Review Reports - Identification of Conserved B and T Cell Epitopes in Glycoprotein S of Mexican Porcine Epidemic Diarrhea Virus (PEDV) Strains via Immunoinformatics Analysis, Molecular Docking, and Immunofluorescence

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This manuscript addresses an important topic and combines immunoinformatics analysis with structural modeling and experimental validation. Although this study has potential relevance, several methodological and interpretative issues must be addressed before the manuscript can be considered for publication.

Major Comments

Immunofluorescence assays demonstrated that sera from PEDV-infected pigs recognized the recombinant construct expressed in PK-15 cells; however, this experiment only confirmed antibody binding and did not establish the functional relevance of the predicted epitopes. As the recombinant construct contains multiple epitopes simultaneously, it is not possible to determine which specific epitope is responsible for the observed antibody recognition. In addition, no neutralization assay was performed to evaluate whether these epitopes induce protective or neutralizing antibodies, and no quantitative validation, such as ELISA using individual synthetic peptides, was performed. Furthermore, specificity controls, including pre-immune sera and peptide competition assays, were not included. The authors should either provide additional functional validation or substantially temper their conclusions regarding the applicability of the vaccine.
Helper T cell epitope prediction was performed using human HLA DRB1 alleles because of the absence of SLA II options in the NetMHCIIpan server. This approach represents a significant methodological limitation because peptide-binding motifs differ between species, and cross-species extrapolation may not reliably reflect swine antigen presentation. The manuscript currently provides limited justification for this substitution. The authors should provide a clear scientific rationale supported by appropriate references, explicitly discuss the limitations of this approach in the Discussion section, and critically evaluate how this substitution may affect the reliability and biological relevance of predicted HTL epitopes.
Molecular docking analysis predicted the binding interactions and free energy values between epitopes and SLA molecules. However, the docking results alone do not demonstrate immunogenicity or T cell activation, and strong predicted binding does not necessarily correlate with biological immune responses. Some statements in the manuscript appear to imply that favorable docking energies support vaccine candidacy, which overextends the conclusions that can be drawn from in silico binding models. The authors should revise the text to avoid implying functional T cell activation based solely on docking results and acknowledge that antigen processing, presentation efficiency, and T cell receptor recognition were not evaluated in this study.
Several previous studies have reported PEDV S glycoprotein epitope predictions using immunoinformatics approaches. The manuscript does not clearly articulate how the present study advances beyond the existing literature, other than focusing on Mexican strains. The authors should provide a direct comparison between the predicted epitopes identified in this study and those previously reported, clarify whether any of the identified epitopes are novel, and compare their findings with epitopes present in currently used vaccine strains. Strengthening this comparative context would improve the manuscript’s scientific contribution.
The statistical analysis of the fluorescence intensity requires additional clarification. The manuscript does not clearly state the number of independent biological or technical replicates performed, nor does it specify the statistical test applied or whether the assumptions regarding data distribution were assessed. For transparency and reproducibility, the authors should clearly report the number of experiments performed, statistical methods used, and criteria for significance.
The relevance of the predicted T cell epitopes depends on the distribution of SLA alleles in the target swine population. The manuscript does not provide a population coverage analysis or a discussion of SLA allele frequencies in Mexican pig populations. The authors should either incorporate a population coverage analysis if feasible or explicitly discuss this limitation and its implications for vaccine development.

Minor Comments

The manuscript should provide additional methodological details regarding the phylogenetic analysis, including the evolutionary model used, bootstrap values, and any alignment trimming procedures applied to ensure reproducibility.
The rationale for selecting the threshold values for BepiPred, VaxiJen, and other prediction tools should be justified with appropriate references or validation studies.
Although structural validation is shown in the supplementary figures, a concise summary of AlphaFold confidence metrics and Ramachandran statistics should be included in the main text to improve clarity.
Several minor grammatical inconsistencies and formatting issues are present in the manuscript. Careful language editing is recommended for this manuscript.

Comments on the Quality of English Language

The manuscript is generally understandable; however, several sentences would benefit from linguistic refinement to improve clarity and precision. Minor grammatical inconsistencies, repetitive phrasing, and occasional ambiguity in methodological descriptions are present. Careful language editing by a professional editing service is recommended to enhance readability.

Author Response

Reviewer 1:

Major Comments

Immunofluorescence assays demonstrated that sera from PEDV-infected pigs recognized the recombinant construct expressed in PK-15 cells; however, this experiment only confirmed antibody binding and did not establish the functional relevance of the predicted epitopes. As the recombinant construct contains multiple epitopes simultaneously, it is not possible to determine which specific epitope is responsible for the observed antibody recognition. In addition, no neutralization assay was performed to evaluate whether these epitopes induce protective or neutralizing antibodies, and no quantitative validation, such as ELISA using individual synthetic peptides, was performed. Furthermore, specificity controls, including pre-immune sera and peptide competition assays, were not included. The authors should either provide additional functional validation or substantially temper their conclusions regarding the applicability of the vaccine.

Response

We thank the reviewer for these valuable comments and suggestions, which are greatly appreciated. We have modified this section of the manuscript for clarity. The use of a construct containing all five epitopes was an initial approach that allowed us to determine whether these B cell epitopes would be recognized by sera from pigs infected with PEDV. This assay indicated that cells transfected with the recombinant construct containing the B cell epitopes are recognized by PEDV-positive sera, and that these sera do not recognize non-transfected cells. This result indicates the presence of antibodies against the B cell epitopes in the sera from PEDV-infected pigs. The neutralization assays will be included in a subsequent manuscript. We decided not to include the full immunological characterization in this manuscript to avoid making it too complex. It is also planned to carry out seroneutralization trials later using sera from immunized animals, Vero E6 cells and a Mexican PEDV strain.

Helper T cell epitope prediction was performed using human HLA DRB1 alleles because of the absence of SLA II options in the NetMHCIIpan server. This approach represents a significant methodological limitation because peptide-binding motifs differ between species, and cross-species extrapolation may not reliably reflect swine antigen presentation. The manuscript currently provides limited justification for this substitution. The authors should provide a clear scientific rationale supported by appropriate references, explicitly discuss the limitations of this approach in the Discussion section, and critically evaluate how this substitution may affect the reliability and biological relevance of predicted HTL epitopes.

Response

Your point is very important and appreciated. Helper T cell epitope prediction was performed using the NetMHCIIpan server, which is designed to predict Helper T cell epitopes in humans and some other mammals. Unfortunately, NetMHCIIpan server does not offer currently the option to predict porcine CD4 T cell epitopes. Nevertheless this server has been previously used in other studies to predict T cell epitopes in different species. Farrell et al (2016) predicted MHC binders in cattle with this server. Peptides found with this server induced IFN-γ in T cells from M. bovis-infected cattle.

Additionally, the NetMHCpan server has been used for predicting CTL epitopes in swines, and the predicted epitopes stimulated CD8 and CD4 T cell responses, as previously reported (Zhai 2024).

One of the prerequisites for activating CD4 T lymphocytes is that the peptides may bind to SLAII. To evaluate if the helper T cell epitopes predicted by the NetMHCIIpan server could bind to SLAII, we performed molecular docking studies of SLAII (modeled and energy-minimized) and the predicted epitopes. These studies revealed that MHCII can dock with low energy to several of the predicted epitopes, which could indicate that some of these epitopes bind to various pig SLAII alleles. Lymphoproliferation studies and ELISPOT will be necessary to verify that these epitopes may activate CD4 T lymphocytes, inducing proliferation, and cytokine production. Such studies are currently under development and will be part of a followed-up manuscript.

Thus, the use of servers like NetMHCIIpan using HLA is an approximation approach to identify MHC binders in other species. We have modified the discussion to include this information.

References

Farrell, D., Jones, G., Pirson, C., Malone, K., Rue-Albrecht, K., Chubb, A. J., Vordermeier, M., & Gordon, S. V. (2016). Integrated computational prediction and experimental validation identifies promiscuous T cell epitopes in the proteome of Mycobacterium bovis. Microbial genomics, 2(8), e000071. https://doi.org/10.1099/mgen.0.000071

Zhai, W., Huang, Y., He, Y., Chu, Y., Tao, C., Pang, Z., Wang, Z., Zhu, H., & Jia, H. (2024). Immunogenicity Analysis and Identification of Potential T-Cell Epitopes in C129R Protein of African Swine Fever Virus. Microorganisms, 12(6), 1056. https://doi.org/10.3390/microorganisms12061056

Molecular docking analysis predicted the binding interactions and free energy values between epitopes and SLA molecules. However, the docking results alone do not demonstrate immunogenicity or T cell activation, and strong predicted binding does not necessarily correlate with biological immune responses. Some statements in the manuscript appear to imply that favorable docking energies support vaccine candidacy, which overextends the conclusions that can be drawn from in silico binding models. The authors should revise the text to avoid implying functional T cell activation based solely on docking results and acknowledge that antigen processing, presentation efficiency, and T cell receptor recognition were not evaluated in this study.

Response

We appreciate your comment. The text has been reviewed and corrected to acknowledge the limitations of the study.

Several previous studies have reported PEDV S glycoprotein epitope predictions using immunoinformatics approaches. The manuscript does not clearly articulate how the present study advances beyond the existing literature, other than focusing on Mexican strains. The authors should provide a direct comparison between the predicted epitopes identified in this study and those previously reported, clarify whether any of the identified epitopes are novel, and compare their findings with epitopes present in currently used vaccine strains. Strengthening this comparative context would improve the manuscript’s scientific contribution.

Response

The main PEDV strain used in vaccines is the one isolated in Colorado USA. In this work we conducted an amino acid alignment analysis with the S glycoprotein from the Colorado PEDV strain, Mexican strains identified in our study and PEDV strains from other countries. This analysis indicated that epitopes identified in other studies such as COE, 2C10 and S1D are not totally conserved among all these strains. The B cell epitopes identified in this work are highly conserved in all the Mexican PEDV strains, as well as in strains isolated in other countries.

The statistical analysis of the fluorescence intensity requires additional clarification. The manuscript does not clearly state the number of independent biological or technical replicates performed, nor does it specify the statistical test applied or whether the assumptions regarding data distribution were assessed. For transparency and reproducibility, the authors should clearly report the number of experiments performed, statistical methods used, and criteria for significance.

Response

Following your pertinent suggestion we have included a new section in material and methods describing in detail how the statistical analysis and fluorescence quantification was conducted.

The relevance of the predicted T cell epitopes depends on the distribution of SLA alleles in the target swine population. The manuscript does not provide a population coverage analysis or a discussion of SLA allele frequencies in Mexican pig populations. The authors should either incorporate a population coverage analysis if feasible or explicitly discuss this limitation and its implications for vaccine development.

Response

This is a very important point to consider in the search for T cell epitopes. In Mexico, the most common pig breeds are Duroc, Landrace, Pietrain, Yorkshire, Hampshire, and Chester White (Government of Mexico). These breeds contain the SLA-1*04:01:01, SLA-2*04:01, and SLA-3*04:01 alleles (Lou et al. 2024), of which only SLA-1*04:01:01 was crystallized, and the SLA-2*04:01 and SLA-3*04:01 were sequenced. These structures were used in our docking analysis to determine if the predicted epitopes could bind to the SLA-I binding groove.

In contrast, the crystallographic structure and genetic information of the SLA-II alleles is very limited, as crystallized structures are not available, the NetMHCIIpan server does not allow direct analysis of SLA-II, and there are only a few annotated genetic sequences. In one study, the HLA alleles DRB1*1101, DRB1*0401, and DRB1*1301 were used to search for helper T cell epitopes in cattle, yielding encouraging results (Farrel 2016). The predicted helper T cell epitopes found in our work using the NetMHCIIpan server were analyzed by molecular docking with the available SLA-II sequences (SLA-DRB1:0401, SLA-DRB1:1101, and SLA-DRB1:1301), which are representative of the swine breeds present in Mexico. We do not have a population coverage analysis, and to our knowledge, no work with this information has been published in Mexico. Nonetheless, it is important to highlight that in Mexico, as in the rest of the world, the most common pig breeds used in the swine industry are basically the same as in other places such as Asia, the USA, and Europe. As with other animals, SLA-I and SLA-II alleles are polymorphic, so the response to infections/vaccines is expected to be heterogeneous.

References

https://www.gob.mx/agricultura/es/articulos/cerdos-miniatura-de-la-granja-a-la-casa#:~:text=En%20M%C3%A9xico%2C%20las%20razas%20que,de%20gran%20volumen%2C%20est%C3%A1n%20alimentados

Luo, T., Xin, C., Liu, H., Li, C., Chen, H., Xia, C., & Gao, C. (2024). Potential SLA Hp-4.0 haplotype-restricted CTL epitopes identified from the membrane protein of PRRSV induce cell immune responses. Frontiers in microbiology, 15, 1404558. https://doi.org/10.3389/fmicb.2024.1404558

Minor Comments

The manuscript should provide additional methodological details regarding the phylogenetic analysis, including the evolutionary model used, bootstrap values, and any alignment trimming procedures applied to ensure reproducibility.

Response

The corrected manuscript has been modified according to this pertinent suggestion. The bootstrap values, evolutionary model, and trimming procedures are found in the Material and Methods section.

The rationale for selecting the threshold values for BepiPred, VaxiJen, and other prediction tools should be justified with appropriate references or validation studies.

Response

The revised manuscript has been modified according to this relevant suggestion. The thresholds used in this work were based on the server's recommendations and published data. For BepiPred, a threshold of 0.5 is strongly recommended, as this value offers the best specificity and sensitivity for each prediction.

For VaxiJen, the standard threshold used for viral antigens is 0.4. Because the Spike glycoprotein is sufficiently large, we decided to use a higher threshold (0.5), which improved the probability of prediction.

For AllerTOP, there is no threshold; this server only indicates whether the peptide is likely to be allergenic or not.

For ToxinPred, the threshold was 0.00. Finally, when epitopes were likely to be allergenic/toxic, further analysis was conducted using BLASTp according to Altschul et al. (1997), using an E value of 0.05 as the threshold.

All these threshold values were included in the revised manuscript.

References

https://services.healthtech.dtu.dk/cgi-bin/webface2.cgi?jobid=69A220FE002EC06250B449DE&wait=20

C. Jespersen, B. Peters, M. Nielsen, and P. Marcatili, “BepiPred-2.0: improving se-quence-based B-cell epitope prediction using conformational epitopes,” Nucleic Acids Res., vol. 45, no. W1, pp. W24–W29, Jul. 2017, doi: 10.1093/nar/gkx346.

https://www.ddg-pharmfac.net/vaxijen/VaxiJen/VaxiJen.html

Irini A Doytchinova and Darren R Flower. VaxiJen: a server for prediction of protective antigens, tumour antigens and subunit vaccines. BMC Bioinformatics. 2007 8:4.

Irini A Doytchinova and Darren R Flower. Identifying candidate subunit vaccines using an alignment-independent method

based on principal amino acid properties. Vaccine. 2007 25:856-866.

Irini A Doytchinova and Darren R Flower. Bioinformatic Approach for Identifying Parasite and Fungal Candidate Subunit Vaccines. Open Vaccines Journal, 2008 1:22-26.

https://www.ddg-pharmfac.net/allertop_test/

Dimitrov, I. Bangov, D. R. Flower, and I. Doytchinova, “AllerTOP v.2—a server for in silico prediction of allergens,” J. Mol. Model., vol. 20, no. 6, p. 2278, Jun. 2014, doi: 10.1007/s00894-014-2278-5.
Altschul, “Gapped BLAST and PSI-BLAST: a new generation of protein database search programs,” Nucleic Acids Res., vol. 25, no. 17, pp. 3389–3402, Sep. 1997, doi: 10.1093/nar/25.17.3389.
Gupta et al., “In Silico Approach for Predicting Toxicity of Peptides and Proteins,” PLoS ONE, vol. 8, no. 9, p. e73957, Sep. 2013, doi: 10.1371/journal.pone.0073957.
Altschul, “Gapped BLAST and PSI-BLAST: a new generation of protein database search programs,” Nucleic Acids Res., vol. 25, no. 17, pp. 3389–3402, Sep. 1997, doi: 10.1093/nar/25.17.3389.

Although structural validation is shown in the supplementary figures, a concise summary of AlphaFold confidence metrics and Ramachandran statistics should be included in the main text to improve clarity.

Response

The revised manuscript has been modified according to this pertinent suggestion. The request metrics for structural validation and additional information can be found in the respective figure captions.

Several minor grammatical inconsistencies and formatting issues are present in the manuscript. Careful language editing is recommended for this manuscript.

Response

We have conducted a detailed revision of grammar and syntax to correct errors in the manuscript.

Comments on the Quality of English Language

Response

We have carried out the pertinent corrections in the revised manuscript.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This study employs a comprehensive immunoinformatics pipeline to identify conserved B-cell and T-cell epitopes within the Spike (S) glycoprotein of Mexican PEDV strains. The authors classify circulating Mexican strains into G1b and G2b genogroups, identifying conserved peptides. They then apply a rigorous series of filters (antigenicity, allergenicity, toxicity, autoimmunity) to shortlist candidates. Key findings include the validation of five B-cell epitopes on the surface of S protein models and their experimental recognition by sera from infected pigs when expressed in PK-15 cells. Furthermore, they identify eight HTL and nine CTL epitopes that demonstrate promiscuous binding and favorable free energy in molecular docking simulations with various Swine Leukocyte Antigen (SLA) molecules. The work is a foundational step toward a multi-epitope vaccine tailored to circulating Mexican strains.

Some Major Weaknesses and Concerns

Lack of Functional Neutralization Data: The title and introduction strongly imply a focus on vaccine development. While the authors show the B-cell epitopes are recognized by immune sera, they do not demonstrate that antibodies against these epitopes are neutralizing. This is the single most important functional gap. An ELISA or immunofluorescence assay confirms binding, but a virus neutralization assay is required to prove protective potential. The authors acknowledge this in the discussion but its absence limits the impact of the study.

Presentation of Data: The manuscript is difficult to follow in places due to the way data is presented.

Tables: The main tables (especially Table 1 in the manuscript text) are extremely dense and list many sequences that were ultimately not selected. It would be clearer to show the initial pool in supplementary tables and only the final, validated epitopes (S02, S03, S04, S06, S07 for B-cell; Table 3 for HTL; Table 5 for CTL) in the main text.

Supplementary Figures: The supplementary figures (especially S1, S2, S3, S4, S5) are referenced but their content is not deeply integrated or discussed. For instance, Figure S4 and S5 show that previously known epitopes are not fully conserved, which is a key finding that supports the authors' rationale. This point is mentioned but could be emphasized more strongly.

Definition of "Promiscuous": The term "promiscuous" is used frequently for T-cell epitopes. For CTLs, only two (FSMSIRTEY and NRVETYIKW) appear to bind all three SLA-I alleles. The definition of promiscuity for HTLs is less clear, as many bind to two of the three alleles. A clearer definition and consistent application of the term is needed.

Over-interpretation of Docking Data: The molecular docking shows binding with low free energy, which is excellent. However, it is an in silico prediction of binding affinity. The manuscript should be careful not to overstate this as proof of immune activation. It is a necessary prerequisite but not sufficient. Factors like TCR repertoire, antigen processing, and presentation kinetics also play a role.

Specific Comments

Introduction: The introduction provides a solid background on PEDV and the S protein. It clearly states the problem of genetic drift and the need for updated vaccines.

Results:

Section 3.2: Table 1 is overwhelming. This data is better suited for a supplementary table.

Section 3.3: The handling of potentially allergenic epitopes is good science. Explaining that they were kept after a BLASTp analysis showed no significant similarity to known allergens demonstrates a nuanced and careful approach.

Section 3.5: Figure 3 is compelling. The immunofluorescence clearly shows surface expression and recognition. Quantification of fluorescence intensity adds robustness.

Section 3.6 & 3.7: The HTL analysis is robust. Figure 4 nicely visualizes the docking into specific pockets. Table 4's binding energy data supports the predictions.

Section 3.8: The CTL analysis mirrors the quality of the HTL section. Figure 5 is informative. Table 6 is a strong summary of the final candidates.

Discussion:

The discussion effectively contextualizes the findings within the existing literature (COE, S1D, 2C10, SS2, SS6). The point that these known epitopes are not fully conserved in Mexican strains (supported by Figure S4 and S5) is a powerful argument for the work presented.

The authors correctly identify the next logical steps: constructing a multi-epitope vaccine and testing for neutralization and protection in a challenge model.

Conclusion: The conclusion is appropriate and summarizes the main achievements of the study without overstatement.

Figure Legends:

The figure legends are generally good but could be more descriptive. For example, in Figure 3 legend, it says "D) Quantification of the anti-PEDV fluorescence intensity..." but then has a second "D)" in the legend text. This should be corrected to "E) Fluorescence intensity profile..."

Mandatory Revisions:

Clarify the title and conclusions to accurately reflect that the B-cell epitopes are antigenic (bind antibodies) but not yet proven to be neutralizing. This is a critical distinction.
Restructure the data presentation by moving large, non-essential tables (like the initial 22-peptide list) to the supplementary materials, allowing the main text to focus on the final, validated candidates.
Refine the discussion to avoid over-interpreting the docking data as definitive proof of in vivo T-cell activation.
Correct the grammatical and stylistic errors noted in the language review to polish the manuscript.

Once these revisions are made, the paper will be an excellent contribution to the field of PEDV vaccinology.

Comments on the Quality of English Language

Overall, the English language is good and understandable. The scientific meaning is clear throughout. However, there are several minor grammatical errors, awkward phrasings, and inconsistencies that should be corrected to improve the manuscript's professional quality.

General Issues:

Subject-Verb Agreement & Pluralization: Minor errors occur, especially with terms like "data" (which is plural).

Article Usage (a/an/the): There are instances of missing or incorrect articles.

Prepositions: Occasionally, the wrong preposition is used (e.g., "on the surface" vs. "in the surface").

Word Choice: A few words are slightly off in context.

Specific Examples and Suggestions:

Abstract:

Original: "...a mortality rate close to 100% in piglets worldwide."

Suggestion: "...mortality rates approaching 100% in piglets worldwide." (Slightly more natural phrasing)

Original: "Five conserved B cell epitopes were found which appear to be exposed on the surface of S glycoprotein..."

Suggestion: "Five conserved B cell epitopes were found, all of which appear to be exposed on the surface of the S glycoprotein..."

Original: "These epitopes could be used for the development of serological tests and still need to be evaluated for their immunogenicity..."

Suggestion: "These epitopes could be used for the development of serological tests and must still be evaluated for their immunogenicity..." ('Still need' is a bit weak; 'must' or 'remain to be' is stronger).

Introduction (Page 2):

Original: "Glycoprotein S forms trimers and is responsible for the entry of the virus into the target cell."

Suggestion: "The S glycoprotein forms trimers and is responsible for viral entry into the host cell." (More concise and standard terminology).

Original: "PEDV GI includes the subgroups GIa, GIb and R."

Suggestion: "The GI genogroup is further divided into the subgroups GIa, GIb, and a recombinant (R) group." (Clarifies what 'R' stands for, as it's not defined until later).

Materials and Methods (Page 3):

Original: "...and the z-score for validates de predicted MHC models."

Suggestion: "...and the z-score used to validate the predicted MHC models." (This appears in a figure legend but is a clear grammatical error).

Results (Page 10):

Original: "The fact that these five predicted epitopes are present on the surface of all Mexican PEDV strains is promising because they may make PEDV more susceptible to antibody neutralization."

Suggestion: "The surface exposure of these five predicted epitopes across all Mexican PEDV strains is promising, as it suggests they may be accessible for antibody binding and potential neutralization." (Slightly more precise wording).

Discussion (Page 22):

Original: "To test whether the PEDV S glycoprotein B cell epitopes were antigenic in vitro, we developed a recombinant protein..."

Suggestion: "To test whether the PEDV S glycoprotein B cell epitopes were antigenic in vitro, we developed a recombinant protein..." (Italics for Latin terms like in vitro).

Original: "It will be important in the future to elucidate the role of the CTLs and humoral immune responses to a better understanding of the protective role of these immune players."

Suggestion: "Future studies should aim to elucidate the roles of both CTL and humoral immune responses to better understand their contributions to protection against PEDV." (More direct and grammatically correct).

Author Response

Reviewer 2:

Some Major Weaknesses and Concerns

Response:

Thank you for this important suggestion. The title describes the procedure conducted in this study. “IDENTIFICATION OF CONSERVED B AND T CELL EPITOPES IN GLYCOPROTEIN S OF MEXICAN PORCINE EPIDEMIC DIARRHEA VIRUS (PEDV) STRAINS BY IMMUNOINFORMATICS ANALYSIS, MOLECULAR DOCKING AND IMMUNOFLUORESCENCE”. Does not imply the generation of a vaccine, only in silico identification of epitopes and immunostaining using sera from PEDV-infected pigs. Since the manuscript is already too long and complex, we decided not to include any immunological characterization and save it for a follow up study. Ongoing neutralization assays and characterization of the immune response will be part of a new manuscript.

Presentation of Data: The manuscript is difficult to follow in places due to the way data is presented.

Response:

Following your pertinent suggestion, we have moved Table 1 to supplementary tables to eliminate irrelevant data. In addition, we have included in the discussion the fact that many epitopes previously identified are not conserved in many PEDV strains, including those isolated in Mexico.

Response: We have amended the manuscript to include a definition of promiscuity for HTLs.

Response:

Thank you for bringing this important point. We have softened the discussion to clearly state that any in silico prediction is only a prediction and not proof of the concept. A detailed immunological characterization of these epitopes is undergoing and will be part of a follow up manuscript.

Specific Comments

Introduction: The introduction provides a solid background on PEDV and the S protein. It clearly states the problem of genetic drift and the need for updated vaccines.

Results:

Section 3.2: Table 1 is overwhelming. This data is better suited for a supplementary table.

Response:

We have moved the table to the supplementary information section.

Section 3.5: Figure 3 is compelling. The immunofluorescence clearly shows surface expression and recognition. Quantification of fluorescence intensity adds robustness.

Section 3.6 & 3.7: The HTL analysis is robust. Figure 4 nicely visualizes the docking into specific pockets. Table 4's binding energy data supports the predictions.

Section 3.8: The CTL analysis mirrors the quality of the HTL section. Figure 5 is informative. Table 6 is a strong summary of the final candidates.

Discussion:

Response:

Thank you for the time devoted to reading our study and your relevant suggestions to improve the manuscript.

The authors correctly identify the next logical steps: constructing a multi-epitope vaccine and testing for neutralization and protection in a challenge model.

Conclusion: The conclusion is appropriate and summarizes the main achievements of the study without overstatement.

Figure Legends:

Response:

We have modified the manuscript figure legends to correct errors and improve clarity.

Mandatory Revisions:

Clarify the title and conclusions to accurately reflect that the B-cell epitopes are antigenic (bind antibodies) but not yet proven to be neutralizing. This is a critical distinction.
Restructure the data presentation by moving large, non-essential tables (like the initial 22-peptide list) to the supplementary materials, allowing the main text to focus on the final, validated candidates.
Refine the discussion to avoid over-interpreting the docking data as definitive proof of in vivo T-cell activation.
Correct the grammatical and stylistic errors noted in the language review to polish the manuscript.

Once these revisions are made, the paper will be an excellent contribution to the field of PEDV vaccinology.

Comments on the Quality of English Language

General Issues:

Subject-Verb Agreement & Pluralization: Minor errors occur, especially with terms like "data" (which is plural).

Article Usage (a/an/the): There are instances of missing or incorrect articles.

Prepositions: Occasionally, the wrong preposition is used (e.g., "on the surface" vs. "in the surface").

Word Choice: A few words are slightly off in context.

Specific Examples and Suggestions:

Abstract:

Original: "...a mortality rate close to 100%100% in piglets worldwide."

Suggestion: "...mortality rates approaching 100% in piglets worldwide." (Slightly more natural phrasing)

Original: "Five conserved B cell epitopes were found which appear to be exposed on the surface of S glycoprotein..."

Suggestion: "Five conserved B cell epitopes were found, all of which appear to be exposed on the surface of the S glycoprotein..."

Original: "These epitopes could be used for the development of serological tests and still need to be evaluated for their immunogenicity..."

Introduction (Page 2):

Original: "Glycoprotein S forms trimers and is responsible for the entry of the virus into the target cell."

Suggestion: "The S glycoprotein forms trimers and is responsible for viral entry into the host cell." (More concise and standard terminology).

Original: "PEDV GI includes the subgroups GIa, GIb and R."

Suggestion: "The GI genogroup is further divided into the subgroups GIa, GIb, and a recombinant (R) group." (Clarifies what 'R' stands for, as it's not defined until later).

Materials and Methods (Page 3):

Original: "...and the z-score for validates de predicted MHC models."

Suggestion: "...and the z-score used to validate the predicted MHC models." (This appears in a figure legend but is a clear grammatical error).

Results (Page 10):

Original: "The fact that these five predicted epitopes are present on the surface of all Mexican PEDV strains is promising because they may make PEDV more susceptible to antibody neutralization."

Discussion (Page 22):

Original: "To test whether the PEDV S glycoprotein B cell epitopes were antigenic in vitro, we developed a recombinant protein..."

Suggestion: "To test whether the PEDV S glycoprotein B cell epitopes were antigenic in vitro, we developed a recombinant protein..." (Italics for Latin terms like in vitro).

Original: "It will be important in the future to elucidate the role of the CTLs and humoral immune responses to a better understanding of the protective role of these immune players."

Response:

We have revised the entire manuscript for grammar and clarity. Your suggestions are very important and helped us to improve readability of the text.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have addressed most of the reviewers’ concerns, and the manuscript has improved substantially since the previous version. This study presents a systematic immunoinformatics analysis to identify conserved B cell and T cell epitopes in the spike glycoprotein of Mexican PEDV strains and combines computational prediction with experimental immunofluorescence validation. The manuscript is generally well organized, and the methodological workflow is clearly presented. However, several points still require minor clarification before the manuscript can be considered for publication.

1. Interpretation of Neutralization Potential

This study demonstrated that the recombinant construct containing the predicted B cell epitopes was recognized by sera from PEDV-infected pigs. However, recognition by immune sera does not necessarily indicate that these epitopes induce neutralizing antibodies. Although the authors clarified that neutralization assays will be performed in future work, several sentences in the manuscript still suggest potential neutralizing activity. The authors are encouraged to further moderate these statements and clearly indicate that the present study demonstrates antibody recognition rather than functional neutralization.

2. Interpretation of Molecular Docking Results

Docking analysis provides useful structural insights into the potential interactions between predicted T cell epitopes and SLA molecules. However, docking results alone cannot demonstrate T cell activation or immunogenicity. The discussion would benefit from explicitly emphasizing that these results represent predicted binding interactions and that antigen processing, peptide presentation efficiency, and T cell receptor recognition were not experimentally evaluated in this study. Addressing this limitation would help to avoid potential overinterpretation of the docking results.

3. Comparison with Previously Reported PEDV Epitopes

The manuscript mentions previously described PEDV epitopes, such as COE, S1D, and 2C10, and briefly discusses sequence conservation among strains. However, the comparison with previously reported epitopes remains largely descriptive. The scientific contribution of the study would be strengthened by providing a clearer comparison between the epitopes identified in this study and those reported in previous literature. For example, the authors could briefly indicate whether the predicted epitopes overlap with or differ from previously reported antigenic regions.

Comments on the Quality of English Language

Although the overall English language has improved, a few sentences remain somewhat repetitive or overly long. Careful language editing would further improve clarity and readability.

Author Response

Reviewer 1

Interpretation of Neutralization Potential

RESPONSE

Thank you for your feedback. In the revised version of our manuscript we have moderated the content of these paragraphs and clarified that some of the epitopes are only found within some neutralizing epitopes, without suggesting neutralizing activities. Furthermore, in the discussion section we have stretched the importance of validating these epitopes experimentally.

Interpretation of Molecular Docking Results

RESPONSE

In the revised version of our manuscript, we have emphasized that the binding of a peptide to MHC is not sufficient to activate a cellular response, and that further experimental assays will be needed to determine if such epitopes are efficient to induce humoral and cellular responses.

Comparison with Previously Reported PEDV Epitopes

RESPONSE

We have described the location of the epitopes identified in our study wwithin previoously reported peptides. Of the seven epitopes found in this work (S01–S07), only five were found on the surface of the PEDV S glycoprotein (S02–S04, S06, and S07). Of these five epitopes, S02 was found within a larger epitope reported in IEDB, as well as within the S1D epitope. The S03 epitope was found partially within the S1D epitope, as well as within three IEDB epitopes. The S04 and S06 epitopes were found in the S2 region with no matches with other IEDB, COE, S1D, or 2C10 peptides. Finally, the S07 epitope contains a 5-amino-acid epitope shared with the SARS-CoV-2 S glycoprotein. We have discuss all this information in the revised version of our manuscript.

Comments on the Quality of English Language

Although the overall English language has improved, a few sentences remain somewhat repetitive or overly long. Careful language editing would further improve clarity and readability.

RESPONSE

We have removed redundant paragraphs from this version of the manuscript and significantly shorten the discussion section.