Patterns of Diversity and Humoral Immunogenicity for HIV-1 Antisense Protein (ASP)

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The research of the immunogenicity of HIV-1 antisense protein is of great interest since it clearly defines the presence of this protein during the HIV infection, shows its participation in immune system responses, which depend on this protein diversity in HIV clades. Hence this manuscript is of high interest for specialists in HIV infection and especially immune responses research. According to this, the manuscript corresponds to the journal Vaccines aims and scope.   However, there are some points that decrease the significance of content, quality of presentation and scientific soundness of the manuscript in its present form and hopefully can be improved.

1. Peptide aa sequences should be presented in the Supplementary material, with defining the HIV-1 clade, to which they correspond, and conserved or variable aa residues. It will improve understanding the results of peptide antigenicities. 

2.  What about putative transmembrane region(s) of ASP? Do it(they) demonstrate any antigenicity that may conflict with their transmembrane state, or not?

3. Patients have not been characterized by HIV-1 clade they are infected with. It may result in decreasing humoral immune responses to certain peptides that have sequences specific for a certain clade because of a low ratio of sera of patients infected by a corresponding HIV clade in the whole serum pool. Why sera of patients infected with a certain HIV-1 pool was not tested with the corresponding peptides separately? This approach could reveal some clade-specific antigenic determinants. 

4. The statement of the distinct regions targeted by IgG and IgM looks not obvious, since it has not been confirmed experimentally. IgG- and IgM-specific B-epitopes can differ in only one or two, but important for the specificity, amino acid residues because of antibody specificity maturation with a further class switch. And the selected set of peptide fragments may not contain a corresponding highly specific peptide. So, the difference in IgG and IgM specificities should be discussed more cautiously. The same is true for an overall discussion of the results of antigenicity studies, since no differential tests for clade-specific peptides with sera of patients infected with the corresponding HIV-1 clades have been performed, while serum pooling can decrease revealing certain clade-specific antibodies.

5. Why immunoreactivity of peptide set 9 is compared only to sets 8 and 10 (lines 294-295)? 

6. The term "reactivity index" is not explained. What was the difference between negative results (control) and minimal positive results? How many immunotest repeats were performed for each serum pool?

7. The first paragraph of the Discussion section repeats the information from the Introduction; it should be excluded or changed. 

8. There are some text errors and unclear phrases. 
a) Lines42-43: "genome genomic elements" -either genome or genomic. 
b) Lines 57-60 - correct the phrase, it is unclear.
c) Lines 84 and 273 - please include references.
d) Line 89: "followed by realignment of those" - it is unclear what are "those".
e) "Fmoc", not "F-moc".
f) Line 132: Abbreviation HICs is not deciphered. 
g) All figures are doubled - typing error?
h) Line 304: "For set 11, the lower response observed for set 10..." - an unclear beginning of the phrase. Amino acid residue names do not require articles.

  

Comments on the Quality of English Language

Amino acid residue names do not require articles. 
See also comments 8,b and 8,d.

Author Response

We would like to thank the reviewer for the evaluation of our manuscript. Below we present our point by point responses to the raised questions. 

The research of the immunogenicity of HIV-1 antisense protein is of great interest since it clearly defines the presence of this protein during the HIV infection, shows its participation in immune system responses, which depend on this protein diversity in HIV clades. Hence this manuscript is of high interest for specialists in HIV infection and especially immune responses research. According to this, the manuscript corresponds to the journal Vaccines aims and scope.   However, there are some points that decrease the significance of content, quality of presentation and scientific soundness of the manuscript in its present form and hopefully can be improved.

Peptide aa sequences should be presented in the Supplementary material, with defining the HIV-1 clade, to which they correspond, and conserved or variable aa residues. It will improve understanding the results of peptide antigenicities.

 Thank you for the observation. We completely agree with the importance of presenting a supplementary table containing peptide list and it was further added to the revised version of the manuscript (Supplementary Table 3). The referred table includes the antigenicity data for each designed peptide, highlights the variable positions considered for each set and describes the residues prevalence in B, C and F1 sequences per peptide.

Unfortunately, due to the complexity of the data, a clear association between most designed peptides and a specific subtype was not possible, since they included complex combinations of subtype specific amino acid signatures and the most prevalent amino acid in specific variable residue. In addition, most of the variations incorporated were shared by at least two subtypes, which also difficult the subtype labelling. Thus, supplementary table 3 highlights and facilitate the identification of the subtype specific amino acid signatures, but a subtype labelling was not added to the table.

What about putative transmembrane region(s) of ASP? Do it(they) demonstrate any antigenicity that may conflict with their transmembrane state, or not?

Thank you for the question. Previous characterization studies based on HIV HXB2 reference suggested that both ASP’s N-terminal and C-terminal domains are intracellular, while the central region of the protein contains an ectodomain delimited by two transmembrane domains. In general, our data agrees with this presumed structure of ASP, as the strongest reactivities were observed for peptides covering the ectodomain region and the lowest responses were observed in the regions covering the N-terminal region, C-terminal region and transmembrane domains. Sets 7-8 and 14-15, which cover the transmembrane domains, a relevant decrease in the mean response can be observed in comparison to their adjacent sets, adding evidence to previous data. We included some points discussing these topics in the revised manuscript, more specifically at lines 397-399 and 436-442.

Patients have not been characterized by HIV-1 clade they are infected with. It may result in decreasing humoral immune responses to certain peptides that have sequences specific for a certain clade because of a low ratio of sera of patients infected by a corresponding HIV clade in the whole serum pool. Why sera of patients infected with a certain HIV-1 pool was not tested with the corresponding peptides separately? This approach could reveal some clade-specific antigenic determinants. 

We understand the reviewer concern and agree with. However, at the present work our main objective was to characterize the humoral immunologic response against ASP aiming to identify general patterns of IgM and IgG reactivity throughout the whole extension of the protein. To achieve this objective, we had to employ Spot synthesis followed by membrane immunoblot as a more generality methodology that provided a great throughput for broad screening and whole protein mapping, in exchange of low resolution to solve problems like the one raised. The proposed analysis that could reveal some clade-specific antigenic determinants is part of the following studies currently in progress by our group. In this context, data from the present manuscript is being used to identify the most relevant peptides that could be synthesized to assess antibody reactivity individually in patients with different clinical settings. We believe that this approach would help to identify not only clade-specific antigenic determinants, but if clinical markers, like time since diagnosis and phase of infection, could be linked to different patterns of ASP immunogenicity.

The statement of the distinct regions targeted by IgG and IgM looks not obvious, since it has not been confirmed experimentally. IgG- and IgM-specific B-epitopes can differ in only one or two, but important for the specificity, amino acid residues because of antibody specificity maturation with a further class switch. And the selected set of peptide fragments may not contain a corresponding highly specific peptide. So, the difference in IgG and IgM specificities should be discussed more cautiously. The same is true for an overall discussion of the results of antigenicity studies, since no differential tests for clade-specific peptides with sera of patients infected with the corresponding HIV-1 clades have been performed, while serum pooling can decrease revealing certain clade-specific antibodies.

We understand the reviewer concerns and agrees that, as for clade-specific associations with peptide antigenicity, IgM and IgG responses could differ for other reasons that are not clearly associated to the peptide sequence. In our study, the use of distinct pools to assess IgM and IgG could also had impacted in this compared analysis. Those questions were now addresed in the last section of the discussions (Lines 449-478), together with the other limitations previously discussed.

Why immunoreactivity of peptide set 9 is compared only to sets 8 and 10 (lines 294-295)? 

The comparation in the referred passage was highlighted to indicate that set 9 was probably an immunogenic core since its reactivity is higher than that verified for sets 8 and 10, which presented an overlap with set 9. A similar comparison is made in the following paragraph for sets 10 and 11.

The term "reactivity index" is not explained. What was the difference between negative results (control) and minimal positive results? How many immunotest repeats were performed for each serum pool?

Thank you for the question. The reactivity index is used as a normalizing measure between the spot signal intensities, considering the strongest signal on the membrane as 100% and all the other signals as a percentage of this maximum. Replicates of the immunotest were not performed, since a completely new membrane synthesis would be needed. The reuse of a stained membrane was also not indicated, since it could introduce bias to the luminescence readings. Furthermore, quantification of spot signals uses algorithms that consider results from background and positive/negative controls to estimate spot intensity. To further clarify this, we included this description in the material and methods section (Lines 150-155), as follows: complete description of the measurement step at lines 160-167, as follows: “Stained membranes were scanned by Odyssey FC imager (LI-COR Bioscience) and signal intensities for each spot were quantified with TotalLab Software (Nonlinear Dynamics, USA), using algorithms that compared the intensity between background, spot area and negative control characterize spot signal intensity. The spot with the strongest signal for each membrane was reported as having 100% intensity, and all other spots had their intensity values expressed a relative percentage to this intensity (IR, Reactivity index).”

The first paragraph of the Discussion section repeats the information from the Introduction; it should be excluded or changed. 

 We understand that this paragraph mirrors some of the information brought at the introduction. This paragraph function as a introductory paragraph to the discussion section and could recapitulate the study background and main objective. This structure of first discussion paragraph is commonly employed in the publish manuscripts of our group and it is suggested in the guidelines provided from several publishing services, like Elsevier. In face of this, we preferred to maintain the referred paragraph unmodified, since we confirmed that it did not repeat sentences from the introduction and believe that its content did not negatively impact the discussion structure.

There are some text errors and unclear phrases. 

a) Lines42-43: "genome genomic elements" -either genome or genomic. 

We apologize for the error. The word “genome” was removed from the sentence.

b) Lines 57-60 - correct the phrase, it is unclear.

The previous sentence was reformulated as follows: “Although ASP’s function is still unknown and a general low expression level of antisense RNAs in infected cells [8–10,15] raises ongoing controversies about their relevance in HIV pathogenesis, some studies have detected ASP expressed associated with viral envelope, and to the plasmatic membrane of infected cells [13,14,16].” (Lines 57-60)

c) Lines 84 and 273 - please include references.

Thank you for the observation. We submitted the manuscript in word format, but we do not note that some inconsistencies were generated during the generation of the PDF version of the manuscript supplied for peer review. This was one of the observed inconsistencies. The proof was checked in the revised version to avoid any of these problems.

d) Line 89: "followed by realignment of those" - it is unclear what are "those".

We altered the text to specify that the realignment was performed on filtered sequence lists. The corrected text is as follows: “Based on this alignment, defective sequences with stop codons or frameshift mutations that compromised alignment in at least 70% of the amino acid sequence were removed from sorted sequence lists, followed by realignment of the filtered sequence.” (Lines 85-87)

e) "Fmoc", not "F-moc".

The term was corrected to Fmoc in the revised version of the manuscript.

f) Line 132: Abbreviation HICs is not deciphered. 

Since the HIC acronym was used only in that section, we removed it from the text and maintained all other references to HIV controllers without the acronym.

g) All figures are doubled - typing error?

Thank you for the observation. This error was also result of the PDF conversion error employed by the journal.

h) Line 304: "For set 11, the lower response observed for set 10..." - an unclear beginning of the phrase. 

Thank you for the suggestion. We reformulated the sentence in the revised version of the manuscript, as it follows: “For set 11, a possible immunogenic core related to this set was identified in the immediate left region that flanks the hypervariable region ASP_V2. Lower reactivities for peptides in set 10 in comparison to set 11 corroborates this hypothesis.” (Lines 293-295)

Reviewer 2 Report

Comments and Suggestions for Authors

Caetano et al. in this study designed and synthesized overlapping peptides spanning the HIV-1 2 Antisense Protein (ASP) using computational tools. They then mapped the immunogenicity of these peptides by testing their binding reactivity with serum samples from HIV-infected patients. The study is well designed, and the data is encouraging. However, a few points need to be addressed to enhance the paper:

 

1. Figure 8 Improvements:

- Include Negative Control: A negative control (serum from HIV-negative healthy individuals) should be included in the peptide testing assay to validate the synthesized peptides. This will help determine if the peptides are cross-reactive to a common antigen recognized by healthy individuals.

- Compare Groups: Serum from recent HIV infections (n=5) and HIV controllers (n=5) were used in the testing. Is there a difference in peptide recognition between these two groups? It is known that HIV patients at different stages of infection tend to have different immunodominance in their peptide recognition.

- Enhance Figure Clarity: Improve the figure to more directly compare both the magnitude and diversity of reactivity among the 16 peptide sets. For example, the "mean immune responses" described as percentages in lines 306-307 should be clearly reflected in the figure.

- Label HIV Subtypes: Clearly label the HIV subtype (B, C, or F1) that each peptide variation belongs to.

 

2. Proofreading:

- Ensure a thorough proofreading of the manuscript. There are several errors, such as figures 1-7 being shown twice in the paper.

- In lines 272 and 273, there is a warning message about the reference: “Error! Reference source not found.” This should be deleted.

 

By addressing these points, the clarity and impact of the paper can be significantly improved.

Author Response

We would like to thank the reviewer for the evaluation of our manuscript. Below we present our point by point responses to the raised questions. 

Caetano et al. in this study designed and synthesized overlapping peptides spanning the HIV-1 2 Antisense Protein (ASP) using computational tools. They then mapped the immunogenicity of these peptides by testing their binding reactivity with serum samples from HIV-infected patients. The study is well designed, and the data is encouraging. However, a few points need to be addressed to enhance the paper:

Figure 8 Improvements:

- Include Negative Control: A negative control (serum from HIV-negative healthy individuals) should be included in the peptide testing assay to validate the synthesized peptides. This will help determine if the peptides are cross-reactive to a common antigen recognized by healthy individuals.

We understand the review concern and apologize for the lack of a negative control sample. For the Spot synthesis technique, the testing of multiple samples individually requires the use of an independent membrane containing the synthesized peptides for each sample or pool of samples evaluated. Due to the high cost of the technique, the testing of a healthy individuals without HIV were not performed, since a completely new membrane synthesis would be needed. The reuse of a stained membrane was also not indicated, since it could introduce bias to the luminescence readings. Despite that, control of experimentation was performed through inclusion of a peptide from an immunogenic region of HIV gp120 (as a HIV-1 antigenicity positive control), a peptide from a immunogenic region of Influenza A (as a general positive control) and a peptide from a sequence of a cowpox virus (as a negative control). The results from the negative control were further added to the Figure 8, following the results from both positive controls.

Moreover, the results obtained with the present study allowed to restrict regions of interest and to select most immunogenic peptides, which will be synthesized and used to assess antibody reactivity individually in patients with different clinical settings. For those experiments, the inclusion of negative control samples obtained from healthy individuals without HIV is predicted and we believe that this approach would help to identify if clinical markers, like time since diagnosis, phase of infection and a viral control phenotype, could be linked to different patterns of ASP response.

 

- Compare Groups: Serum from recent HIV infections (n=5) and HIV controllers (n=5) were used in the testing. Is there a difference in peptide recognition between these two groups? It is known that HIV patients at different stages of infection tend to have different immunodominance in their peptide recognition.

Thank you for the question. We were not able to assess the differences between the recently infected individuals and HIV controllers, since samples from these two groups were mixed to generate the pool sample employed in the immunoblots. As previously explained, the individual testing of both these groups would require the synthesis of at least two membranes for each group (one for IgM testing and other for IgG testing). As discussed in the question above, these evaluations will be performed in other study currently being developed by our group.

- Enhance Figure Clarity: Improve the figure to more directly compare both the magnitude and diversity of reactivity among the 16 peptide sets. For example, the "mean immune responses" described as percentages in lines 306-307 should be clearly reflected in the figure.

Thank you for the suggestion. Due to the complexity of data, several drafts of figure 8 were generated to improve the figure clarity. For this revised version, we improved the highlight of the variable positions in the peptides description that accompany the bar plots.

We tried to include the mean responses per set as a new summary bar or as a text for each set in the figure 8, however it resulted in a figure with reduced clarity. To further aid figure data interpretation, a supplementary figure was added to the new version of the manuscript, which includes the reactivities for set displayed in the format of boxplots (Supplementary Figure 1).

 

- Label HIV Subtypes: Clearly label the HIV subtype (B, C, or F1) that each peptide variation belongs to.

Thank you for the suggestions. For the present study, our goal was to assess peptides that incorporated all the relevant variations concerning HIV subtypes and those with high representativity at variable position that were identified in the analyses. Although, we were able to identify some subtype-specific signatures and designed some peptides that clearly represent one or more HIV subtypes, most of the synthesized peptides included complex combinations of residues that make it difficult to clearly associate the whole peptide sequence to a specific subtype.

Despite that, we believed that the strategy employed was necessary to further assess the impact of the protein diversity in the immune response. For the most complex peptide sets, like set 13, we were not able to find fully matched sequences in our dataset, but the comparative analyses of the different peptides allowed to assess the impact of the amino acid variation in the motif recognition.

Moreover, we agree that the data could be more discriminated to further aid the interpretation of the results. In face of this, we are including in the supplementary material a table (Supplementary table 3) detailing the residues prevalence in B, C and F1 sequences for each designed peptide. We hope that it could improve data interpretation.

2. Proofreading:

- Ensure a thorough proofreading of the manuscript. There are several errors, such as figures 1-7 being shown twice in the paper.

- In lines 272 and 273, there is a warning message about the reference: “Error! Reference source not found.” This should be deleted.

 Thank you for the observation. We submitted the manuscript in word format (.docx), but noted that some inconsistencies were generated during the generation of the PDF version of the manuscript supplied for peer review. The doubling of figures and warning about lack of references in some points of the text resulted from it. The text was checked in the revised version to avoid any of these problems.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Supplementary table 3 should be mentioned in the Materials and Methods section, subsection 2.2. 

 

Comments on the Quality of English Language

Please check the correct use of singular and plural forms of the verb "to be". Also the use of an additional ending  " 's " is incorrect in some cases.    

Author Response

Thank you for the second round of review. As requested, supplementary table 3 is now cited in the suggested section. We also reviewed the text and corrected some typos, mispellings or other gramatical errors and hopes that english language is fine. 

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have addressed the questions and comments from my initial review, and there has been noticeable improvement since then. Therefore, I support the full publication of this paper.

Author Response

We are glad that the reviewer is satisfied with the modifications and improvements made to the manuscript. Thank you for the approval and recommendation for publication.