the context of AMPK and related kinases in discussion.

Advance - The study makes a pre-clinical advance, and has potential to make it fundamentally or conceptually sound.

Audience - Virologists and Clinicians.

Expertise - SARS-CoV-2 infection biology, Cell biology an Molecular virology

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

Summary:

Provide a short summary of the findings and key conclusions (including methodology and model system(s) where appropriate).

The study by Cottignies-Calamarte et al. describes that AMP-activated protein kinase (AMPK) regulates cell energy balance by suppressing energy-consuming pathways like lipid and protein synthesis and promoting nutrient availability through autophagy. These pathways contribute to SARS-CoV-2 infection by hijacking autophagy and accumulating lipid droplets for viral replication. The antiviral activity of MK-8722, a direct pan-AMPK allosteric activator, was evaluated in vitro. MK-8722 effectively inhibited Alpha and Omicron SARS-CoV-2 variants in Vero76 and human bronchial epithelial Calu-3 cells at micromolar concentrations. This inhibition restored autophagic flux, degrading newly synthesized viral proteins, and reduced lipid metabolism, affecting viral factories. Additionally, MK-8722 treatment increased the type I interferon (IFN-I) response. Post-infection treatment with MK-8722 efficiently suppressed viral replication and restored the IFN-I response without altering the SARS-CoV-2-specific CD8+ T cell response elicited by Spike vaccination. The authors concluded that, MK-8722 acts as an effective antiviral against SARS-CoV-2 infection, even when applied post-exposure, suggesting potential for preclinical tests to inhibit viral replication and alleviate patient symptoms.

Major comments:

• Are the key conclusions convincing?

Partially. See comments below! - Should the authors qualify some of their claims as preliminary or speculative, or remove them altogether?

From my perspective, the title "Direct pharmacological AMPK activation inhibits mucosal SARS-CoV-2 infection by reducing lipid metabolism, restoring autophagy flux and the type I IFN response" is a clear overstatement. In no way can the authors make statements about the autophagic flux, as it simply was not measured. The study would greatly benefit from conducting an autophagy/autophagic flux assay. See specific comments below! - Would additional experiments be essential to support the claims of the paper? Request additional experiments only where necessary for the paper as it is, and do not ask authors to open new lines of experimentation.

See below specific comments regarding cell line consistency and autophagy measurements. - Are the suggested experiments realistic in terms of time and resources? It would help if you could add an estimated cost and time investment for substantial experiments.

This question depends on various factors such as access to relevant biosafety labs, availability of required reagents, etc. In my estimation, experiments involving WT viruses and autophagy measurements could be conducted within 3-4 months. The proposed experiments with the delta-N-SARS-CoV-2 mutants, of course, depend on access to such viruses. Overall, I believe all experiments could be completed within 6 months. The costs of those assays are not very high. - Are the data and the methods presented in such a way that they can be reproduced?

In the present study, a total of 3 cell lines and human PBMCs were utilized for various experiments. Please indicate why each cellular model was chosen and highlight the differences between these models considering what is known for SARS-CoV-2 infection and autophagy! Furthermore, the study would greatly benefit if the key findings were consistently demonstrated in a single cell line.

The authors conclude that selective activation of AMPK has a pro-autophagic effect which in turn leads to a reduced SARS-CoV-2 replication. I would generally agree with this statement, but throughout the entire manuscript, no real autophagy assays are shown. This should definitely be rectified. It is important to demonstrate that (1) MK-8722 is capable of increasing autophagy and particularly autophagic flux in the cell models used, (2) that in the cell models employed, SARS-CoV-2 infection leads to modulation of autophagy, and (3) that SARS-CoV-2 infected cells, when co-treated with MK-8722, lead to a re-established autophagy. The autophagy assays should be performed according to the expert-curated guidelines by Klionsky et al. This is extremely important so that the results can be compared with the now vast number of existing autophagy-SARS-CoV-2 studies. - Are the experiments adequately replicated and statistical analysis adequate?

Minor comments: - Specific experimental issues that are easily addressable.

Typo: Line 288: It should be "MK-8722" instead of "MK-7288". In general, a space between value and unit is not consistently used. Please indicate always the phosphosite of substrate proteins when phosphorylation is described. E.g. line 288 and throughout the manuscript: regarding ACC phosphorylation. - Are prior studies referenced appropriately?

See comment below. Fundmental work that describes the virus-autophagy relationship, such as the work by the Beth Levine lab would be important to add. Also the work bei Konstantin Sparrer and colleagues is important and leads to the current work presented here. - Are the text and figures clear and accurate?

Introduction:

In general, for some of the statements claimed in the introduction, which is in sum nicely written, informative and well structured, citations are required. For example - line 59 "...autophagy is sequentially activated and inhibited." Lines 59-65. In the introduction, the interaction between autophagy and SARS-CoV-2 is primarily described in a one-sided manner. There are now several studies demonstrating that both inhibition of autophagy and also induction of autophagy in context of coronaviral infection. Both aspects should be illuminated and introduced here.

Discussion: The selectivity of the compound should be discussed. - Do you have suggestions that would help the authors improve the presentation of their data and conclusions?

The presentation of the data is understandable. For reader with a less mechanistic background it would be helpful to present a schematic figure like a graphical abstract.

Significance

• Describe the nature and significance of the advance (e.g. conceptual, technical, clinical) for the field.

Even though there is now a plethora of studies on autophagy and SARS-CoV-2, this study is important and of great interest to a broad readership. Not only virologists, immunologists, and autophagy researchers will eagerly anticipate this study, but especially researchers focusing on pharmacology around AMPK and autophagy will recognize the importance of the data presented here. - Place the work in the context of the existing literature (provide references, where appropriate).

The work builds upon a variety of studies on the interplay between coronaviruses and the mechanism of autophagy. Foundational contributions to this research stem from groundbreaking preliminary work conducted in the laboratories of Gassen and Müller (Gassen et al. 2019 and Gassen et al. 2021), as well as general virus-autophagy studies from the laboratory of Beth Levine. Additionally, recent work by Konstantin Sparrer and colleagues would be important to cite, as it underscores the insights gained in this manuscript. - State what audience might be interested in and influenced by the reported findings.

See comment above! - Define your field of expertise with a few keywords to help the authors contextualize your point of view. Indicate if there are any parts of the paper that you do not have sufficient expertise to evaluate.

My expertise is limited to the mechanistic aspects of autophagy, metabolism, and signaling cascades related to autophagy and endosomal-exosomal mechanisms. In some studies, I have been able to investigate the interplay between CoV and autophagy in collaborations with coronavirus experts. I can only provide a superficial assessment of the purely virological (methodological) aspects of the present study.

Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

Learn more at Review Commons

---

Reply to the reviewers

Answers to reviewers

---

Reviewer #1

Sagia et al. present a manuscript using A. nidulans as model to study different transport routes of membrane proteins from the ER to the plasma membrane. They showed in earlier work that apparently at least two different transport routes exist, one involving the classical ER-ERES-ERGIC-Golgi route, one bypassing the Golgi. Unpolarized membrane proteins use the former, apically sorted membrane proteins the latter route. The study here confirms their earlier findings, uses a better model (co-expression of representatives for both routes in the same cell) and provides additional mechanistic insights about the roles of rabs, SNARES and other important proteins of the secretory pathway. The study is thoroughly done, figures are of high quality, data and methods well described and adequately replicated.

Thank you for your positive comments

I do have, however, a number of comments that could help to improve the manuscript.

-I suggest using the term polarized or apical rather than polar. Polar alone to me refers more to physico-chemical properties like water-solubility.

Amended in most parts of the revised text.

-introduction and discussion: I don’t think the literature about unconventional secretion bypassing the Golgi is complete, for example studies about TMED10 like Zhang, M. et al. Cell 181, 637-652 e615 (2020) or Zhang et al. Elife 4 (2015) are missing, there might be others. Is UapA a leader-less cargo that could be inserted via TMED10 translocation?

Thank you for letting us know, we have missed these articles. More references on UPS are now added, including the Zhang et all publications. UapA, as all transporters, is a multispan transmembrane protein with no leader peptide. In fact, we have checked the role of p24 family proteins (homologous to TMED10) in UapA trafficking. The knock-out of key p24 proteins does not affect UapA sorting to the PM (please consider this as confidential unpublished results)

-Fig. 1C. Can these intracellular structures be characterized in more detail?

As explained briefly to the handling editor above, and following the reviewer’s suggestion, we performed new experiments to better characterize the identity of the cargo-labeled fluorescent puncta. To do so, we used co-expression of a standard ERES marker, Sec16, in cells expressing either UapA or SynA, tagged with different fluorescent tags. More specifically, we constructed and analyzed strains co-expressing UapA-GFP/Sec16-mCherry or GFP-SynA/mCherry-Sec16 in the sec31ts genetic background, which allows synchronization and better analysis of ER exit, as described in our text. The new findings appear as Figure 5C __in the revised manuscript. Notice that sec16-mCherry introduced in the native sec16 locus by standard knock-in reverse genetics of A. nidulans (see Materials and methods) does not affect Aspergillus growth or secretion. Experiments depicted in __5C show that both cargoes, UapA and SynA, co-localize significantly (PCC ≈ 0.6), with Sec16, suggesting that most of these puncta are indeed ERES structures. Given that the puncta marked with UapA or SynA are clearly distinct (see Figures 1C,2A, 3A, 5B), this new experiment strongly suggests that there are indeed two distinct ERES, one populated mostly by UapA and the other by SynA. Notice, as we already outline in our response to the editor above, a three-colored approach using Sec16-BFP (or Sec13-BFP) for showing directly the existence of these two populations of cargo-specific ERES in the same cell failed as the BFP signal was problematic for colocalization studies.

Where is the Golgi localized in A. nidulans, is it decentralized like in yeast?

Yes, as in S. cerevisiae, A. nidulans Golgi cisternae are individually scattered throughout the cytoplasm, also similarly to other filamentous fungi. Notice that in A. nidulans Golgi structures are moderately polarized (Pantazopoulou and Penalva 2009).

Is the UapA at the time points shown in Fig. 1C in some sub-PM structures? To me the distribution at or near the PM is more punctate than in the steady state image shown in 1B

The punctuate appearance of PM transporters at the periphery of fungal cells is a common theme when these do not reach high, steady-state, levels of accumulation. In fact, several transporters mark specific subdomains of the PM, more evident before achieving their steady-state levels. For example, in yeast several amino acid and nucleobase transporters mark punctuate structures that colocalize with eisosomes markers (caveolin-like PM subdomains), while the proton pump ATPase Pma1 marks distinct punctuate domains. Similarly, UapA and other solute transporters mark punctuate structures before reaching their state-state accumulation in the PM. Figure 1C shows the de novo synthesis of cargoes after 100 min of transcription, while Figure 1B depicts the steady-state localization of UapA and SynA after 4h. In the latter case, the PM is ‘saturated’ with UapA molecules and thus the fluorescent signal of distinct puncta ‘fuses’, creating continuous fluorescent labeling. Notice also that in several cases, in our work, we have also performed UapA transport assays, which provide a direct tool to test and confirm the presence of UapA in the PM (see Figures 4D or 6C).

-Fig. 3A. To me it looks like there is actually a lot of colocalization of UapA and SynA, especially at or near the PM, where there is quite some white, punctate staining. The green fluorescence is just much stronger, overlaying the violet. Can you show separate channels and explain?

We think the reviewer means Figure 2A, which compares UapA and SynA (Figure 3A compares UapA with Golgi markers). If so, we have quantitatively estimated and performed statistical analysis (PCC) which indicates that this, visually apparent colocalization, is not significant (right panel in Figure 2A). Notice also that we cannot totally exclude very minimal colocalization of UapA and SynA signals as both cargoes mark very proximal early secretory domains (i.e., ERES or ERGIC), especially in fungal cells. Anyhow, in the revised Figure 2 we also added a panel depicting separate channels, as the reviewer asks.

-Fig. 3: In my opinion the statement that UapA "is probably sorted from an early secretory compartment, ultimately bypassing the need for Golgi maturation" is too strong at that point. You say for both UapA and SynA you don’t get significant colocalization with early Golgi/ERGIC marker, then you cannot conclude that one takes the conventional route via early-late Golgi and the other does not. What you can say is that UapA is apparently not going through late Golgi.

The reviewer is in principle correct. However, significant colocalization with the late Golgi marker, as SynA shows, strongly suggests that this cargo has passed via the early Golgi compartment. The fact we failed to detect significant colocalization of any cargo tested with early Golgi/ERGIC markers (e.g., SedV) is very probably due to very rapid passage of cargoes from these compartments, which conventional widefield or confocal microscopy cannot detect. To achieve this, ultra-fast fluorescent microcopy, as Lattice Light Sheet Microscopy (LLSM), should be used. In fact, we are currently initiating these studies, which will appear in the near future elsewhere.

-Fig. 4C: UapA does not seem to accumulate in the ER in the Sec24 and 13 mutants but in punctate structures. This for me is unexpected, any explanations? Can you characterize that punctate staining?

This is an interesting observation. Notice that UapA is a large homodimeric protein (e.g., 28 transmembrane domains) that oligomerizes further upon translocation into the ER membrane. Repression of Sec24, and to a less extent of Sec13, leads to inability to exit the ER properly. Consequently, this will lead to UapA overaccumulation in the ER, which might in turn lead to ER stress and turnover, reflected in UapA aggregates. In line with this, we have previously shown that specific mutants of UapA unable to exit the ER are indeed degraded by selective autophagy (Evangelinos et al., 2016). In contrast to UapA, SynA partitions in the entire ER without forming aggregates when sec24 or sec13 are repressed. This might be due to the fact that is a single-pass, much smaller, membrane protein compared to UapA and one that is not known to form oligomers. Thus, its overaccumulation in the ER might not lead to aggregation, allowing it to diffuse laterally in the membrane of the ER. A note on this is included in the Figure legend of the revised manuscript.

-Fig. 6D: You state that BFA "has only a very modest effect on UapA translocation to the PM". To me the PM (or very near PM) staining of UapA looks very different in the PFA treated cells, more uneven/punctate. Is there an explanation for that?

Our explanation is the following. When BFA is added, conventional secretion is blocked and Golgi collapses. We believe that this might have a moderate indirect effect also on cargoes bypassing the late Golgi/TGN, as UapA (i.e., lower levels of UapA present in the PM). This is based on the fact that UapA, in addition to conventional cargoes, requires the Q-SNARE complex SsoA/Sec9 to translocate to the PM. SsoA, being a membrane protein cargo itself, also needs to traffic to the PM. Interestingly, we have previously obtained evidence suggesting that SsoA traffics to the PM by both conventional and a Golgi-bypass routes (Dimou et al 2020). Thus, UapA translocation to the PM might indeed be partially impeded or delayed due to repression of proteins, such as SsoA (and probably Sec9), needed for its final integration into the PM bilayer. Importantly, in line with an indirect effect of BFA on the levels of UapA localized in the PM, notice that, unlike SynA, UapA was never trapped in brefeldin bodies (i.e., Golgi aggregates).

Reviewer #1 (Significance):

One strength of the study is the use of a model organism, A. nidulans, not cell cultures. Also, the use of both reporters, UapA and SynA, in the same cell is an advantage over previous studies using different lines and different promotors. Limitation of the study might be that it remains unclear to what extend the basic mechanism (UapA and SynA are transported to PM in different carrier and via different routes) can be generalized to other polarized (apically?) membrane proteins versus non-polarized membrane proteins in A. nidulans and whether a similar mechanism exists in other organisms. Some of the basic findings of the study are not new but were published by the same group. However, as the authors point out, the current study uses improved assays and extends their previous studies, advancing our understanding of the mechanistics of transport in the conventional secretory pathway and novel alternative routes. The study will be of interest for basic researchers in the trafficking field. My own expertise is transport through the secretory pathway in mammalian cells, many years ago more post-Golgi, now mostly ER-Golgi and ER itself.

We thank the reviewer for his positive comments.

__Reviewer #2 __

__ __The idea that transmembrane proteins of the plasma membrane move from the ER to the Golgi and then to the cell surface is firmly entrenched, and the mechanisms and components of this secretory pathway have been extensively characterized. Secretory vesicles are often delivered from the Golgi to sites of polarized growth. This paper builds on previous work by the same group to provide evidence that in Aspergillus nidulans, some non-polarly localized plasma membrane proteins follow a very different pathway, which bypasses components of the conventional secretory machinery such as SNAREs that have been implicated in secretion as well as the exocyst. In particular, they systematically compare the trafficking of the SNARE SynA, which follows the conventional secretory pathway, with that of the purine transporter UapA, which apparently does not. The two proteins were co-expressed in the same cells using the same promoter. A variety of genetic and microscopy methods are used to support the conclusion that UapA reaches the plasma membrane by a route distinct from that followed by SynA.

In my view, the authors present a convincing case. The individual experimental results are sometimes ambiguous, but the combined results favor the conclusion that UapA follows a novel pathway to the plasma membrane. I have only a few relatively minor comments.

Thank you for your positive comments

1. In the Introduction and elsewhere: to my knowledge, there is no clear evidence that AP-1-containing clathrin-coated vesicles carry cargoes from the Golgi to the plasma membrane. On the contrary, as recently reported by Robinson (https://pubmed.ncbi.nlm.nih.gov/38578286/), AP-1-containing vesicles likely mediate retrograde traffic in the late secretory pathway.

Thank you for this comment and the relative reference. We are aware that AP-1 is likely to also mediate retrograde traffic in the late secretory pathway or/and intra-Golgi recycling, as also reported by the group of Benjamin Glick. Thus, in the revised version we added a short comment on this plus relative references. Along this line, our previous work has shown that transcriptional repression of AP-1 arrests the polar localization of several apical markers in A. nidulans and we reported that this might be due to an effect on both anterograde and retrograde trafficking. Please see “Secretory Vesicle Polar Sorting, Endosome Recycling and Cytoskeleton Organization Require the AP-1 Complex in Aspergillus nidulans”. Martzoukou O, Diallinas G, Amillis S. Genetics. 2018 Aug;209(4):1121-1138. Overall, the fact that AP-1 was found absolutely dispensable for UapA trafficking, further strengthens our conclusion that UapA bypasses the Golgi.

1. In Figure 2, is there any known significance to the presence of UapA in "cytoplasmic oscillating thread structures decorated by pearl-like foci as well as a very faint vesicular/tubular network"?

At present we cannot answer this question. In order to understand what these structures represent and answer what is their role, we will need to employ super-resolution and ultra-fast microscopy and additional markers, which we envision to do. We suspect that they might be tubular networks, but this extends beyond the present work.

1. SynA is related to S. cerevisiae Snc1/2, which are known to be present in late Golgi compartments due to repeated rounds of endocytosis to the Golgi and exocytosis to the plasma membrane. The SynA shown here to colocalize with PHosbp is probably present in a similar recycling loop rather than being en route to the plasma membrane for the first time. Therefore, the differential colocalization of UapA and SynA with PHosbp does not by itself provide "strong evidence that the two cargoes studied traffic via different routes" as stated in the text but might instead indicate that only SynA undergoes frequent endocytosis. The text should be amended accordingly.

The reviewer is in principle correct. However, given that colocalization of SynA and PHosbp occurred all over the cytoplasm of hyphae and not only at the apical region, and because we record colocalization of cargoes before their steady-state accumulation to the PM, thus at a stage where recycling must be minimal, the recorded colocalization should reflect anterograde transport rather than recycling. We added this reasoning it the revised text.

1. A missing piece of the story is a test of whether the puncta visualized for the two cargoes in Figure 5B are indeed distinct populations of COPII-containing ER exit sites. The relevant experiment would involve co-labeling of the cargoes together with a COPII marker. Three-color labeling would presumably be needed.

This point was also raised by reviewer 1 (and review 3) and thus performed new experiments to better characterize the identity of the cargo-labeled fluorescent puncta. To do so, we used co-expression of a standard ERES marker, Sec16, in cells expressing either UapA or SynA, tagged with different fluorescent tags. More specifically, we constructed and analyzed strains co-expressing UapA-GFP/Sec16-mCherry or GFP-SynA/Sec16-mCherry in the sec31ts genetic background, which allows synchronization and better analysis of ER exit, as described in our text. The new findings appear as Figure 5C __in the revised manuscript. Notice that sec16-mCherry introduced in the native sec16 locus by standard knock-in reverse genetics of A. nidulans (see Materials and methods) does not affect Aspergillus growth or secretion. Experiments depicted in __5C show that both cargoes, UapA and SynA, co-localize significantly (PCC ≈ 0.6), with Sec16, suggesting that most of these puncta are indeed ERES structures. Given that the puncta marked with UapA or SynA are clearly distinct (see Figures 1C,2A, 3A, 5B), this new experiment strongly suggests that there are indeed two distinct ERES, one populated mostly by UapA and the other by SynA. Notice, as we already outline in our response to the editor above, a three-colored approach using Sec16-BFP (or Sec13-BFP) for showing directly the existence of these two populations of cargo-specific ERES in the same cell failed as the BFP signal was problematic for colocalization studies.

Reviewer #2 (Significance):

This study provides compelling evidence that in the fungus Aspergillus nidulans, some transmembrane transporter proteins reach the plasma membrane by a pathway that bypasses much of the conventional machinery associated with the Golgi apparatus and secretory vesicles. Although previous publications pointed toward a similar conclusion, the present work tackles the problem in a more rigorous and systematic way. These findings are important for cell biologists who study membrane traffic, it remains to be determined how prevalent this type of non-canonical secretion might be in other organisms.

We thank the reviewer for his positive comments

Reviewer #3

The manuscript by Sagia et al compares the trafficking of a polarized (SynA) with a non-polarized (UapA) transmembrane protein. In agreement with previous work of the same lab, they find that UapA reaches the plasma membrane through a Golgi-bypass route, which they characterize to some extent. Overall, the data are of good quality and the story is interesting and timely. Understanding trafficking routes that bypass the Golgi is highly interesting. Nevertheless, there are several points of criticism that I have and below is a list where I combine major and minor points together:

Thank you for your positive comments

Major Comments:

1- Is it possible that the polarized phenotype of SynA is caused by selective removal, i.e. SynA is delivered to the entire plasma membrane, but endocytosed rapidly from all areas except the tip of the hyphae. This would also result in a polarized distribution.

This is in principle possible, but here this is not the case. SynA is polarized due to rapid local endocytosis and immediate recycling at the subapical region, known as the subapical collar. Please see:

Taheri-Talesh N, Horio T, Araujo-Bazán L, Dou X, Espeso EA, Peñalva MA, Osmani SA, Oakley BR. The tip growth apparatus of Aspergillus nidulans. Mol Biol Cell. 2008 Apr;19(4):1439-49. doi: 10.1091/mbc.e07-05-0464.

Hernández-González M, Bravo-Plaza I, Pinar M, de Los Ríos V, Arst HN Jr, Peñalva MA. Endocytic recycling via the TGN underlies the polarized hyphal mode of life. PLoS Genet. 2018;14(4):e1007291. Published 2018 Apr 2. doi:10.1371/journal.pgen.1007291

This applies to all apical markers; they remain polarized by continuous local recycling after the diffuse laterally to the subapical collar.

2- The authors describe the distribution of SynA and UapA in cells deficient of various COPII/ERES proteins. However, these data are not shown, and it is not clear how they were quantified. It would be important to add quantitative data here.

Quantitative data are included in Figure 4C, displaying the percentages of cells with UapA either retained in the ER or reaching the PM for each background deficient in a COPII protein. Repression of SarA and Sec31 resulted in UapA retention in the ER in all analyzed cells (100%). However, repression of Sec12, Sec24, or Sec13 had a differential effect across the cell population, with UapA reaching the PM in some cells, while remaining trapped in the ER in others. To quantify these data and determine which cargo localization pattern prevails, we measured the number of cells in each category and represented them as percentages. A similar approach was used to examine the role of Golgi proteins in the trafficking of UapA and SynA (Figure 6).

3- on page 8, the authors discuss the discrepancy regarding the role of Sec13. They offer as an explanation that the previous studies have been performed in strains that separately expressed the two cargoes. However, I am unable to see why and how this would be a valid explanation.

Given that Sec13 has a variable/partial effect on UapA, we have previously been biased towards images that showed an effect on localization, as expected, and considered that the lack of an effect might have been due to inefficient repression in a fraction of cells. In our new system, we were able to directly compare UapA to SynA and find out that while SynA was always affected under our conditions, the effect of UapA was still variable. Thus, the partial effect of Sec13 on UapA is physiologically valid and not a matter of insufficient repression in a fraction of cells. This shows the importance of our new improved system where we follow the synchronous expression of two cargoes in the same cells.

4- Why is the effect of Sec24 depletion so much stronger than of Sec12 depletion? Sec12 is the GEF for SarA, without which Sec24 should not be recruited to ERES. The explanation that low amounts of Sec12 are still present and sufficient to carry out the role of this protein. What is the evidence for that?

Sec24 is the principal receptor of cargoes responsible for their recruitment to ERES. Sec12 is the catalytic effector for SarA required for the initiation of COPII vesicle formation. The question of the reviewer is thus logical.

However, Sec12 is indeed present at extremely very low levels when expressed from its native promoter under the condition of our experiment (minimal media). This is supported by our recent proteomic analysis, performed under similar conditions, which failed to detect the Sec12 protein, unlike all other COPII components (see Dimou et al., 2021, doi; 10.3390/jof7070560), but also by cellular studies of the group of M.A. Peñalva, who failed to detect Sec12 tagged with GFP (Bravo-Plaza et al., 2019, doi: 10.1016/j.bbamcr.2019.118551). Additionally, in yeast, immune detection of Sec12 has been possible only in cells harboring sec12 on a multicopy plasmid, suggesting its low abundance in wild-type cells (Nakano et al., 1988, doi:10.1083/jcb.107.3.851).

Given that repression of sec12 transcription via the thiAp promoter still allows 68% of cells to secrete normally both SynA and UapA, while 32% of cells are blocked in the trafficking of both cargoes, suggests that in most cells either SarA can catalyze the exchange of GDP for GTP without Sec12, maybe through a cryptic guanine nucleotide exchange factor (GEF), or that very small amounts of Sec12 remaining after repression are sufficient for significant SarA activation. Whichever scenario is true, Sec12, similarly to SarA, is not critical for distinguishing Golgi-dependent from Golgi-independent routes, as both cargoes are affected similarly. In the revised text we added a not on this issue.

5- In Figure 5, it would help readers who are not so familiar with Aspergillus organelle morphology to explain the figure a bit better. This might appear trivial for experts, but anyone from outside this field is slightly lost.

In the revised manuscript we added a figure panel depicting a schematic representation of A. nidulans key secretory compartments.

6- The authors write that not seeing UapA in Golgi membranes is evidence that it does not pass through this organelle. However, when they write that SynA is never seen in cis-Golgi elements, they do not conclude that SynA bypasses the cis-Golgi.

The fact that SynA, unlike UapA, colocalized significantly with late-Golgi/TGN and follows conventional secretion in general, strongly suggests that SynA also passes from the early-Golgi. Cargo traffic through the Golgi is mediated by cisternal maturation, where an individual cisterna gradually changes its nature from an earlier to a later one, while the cargo remains inside. UapA, unlike SynA, never colocalized with any Golgi marker used and was not affected by BFA. We agree with the reviewer that we did not have direct proof for passage of UapA or SynA from the early-Golgi in the wt background, which allows for the alternative, but rather unlikely hypothesis, that none of the two cargos is sorted to the early Golgi and that SynA traffics directly to late-Golgi/TGN. Our inability to detect sorting of any cargo to the early-Golgi is seemingly due to ultra-fast passage of cargoes from very early secretory compartments, such as ERGIC/early-Golgi. In fact, we have obtained evidence of this using Lattice Light Sheet microscopy (results in progress, to appear elsewhere).

7- Figure 5C: the authors claim that the CopA and ArfA affects trafficking of UapA and SynA from ER to plasma membrane and assign CopA and ArfA as regulators for anterograde trafficking. I think this interpretation is not justified by the data. Depletion of CopA and ArfA will affect the Golgi apparatus in structure and function. The more straight-forward interpretation is that repression of the COPI machinery results in a defect in Golgi exit and therefore retention in pre-Golgi compartments (including the ER and maybe the ERGIC should it exist in Aspergillus). The same is true for BFA treatment where there are also negative effects on ER export, which are rather indirect consequences of alterations of Golgi function and integrity. Likewise, the interpretation of the papers by Weigel et al and Shomron et al is not correct. It is more likely that COPI is recruited to the growing ERES-derived tubule (or ERGIC) to recycle proteins back to the ER. This is not necessarily a proof that COPI regulates anterograde trafficking

This is a highly debatable issue which our work cannot address. However, we amended the text accordingly.

8- Figure 6: The images look like in Figure 5, yet here you don't call them ER-associated.

The two images are not alike. In Figure 5 upon activation of Sec31 (permissive temperature) we detect mostly punctual structures resembling ERES, whereas at the nonpermissive temperature we detect a membranous network typical of the ER. Upon repression of CopA we also detect punctual structures similar to ERES. In Figure 6, we mostly detect an effect on SynA. Repression of early secretory steps (SedV, GeaA) lead to collapse of SynA in the entire ER network. Repression at later stages of Golgi maturation and post-Golgi secretion (RabO, HypB, RabE, AP-1) lead to the appearance of punctual structures, most probably Golgi aggregates.

9- Figure 6D: How long was the BFA treatment. I am surprised that the pool of SynA preexisting at the plasma membrane seems to also be sensitive to BFA.

Cells were grown overnight under repressed conditions for both UapA and SynA. After 12-14h cells were shifted to derepressed conditions using fructose as carbon source. BFA was added after 90min of cargo derepression, while both cargoes were still in cytoplasmic structures so there was not preexisting SynA or UapA at the PM (see also Figure 1C). Subcellular localization of both cargoes was studied for 60min after BFA treatment.

10- This might be beyond the scope of this study, but as far as I know UapA is not N-glycosylated. Would the introduction of an N-glycosylation site shift it towards the Golgi-based route?

Thank you for this suggestion. We have performed this experiment, adding a glycosylation site on UapA, based on the glycosylation sites found in tis mammalians homologues. We did not detect any effect on UapA trafficking route or its activity. As the reviewer recognizes this goes beyond the scope of this study and thus, we did not include it the manuscript. Differential cargo glycosylation is however an important issue to be studied systemically in respect to different trafficking routes, and we envision to investigate it systematically.

Minor Comments

1- This might be just a personal preference, but I think that the term polar is misleading, because it implies something about the polarity of the amino acids. I think "polarized" might be the more common term. Anyway, this is just a minor point and just a suggestion from my side.

Amended in the revised text.

2- The paper by the Saraste lab should be mentioned and discussed (PMID: 16421253), which I think is very relevant to the current story.

We thank the reviewer for pointing out this important publication. In that case, the Rab1 GTPase defined a pathway connecting a pre-Golgi intermediate compartment with the PM in mammalians nerve cells. Thus, the Saraste lab publication is indeed along the lines of findings supporting that Golgi-independent unconventional cargo trafficking routes initiate at very early secretory compartments. Notice, however, that RabO, the A. nidulans homologue of Rab1, which in their case was essential for direct cargo sorting from the ERES/ERGIC to the PM, in or system, was dispensable for Golgi bypass. The Saraste lab article is now mentioned and discussed.

3- Having worked with ERES for over two decades, I find it strange to see it written ERes. I see no reason why ER exit sites in Aspergillus should be abbreviated differently from all other types of cells (yeast, drosophila, worms, mammals). I think that the entire acronym should be capitalized.

Amended in the text

4- When discussing the data about the partial effect of Sec13, it would be good to refer to a previous paper by the Stephens lab that showed that silencing Sec13/31 results in a defect in trafficking of collagen, but not of VSVG (PMID: 18713835).

We thank the reviewer for also pointing out the publication of the Stephens lab, now mentioned in the revised text. Noticeably, in that case silencing of both Sec13 and Sec31 has no effect on the trafficking of specific cargoes, whereas in our case Sec31 is still absolutely needed for both conventional and Golgi-independent secretion of SynA and UapA, respectively.

Reviewer #3 (Significance):

Overall, the data are of good quality and the story is interesting and timely. Understanding trafficking routes that bypass the Golgi is highly interesting. The main weakness is the lack of mechanistic understanding of the Golgi-bypass pathway. In addition, the study is limited to two proteins as representatives of polarized vs. non-polarized proteins. The main target audience for this paper are scientists working in the area of secretion and trafficking in the secretory pathway.

We thank the reviewer for his positive comments.

We are aware that the mechanistic details of Golgi bypass are missing and this is our next goal, dissecting those via various approaches genetic and biochemical approaches and employment of super resolution and ultra-fast microscopy.

__Reviewer #4 __

In this study, Sagia et al investigate the trafficking of different secretory cargo in Aspergillus nidulans under conditions that repress expression of transport factors or block stages in membrane trafficking. The primary approach is to conduct dual live-cell imaging of GFP-tagged UapA (plasma membrane localized purine transporter) and SynA (plasma membrane R-SNARE) after their simultaneous derepression to monitor trafficking routes. In germlings, both secretory proteins are detected in non-overlapping intracellular compartments and puncta after 60-90 min of derepression. After 4-6 hrs, SynA localizes to hyphal tips whereas UapA localizes to non-polar regions of the PM. Colocalization studies do not show UapA overlap with Golgi markers (SedV, PH-OSBP) during its biogenesis whereas SynA displays significant co-localization. Repression of COPII and COPI components generally block transport of both cargos to the PM and cause accumulation in ER compartments, although there are some differential effects on UapA and SynA localization. Finally, repression of other transport factors (ER-Golgi SNAREs, Golgi transport factors, and exocytic machinery) had differential effects on UapA and SynA localization over time with UapA reaching the plasma membrane in many instances and SynA accumulating in intracellular compartments.

Based on these observations, the authors conclude that UapA and SynA follow distinct trafficking routes to the plasma membrane where SynA uses a canonical SNARE-dependent secretory pathway route and UapA follows a non-canonical route that may bypass Golgi compartments. The study is extensive and supports the model that biogenesis of SynA and UapA follow distinct processes. However, there are some complexities that may limit interpretation. First, the cargo studied are targeted to the ER differently. UapA is a multispanning transmembrane protein that is likely dependent on the Sec61 translocon for co-translational membrane insertion and will involve ER chaperones and quality control machinery for its biogenesis. SynA will depend on the tail-anchored machinery (GET/TRC pathway) for insertion into the ER and is processed by cytosolic factors/chaperones. Therefore, the sites of ER insertion and the rates of biogenesis of these cargoes will be different. In addition, the repression of trafficking machinery used in this study appears to be variable and may exert partial blocks on intracellular transport stages. Regardless, the study clearly documents that SynA and UapA follow distinct biogenesis and transport processes when co-expressed in cells under experimentally controlled conditions.

Thank you for your positive comments.

To our knowledge there is no evidence suggesting that SynA translocates via a tail-anchored machinery (GET/TRC pathway) and not through the translocase. Despite this, we agree with the reviewer that translocation to the ER, as well as exit from it, might be cargo-dependent, especially when it concerns proteins with very different size, structures and oligomerization. Thus, the rate of biogenesis of UapA and SynA is probably quite different. However, this still does not dismiss our basic conclusion that the two cargoes follow distinct routes to traffic to the PM. The ‘problem’ of variable transcriptional repression of some trafficking-related proteins is solved by comparing the relative effect on the two cargoes in the same cells, and this is in fact the advantage of our new system. Importantly, notice that we took care to use conditions of repression where SynA trafficking by the conventional path was totally abolished and compared it to UapA.

1. It was not clear if the translation, ER insertion and folding of UapA and SynA are fully synchronous. Is it possible that the rate of UapA synthesis and transport to the plasma membrane is substantially faster than for SynA? The imposition of transport blocks could trap SynA and not UapA if this cargo was at later transport stages.

As already discussed above translation, ER insertion and folding of UapA and SynA might indeed by different. This might somehow affect the trafficking path followed, but this issue is beyond the scope of this work. Notice, however, that the transcription of both cargoes is kept fully repressed during establishment of repression of secretion. Only when repression and blocking of secretion is established (12-14 h germination), as verified by Western blot analysis, we derepress the transcription of UapA and SynA, expressed from the same promoter, and follow their dynamic subcellular localization. Hence, this system ensures that both cargoes start from the earliest transport stage, the ER, upon imposition of transport blocks.

1. In repressing transport factors (e.g., SarA, Sec12, Sec24, Sec13, SedV, RabE), it is clear that under thiamine repressing conditions these cells do not grow or have greatly reduced growth rates. However, it was not clear if proteins are depleted to the same extent in cells after repression for 12-14 hr or 16-22 hr. as mentioned in the methods. Indeed, in some cases depleted cells display different cargo localization patterns, for example 67% of cells show normal localization of UapA and SynA after sec12 repression and 33% show ER accumulation of both cargoes. There is differential localization of UapA and SynA in many cases where transport factors are repressed, but this could be due to partial inhibition and not complete blocks. It would be helpful to clearly indicate the time points and conditions in each of the figure legends as in points 3-5 below.

In the revised manuscript we did our best to clearly indicate the time points and conditions in each of the figure legends. Differential localization of UapA and SynA in many cases where trafficking factors are repressed is indeed an interesting outcome. Inefficient repression was dismissed based on the lack of colony growth (see relative growth tests of SarA, Sec24, Sec13, Sec31, SedV, GeaA, RabO, RabE, Ykt6, Sft1, SsoA and Sec9), but also by western blots (e.g., Sec24, Sec13, Sec31 or Sec9 shown in the present manuscript, or other trafficking proteins studied previously. Martzoukou et al., 2018; Dimou et al., 2020). Repression of Sec12 and HypB, and to lower degree AP-1, allowed formation of small and/or compact colonies, but even in these cases relative protein levels could not be detected in western blots, guaranteeing efficient repression.

1. In Fig 4A immunoblot, HA-tagged proteins are not detected after thiamine repression. Please state the time of thiamine repression used before protein extraction and blot. Is this for the same length of time as for cells shown in panel 4C? It would also be helpful to state the time of cargo derepression before capturing images in 4C. The methods section mentions 12-14 hr or 16-22 hr of growth, presumably with thiamine in the culture, and then 1-8 hr or 60 min to 4 hr of cargo derepression before imaging. Please specify.

The time of thiamine repression before protein extraction was 16-18h. The same repression time was used for experiments shown in Figures 4C and 6C (ER/COPII and Golgi/post-Golgi repression respectively). More specifically, for microscopy experiments cells were grown in the presence of glucose and thiamine for 12-14h (repressed UapA/SynA and thiAp expressed gene). After this time, cells were shifted to fructose and thiamine for 4h (derepression of UapA/SynA and repression of thiAp expressed gene). In both cases (protein extraction and microscopy experiments) the total time of thiamine repression was 16-18h.

1. For the thiA-copA and thiA-arfA repression experiments (Fig 5C), the methods section states that thiamine was not added ab initio in the culture, but after an 8 h time window without thiamine at the start of spore incubation. This is interpreted to mean that repression was for a shorter period to time than the 12-14 hr overnight growth. However, the figure legend states that De novo synthesis of cargos takes place after full repression of CopA and ArfA is achieved (>16 hr). Please clarify.

We think that the review was confused with repression of cargo synthesis (via alcAp+glucose) versus repression of trafficking proteins (via thiAp+thiamine). Please see Materials and methods. We clarify our protocol also here:

For the thiAp-copA and thiAp-arfA repression experiments addition of thiamine ab initio in the culture leads to total arrest of spore germination and germling formation. Thus, we added an 8-hour time window without thiamine to allow conidiospores to germinate until the stage of young germlings, under conditions where cargo expression via the alcAp was repressed by glucose. Subsequently, thiamine was added in the media (16-18 h) to repress CopA and ArfA, while cargo expression remained glucose-repressed. The transcriptional repression of the cargoes UapA and SynA was maintained for a longer period (24-26 h) compared to other repression experiments, but longer times of repression of cargoes do not make any difference, as full repression is achieved already at 12 h. De novo cargo trafficking was followed next day by eliciting depression, via a shift to fructose media, while still maintaining thiamine to repress CopA or ArfA.

1. In Fig 6D, BFA treatment is shown to trap SynA in Golgi aggregates while UapA still reaches the plasma membrane. Please state the time of BFA treatment before collecting these images. Do longer treatments with BFA before cargo derepression cause accumulation of UapA in intracellular compartments?

As mentioned above (response to Reviewer’s #3 comment 9) cells were grown overnight under repressed conditions for both UapA and SynA. After 12-14h cells were shifted to derepressed conditions using fructose as carbon source. BFA was added after 90min of cargo derepression, while both cargoes were still in cytoplasmic structures so there was not preexisting SynA or UapA at the PM (see also Figure 1C). We have not noticed any different effect on UapA trafficking after a max of 1h of BFA treatment.

1. A minor point, but on page 21 the methods state that "cells were shifted down to the permissive temperature (25 C), to restore the secretory block...". Suggest changing to "to reverse the secretory block..."

Modified accordingly

Reviewer #4 (Significance):

This manuscript nicely builds on a developing line of investigation in the Aspergillus nidulans model that specific plasma membrane proteins are efficiently delivered to the cell surface in a pathway that is distinct from the canonical secretory pathway. Previous work from this lab has suggested that a subpopulation of COPII carriers can bypass the Golgi for delivery of specific cargo to the plasma membrane. The current study uses dual expression of UapA-GFP and mCherry-SynA to provide further support for this model. Molecular definition of a direct ER to PM transport pathway for secretory cargo would be a significant advance to a broad audience. This study provides additional depth and support that such a pathway exists but does not define how COPII vesicles or related intermediates are transported to the PM.

Again, thank you for your positive comments.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #4

Evidence, reproducibility and clarity

In this study, Sagia et al investigate the trafficking of different secretory cargo in Aspergillus nidulans under conditions that repress expression of transport factors or block stages in membrane trafficking. The primary approach is to conduct dual live-cell imaging of GFP-tagged UapA (plasma membrane localized purine transporter) and SynA (plasma membrane R-SNARE) after their simultaneous derepression to monitor trafficking routes. In germlings, both secretory proteins are detected in non-overlapping intracellular compartments and puncta after 60-90 min of derepression. After 4-6 hrs, SynA localizes to hyphal tips whereas UapA localizes to non-polar regions of the PM. Colocalization studies do not show UapA overlap with Golgi markers (SedV, PH-OSBP) during its biogenesis whereas SynA displays significant co-localization. Repression of COPII and COPI components generally block transport of both cargos to the PM and cause accumulation in ER compartments, although there are some differential effects on UapA and SynA localization. Finally, repression of other transport factors (ER-Golgi SNAREs, Golgi transport factors, and exocytic machinery) had differential effects on UapA and SynA localization over time with UapA reaching the plasma membrane in many instances and SynA accumulating in intracellular compartments.

Based on these observations, the authors conclude that UapA and SynA follow distinct trafficking routes to the plasma membrane where SynA uses a canonical SNARE-dependent secretory pathway route and UapA follows a non-canonical route that may bypass Golgi compartments. The study is extensive and supports the model that biogenesis of SynA and UapA follow distinct processes. However, there are some complexities that may limit interpretation. First, the cargo studied are targeted to the ER differently. UapA is a multispanning transmembrane protein that is likely dependent on the Sec61 translocon for co-translational membrane insertion and will involve ER chaperones and quality control machinery for its biogenesis. SynA will depend on the tail-anchored machinery (GET/TRC pathway) for insertion into the ER and is processed by cytosolic factors/chaperones. Therefore, the sites of ER insertion and the rates of biogenesis of these cargo will be different. In addition, the repression of trafficking machinery used in this study appear to be variable and may exert partial blocks on intracellular transport stages. Regardless, the study clearly documents that SynA and UapA follow distinct biogenesis and transport processes when co-expressed in cells under experimentally controlled conditions.

1. It was not clear if the translation, ER insertion and folding of UapA and SynA are fully synchronous. Is it possible that the rate of UapA synthesis and transport to the plasma membrane is substantially faster than for SynA? The imposition of transport blocks could trap SynA and not UapA if this cargo was at later transport stages.
2. In repressing transport factors (e.g. sarA, sec12, sec24, sec13, sedV, rabE), it is clear that under thiamine repressing conditions these cells do not grow or have greatly reduced growth rates. However, it was not clear if proteins are depleted to the same extent in cells after repression for 12-14 hr or 16-22 hr as mentioned in the methods. Indeed, in some cases depleted cells display different cargo localization patterns, for example 67% of cells show normal localization of UapA and SynA after sec12 repression and 33% show ER accumulation of both cargo. There is differential localization of UapA and SynA in many cases where transport factors are repressed, but this could be due to partial inhibition and not complete blocks. It would be helpful to clearly indicate the time points and conditions in each of the figure legends as in points 3-5 below.
3. In Fig 4A immunoblot, HA-tagged proteins are not detected after thiamine repression. Please state the time of thiamine repression used before protein extraction and blot. Is this for the same length of time as for cells shown in panel 4C? It would also be helpful to state the time of cargo derepression before capturing images in 4C. The methods section mentions 12-14 hr or 16-22 hr of growth, presumably with thiamine in the culture, and then 1-8 hr or 60 min to 4 hr of cargo derepression before imaging. Please specify.
4. For the thiA-copA and thiA-arfA repression experiments (Fig 5C), the methods section states that thiamine was not added ab initio in the culture, but after an 8 h time window without thiamine at the start of spore incubation. This is interpreted to mean that repression was for a shorter period to time than the 12-14 hr overnight growth. However, the figure legend states that De novo synthesis of cargos takes place after full repression of CopA and ArfA is achieved (>16 hr). Please clarify.
5. In Fig 6D, BFA treatment is shown to trap SynA in Golgi aggregates while UapA still reaches the plasma membrane. Please state the time of BFA treatment before collecting these images. Do longer treatments with BFA before cargo derepression cause accumulation of UapA in intracellular compartments?
6. A minor point, but on page 21 the methods state that "cells were shifted down to the permissive temperature (25 C), to restore the secretory block...". Suggest changing to "to reverse the secretory block..."

Significance

This manuscript nicely builds on a developing line of investigation in the Aspergillus nidulans model that specific plasma membrane proteins are efficiently delivered to the cell surface in a pathway that is distinct from the canonical secretory pathway. Previous work from this lab has suggested that a subpopulation of COPII carriers can bypass the Golgi for delivery of specific cargo to the plasma membrane. The current study uses dual expression of UapA-GFP and mCherry-SynA to provide further support for this model.

Molecular definition of a direct ER to PM transport pathway for secretory cargo would be a significant advance to a broad audience. This study provides additional depth and support that such a pathway exists but does not define how COPII vesicles or related intermediates are transported to the PM.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #3

Evidence, reproducibility and clarity

The manuscript by Sagia et al compares the trafficking of a polarized (SynA) with a non-polarized (UapA) transmembrane protein. In agreement with previous work of the same lab, they find that UapA reaches the plasma membrane through a Golgi-bypass route, which they characterize to some extent. Overall, the data are of good quality and the story is interesting and timely. Understanding trafficking routes that bypass the Golgi is highly interesting. Nevertheless, there are several points of criticism that I have and below is a list where I combine major and minor points together:

Major Comments:

1. Is it possible that the polarized phenotype of SynA is caused by selective removal, i.e. SynA is delivered to the entire plasma membrane, but endocytosed rapidly from all areas except the tip of the hyphae. This would also result in a polarized distribution.
2. The authors describe the distribution of SynA and UapA in cells deficient of various COPII/ERES proteins. However, these data are not shown, and it is not clear how they were quantified. It would be important to add quantitative data here.
3. on page 8, the authors discuss the discrepancy regarding the role of Sec13. They offer as an explanation that the previous studies have been performed in strains that separately expressed the two cargoes. However, I am unable to see why and how this would be a valid explanation.
4. Why is the effect of Sec24 depletion so much stronger than of Sec12 depletion? Sec12 is the GEF for SarA, without which Sec24 should not be recruited to ERES. The explanation that low amounts of Sec12 are still present and sufficient to carry out the role of this protein. What is the evidence for that?
5. In Figure 5, it would help readers who are not so familiar with Aspergillus organelle morphology to explain the figure a bit better. This might appear trivial for experts, but anyone from outside this field is slightly lost.
6. The authors write that not seeing UapA in Golgi membranes is evidence that it does not pass through this organelle. However, when they write that SynA is never seen in cis-Golgi elements, they do not conclude that SynA bypasses the cis-Golgi.
7. Figure 5C: the authors claim that the CopA and ArfA affects trafficking of UapA and SynA from ER to plasma membrane and assign copA and ArfA as regulators fo anterograde trafficking. I think this interpretation is not justified by the data. Depletion of CopA and ArfA will affect the Golgi apparatus in structure and function. The more straight-forward interpretation is that repression of the COPI machinery results in a defect in Golgi exit and therefore retention in pre-Golgi compartments (including the ER and maybe the ERGIC should it exist in Aspergillus). The same is true for BFA treatment where there are also negative effects on ER export, which are rather indirect consequences of alterations of Golgi function and integrity. Likewise, the interpretation of the papers by Weigel et al and Shomron et al is not correct. It is more likely that COPI is recruited to the growing ERES-derived tubule (or ERGIC) to recycle proteins back to the ER. This is not necessarily a proof that COPI regulates anterograde trafficking
8. Figure 6: The images look like in Figure 5, yet here you don't call them ER-associated.
9. Figure 6D: How long was the BFA treatment. I am surprised that the pool of SynA preexisting at the plasma membrane seems to also be sensitive to BFA.
10. This might be beyond the scope of this study, but as far as I know UapA is not N-glycosylated. Would the introduction of an N-glycosylation site shift it towards the Golgi-based route?

Minor Comments

1. This might be just a personal preference, but I think that the term polar is misleading, because it implies something about the polarity of the amino acids. I think "polarized" might be the more common term. Anyway, this is just a minor point and just a suggestion from my side.
2. The paper by the Saraste lab should be mentioned and discussed (PMID: 16421253), which I think is very relevant to the current story.
3. Having worked with ERES for over two decades, I find it strange to see it written ERes. I see no reason why ER exit sites in Aspergillus should be abbreviated differently from all other types of cells (yeast, drosophila, worms, mammals). I think that the entire acronym should be capitalized.
4. When discussing the data about the partial effect of Sec13, it would be good to refer to a previous paper by the Stephens lab that showed that silencing Sec13/31 results in a defect in trafficking of collagen, but not of VSVG (PMID: 18713835)

Significance

Overall, the data are of good quality and the story is interesting and timely. Understanding trafficking routes that bypass the Golgi is highly interesting. The main weakness is the lack of mechanistic understanding of the Golgi-bypass pathway. In addition, the study is limited to two proteins as representatives of polarized vs. non-polarized proteins. The main target audience for this paper are scientists working in the area of secretion and trafficking in the secretory pathway.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

The idea that transmembrane proteins of the plasma membrane move from the ER to the Golgi and then to the cell surface is firmly entrenched, and the mechanisms and components of this secretory pathway have been extensively characterized. Secretory vesicles are often delivered from the Golgi to sites of polarized growth. This paper builds on previous work by the same group to provide evidence that in Aspergillus nidulans, some non-polarly localized plasma membrane proteins follow a very different pathway, which bypasses components of the conventional secretory machinery such as SNAREs that have been implicated in secretion as well as the exocyst. In particular, they systematically compare the trafficking of the SNARE SynA, which follows the conventional secretory pathway, with that of the purine transporter UapA, which apparently does not. The two proteins were co-expressed in the same cells using the same promoter. A variety of genetic and microscopy methods are used to support the conclusion that UapA reaches the plasma membrane by a route distinct from that followed by SynA.

In my view, the authors present a convincing case. The individual experimental results are sometimes ambiguous, but the combined results favor the conclusion that UapA follows a novel pathway to the plasma membrane. I have only a few relatively minor comments.

1. In the Introduction and elsewhere: to my knowledge, there is no clear evidence that AP-1-containing clathrin-coated vesicles carry cargoes from the Golgi to the plasma membrane. On the contrary, as recently reported by Robinson (https://pubmed.ncbi.nlm.nih.gov/38578286/), AP-1-containing vesicles likely mediate retrograde traffic in the late secretory pathway.
2. In Figure 2, is there any known significance to the presence of UapA in "cytoplasmic oscillating thread structures decorated by pearl-like foci as well as a very faint vesicular/tubular network"?
3. SynA is related to S. cerevisiae Snc1/2, which are known to be present in late Golgi compartments due to repeated rounds of endocytosis to the Golgi and exocytosis to the plasma membrane. The SynA shown here to colocalize with PH-osbp is probably present in a similar recycling loop rather than being en route to the plasma membrane for the first time. Therefore, the differential colocalization of UapA and SynA with PH-osbp does not by itself provide "strong evidence that the two cargoes studied traffic via different routes" as stated in the text, but might instead indicate that only SynA undergoes frequent endocytosis. The text should be amended accordingly.
4. A missing piece of the story is a test of whether the puncta visualized for the two cargoes in Figure 5B are indeed distinct populations of COPII-containing ER exit sites. The relevant experiment would involve co-labeling of the cargoes together with a COPII marker. Three-color labeling would presumably be needed.

Significance

This study provides compelling evidence that in the fungus Aspergillus nidulans, some transmembrane transporter proteins reach the plasma membrane by a pathway that bypasses much of the conventional machinery associated with the Golgi apparatus and secretory vesicles. Although previous publications pointed toward a similar conclusion, the present work tackles the problem in a more rigorous and systematic way. These findings are important for cell biologists who study membrane traffic, although it remains to be determined how prevalent this type of non-canonical secretion might be in other organisms.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

Sagia et al. present a manuscript using A. nidulans as model to study different transport routes of membrane proteins from the ER to the plasma membrane. They showed in earlier work that apparently at least two different transport routes exist, one involving the classical ER-ERES-ERGIC-Golgi route, one bypassing the Golgi. Unpolarized membrane proteins use the former, apically sorted membrane proteins the latter route. The study here confirms their earlier findings, uses a better model (co-expression of representatives for both routes in the same cell) and provides additional mechanistic insights about the roles of rabs, SNARES and other important proteins of the secretory pathway. The study is thoroughly done, figures are of high quality, data and methods well described and adequately replicated.

I do have, however, a number of comments that could help to improve the manuscript.

• I suggest to use the term polarized or apical rather than polar. Polar alone to me refers more to physico-chemical properties like water-solubility.
• introduction and discussion: I don´t think the literature about unconventional secretion bypassing the Golgi is complete, for example studies about TMED10 like Zhang, M. et al. Cell 181, 637-652 e615 (2020) or Zhang et al. Elife 4 (2015) are missing, there might be others. Is UapA a leader-less cargo that could be inserted via TMED10 translocation?
• Fig. 1C. Can these intracellular structures be characterized in more detail? Where is the Golgi localized in A. nidulans, is it decentralized like in yeast? Is the UapA at the time points shown in Fig. 1C in some sub-PM structures? To me the distribution at or near the PM is more punctate than in the steady state image shown in 1B.
• Fig. 3A. To me it looks like there is actually a lot of colocalization of UapA and SynA, especially at or near the PM, where there is quite some white, punctate staining. The green fluorescence is just much stronger, overlaying the violet. Can you show separate channels and explain?
• Fig. 3: In my opinion the statement that UapA "is probably sorted from an early secretory compartment, ultimately bypassing the need for Golgi maturation" is too strong at that point. You say for both UapA and SynA you don´t get significant colocalization with early Golgi/ERGIC marker, then you cannot conclude that one takes the conventional route via early-late Golgi and the other does not. What you can say is that UapA is apparently not going through late Golgi.
• Fig. 4C: UapA does not seem to accumulate in the ER in the Sec24 and 13 mutants but in punctate structures. This for me is unexpected, any explanations? Can you characterize that punctate staining?
• Fig. 6D: You state that BFA "has only a very modest effect on UaPA translocation to the PM". To me the PM (or very near PM) staining of UaPA looks very different in the PFA treated cells, more uneven/punctate. Is there an explanation for that?

Significance

One strength of the study is the use of a model organism, A. nidulans, not cell cultures. Also the use of both reporters, UapA and SynA, in the same cell is an advantage over previous studies using different lines and different promotors. Limitation of the study might be that it remains unclear to what extend the basic mechanism (UapA and SynA are transported to PM in different carrier and via different routes) can be generalized to other polarized (apically?) membrane proteins versus non-polarized membrane proteins in A. nidulans and whether a similar mechanism exists in other organisms.

Some of the basic findings of the study are not new but were published by the same group. However, as the authors point out, the current study uses improved assays and extends their previous studies, advancing our understanding of the mechanistics of transport in the conventional secretory pathway and novel alternative routes. The study will be of interest for basic researchers in the trafficking field.

My own expertise is transport through the secretory pathway in mammalian cells, many years ago more post-Golgi, now mostly ER-Golgi and ER itself.

Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

Learn more at Review Commons

---

Reply to the reviewers

Reviewer #1 (Evidence, reproducibility and clarity (Required)):

Summary: MIC26 is a subunit of the 'mitochondrial contact site and cristae organizing system' (MICOS) complex required for crista junction (CJ) formation and was functionally linked to diabetes and modulation of lipid metabolism. In order to understand the role of MIC26 in metabolism, the authors generated MIC26-KO HepG2 cells and investigated the pathways regulated by MIC26 under normo- and hyper-glycemic culture conditions. They employed a multi-omics approach that include transcriptomics, proteomics, targeted metabolomics, and functional assays to document the changes in mRNAs, proteins, and metabolites as a result of MIC26 deletion. Through bioinformatic analyses, they showed that the function of MIC26 is critical in various pathways regulating fatty acid synthesis, oxidation, cholesterol metabolism, and glycolysis. Interestingly, they found an entirely antagonistic effect of lipogenesis in MIC26-KO cells compared to WT cells depending on the glucose concentration of the culture media. In addition, they showed that MIC26 deletion led to a major metabolic rewiring of glutamine utilization as well as oxidative phosphorylation.

Major comments: 1) This is basically a descriptive study that document the transcriptomic, proteomic, and metabolic consequences of lacking MIC26 in normal or high glucose environment. It is data rich but insight poor. The connections between MIC26 as a subunit of MICOS complex and all those metabolic pathways are so tenuous that it is hard to see what to follow up after.

__Response: __We respectfully differ from the reviewer’s opinion that the manuscript is data rich and insight poor. Our study provides significant insights by demonstrating how MIC26, strategically residing in the mitochondrial inner membrane (IM), regulates major cellular pathways.

MIC26 operates in a dual manner:

A) Depending on the nutritional status, normoglycemia or hyperglycemia, MIC26 regulates the glycolysis, lipid, cholesterol metabolism and TCA cycle intermediates in an antagonistic manner. B) Independent of the nutritional status, it regulates glutamine and OXPHOS metabolism.

In addition, based on the suggestions by Reviewer #2, we have tested whether other proteins of MICOS (MIC27 and MIC19) present in two different sub-complexes regulate important metabolic pathways. Using the experimental results achieved (See reply to comments from Reviewer #2), we conclude that MIC26 plays a unique role as metabolic regulator in the IM and this study is therefore important to the general field of metabolism.

2) In the Results section (page 5, line 114-123), the description of the Western blot (WB) analysis appears inconsistent with several blot images of Fig. 1A, which makes the result unconvincing. The authors should select appropriate representative WB images, assuming they have them, to support their claim.

Response: Thank you for the suggestion. Firstly, we have performed additional experiments in this regard, and included the relevant quantification. Secondly, we have replaced some WBs which depict the appropriate quantification.

Further, we also modified the relevant lines in the manuscript stating that ‘Mitochondrial apolipoproteins, MIC26, MIC27, and MIC25 are increased in cells exposed to hyperglycemia’ and not only MIC26 as stated before.

Minor comments:

3) As the functional role of MIC26 in metabolism is the primary focus, the authors should present the results in the figures in the order of WT-N, MIC26KO-N, WT-H, MIC26KO-H for easier comparison.

Response: We understand the reasoning to interchange the two conditions. However, such an endeavour will involve cropping images of WBs, BN-PAGE and Clear native PAGE to better represent the corresponding quantification. This will also involve modifying all figures (data-sets and functional assays) in the whole manuscript. Overall, considering that the benefits of interchanging the order, of the WT-Hyperglycemia and MIC26 KO-Normoglycemia, are relatively minor, we have decided to stick to the original representation of the figures.

Reviewer #1 (Significance (Required)):

Mitochondria play important roles in metabolism and metabolic disorders. This study generated large amount of data relating to the role of a mitochondrial protein MIC26 in metabolism. Mutations in MIC26 have been associated with mitochondrial myopathy, lactic acidosis, and cognition defects (Beninca et al. 2021) and a lethal progeria-like condition (Peifer-Weib et al. 2023). There is also a connection between MIC26 and metabolic disorders. The results of this study will be of interest to researchers in the fields of mitochondrial diseases and metabolic disorders. My field of expertise is mitochondrial disease, proteomics, lipidomics, phospholipid biochemistry.

Reviewer #2 (Evidence, reproducibility and clarity (Required)):

Summary

This study determines the role of MIC26, a mitochondrial component of the MICOS complex, in influencing cellular metabolic status. Using a variety of multiomic profiling techniques, as well as functional assays such as Seahorse-based respirometry, the authors propose that MIC26 regulates a variety of metabolic processes, including lipid and cholesterol homeostasis, glycolysis, fatty acid oxidation, fatty acid synthesis, TCA cycle homeostasis, glutamine metabolism, and general mitochondrial bioenergetics via OxPhos activity and supercomplex formation. The data were generated in MIC26 knockout HepG2 hepatocellular carcinoma cells grown in two nutrient conditions: high glucose (which they term "hyperglycemia") and low glucose (which they term "normoglycemia") DMEM. While the data support the authors' conclusion that loss of MIC26 causes braod metabolic changes across these conditions, the authors do not distinguish whether MIC26 knockout affects metabolism due to its canonical role in MICOS, or whether it acts as "a metabolic rheostat" to directly regulate central cellular fuel pathways, as is claimed in the title of this manuscript. Given the breadth of metabolic alterations seen in the MIC26 KO cells, it seems likely that at least a subset of these changes are indirect, rather than that MIC26 plays a direct regulatory role in the eight distinct metabolic pathways outlined above. Thus, while the data generally support the conclusion that expression of MIC26 is important for metabolic homeostasis, the mechanism(s) by which MIC26 influences cellular metabolism remains unclear and should be further addressed.

Major Comments

1) The authors infer that MIC26 influences cellular metabolism by referencing a handful of papers on MIC26 transcript differences in select metabolic models, metabolic alterations seen in a MIC26 transgenic/overexpression mouse model, or metabolic effects seen due to MIC26 tissue specific KO models. They thus hypothesize that "MIC26 has an unidentified regulatory role under nutrient-enriched conditions" (line 88). They support this observation by showing that MIC26 is increased in high glucose DMEM relative to low glucose DMEM in Figure 1. However, the authors claim, in both the figure legend title as well as the results section header, that "MIC26 is selectively increased in cells exposed to hyperglycemia" (lines 103-4), when their data demonstrate that this is not true. Figure 1B shows that MIC27, MIC26, and MIC25 increase in high glucose relative to low glucose conditions, indicating that MIC26 is not selectively increased, but rather that multiple subunits of MICOS are increased under nutrient-enriched conditions.

Response: We agree to the reviewer’s comment and have now replaced the title in the results section in the manuscript accordingly - ‘Mitochondrial apolipoproteins, MIC26, MIC27, and MIC25 are increased in cells exposed to hyperglycemia’ and not only MIC26 as stated before. Further, in order to strengthen the WB data from Fig 1A & B, we also increased the number of experiments and updated the Fig 1B and also some WBs in Fig 1A to better represent the quantification.

2) This does not suggest that MIC26 does not have an important role in maintaining metabolism under such nutrient conditions, but rather supports a model in which MICOS itself dynamically responds to altered nutrient conditions in cell culture. Interestingly, other MICOS subunits do not change in abundance (e.g., MIC19 and MIC60), which have previously been associated with a separate MICOS subcomplex than MIC26 in yeast (PMID: 33053165). These data may suggest that the nutrient responsive behavior of MIC26 may be due to its assembly within this specific MICOS subcomplex, rather than an independent "unidentified regulatory role under nutrient-enriched conditions".

To test this, the authors should repeat a subset of functional assays (perhaps the Seahorse metabolic assays) in other MICOS deletion cell lines, including one that dynamically changes in expression in a similar manner to MIC26 (e.g., MIC27 KO or MIC25 KO) and one that does not dynamically change in expression in high glucose conditions (e.g., MIC60 or MIC13). In these sets of experiments the authors will be able to distinguish three possibilities:

Model 1: if MICOS in general affects metabolic pathways, similar results will be seen for all MICOS subunit KOS.

Model 2: if the MIC26-specific subcomplex is dynamically regulated to influence cellular metabolism, KO of MIC26 and other subunits of this subcomplex should show similar results, but KO of subunits of the non-dynamic subcomplex (e.g., MIC60) would not show similar phenotypes.

Model 3: If only MIC26 KO, but not other MICOS subunits, show metabolic phenotypes, this would support a MICOS-independent role for MIC26 in influencing cellular metabolism.

Importantly, any of these models are interesting, and testing these models does not invalidate any of the phenotypes presented in the manuscript. Rather, these experiments would assist the reader in understanding the underlying mechanism by which MIC26 loss causes cellular metabolic defects. Furthermore, it is worth stating that performing all experiments with multiple other MICOS cell lines is beyond the scope of the manuscript, but testing effects in select (preferably functional) assays, such as the glycolysis stress test (Fig 3E), FAO Seahorse (Fig 3H-J) glutamine oxidation (Fig 6B), and general stress test (Fig 7E) would be appropriate. Other more defined and easily achievable experiments could also be used to support these claims (e.g., western blots probing for levels of key metabolic regulators).

__Response: __We appreciate the balanced comments of the reviewer who has carefully read and appreciated our manuscript. We also appreciate the constructive criticism of the reviewer who suggested various possible models considering the MIC26 role.

In this endeavour, we have performed the following extensive experiments:

1. Generated MIC19 KOs in HepG2 cells. We had already generated MIC27 KOs during the course of a previous publication (Lubeck et al, 2023) and used them for this publication (Fig S4). WB analyses was performed using the MIC27 KO and MIC19 KO cells.
2. Measured glycolysis function using the glycolysis stress test in MIC27 and MIC19 KOs (Fig S5).
3. Analysed lipid metabolism by imaging and extensively quantifying LD number and BODIPY fluorescence intensities in MIC27 and MIC19 KO cells (Fig S6).
4. Measured glutamine oxidation using Mito Flex fuel tests in MIC27 and MIC19 KOs (Fig S10F).
5. Analysed general mitochondrial respiration by using mito stress kit in MIC27 and MIC19 KOs (Fig S11-E-H). We provide a summary of the above results:

#

Metabolic Pathway

Experiment

MIC26____ KO

MIC27____ KO

MIC19____ KO

1

Glycolysis

Glycolysis stress kit

Glycolytic reserve increased

Glycolytic reserve unchanged

Glycolytic reserve unchanged

2A

Lipid Metabolism

LD number

General increase

No consistent increase in treatment with and without palmitate in normoglycemia and hyperglycemia

No consistent increase in treatment with and without palmitate in normoglycemia and hyperglycemia

2B

Lipid Metabolism

LD (BODIPY) intensities

Presence of antagonistic regulation of LD content

Presence of antagonistic regulation of LD content

Absence of antagonistic regulation of LD content

3

Glutamine oxidation

Mito flex fuel test

No dependency

Dependent

Dependent

4

Steady-state respiration

Mito stress test

Basal respiration increased

Basal respiration unchanged

Basal respiration unchanged

5

Steady-state respiration

Mito stress test

SRC decreased in normoglycemia

SRC unchanged

SRC unchanged

Taking the above summary, we investigated three possibilities considering the role of MIC26:

1. General role of MICOS – whether deletion of any MICOS protein leads to similar phenotype as MIC26 deletion

2. Specific role of MIC26/27/10 subcomplex – whether deletion of any other protein in the MIC26-subcomplex like MIC27 leads to similar results, accompanied by dissimilar results in KOs of any protein belonging to the other MICOS subcomplex (MI19/MIC25/MIC60) and whose protein levels were not changed upon hyperglycemia treatment (MIC19 or MIC60 KOs).

3. MICOS-independent role of MIC26. Considering the various metabolic pathways analysed, we conclude that MIC26 has a MICOS-independent role in regulating major cellular pathways.

Minor Comments 1. The multiomics data as presented in Figure 2 is difficult to interpret. This is mainly driven by the fact that there are mulitpe comparisons that should be communicated (KO v WT, normoglycemia v hyperglycemia, upregulated v. downregulated), but only select enrichment values are shown (e.g., normoglycemia upregulated in 2C and hyperglycemia downregulated in 2D). It took me a long time as a reader to understand what I was looking at because only select analyses are presented. What pathways are upregulated in hyperglycemia in MIC26 KO v. WT?

__Response: __Thank you for pointing this out. We have now represented all the four conditions regarding enrichment analysis as suggested. The antagonistic metabolic regulation is only observed in MIC26 KO cells cultured in normoglycemia (upregulated) when compared to MIC26 KOs cells cultured in hyperglycemia (downregulated). When MIC26 KOs cultured in hyperglycemia (upregulated) were compared with MIC26 KOs cultured in normoglycemia (downregulated), there were also few pathways which showed an antagonistic regulation, but not directly involved with metabolism, relating to apoptosis etc. Hence, we did not focus on these pathways in the current manuscript.

The Treemaps have been shifted to Fig S1. All four Treemaps have been represented instead of the two shown before. In the process, the previous proteomics Fig S1B and S1C depicting antagonism in different pathways have been excluded.

2. Figure 3 would be stronger if expression from all glycolytic proteins were shown instead of only a subset. If the authors are making the claim that MIC26 KO increases glycolytic flux via protein-level upregulation of glycolysis, this could be substantiated at a pathway level. These data could be included in supplemental data if they are difficult to fit into the figure.

__Response: __Thank you. We have now included the data of proteins regulating glycolysis as a new figure (Fig S3C). MIC26 KO cells cultured in normoglycemia had increased levels of aldolase (ALDOA & ALDOC), phosphoglycerate kinase (PGK1) and pyruvate kinase (PKM & PKLR) when compared to control cells. We have included this data in the manuscript.

3. In Figure 4H-J - the raw data for the Seahorse traces should be shown, and OCR should be reported in pmol/min rather than relative percentages so as to help the reader more critically evaluate the data.

__Response: __For the FAO assays, we treat the cells with palmitate or mock (BSA serving as control). The histograms (Fig 4H-J) are represented as such because we normalised the oxygen consumption after palmitate treatment with oxygen consumption of mock-treated cells. We understand the reviewer’s concern and have now included the absolute values of oxygen consumption of FAO assay in an excel sheet (Supplementary Table S5). In addition, we have also included the absolute values for mito stress test and glycolysis assays where the oxygen consumption has been normalised to WT-normoglycemia condition (Supplementary Table S5). The original oxygen consumption curves for glycolysis stress kit (Fig 3E, S5A) and mito stress kits (Fig 7E, S11E) are shown as figures.

4. The majority of the plots are shown with 4 comparisons, but statistical comparisons are often only provided for a subset of comparisons. It is unclear whether statistics were compared across all comparisons and the non-annotated comparisons are not significant, or whether those calculations were not performed. Defining this in the figure, or, better, annotating all relevant comparisons on each graph with "ns" for not significant, would assist the reader with interpreting the data.

__Response: __Thanks for pointing this out. After comparing all meaningful conditions (except WT-N to MIC26 KO-H and WT-H to MIC26 KO-N), only those that were significant were represented using asterisks. We have now mentioned this information in the respective figure legends. We avoided using ‘non-significant (ns)’ in the figure as it would make some figures very crowded as seen from some of our trials.

Reviewer #2 (Significance (Required)):

This study broadly profiles the metabolic defects associated with loss of the MICOS subunit MIC26 in hepatocellular carcinoma cells in variable nutrient conditions (e.g., high glucose and low glucose). As a reviewer with expertise in multiomic profiling of metabolic models, I found the breadth of pathways studied in this manuscript to be impressive. Furthermore, the authors use a variety of techniques, including multiomic profiling, isotopic flux analysis, and functional Seahorse assays to support their conclusions. The study provides a comprehensive analysis of metabolic changes associated with MIC26, and is thus an important advance in profiling how loss of MIC26 (or MICOS in general; see below) affects cellular metabolism in the context of dynamic nutrient changes. However, the claim that "MIC26 is a metabolic rheostat regulating central cellular fuel pathways", as is proposed in the title of the manuscript, is unsubstantiated, as the authors do not test whether loss of MIC26 specifically influences cellular metabolism independent of its role in MICOS. This paper would be significantly strengthened if a subset of functional assays across metabolic pathways were repeated with other MICOS KO cell lines to delineate whether these metabolic effects are direct or indirect.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

Summary

This study determines the role of MIC26, a mitochondrial component of the MICOS complex, in influencing cellular metabolic status. Using a variety of multiomic profiling techniques, as well as functional assays such as Seahorse-based respirometry, the authors propose that MIC26 regulates a variety of metabolic processes, including lipid and cholesterol homeostasis, glycolysis, fatty acid oxidation, fatty acid synthesis, TCA cycle homeostasis, glutamine metabolism, and general mitochondrial bioenergetics via OxPhos activity and supercomplex formation. The data were generated in MIC26 knockout HepG2 hepatocellular carcinoma cells grown in two nutrient conditions: high glucose (which they term "hyperglycemia") and low glucose (which they term "normoglycemia") DMEM. While the data support the authors' conclusion that loss of MIC26 causes braod metabolic changes across these conditions, the authors do not distinguish whether MIC26 knockout affects metabolism due to its canonical role in MICOS, or whether it acts as "a metabolic rheostat" to directly regulate central cellular fuel pathways, as is claimed in the title of this manuscript. Given the breadth of metabolic alterations seen in the MIC26 KO cells, it seems likely that at least a subset of these changes are indirect, rather than that MIC26 plays a direct regulatory role in the eight distinct metabolic pathways outlined above. Thus, while the data generally support the conclusion that expression of MIC26 is important for metabolic homeostasis, the mechanism(s) by which MIC26 influences cellular metabolism remains unclear and should be further addressed.

Major Comments

1. The authors infer that MIC26 influences cellular metabolism by referencing a handful of papers on MIC26 transcript differences in select metabolic models, metabolic alterations seen in a MIC26 transgenic/overexpression mouse model, or metabolic effects seen due to MIC26 tissue specific KO models. They thus hypothesize that "MIC26 has an unidentified regulatory role under nutrient-enriched conditions" (line 88). They support this observation by showing that MIC26 is increased in high glucose DMEM relative to low glucose DMEM in Figure 1. However, the authors claim, in both the figure legend title as well as the results section header, that "MIC26 is selectively increased in cells exposed to hyperglycemia" (lines 103-4), when their data demonstrate that this is not true. Figure 1B shows that MIC27, MIC26, and MIC25 increase in high glucose relative to low glucose conditions, indicating that MIC26 is not selectively increased, but rather that multiple subunits of MICOS are increased under nutrient-enriched conditions. This does not suggest that MIC26 does not have an important role in maintaining metabolism under such nutrient conditions, but rather supports a model in which MICOS itself dynamically responds to altered nutrient conditions in cell culture. Interestingly, other MICOS subunits do not change in abundance (e.g., MIC19 and MIC60), which have previously been associated with a separate MICOS subcomplex than MIC26 in yeast (PMID: 33053165). These data may suggest that the nutrient responsive behavior of MIC26 may be due to its assembly within this specific MICOS subcomplex, rather than an independent "unidentified regulatory role under nutrient-enriched conditions".

To test this, the authors should repeat a subset of functional assays (perhaps the Seahorse metabolic assays) in other MICOS deletion cell lines, including one that dynamically changes in expression in a similar manner to MIC26 (e.g., MIC27 KO or MIC25 KO) and one that does not dynamically change in expression in high glucose conditions (e.g., MIC60 or MIC13). In these sets of experiments the authors will be able to distinguish three possibilities:

Model 1: if MICOS in general affects metabolic pathways, similar results will be seen for all MICOS subunit KOS.

Model 2: if the MIC26-specific subcomplex is dynamically regulated to influence cellular metabolism, KO of MIC26 and other subunits of this subcomplex should show similar results, but KO of subunits of the non-dynamic subcomplex (e.g., MIC60) would not show similar phenotypes.

Model 3: If only MIC26 KO, but not other MICOS subunits, show metabolic phenotypes, this would support a MICOS-independent role for MIC26 in influencing cellular metabolism.

Importantly, any of these models are interesting, and testing these models does not invalidate any of the phenotypes presented in the manuscript. Rather, these experiments would assist the reader in understanding the underlying mechanism by which MIC26 loss causes cellular metabolic defects. Furthermore, it is worth stating that performing all experiments with multiple other MICOS cell lines is beyond the scope of the manuscript, but testing effects in select (preferably functional) assays, such as the glycolysis stress test (Fig 3E), FAO Seahorse (Fig 3H-J) glutamine oxidation (Fig 6B), and general stress test (Fig 7E) would be appropriate. Other more defined and easily achievable experiments could also be used to support these claims (e.g., western blots probing for levels of key metabolic regulators).

Minor Comments

1. The multiomics data as presented in Figure 2 is difficult to interpret. This is mainly driven by the fact that there are mulitpe comparisons that should be communicated (KO v WT, normoglycemia v hyperglycemia, upregulated v. downregulated), but only select enrichment values are shown (e.g., normoglycemia upregulated in 2C and hyperglycemia downregulated in 2D). It took me a long time as a reader to understand what I was looking at because only select analyses are presented. What pathways are upregulated in hyperglycemia in MIC26 KO v. WT?
2. Figure 3 would be stronger if expression from all glycolytic proteins were shown instead of only a subset. If the authors are making the claim that MIC26 KO increases glycolytic flux via protein-level upregulation of glycolysis, this could be substantiated at a pathway level. These data could be included in supplemental data if they are difficult to fit into the figure.
3. In Figure 4H-J - the raw data for the Seahorse traces should be shown, and OCR should be reported in pmol/min rather than relative percentages so as to help the reader more critically evaluate the data.
4. The majority of the plots are shown with 4 comparisons, but statistical comparisons are often only provided for a subset of comparisons. It is unclear whether statistics were compared across all comparisons and the non-annotated comparisons are not significant, or whether those calculations were not performed. Defining this in the figure, or, better, annotating all relevant comparisons on each graph with "ns" for not significant, would assist the reader with interpreting the data.

Significance

This study broadly profiles the metabolic defects associated with loss of the MICOS subunit MIC26 in hepatocellular carcinoma cells in variable nutrient conditions (e.g., high glucose and low glucose). As a reviewer with expertise in multiomic profiling of metabolic models, I found the breadth of pathways studied in this manuscript to be impressive. Furthermore, the authors use a variety of techniques, including multiomic profiling, isotopic flux analysis, and functional Seahorse assays to support their conclusions. The study provides a comprehensive analysis of metabolic changes associated with MIC26, and is thus an important advance in profiling how loss of MIC26 (or MICOS in general; see below) affects cellular metabolism in the context of dynamic nutrient changes. However, the claim that "MIC26 is a metabolic rheostat regulating central cellular fuel pathways", as is proposed in the title of the manuscript, is unsubstantiated, as the authors do not test whether loss of MIC26 specifically influences cellular metabolism independent of its role in MICOS. This paper would be significantly strengthened if a subset of functional assays across metabolic pathways were repeated with other MICOS KO cell lines to delineate whether these metabolic effects are direct or indirect.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

Summary:

MIC26 is a subunit of the 'mitochondrial contact site and cristae organizing system' (MICOS) complex required for crista junction (CJ) formation and was functionally linked to diabetes and modulation of lipid metabolism. In order to understand the role of MIC26 in metabolism, the authors generated MIC26-KO HepG2 cells and investigated the pathways regulated by MIC26 under normo- and hyper-glycemic culture conditions. They employed a multi-omics approach that include transcriptomics, proteomics, targeted metabolomics, and functional assays to document the changes in mRNAs, proteins, and metabolites as a result of MIC26 deletion. Through bioinformatic analyses, they showed that the function of MIC26 is critical in various pathways regulating fatty acid synthesis, oxidation, cholesterol metabolism, and glycolysis. Interestingly, they found an entirely antagonistic effect of lipogenesis in MIC26-KO cells compared to WT cells depending on the glucose concentration of the culture media. In addition, they showed that MIC26 deletion led to a major metabolic rewiring of glutamine utilization as well as oxidative phosphorylation.

Major comments:

This is basically a descriptive study that document the transcriptomic, proteomic, and metabolic consequences of lacking MIC26 in normal or high glucose environment. It is data rich but insight poor. The connections between MIC26 as a subunit of MICOS complex and all those metabolic pathways are so tenuous that it is hard to see what to follow up after.

In the Results section (page 5, line 114-123), the description of the Western blot (WB) analysis appears inconsistent with several blot images of Fig. 1A, which makes the result unconvincing. The authors should select appropriate representative WB images, assuming they have them, to support their claim.

Minor comments:

As the functional role of MIC26 in metabolism is the primary focus, the authors should present the results in the figures in the order of WT-N, MIC26KO-N, WT-H, MIC26KO-H for easier comparison.

Significance

Mitochondria play important roles in metabolism and metabolic disorders. This study generated large amount of data relating to the role of a mitochondrial protein MIC26 in metabolism. Mutations in MIC26 have been associated with mitochondrial myopathy, lactic acidosis, and cognition defects (Beninca et al. 2021) and a lethal progeria-like condition (Peifer-Weib et al. 2023). There is also a connection between MIC26 and metabolic disorders. The results of this study will be of interest to researchers in the fields of mitochodrial diseases and metabolic disorders.

My field of expertise is mitochondrial disease, proteomics, lipidomics, phospholipid biochemistry.