On 2014 Jan 25, Ferenc Zsila commented:

According to my comments outlined below, the peer-review procedure of this poorly written paper was seriously incomplete (page numbering refers to the PDF version).

Ferenc Zsila M.D., Ph.D. Institute of Molecular Pharmacology, Research Centre for Natural Sciences Budapest, Hungary

Missing experimental data (Materials and Methods, pp. 2-3.)

Purity and fatty acid content of the HSA sample used are unknown. The source and chemical character (salt or free acid, racemic or not) of ibuprofen, warfarin, and digitoxin used in site probe studies are unknown. Experimental details of the site probe measurements (see on p. 7) are completely missing. There is not even a word to describe how the protein and LMF molecules were prepared for the docking procedures (p. 3). Docking type information are also missing (blind? targeted? rigid? flexible?). What are the parameters of the docking box used for the proteins? How much docking runs were performed? Where is the docking energy list? Where are the affinity constants calculated from the binding free energy values? How do they relate to the experimental values?

Errors, speculations, and unsupported claims

Page 1: "Based on these characteristics, with increasing amounts of PPIX in the blood, the probability of HSA-PPIX formation increases accordingly." It is a mere speculation which is not supported by research data. Hemopexin is the primary heme binding protein in the serum. "Methemalbumin is an abnormal component of plasma and has been found only in diseases associated with massive hemolysis, when hemopexin-binding capacity is exceeded." (Tolosano et al., Antioxid. Redox. Signal. 2010 (12) 305-320).

Page 4: "When HSA is excited at 285 nm, it radiates strong intrinsic fluorescence at 334 nm." No emission band at 334 nm can be seen either in Fig. 1A or 1B. Fluorescence contribution of the bound PPIX is totally ignored. Origin and nature of the large emission peak centered around 430 nm remain obscure.

Page 4: "To investigate the effect of TF on the binding of LMF to the (HSA-PPIX)-TF complex, steady-state fluorescence spectra of (HSA-PPIX), shown in Fig. 1A, and of (HSA-PPIX)-TF, displayed in Fig. 1B, were run and showed that the fluorescence intensity of the two systems became consistently weakened when increasing the LMF concentration." Just an opposite situation is shown in Fig. 1A and 1B: increase of the LMF concentration enhances the fluorescence intensity in both cases.

In contrast to HSA, transferrin contains 8 Trp residues. It remains a mystery, how the fluorescence contribution of these residues was considered during evaluation of the emission spectra of (HSA-PPIX)-TF.

Page 5: "Fig. 3 shows second-derivative fluorescence spectra of the (HSA-PPIX)-LMF and [(HSA-PPIX)-TF]-LMF complexes (inset). As can be seen, a negative band was observed at 310 nm and 316 nm in the (HSA-PPIX)-LMF and [(HSA-PPIX)-TF]-LMF spectra, respectively. These changes in the protein spectra should be assigned to the combination of environments of the aromatic residues in the protein and indicate that the Trp residue was in a relatively hydrophobic microenvironment." What Trp residue? There are 9 Trp in the [(HSA-PPIX)-TF]-LMF complex.

Page 7: "The corresponding binding constants were evaluated and are listed in Table 3." These are not binding but quenching constants (cf. Table 1). No association constants (Ka) are shown in Table 3. HSA affinity constants as well as the amounts of the site probes added are unknown.

Page 7: "Based on these results, it was concluded that ibuprofen had little effect on the binding of LMF to (HSA-PPIX)-TF while digitoxin removed LMF from the bindings site." Changing of the quenching constants shown in Table 3 contradicts to this claim. The Ksv value of [(HSA-PPIX) LMF] TF was reduced by an order of magnitude upon addition of both ibuprofen and digitoxin. The exact HSA binding location of digitoxin is unknown so it can not be used as a site probe. It may occur that these site probes bind to transferrin as well. Where are the controll experiments?

Page 9: "In summary, this study provides useful information for the pharmacological and pharmacodynamic behavior of LMF and should be taken into account when calculating its dosage." This work deals with one aspect of the pharmakokinetic behavior of LMF (serum protein binding). No data are presented on the pharmacodynamic behavior of the drug.

Confusing binding site designations

Page 7: "In our previous work [18], we showed that PPIX occupied site I in the HSA and LMF bound to site IIA." As it is defined on page 1, site I corresponds to subdomain IIA ("Aromatic and heterocyclic ligands have been found to bind with two hydrophobic pockets in subdomains IIA and IIIA, namely site I and site II"). Therefore, according to the above sentence, both PPIX and LMF is bound in subdomain IIA.

Page 7: "... we believe that the lower-affinity site of LMF, corresponding to the binding site, was switched in the presence of TF, causing LMF to bind to site III in HSA." No definition of site III is given.

Page 9: "For the (HSA-PPIX)-TF complex in solution, the binding site of LMF to HSA-PPIX changed from Sudlow’s site IIA to Sudlow’s site IIIB with low affinity." No such site designations exist in the literature. Moreover, subdomain IIIB binding of LMF is very doubtful since it is not supported by experimental data.

Wrong references

Page 1: "The direct interaction of drugs with TF has been widely studied [9-12]." Ref. 11 is a drug-serum albumin and not a drug-TF binding study.

Page 1: "LMF is very active against both Gram (+) and Gram (2) bacteria through inhibition of their DNA gyrase and is widely used for the clinical treatment of severe systematic infections, such as soft tissue infection, typhoid fever, bone and joint infections, prostatitis blood poisoning and sinusitis [13]." Ref. 13 is a spectroscopic study on the ion-association complex formation between fluoroquinolones and a synthetic dye (erythrosine). Thus, it does not cover the biological activities of lomefloxacin.

Page 7: "In our previous work [18], we showed that PPIX occupied site I in the HSA and LMF bound to site IIA." Ref. 18 is not the authors' paper.

Page 9: "The literature has numerous reports of LMF being located within an alpha-helix and a beta-sheet in the N-lobe of transferrin [46]." Ref. 46 is not about the LMF-transferrin interaction.

On 2014 Jan 25, Ferenc Zsila commented:

According to my comments outlined below, the peer-review procedure of this poorly written paper was seriously incomplete (page numbering refers to the PDF version).

Ferenc Zsila M.D., Ph.D. Institute of Molecular Pharmacology, Research Centre for Natural Sciences Budapest, Hungary

Missing experimental data (Materials and Methods, pp. 2-3.)

Purity and fatty acid content of the HSA sample used are unknown. The source and chemical character (salt or free acid, racemic or not) of ibuprofen, warfarin, and digitoxin used in site probe studies are unknown. Experimental details of the site probe measurements (see on p. 7) are completely missing. There is not even a word to describe how the protein and LMF molecules were prepared for the docking procedures (p. 3). Docking type information are also missing (blind? targeted? rigid? flexible?). What are the parameters of the docking box used for the proteins? How much docking runs were performed? Where is the docking energy list? Where are the affinity constants calculated from the binding free energy values? How do they relate to the experimental values?

Errors, speculations, and unsupported claims

Page 1: "Based on these characteristics, with increasing amounts of PPIX in the blood, the probability of HSA-PPIX formation increases accordingly." It is a mere speculation which is not supported by research data. Hemopexin is the primary heme binding protein in the serum. "Methemalbumin is an abnormal component of plasma and has been found only in diseases associated with massive hemolysis, when hemopexin-binding capacity is exceeded." (Tolosano et al., Antioxid. Redox. Signal. 2010 (12) 305-320).

Page 4: "When HSA is excited at 285 nm, it radiates strong intrinsic fluorescence at 334 nm." No emission band at 334 nm can be seen either in Fig. 1A or 1B. Fluorescence contribution of the bound PPIX is totally ignored. Origin and nature of the large emission peak centered around 430 nm remain obscure.

Page 4: "To investigate the effect of TF on the binding of LMF to the (HSA-PPIX)-TF complex, steady-state fluorescence spectra of (HSA-PPIX), shown in Fig. 1A, and of (HSA-PPIX)-TF, displayed in Fig. 1B, were run and showed that the fluorescence intensity of the two systems became consistently weakened when increasing the LMF concentration." Just an opposite situation is shown in Fig. 1A and 1B: increase of the LMF concentration enhances the fluorescence intensity in both cases.

In contrast to HSA, transferrin contains 8 Trp residues. It remains a mystery, how the fluorescence contribution of these residues was considered during evaluation of the emission spectra of (HSA-PPIX)-TF.

Page 5: "Fig. 3 shows second-derivative fluorescence spectra of the (HSA-PPIX)-LMF and [(HSA-PPIX)-TF]-LMF complexes (inset). As can be seen, a negative band was observed at 310 nm and 316 nm in the (HSA-PPIX)-LMF and [(HSA-PPIX)-TF]-LMF spectra, respectively. These changes in the protein spectra should be assigned to the combination of environments of the aromatic residues in the protein and indicate that the Trp residue was in a relatively hydrophobic microenvironment." What Trp residue? There are 9 Trp in the [(HSA-PPIX)-TF]-LMF complex.

Page 7: "The corresponding binding constants were evaluated and are listed in Table 3." These are not binding but quenching constants (cf. Table 1). No association constants (Ka) are shown in Table 3. HSA affinity constants as well as the amounts of the site probes added are unknown.

Page 7: "Based on these results, it was concluded that ibuprofen had little effect on the binding of LMF to (HSA-PPIX)-TF while digitoxin removed LMF from the bindings site." Changing of the quenching constants shown in Table 3 contradicts to this claim. The Ksv value of [(HSA-PPIX) LMF] TF was reduced by an order of magnitude upon addition of both ibuprofen and digitoxin. The exact HSA binding location of digitoxin is unknown so it can not be used as a site probe. It may occur that these site probes bind to transferrin as well. Where are the controll experiments?

Page 9: "In summary, this study provides useful information for the pharmacological and pharmacodynamic behavior of LMF and should be taken into account when calculating its dosage." This work deals with one aspect of the pharmakokinetic behavior of LMF (serum protein binding). No data are presented on the pharmacodynamic behavior of the drug.

Confusing binding site designations

Page 7: "In our previous work [18], we showed that PPIX occupied site I in the HSA and LMF bound to site IIA." As it is defined on page 1, site I corresponds to subdomain IIA ("Aromatic and heterocyclic ligands have been found to bind with two hydrophobic pockets in subdomains IIA and IIIA, namely site I and site II"). Therefore, according to the above sentence, both PPIX and LMF is bound in subdomain IIA.

Page 7: "... we believe that the lower-affinity site of LMF, corresponding to the binding site, was switched in the presence of TF, causing LMF to bind to site III in HSA." No definition of site III is given.

Page 9: "For the (HSA-PPIX)-TF complex in solution, the binding site of LMF to HSA-PPIX changed from Sudlow’s site IIA to Sudlow’s site IIIB with low affinity." No such site designations exist in the literature. Moreover, subdomain IIIB binding of LMF is very doubtful since it is not supported by experimental data.

Wrong references

Page 1: "The direct interaction of drugs with TF has been widely studied [9-12]." Ref. 11 is a drug-serum albumin and not a drug-TF binding study.

Page 1: "LMF is very active against both Gram (+) and Gram (2) bacteria through inhibition of their DNA gyrase and is widely used for the clinical treatment of severe systematic infections, such as soft tissue infection, typhoid fever, bone and joint infections, prostatitis blood poisoning and sinusitis [13]." Ref. 13 is a spectroscopic study on the ion-association complex formation between fluoroquinolones and a synthetic dye (erythrosine). Thus, it does not cover the biological activities of lomefloxacin.

Page 7: "In our previous work [18], we showed that PPIX occupied site I in the HSA and LMF bound to site IIA." Ref. 18 is not the authors' paper.

Page 9: "The literature has numerous reports of LMF being located within an alpha-helix and a beta-sheet in the N-lobe of transferrin [46]." Ref. 46 is not about the LMF-transferrin interaction.