Identification of T-Cell Epitopes Using a Combined In-Silico and Experimental Approach in a Mouse Model for SARS-CoV-2
, Hagit Achdout 1, Yfat Yahalom-Ronen 1, Shimrit Adutler-Lieber 1, Liat Bar-On 2, Erez Bar-Haim 2, Boaz Politi 1, Einat B. Vitner 1, Hadas Tamir 1, Sharon Melamed 1, Nir Paran 1 and Tomer Israely 1,*Round 1
Reviewer 1 Report
The authors aim to develop a specific and sensitive assay that accommodates various scenarios and routes of exposure, to assess cellular response after SARS-CoV-2 immunization or exposure. In this work the authors describe two different and complementary approaches for identifying epitopes and determining their hierarchy. The results are well presented and the manuscript is fine. Only some modifications/improvements are needed.
1- In the introduction line 72, the authors write MIP1 (Macrophage inflammatory protein), this is not the official name of the protein.
2- In material and methods, line 101 “dose in a volume of 50 l” need correction.
3- When you write where you purchased the reagents, this should be uniform (company, city and country) In line 121 you write only country, in line 124-125 you write company, province, city and country, in line 141 you write only the company (eBioscience)
4- Write the Cat. Number for the protein transport inhibitor cocktail.
5- Some reagents you not mention from where are purchased (exp. Ficoll Paque, 96-well ELISpot plates…etc.)
6- For the mAb clones in line 147-148 describe the antibodies and the Fluorescent Dye used
7- In line 168, 208, 210 the concentration needs to be corrected (1 g/ml to 10 g/ml, 2 g/ml, 1 g/ml), check the whole manuscript.
8- I suggest to remove the sup. Figure 1 as this not improve the quality of the manuscript
9- In the infection with SARS-CoV-2 virus, you wait 3 weeks, but in the previous assay (VSV G-spike immunization) why you not take the same time point in both assays, can explain why?
10- Your statement that “…the ELISpot assay was significantly more robust following SARS-CoV-2 infection than after immunization with VSV-G-spike” is not correct, because the time after infection/immunization is not the same (3 weeks vs 1 week).
11- In the discussion line 316, the sentence “ …However, this approach induced a….” is not complete
12- In line 319, specify which IFN
13- In the analysis of the IFN expression, why you don’t use other methods like real-time PCR which is counted as gold standard for studying the expression?
non comments
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
This paper is interesting in that it performs a primary screening of peptides from the viewpoint of H-2Kb binding affinity by using the prediction method NetMHCpan EL4.1 tool. Using this method, they were able to narrow down the peptides to be studied to 10 peptides. The authors state that this method was able to efficiently identify peptides recognized by T cells that respond to SARS-CoV-2, but I have to say that there is the lack of sufficient comparison with the report by Zhuang et al (JEM 2021; 218(4): e20202187). Zhuang et al have reported study on comprehensive peptide mapping of T cells responsive to SARS-CoV-2, previously. They comprehensively investigated peptides centered on the antigen stimulation site in the Spike protein, and analyzed antigenic peptides recognized by virus-specific CD8 or CD4-positive T cells using almost the same method as the authors. In particular, it is necessary to mention the significance of S915, which was newly identified this time, and the significance and interpretation of S471, S510, and S820, which were listed as recognition peptides in the report by Zhuang et al., but not listed or recognized in this study.
While the study on the 8aa peptide bound by H-2Kb is well explained, the explanation on the 15aa peptide shown in Table 2 is insufficient and difficult to understand. Those that react with the 15aa peptide are probably assumed to be CD4-positive T cells, but the same data as in Fig. 3 should be included. It seems that the description of the method for selecting the 15aa peptide is insufficient.
In Table 2, the number of reaction spots per well and the number of reaction spots per 106 cells are listed together, but since there are 4*105 cells per well, there is no need to write them together, but rather the number of reaction spots for the negative control should be listed together.
Minor points:
There are many places where letters and sentences are missing.
Line 100, 114,192,239, 255, 271, 341,345: VSV- G-spike ➡ VSV-⊿G-spike
Line 101: 50 l ➡ 50μl
Line 122,169, 208, 238,308 : g/ml ➡ μg/ml
Line 168, 198: IFN ➡ IFN₋γ
Line 317: induced a ・・・・・・・?????
In Figure 2, B: The units on the horizontal axis are missing.
The cited references list is incomplete. （Especially, page number is missing.）
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
It is my pleasure to review the revised version of the manuscript.
I would like to thank the authors for their sincere and appropriate responses to my questions.
I have just one question, so please give me a minor revision.
Although the S539 and S915 peptides of 11aa have strong INF-g-inducing properties, in the case of 15aa, which contains the same 11aa peptide sequence, the former induced significant IFN-g but the latter did not (revised version in line 274-294). The reason for this should be discussed. Please add it in the discussion.
Author Response
We highly appreciate the positive response of the Reviewer.
We appreciate the reviewer comment and add accordingly the following to the Discussion section (lines 347-355):
Our in-silico analysis of CD8 T-cell epitopes identified a set of 10 potential immunodominant epitopes, each consisting of 8 amino acids. Subsequent experimental validation using the ELISpot assay confirmed the authenticity of four of these epitopes, with S539 and S915 ranking highest in the hierarchy. Interestingly, the screening of an overlapping peptide library revealed peptides containing the sequence S539-546, while notably lacking the sequence S915-922. Although we haven’t addressed this discrepancy any further, it is conceivable that the inclusion of additional flanking amino acids in the library’s peptides (each consisting of 15 amino acids) may have impeded the binding affinity of some peptides but not others.
