Non-Dominant Genotypes (GII, GIV and GV) of Japanese Encephalitis Virus Exhibit an Elevated Evolutionary Rate in Nature

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

In this study from Wang et al., a collection of JEV sequences (126 total, comprising all 5 genotypes), was subjected to phylogenetic analysis. As this sequence database was larger than those used in prior analyses, novel conclusions about the evolutionary appearance and the mutation rates of distinct genotypes could be made. The manuscript is clearly written, and this reviewer is always appreciative of publications that present updated virus phylogenetic analyses, as many commonly cited studies can be quite old.

I have a few minor comments:

1. Throughout the manuscript (Line 138, and in Figure 4, perhaps other places I missed), what is referred to as the M protein should be changed to prM, as that encompasses the entire sequence region.
2. For the first introduction to tMRCA, it would be useful to non-phylogeny experts to spell out the abbreviation.
3. Figure 4 legend should note that % identity values are in reference to the P3 vaccine strain (which is GIII).
4. Section 3.6 (lines 265-276) and Figure 6. While I could be mistaken, essentially all flavivirus structure papers I have read do not have as many linker regions identified between the various domain segments. The longer DI-DIII linker, and DIII-stem linker regions are commonly denoted, but the shorter 1-3 amino acid linkers colored green in figure 6 are not.
5. Figure 6 (and 7a): Domain coloring of the flavivirus E protein is almost universally shown as red, yellow, blue for DI, DII, and DIII, respectively.
6. Section 3.7 and Figure 7- This section is speculative and the importance of residues that may impact surface charge or structure is unknown without functional studies. What are the criteria for “significant physicochemical changes” or “notable hydrophobicity/ hydrophilicity alterations”? The figure panels are not particularly informative as is, shown as individual genotypes. In 7a, it would be more helpful for readers familiar with E protein structure to have the domains labelled in different colors. In 7b, it is hard to tell the difference between the shading, and it is only presented in the text as a list of residues with charge changes. Overall, this data could be removed, or perhaps a table listing residues may be more succinct. As is, the data takes up a lot of figure space without being brought up in the discussion.

Author Response

Response to reviewer 1

 

Comments and Suggestions for Authors

In this study from Wang et al., a collection of JEV sequences (126 total, comprising all 5 genotypes), was subjected to phylogenetic analysis. As this sequence database was larger than those used in prior analyses, novel conclusions about the evolutionary appearance and the mutation rates of distinct genotypes could be made. The manuscript is clearly written, and this reviewer is always appreciative of publications that present updated virus phylogenetic analyses, as many commonly cited studies can be quite old.

Response: We sincerely thank the reviewer for the positive and encouraging comments. We are glad that you found the manuscript clearly written and appreciated the inclusion of a larger, updated dataset of 126 JEV sequences covering all five genotypes. Your acknowledgment of the novel insights regarding the evolutionary appearance and mutation rates of distinct genotypes is greatly appreciated. We hope that our study provides a valuable and timely contribution to the understanding of JEV phylogenetics and evolution.

• Throughout the manuscript (Line 138, and in Figure 4, perhaps other places I missed), what is referred to as the M protein should be changed to prM, as that encompasses the entire sequence region.

Response: We express our profound appreciation for your comprehensive review and invaluable suggestions. We have conducted a meticulous examination of the entire manuscript, specifically addressing Line 138 (now located on Line 152, highlighted in yellow, and in the revised Figure 4). We have amended all references from "M protein" to "prM" to accurately reflect the entire sequence region. Your insightful feedback has been instrumental in enhancing the precision of our manuscript.

• For the first introduction to tMRCA, it would be useful to non-phylogeny experts to spell out the abbreviation.

Response: We extend our sincere gratitude for your valuable suggestion. In response, we have elucidated the initial introduction of the term tMRCA (located in Lines 22–23 of the abstract, highlighted in yellow) by providing its full form, "time to the most recent common ancestor," to enhance comprehension for readers who may not be familiar with phylogenetic terminology. Your insightful feedback has substantially enhanced the clarity of our manuscript.

• Figure 4 legend should note that % identity values are in reference to the P3 vaccine strain (which is GIII).

Response: We sincerely appreciate your insightful suggestion regarding the legend of Figure 4. In response, we have incorporated a clarifying note (Page 9, Lines 265-269, highlighted in yellow) specifying that the percentage identity values (% identity) are calculated with reference to the P3 vaccine strain (GIII). Your valuable feedback has significantly improved the accuracy of the figure.

• Section 3.6 (lines 265-276) and Figure 6. While I could be mistaken, essentially all flavivirus structure papers I have read do not have as many linker regions identified between the various domain segments. The longer DI-DIII linker, and DIII-stem linker regions are commonly denoted, but the shorter 1-3 amino acid linkers colored green in figure 6 are not.

Response: We sincerely appreciate your insightful comments on Section 3.6 and Figure 6. After careful consideration, we fully agree that short linkers consisting of 1–3 amino acids are generally not emphasized in flavivirus structural studies. Although these regions were initially annotated to provide finer visualization of domain boundaries, we acknowledge that this practice is inconsistent with conventional annotation standards in previous flavivirus structures. Accordingly, we have revised Figure 6 (on page 11) and Section 3.6 (formerly Lines 265–276, now Lines 290–291 and 312 on pages 10–11, highlighted in yellow) by removing the green-colored short linker annotations and retaining only the well-established DI–DIII and DIII–stem linker regions. Your valuable suggestion has greatly enhanced the accuracy and consistency of our structural representations, and we are deeply grateful for your guidance.

• Figure 6 (and 7a): Domain coloring of the flavivirus E protein is almost universally shown as red, yellow, blue for DI, DII, and DIII, respectively.

Response: We express our sincere gratitude for your insightful comments concerning Figures 6 and 7a. In response, we have revised the domain coloring of the flavivirus E protein in both figures to align with the widely accepted convention: red for Domain I (DI), yellow for Domain II (DII), and dark blue for Domain III (DIII). This modification enhances consistency with existing literature and improves the clarity of the figures. We deeply appreciate your valuable suggestion.

• Section 3.7 and Figure 7- This section is speculative and the importance of residues that may impact surface charge or structure is unknown without functional studies. What are the criteria for “significant physicochemical changes” or “notable hydrophobicity/ hydrophilicity alterations”? The figure panels are not particularly informative as is, shown as individual genotypes. In 7a, it would be more helpful for readers familiar with E protein structure to have the domains labelled in different colors. In 7b, it is hard to tell the difference between the shading, and it is only presented in the text as a list of residues with charge changes. Overall, this data could be removed, or perhaps a table listing residues may be more succinct. As is, the data takes up a lot of figure space without being brought up in the discussion.

Response: We sincerely appreciate your valuable comments on Section 3.7 and Figure 7. In response to your concerns regarding the speculative nature of the content and the unclear impact of residues on surface charge or structure, we have revised the relevant section (Page 12, Lines 329-344, highlighted in yellow) and explicitly defined the criteria for "significant physicochemical changes": amino acid mutations meeting any of the following conditions were considered significant: 1) charge alterations (conversion between positive/negative charges or charged/neutral states); 2) hydrophobicity/hydrophilicity changes (Kyte-Doolittle hydropathy index change ≥1); or 3) steric effects (side-chain volume change ≥50 ų). Only sites satisfying at least one criterion were included in subsequent analyses. We emphasize in the manuscript that these analyses are theoretical predictions, and their actual biological functional impacts require experimental validation. Furthermore, we have relocated the original Figure 7b to the Supplementary Materials (Figure S3) and added a corresponding table (Table S2) for Figure 7 to enhance data accessibility and interpretability.

In accordance with your suggestions, we have provided the P3 vaccine strain structural model in the Supplementary Materials (Figure S2) using the standard domain coloring scheme (DI: red, DII: yellow, DIII: blue) to facilitate precise domain-level comparison. We agree that this approach significantly improves the recognizability of structural features, thereby substantially enhancing the clarity and accessibility of data presentation. The results of Figure S2 and Table S2 are presented below. The supplementary figures/tables are provided in PDF file.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript presents an in-depth phylogenetic analysis of 126 complete JEV genomes, with an emphasis on non-dominant genotypes. The study is timely and relevant for epidemiological surveillance and vaccine development, especially given the reemergence of GIV and GV.

The work has publication potential but requires significant improvements:

Major comments

The manuscript states that GII, GIV, and GV have higher evolutionary rates and recent epidemiological relevance; however, historical data or meta-analyses are not integrated to assess whether these rates correlate with greater virulence, transmissibility, or immune escape.
MCMC parameters are reported, but there is no evidence of convergence validation (ESS values, burn-in, replicates). Furthermore, only BEAST is used, which limits robustness. Report ESS >200 for key parameters. Include a machine learning tree with IQTREE or RAxML for validation.

It is concluded that mutations in E justify low vaccine protection, but no neutralization data are presented; only in silico analyses are shown. Change the conclusions to a conditional tone and avoid making direct causal claims without experimental data.

The table shows 126 complete sequences without logical grouping. Move the entire table to Supplementary Materials. In the text, include summarized versions or a figure showing distributions by region, year, and host.
Rates are reported, but not statistically compared between genotypes (p-values, HPD overlaps, Bayes factors).

Minor comments
In Figure 1, I suggest removing the Americas and showing Asia, Oceania, and Africa in greater detail. Another option is to display a small world map and zoom in on these regions.
In Figure 2, I suggest cropping the tree scale to reduce the dendrogram size and allow the strain section to be enlarged.

Author Response

Response to reviewer 2

 

Comments and Suggestions for Authors

The manuscript presents an in-depth phylogenetic analysis of 126 complete JEV genomes, with an emphasis on non-dominant genotypes. The study is timely and relevant for epidemiological surveillance and vaccine development, especially given the reemergence of GIV and GV. The work has publication potential but requires significant improvements:

Response: We express our sincere gratitude for your insightful comments and constructive feedback. We are pleased to learn that you consider the phylogenetic analysis of 126 complete Japanese Encephalitis Virus (JEV) genomes, particularly with an emphasis on non-dominant genotypes, to be both timely and pertinent to epidemiological surveillance and vaccine development. We acknowledge your recommendation for substantial improvements and have meticulously revised the manuscript to address concerns related to the clarity of methodology, data interpretation, and presentation of results. Your input has been instrumental in enhancing the scientific rigor, clarity, and overall quality of our manuscript.

• The manuscript states that GII, GIV, and GV have higher evolutionary rates and recent epidemiological relevance; however, historical data or meta-analyses are not integrated to assess whether these rates correlate with greater virulence, transmissibility, or immune escape.

Response: We sincerely thank you for your insightful comment. We acknowledge that a comprehensive meta-analysis would be the ideal approach to rigorously correlate evolutionary rates with phenotypes like virulence and transmissibility. We plan to undertake such a meta-analysis in our future research endeavors and appreciate the reviewer's suggestion in highlighting this significant research direction. In response to your concerns regarding the elevated evolutionary rates and epidemiological significance of genotypes GII, GIV, and GV, we have enhanced the discussion by incorporating historical epidemiological data and pertinent experimental research literature. This addition aims to evaluate whether these increased rates are linked to viral virulence, transmissibility, or immune evasion. The revised discussion is highlighted in yellow in the manuscript (Page 14, Lines 413-424, highlighted in yellow). The modifications are as follows: "Moreover, it is important to highlight that the rapidly evolving GIV genotype, with an estimated evolutionary rate of 1.7×10^-3 substitutions per site per year (HPD, 4.5×10^-4 to 3.9×10^-3), was implicated in a significant JE outbreak in southeastern Australia in 2022, resulting in 45 human cases and 7 fatalities. Additionally, the GV genotype, which re-emerged in mosquito samples from Tibet, China (78°25°E - 99°06°E, 26°50°N - 36°53°N), in 2009, has, within a span of just 16 years, disseminated from its initial location to South Korea (125.1°E - 129.3°E, 34.3°N - 38.6°N), where it has swiftly become the predominant circulating JEV genotype, causing multiple cases of JE encephalitis. Our team's previous experimental findings indicate that conventional vaccines developed against the GIII genotype provide limited immune protection against the GV genotype. Collectively, these findings suggest that an increase in evolutionary rate may be associated with the acquisition of new transmission advantages or adaptation to novel hosts and environments.”

We sincerely thank you for your suggestions, which have significantly enhanced the objectivity and rigor of our manuscript.

• MCMC parameters are reported, but there is no evidence of convergence validation (ESS values, burn-in, replicates). Furthermore, only BEAST is used, which limits robustness. Report ESS >200 for key parameters. Include a machine learning tree with IQTREE or RAxML for validation.

Response: We sincerely appreciate the reviewer's constructive comments. Regarding the convergence of MCMC analysis, we have supplemented the Methods section (Page 3, Lines 127-129 and 133-136, highlighted in yellow) with the following details: "Three independent MCMC runs were conducted, each with a chain length of 5×10⁸ generations, for reproducibility and convergence validation and to ensure adequate mixing. Parameter convergence and evolutionary rate estimation were evaluated utilizing Tracer software, with a criterion that the Effective Sample Size (ESS) values must meet or exceed 200 to be considered acceptable. The MCC tree was annotated with TreeAnnotator after discarding the initial 10% of samples as burn-in." Effective Sample Sizes of Key Parameters have been provided as supplementary material, with the parameters shown in Table S. To enhance the robustness of phylogenetic inference, we have added maximum likelihood tree analysis using IQ-TREE (Page 3-4, Lines 141-146 and Page 7, Lines 219-222, highlighted in yellow). This analysis selected the GTR+G4 model as the best-fitting nucleotide substitution model through ModelFinder in PhyloSuite, and node support was assessed using the Ultrafast Bootstrap method with 1,000 replicates. The topological consistency between the maximum likelihood tree and the BEAST time tree further validates the reliability of our analytical results. These revisions have significantly improved the scientific rigor of our study. The constructed maximum likelihood tree has been included as Figure S1 in the Supplementary Materials, and the results are presented in the figure below.

• It is concluded that mutations in E justify low vaccine protection, but no neutralization data are presented; only in silico analyses are shown. Change the conclusions to a conditional tone and avoid making direct causal claims without experimental data.

Response: We greatly appreciate your insightful comment. In response to your feedback regarding the conclusions about E protein mutations and vaccine protection, we have revised the relevant section (Page 15, Lines 457-460, highlighted in yellow) to adopt a more conditional tone. Specifically, we now indicate that the “It is crucial to acknowledge that the suggested association between E protein mutations and reduced vaccine efficacy is predominantly based on in silico analyses. Consequently, this hypothesis requires further validation through neutralization assays and other experimental methodologies.”

• The table shows 126 complete sequences without logical grouping. Move the entire table to Supplementary Materials. In the text, include summarized versions or a figure showing distributions by region, year, and host.

Response: We acknowledge and appreciate your feedback regarding the logical organization of the table containing the 126 complete sequences. In response, we have relocated the complete table to the Supplementary Materials (Supplementary Table S1) and provided a streamlined version in the main text (Page 6,lines 204-205, highlighted in yellow,Table 1), which clearly displays the distribution of sequences by region, year, and host. This adjustment not only improves the logical structure of the data but also significantly enhances readability, enabling readers to more intuitively grasp the distribution characteristics of the sequence data.

• Rates are reported, but not statistically compared between genotypes (p-values, HPD overlaps, Bayes factors).

Response: We thank the reviewer for raising this important point regarding the statistical comparison of evolutionary rates. To address this, we have now included an analysis of the overlaps between the 95% HPD intervals across genotypes. The results are presented in Figure S entitled "Comparison of 95%HPD intervals of Bayesian evolutionary rate estimates across the five genotypes". As shown in figure S, JEV GII and GIV demonstrate a lack of overlap in their 95% HPD intervals with GIII, thereby providing statistical support for significant differences in their evolutionary rates. It is noteworthy that GI is well recognized as a rapidly evolving lineage and has replaced GIII to be the dominant genotype across Asia since 2000, its mean evolutionary rate remains numerically lower than those of GII and GIV. Special attention should be given to GV, as its estimated mean evolutionary rate is 2.2×10⁻³, which is approximately on the order of 10⁻³ and significantly higher than that of GIII. However, the limited availability of viral sequences currently accessible contributes to a wide 95% highest posterior density (HPD) interval. We have explicitly stated in the manuscript (Page 14, Lines 403-407, highlighted in yellow) " The observed high evolutionary rate in GV (GII & GIV)merits attention; however, it is important to acknowledge that this estimate is based on limited sequence data currently available. Achieving a more accurate comprehension of its evolutionary dynamics will necessitate the future collection and analysis of additional natural JEV strains." This supplementary clarification aims to ensure the rigor of our conclusions.

• In Figure 1, I suggest removing the Americas and showing Asia, Oceania, and Africa in greater detail. Another option is to display a small world map and zoom in on these regions.

Response: We are grateful for your valuable suggestion regarding Figure 1. In response, we have revised the figure on page 5 by removing the Americas section, thereby enlarging the display area and significantly improving the visual clarity of the geographic distributions across Asia, Oceania, and Africa.

• In Figure 2, I suggest cropping the tree scale to reduce the dendrogram size and allow the strain section to be enlarged.

Response: We sincerely appreciate your insightful suggestion regarding Figure 2. Following your advice, we have revised the figure on page 7 by adding detailed enlarged sections to the original phylogenetic tree that highlight the strains of the five genotypes. This modification has significantly improved the visual clarity of the strain information, making the relevant details more accessible and easier to interpret.  The supplementary figures/tables are provided in PDF file.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have responded to my comments, and the article is ready for publication.