COVID-19 coronavirus vaccine T cell epitope prediction analysis based on distributions of HLA class I loci (HLA-A, -B, -C) across global populations

ABSTRACT

T cell immunity, such as CD4 and/or CD8 T cell responses, plays a vital role in controlling the virus infection and pathological damage. Several studies have reported SARS-CoV-2 proteins could serve as ideal vaccine candidates against SARS-CoV-2 infection by activating the T cell responses. In the current study, based on the SARS-CoV-2 sequence and distribution of host human leukocyte antigen (HLA), we predicted the possible epitopes for the vaccine against SARS-CoV-2 infections. Firstly, the current study retrieved the SARS-CoV-2 S and N protein sequences from the NCBI Database. Then, using the Immune Epitope Database Analysis Resource, we predicted the CTL epitopes of the SARS-CoV-2 S and N proteins according to worldwide frequency distributions of HLA-A, -B, and -C alleles (>1%). Our results predicted 90 and 106 epitopes of N and S proteins, respectively. Epitope cluster analysis showed 16 and 34 respective clusters of SARS-CoV-2 N and S proteins, which covered 95.91% and 96.14% of the global population, respectively. After epitope conservancy analysis, 8 N protein epitopes and 6 S protein epitopes showed conservancy within two SARS-CoV-2 types. Of these 14 epitopes, 13 could cover SARS coronavirus and Bat SARS-like coronavirus. The remaining epitope (KWPWYIWLGF_1211-1220) could cover MERS coronavirus. Finally, the 14-epitope combination could vaccinate 89.60% of all individuals worldwide. Our results propose single or combined CTL epitopes predicted in the current study as candidates for vaccines to effectively control SARS-CoV-2 infection and development.

KEYWORDS:

• SARS-CoV-2
• Human Leukocyte Antigen (HLA)
• T cell epitopes
• population coverage
• vaccine

Introduction

Beginning December 2019, a cluster of acute respiratory disease, known as novel coronavirus-infected pneumonia (COVID-19), occurred in Wuhan, Hubei Province, China.^1–3

In January 2019, SARS-CoV-2 was identified and confirmed as the cause of COVID-19.⁴ Then, the full genome sequences of SARS-CoV-2 (NC_045512.2) were published in the National Center for Biotechnology Information (NCBI) website. The full-genome sequence and phylogenetic analysis indicated that SARS-CoV-2 forms a clade that is distinct from the beta coronaviruses associated with human severe acute respiratory syndrome (SARS) and Middle East respiratory syndrome (MERS).^3,4 Phylogenetic analysis also revealed that the gene sequence of SARS-CoV-2 is about 88% identical to that of bat SARS-like coronavirus ZXC21 (bat-SL-CoVZXC21, accession no. MG772934.1) and ZC45 (MG772933.1),⁵ which indicated SARS-CoV-2 has closer homology to bat SARS-like coronavirus and bats could be the primary source of SARS-CoV-2.^6,7

The antigenicity of coronaviruses, such as SARS-CoV, seems to be largely dependent upon two viral proteins that comprise the nucleocapsid protein (N) and the spike protein (S). The N protein, which is a nucleocapsid phosphoprotein, has been demonstrated to be highly immunogenic and seems to be an important component of the humoral response to SARS-CoV.^8,9 The S protein, which is a surface glycoprotein, is a large type-I transmembrane glycoprotein that is not only responsible for receptor binding and membrane fusion, but also serves as a potent immunogen that induces neutralizing antibodies.^8,10 Whole-sequence alignment showed about 79% sequence identity between SARS-CoV-2 and SARS-CoV.⁵ Thus, the SARS-CoV-2 S and N proteins could serve as target antigens for immune intervention protocols against coronaviruses.

T cell immunity, such as CD4 and/or CD8 T cell responses, plays a vital role in controlling the SARS-CoV infection and/or pathological damage after infection with MERS-CoV.^11,12 T cell responses have been shown to provide long-term protection.^13–15 Several studies reported that T cells had the strongest immunogenicity to structural proteins in peripheral blood mononuclear cells of convalescent SARS-CoV patients.^16,17 Based on the high avidity of cytotoxic T lymphocytes (CTLs), CTLs are considered the major eradicators of viral infections through adaptive immune response. Several studies showed that the presented CTL epitopes of S and N proteins in the context of human leukocyte antigen (HLA) alleles will aid in characterizing the virus control mechanisms and immunopathology of SARS-CoV infection and could provide a new strategy to develop an epitope-based vaccine for SARS.^18–22 For example, in 2007 and 2008, Cheung et al. predicted the SARS N protein peptide sequence for human MHC class I binding and screened for potential CTL epitopes to control the SARS-associated coronavirus infection in vitro and in vivo.^18,23 In 2020, Ibrahim and Kafi used a computational approach for vaccine design to search for candidate epitopes to control MERS-CoV infections.²⁴ But the epitope vaccine method is only in the research stage, and there is no relevant clinical trial report for coronavirus. However, for other viruses, the peptides vaccines have entered clinical trial, such as HPV peptide vaccines.²⁵

When a virus infects cells, the viral antigens are presented to the host immune system through the antigen processing machine (APM). The APM is composed of a proteasome (where the antigens are degraded into peptides), transporters associated with antigen presentation (TAPs, which are responsible for translocating peptide precursors), endoplasmic reticulum aminopeptidases (ERAPs, which trim peptides to fit HLA molecules), and the major histocompatibility complex (HLA, which presents antigen peptides to the cell surface).^26,27 The CTL epitopes bind to the cleft of various HLA-I molecules through features embedded in the peptide sequence and, more specifically, in anchor residues of HLA-I molecules.²⁸ Then, the HLA-I antigen processing system plays important roles in eliminating the infected cells. However, epitope-based vaccines are limited by HLA specificity, as HLA molecules are highly polymorphic. Therefore, it may be difficult to produce an epitope-based vaccine that is effective in patients with different HLA molecules, thus making it impractical for large-scale vaccination programs.²⁸

In the current study, based on the distribution characteristics of HLA alleles across all populations, we predicted putative CTL epitopes of the novel coronavirus (SARS-CoV-2) N and S proteins using immunoinformatic methods, which combine predictors of proteasomal processing, TAP transport, and MHC binding to produce an overall score indicating the intrinsic potential of each peptide as a T cell epitope. Our results provide likely candidate CTL epitopes or combinations thereof for vaccine development to effectively control SARS-CoV-2.

Methods

Sequence retrieval of SARS-CoV-2 N and S proteins

The ID of SARS-CoV-2 (NC_045512.2) was retrieved from NCBI (https://www.ncbi.nlm.nih.gov/nuccore/NC_045512.2).

Epitope prediction

T cell CTL epitopes of the SARS-CoV-2 N and S proteins were predicted using the Proteasomal cleavage/TAP transport/MHC class I combined predictor, which combines predictors of proteasomal processing, TAP transport, and MHC binding to produce a total score indicating the intrinsic potential of each peptide as a T cell epitope. Based on HLA allele data from the Immune Epitope Database (IEDB) Analysis Resource (http://tools.iedb.org/processing/), only HLA alleles occurring in at least 1% of the human population or with an allele frequency of 1% or more were selected. This prediction method is recommended by IEDB. The total score combines the proteasomal cleavage, TAP transport, and MHC binding predictions, which predicts a quantity proportional to the amount of peptide presented by MHC molecules on the cell surface. As the prediction method recommended, high score equal to the high efficiency of the epitope presented by MHC molecules. The total score was used as the cut off value for epitope selection in the present study.

Epitope cluster analysis

Based on sequence identity, the predicted epitopes were grouped into clusters using Epitope Cluster Analysis (http://tools.iedb.org/cluster/). A cluster is defined as a group of sequences having a sequence similarity greater than the specified minimum sequence identity threshold. An identity set percentage means that any member of the cluster will be at least the set percentage identical to at least one member of the cluster.²⁹ In the current study, a cluster was defined as a group of sequences that have a sequence similarity >70% minimum sequence identity threshold, which is the default of the cluster analysis (http://tools.iedb.org/cluster/help/)

Population coverage

The population coverage method calculates the fraction of individuals predicted to respond to a given set of epitopes with known MHC restrictions (http://tools.iedb.org/population/). This calculation is based on HLA genotypic frequencies assuming non-linkage disequilibrium between HLA loci. According to the results, we selected >95% population coverage as the threshold value.

The blast analysis of the 13 coronavirus N and S proteins

Blast method was used to analyze identity proportion of the N and S protein sequences between NC_045512.2 (Wuhan seafood market pneumonia virus isolate Wuhan-Hu-1) and other 12 coronaviruses which are MT007544.1 (Severe acute respiratory syndrome coronavirus 2 isolate Australia/VIC01/2020), FJ882944.1 (SARS coronavirus ExoN1 isolate P3pp23), NC_004718.3 (SARS coronavirus Tor2), MG987420.1 (Middle East respiratory syndrome-related coronavirus isolate NL13892) and MG021451.1 (Middle East respiratory syndrome-related coronavirus isolate NL13845), NC_006213.1 (Human coronavirus OC43 strain ATCC VR-759), NC_006577.2 (Human coronavirus HKU1), KY983587.1 (Human coronavirus 229E strain HCoV_229E/Seattle/USA/SC3112/2015), NC_005831.2 (Human Coronavirus NL63) which contain all 7 known human coronaviruses, and MG772934.1 (Bat SARS-like coronavirus isolate bat-SL-CoVZXC21), KY417144.1 (Bat SARS-like coronavirus isolate Rs4084), KT444582.1 (SARS-like coronavirus WIV16) which contain none human coronaviruses.

Epitope conservancy analysis

Epitope conservancy analysis was used to calculate the degree of conservancy of an epitope within N and S protein sequence of 13 coronaviruses set at different degrees of sequence identity (http://tools.iedb.org/conservancy/). The degree of conservancy is defined as the fraction of protein sequences containing the epitope at a given identity level that the selected epitopes completely matched at least two human coronaviruses above.

Results

The flow chart of epitope prediction analysis and results is illustrated in Figure 1.

Figure 1. The flow chart of epitope prediction analysis and results

HLA allele analysis

The average frequency of HLA alleles (>1%) across all population samples was selected in the current study (http://tools.iedb.org/processing/help/). In total, 70 HLA alleles were selected for analysis according to the IEDB Analysis Resource (Table 1). There were 18 HLA-A alleles, 32 HLA-B alleles, and 20 HLA-C alleles.

Table 1. List of 70 HLA alleles (18 HLA-A alleles, 32 HLA-B alleles, 20 HLA-C alleles) selected occurring in at least 1% of the human population

HLA Allele Loci	Alleles
HLA-A	A*02:01, A*01:01, A*02:06, A*03:01, A*11:01, A*23:01, A*24:02, A*25:01, A*26:01, A*29:02, A*30:01, A*30:02, A*31:01, A*32:01, A*33:03, A*68:01, A*74:01, A*68:02
HLA-B	B*07:02, B*08:01, B*13:01, B*13:02, B*14:02, B*15:01, B*15:02, B*15:25, B*18:01, B*27:02, B*27:05, B*35:01, B*35:03, B*37:01, B*38:01, B*39:01, B*40:01, B*40:02, B*44:02, B*44:03, B*46:01, B*48:01, B*49:01, B*50:01, B*51:01, B*52:01, B*53:01, B*55:01, B*56:01, B*57:01, B*58:01, B*58:02
HLA-C	C*01:02, C*02:02, C*02:09, C*03:02, C*03:03, C*03:04, C*04:01, C*05:01, C*06:02, C*07:01, C*07:02, C*07:04, C*08:01, C*08:02, C*12:02, C*12:03, C*14:02, C*15:02, C*16:01, C*17:01

Epitope prediction

A total of 200,410 epitopes of SARS-CoV-2 N protein were predicted against the 70 alleles of the HLA-A, -B, and -C loci. We selected 0.60, 0.55, 0.50, 0.47, 0.45, 0.40, and 0.30 as the cut off values of total scores for epitope selection, and 71, 82, 90, 99, 103, 110, and 130 epitopes were found within these cut offs. According to the epitope cluster analysis and population coverage results, we chose 0.50 as the cut off value for epitope selection (90 epitopes), as this value identified few epitopes while reaching more than 95% population coverage (Table 2).

Table 2. The epitopes number, cluster numbers, and population coverage results of the predicted SARS-CoV-2 N and S proteins

Protein	Epitopes Number	Total Score	Population Coverage			Clustering
			coverage^a	Average hit^b	PC90^c
N protein	130	0.30	0.9910	8.18	2.44	21
	110	0.40	0.9903	7.25	2.28	17
	103	0.45	0.9903	6.57	2.13	17
	99	0.47	0.9665	5.99	1.64	17
	90	0.50	0.9591	5.57	1.51	16
	82	0.55	0.9458	4.97	1.32	15
	71	0.60	0.9228	4.39	1.14	15
S protein	313	0.70	0.9901	19.77	4.78	57
	247	0.80	0.9864	16.12	4.33	49
	179	0.90	0.9810	12.09	3.31	45
	158	0.95	0.9788	10.65	3.23	40
	133	1.00	0.9632	8.65	1.60	35
	113	1.05	0.9614	7.35	1.56	33
	106	1.07	0.9614	7.08	1.56	34
	95	1.10	0.9319	6.17	1.28	33
	80	1.15	0.9242	5.25	1.17	30

^aprojected population coverage

^baverage number of epitope hits/HLA combinations recognized by the population

^cminimum number of epitope hits/HLA combinations recognized by 90% of the population

For the SARS-CoV-2 S protein, 618,871 total epitopes were predicted against the 70 alleles of the HLA-A, -B and -C loci. We selected 0.70, 0.80, 0.90, 0.95, 1.00, 1.05, 1.07, 1.10, and 1.15 as cut off values for epitope selection, and 80, 95, 106, 113, 133, 158, 179, 247, and 313 epitopes were found. According to the epitope cluster analysis and population coverage results, 1.07 was chosen as the cut off value for epitope selection (106 epitopes), as this value indicated few epitopes reaching more than 95% population coverage (Table 2).

Epitope cluster analysis

Using the selected epitopes (90 epitopes in N protein and 106 epitopes in S protein), the predicted epitopes were grouped into 16 and 34 clusters, respectively. Among the 16 N protein clusters, cluster 1 contained the most epitopes (11 epitopes), which was represented by HLA-A*30:02 (8/11), A*29:02 (8/11), A*30:01 (1/11), A*31:01 (1/11), B*15:25 (1/11), B*15:01 (1/11), A*03:01 (1/11), B*07:02 (1/11), and B*08:01 (1/11). Clusters 9 to 16 contained only one epitope each (Table 3). Among the 34 S protein clusters, clusters 1 and 2 contained the most epitopes (5 epitopes each). Cluster 1 was represented by HLA-B*15:25 (2/5), B*15:01 (1/5), A*03:01 (1/5), A*23:01 (1/5), A*24:02 (1/5), and A*30:02 (1/5). Cluster 2 included HLA- A*23:01 (2/5), A*29:02 (1/5) and A*02:01 (1/5). Clusters 16 to 34 contained only one epitope each (Table 4).

Table 3. The epitopes sequences, amino acid position and its presented HLA alleles of SARS-CoV-2 N protein by cluster analysis

Cluster Number	Epitope Number	Alignment	Epitope	Amino-acid Position	HLA Alleles
1.1	Consensus	DGKMKDLSPRWYFYYL	-	98–113
	1	DGKMKDLSPRWYFY–	DGKMKDLSPRWYFY	98–111	A*30:02
	2	-GKMKDLSPRWYFY–	GKMKDLSPRWYFY	99–111	A*30:02, A*29:02
	3	–KMKDLSPRWYFYY-	KMKDLSPRWYFYY	100–112	A*30:02, A*29:02
	4	–KMKDLSPRWYFY–	KMKDLSPRWYFY	100–111	A*30:02, A*29:02, A*30:01, A*31:01
	5	–KMKDLSPRWY – -	KMKDLSPRWY	100–109	A*30:02, B*15:25, B*15:01
	6	– MKDLSPRWYFY–	MKDLSPRWYFY	101–111	A*29:02, A*30:02
	7	– -KDLSPRWYFY–	KDLSPRWYFY	102–111	A*30:02, A*03:01, A*29:02
	8	– –DLSPRWYFYY-	DLSPRWYFYY	103–112	A*29:02
	9	– –DLSPRWYFY–	DLSPRWYFY	103–111	A*29:02
	10	– – LSPRWYFYY-	LSPRWYFYY	104–112	A*29:02, A*30:02
	11	– – -SPRWYFYYL	SPRWYFYYL	105–113	B*07:02, B*08:01
2.1	Consensus	IAQFAPSASAFF	-	304–315
	1	IAQFAPSASAF-	IAQFAPSASAF	304–314	B*15:25
	2	-AQFAPSASAFF	AQFAPSASAFF	305–315	B*15:01, B*15:25
	3	-AQFAPSASAF-	AQFAPSASAF	305–314	B*15:25, B*15:01, B*15:02
	4	–QFAPSASAFF	QFAPSASAFF	306–315	A*23:01
	5	–QFAPSASAF-	QFAPSASAF	306–314	C*14:02
	6	– FAPSASAFF	FAPSASAFF	307–315	C*12:03, B*35:01, C*03:03, C*03:04, C*03:02, C*14:02, C*16:01
3.1	Consensus	MSRIGMEVTPSGTWLTY	-	317–333
	1	MSRIGMEVTPSGTW –	MSRIGMEVTPSGTW	317–330	B*58:01
	2	– –MEVTPSGTWLTY	MEVTPSGTWLTY	322–333	B*35:01, B*18:01, B*44:03
	3	– –MEVTPSGTW –	MEVTPSGTW	322–330	B*44:02
	4	– – -VTPSGTWLTY	VTPSGTWLTY	324–333	A*29:02, B*35:01, A*30:02
	5	– – –TPSGTWLTY	TPSGTWLTY	325–333	B*35:01, B*53:01
4.1	Consensus	NTNSSPDDQIGYY	-	75–87
	1	NTNSSPDDQIGYY	NTNSSPDDQIGYY	75–87	A*01:01
	2	–NSSPDDQIGYY	NSSPDDQIGYY	77–87	A*01:01
	3	– SSPDDQIGYY	SSPDDQIGYY	78–87	A*01:01
	4	– -SPDDQIGYY	SPDDQIGYY	79–87	B*35:01
5.1	Consensus	FYYLGTGPEAGLPY	-	110–123
	1	FYYLGTGPEAGLPY	FYYLGTGPEAGLPY	110–123	A*29:02, C*14:02
	2	-YYLGTGPEAGLPY	YYLGTGPEAGLPY	111–123	A*29:02
6.1	Consensus	LPQGTTLPKGFY	-	161–172
	1	LPQGTTLPKGFY	LPQGTTLPKGFY	161–172	B*35:01
	2	– GTTLPKGFY	GTTLPKGFY	164–172	A*30:02
7.1	Consensus	ILLNKHIDAY	-	351–360
	1	ILLNKHIDAY	ILLNKHIDAY	351–360	B*15:25, B*15:01
	2	-LLNKHIDAY	LLNKHIDAY	352–360	B*15:25, B*15:01, B*15:02
8.1	Consensus	KFPRGQGVPI	-	65–74
	1	KFPRGQGVPI	KFPRGQGVPI	65–74	B*07:02
	2	-FPRGQGVPI	FPRGQGVPI	66–74	B*07:02
9.1	Singleton	NTASWFTAL	NTASWFTAL	48–56	A*68:02
10.1	Singleton	RQKRTATKAY	RQKRTATKAY	259–268	B*15:01, B*15:25, A*30:02
11.1	Singleton	KAYNVTQAF	KAYNVTQAF	266–274	C*03:02, B*15:25, C*12:03, A*32:01, C*16:01, C*03:03, C*03:04, B*15:01, C*14:02, B*58:01, C*12:02, B*15:02, B*35:01, C*02:09, C*02:02
12.1	Singleton	LPAADLDDF	LPAADLDDF	395–403	B*35:01
13.1	Singleton	LPNNTASWF	LPNNTASWF	45–53	B*35:01, B*53:01
14.1	Singleton	LLLDRLNQL	LLLDRLNQL	222–230	A*02:01, A*02:06
15.1	Singleton	IGYYRRATR	IGYYRRATR	84–92	A*31:01
16.1	Singleton	NQRNAPRITF	NQRNAPRITF	8–17	B*15:01

Table 4. The epitopes sequences, amino acid position and its presented HLA alleles of SARS-CoV-2 S protein by cluster analysis

Cluster Number	Epitope Number	Alignment	Epitope	Amino-acid Position	HLA Alleles
1.1	Consensus	YSVLYNSASFSTFKCY	-	365–380
	1	YSVLYNSASF – –	YSVLYNSASF	365–374	B*15:01
	2	-SVLYNSASF – –	SVLYNSASF	366–374	B*15:25
	3	–VLYNSASFSTFKCY	VLYNSASFSTFKCY	367–380	A*03:01
	4	– LYNSASFSTF –	LYNSASFSTF	368–377	A*23:01, A*24:02
	5	– – -ASFSTFKCY	ASFSTFKCY	372–380	B*15:25, A*30:02
2.1	Consensus	AYYVGYLQPRTFLLKY	-	264–279
	1	AYYVGYLQPRTF – -	AYYVGYLQPRTF	264–275	A*23:01
	2	-YYVGYLQPRTF – -	YYVGYLQPRTF	265–275	A*23:01
	3	– -GYLQPRTFLLKY	GYLQPRTFLLKY	268–279	A*29:02
	4	– –YLQPRTFLLKY	YLQPRTFLLKY	269–279	A*29:02
	5	– –YLQPRTFLL–	YLQPRTFLL	269–277	A*02:01
3.1	Consensus	RISNCVADYSVLY	-	357–369
	1	RISNCVADYSVLY	RISNCVADYSVLY	357–369	A*30:02
	2	RISNCVADY – -	RISNCVADY	357–365	A*30:02
	3	– NCVADYSVLY	NCVADYSVLY	360–369	A*29:02
	4	– -CVADYSVLY	CVADYSVLY	361–369	A*29:02, B*35:01, A*26:01, A*01:01
4.1	Consensus	LQIPFAMQMAYRF	-	894–906
	1	LQIPFAMQMAY–	LQIPFAMQMAY	894–904	B*35:01, B*15:25, B*15:01
	2	-QIPFAMQMAY–	QIPFAMQMAY	895–904	B*35:01, A*29:02
	3	–IPFAMQMAY–	IPFAMQMAY	896–904	B*35:01, B*53:01
	4	– -FAMQMAYRF	FAMQMAYRF	898–906	B*35:01, C*03:02, B*53:01, C*03:04, C*03:03, B*58:01, A*23:01
5.1	Consensus	RVYSSANNCTFEY	-	158–170
	1	RVYSSANNCTFEY	RVYSSANNCTFEY	158–170	A*30:02
	2	-VYSSANNCTF–	VYSSANNCTF	159–168	A*24:02
	3	–YSSANNCTF–	YSSANNCTF	160–168	C*03:02, C*03:04, C*03:03, B*35:01, C*16:01, C*12:03
	4	– -SANNCTFEY	SANNCTFEY	162–170	B*35:01, A*29:02
6.1	Consensus	AYTMSLGAENSVAY	-	694–707
	1	AYTMSLGAENSVAY	AYTMSLGAENSVAY	694–707	A*29:02
	2	-YTMSLGAENSVAY	YTMSLGAENSVAY	695–707	A*29:02, B*15:25
	3	– -SLGAENSVAY	SLGAENSVAY	698–707	B*15:25
	4	– –LGAENSVAY	LGAENSVAY	699–707	B*35:01
7.1	Consensus	SWMESEFRVY	-	151–160
	1	SWMESEFRVY	SWMESEFRVY	151–160	A*29:02
	2	-WMESEFRVY	WMESEFRVY	152–160	B*15:25, B*15:02
	3	–MESEFRVY	MESEFRVY	153–160	B*18:01
8.1	Consensus	YTNSFTRGVYY	-	28–38
	1	YTNSFTRGVYY	YTNSFTRGVYY	28–38	A*01:01, A*29:02
	2	YTNSFTRGVY-	YTNSFTRGVY	28–37	A*30:02, A*01:01
	3	–NSFTRGVYY	NSFTRGVYY	29–38	C*12:03, A*29:02
9.1	Consensus	HWFVTQRNFY	-	1101–1110
	1	HWFVTQRNFY	HWFVTQRNFY	1101–1110	A*29:02
	2	-WFVTQRNFY	WFVTQRNFY	1102–1110	A*29:02
10.1	Consensus	FQFCNDPFLGVY	-	133–144
	1	FQFCNDPFLGVY	FQFCNDPFLGVY	133–144	B*15:25, B*15:01
	2	FQFCNDPFL –	FQFCNDPFL	133–141	A*02:06
11.1	Consensus	SVASQSIIAY	-	686–695
	1	SVASQSIIAY	SVASQSIIAY	686–695	B*15:25
	2	-VASQSIIAY	VASQSIIAY	687–695	B*35:01, B*15:25, C*03:02
12.1	Consensus	FLPFFSNVTW	-	55–64
	1	FLPFFSNVTW	FLPFFSNVTW	55–64	B*53:01
	2	-LPFFSNVTW	LPFFSNVTW	56–64	B*53:01
13.1	Consensus	KRSFIEDLLF	-	814–823
	1	KRSFIEDLLF	KRSFIEDLLF	814–823	B*58:01
	2	-RSFIEDLLF	RSFIEDLLF	815–823	B*58:01, B*57:01
14.1	Consensus	LLTDEMIAQY	-	864–873
	1	LLTDEMIAQY	LLTDEMIAQY	864–873	A*01:01
	2	-LTDEMIAQY	LTDEMIAQY	865–873	A*01:01
15.1	Consensus	RVYSTGSNVF	-	634–643
	1	RVYSTGSNVF	RVYSTGSNVF	634–643	B*15:25, B*15:01, A*32:01
	2	-VYSTGSNVF	VYSTGSNVF	635–643	C*14:02
16.1	Singleton	WTAGAAAYY	WTAGAAAYY	258–266	A*29:02, A*26:01, A*68:01
17.1	Singleton	KWPWYIWLGF	KWPWYIWLGF	1211–1220	A*23:01, A*24:02
18.1	Singleton	KSNIIRGWIF	KSNIIRGWIF	97–106	B*58:01
19.1	Singleton	CYFPLQSYGF	CYFPLQSYGF	488–497	A*23:01, A*24:02
20.1	Singleton	FEYVSQPFL	FEYVSQPFL	168–176	B*40:01
21.1	Singleton	FVFKNIDGY	FVFKNIDGY	192–200	B*35:01, A*29:02
22.1	Singleton	LMSFPQSAPHGVVF	LMSFPQSAPHGVVF	1057–1070	B*15:01, B*15:25
23.1	Singleton	NATRFASVY	NATRFASVY	151–159	B*35:01
24.1	Singleton	KSFTVEKGIY	KSFTVEKGIY	312–321	A*30:02
25.1	Singleton	LPFNDGVYF	LPFNDGVYF	84–92	B*35:01, B*53:01
26.1	Singleton	TLLALHRSY	TLLALHRSY	248–256	B*15:25
27.1	Singleton	NDLCFTNVY	NDLCFTNVY	396–404	B*18:01
28.1	Singleton	YEQYIKWPWY	YEQYIKWPWY	1214–1223	B*18:01
29.1	Singleton	KRFDNPVLPF	KRFDNPVLPF	77–86	B*27:05
30.1	Singleton	FPNITNLCPF	FPNITNLCPF	337–346	B*35:01
31.1	Singleton	NVYADSFVIR	NVYADSFVIR	402–411	A*68:01
32.1	Singleton	QLTPTWRVY	QLTPTWRVY	636–644	B*15:25, B*15:02
33.1	Singleton	KVGGNYNYLY	KVGGNYNYLY	452–461	A*30:02
34.1	Singleton	WTFGAGAAL	WTFGAGAAL	894–902	C*03:04, C*03:03

Population coverage

The HLA allele frequencies and associated data for different individual populations from worldwide studies were provided by the allelefrequencies.net database (http://www.allelefrequencies.net/), which covered 10,656,469 individuals from 1,270 populations in 16 different geographical areas including 115 countries and 21 different ethnicity groups.

The worldwide population coverage of the 90 and 106 epitopes of N and S proteins were 95.91% and 96.14%, respectively (Table 2). For the 90 N protein epitopes, the average number of epitope hits/HLA combinations recognized by the population was 5.57, and the minimum number recognized by 90% of the population was 1.51. For the 106 S protein epitopes, the average number of epitope hits/HLA combinations recognized by the population was 7.08, and the minimum number recognized by 90% of the population was 1.56.

The blast analysis of the 13 coronavirus N and S proteins

The BLAST results of N and S proteins between NC_045512.2 and 12 other sequences (MT007544.1, MG772934.1, KY417144.1, KT444582.1, FJ882944.1, NC_004718.3, MG987420.1, MG021451.1, NC_006213.1, NC_006577.2, KY983587.1, and NC_005831.2) selected in the current study are shown in Table 5. The percent identities of other human coronaviruses except SARS-CoV-2 and SARS were relatively low (29.43%~50.97% for N protein; 30.78%~37.63% for S protein).

Table 5. The results of comparing N and S protein of SARS-CoV-2 with Bat coronavirus and other human coronaviruses (SARS, MERS, HCoV-OC43, HCoV-HKU1, HCoV-229E and HCoV-NL63)

Serial number	complete genome^a	Genus	Subgenus	Description	Max Score	Total Score	Query Cover^b	E value^c	Per. Ident^d	Accession
1	NC_045512.2	Betacoronavirus	Sarbecovirus	nucleocapsid phosphoprotein	854	854	100%	0	100.00%	YP_009724397.2
				surface glycoprotein	2637	2637	100%	0	100.00%	YP_009724390.1
2	MT007544.1	Betacoronavirus	Sarbecovirus	nucleocapsid phosphoprotein	854	854	100%	0	100.00%	QHR84456.1
				surface glycoprotein	2634	2634	100%	0	99.92%	QHR84449.1
3	MG772934.1	Betacoronavirus	Sarbecovirus	nucleocapsid protein	715	715	100%	0	94.27%	AVP78049.1
				spike protein	2105	2105	99%	0	80.32%	AVP78042.1
4	KT444582.1	Betacoronavirus	Sarbecovirus	nucleocapsid protein	669	669	100%	0	90.28%	ALK02467.1
				spike protein	2065	2065	100%	0	77.07%	ALK02457.1
5	KY417144.1	Betacoronavirus	Sarbecovirus	nucleocapsid protein	666	666	100%	0	90.05%	ATO98142.1
				spike protein	2049	2049	99%	0	77.23%	ATO98132.1
6	FJ882944.1	Betacoronavirus	Sarbecovirus	nucleocapsid protein	672	672	100%	0	90.52%	ACZ72030.1
				spike glycoprotein precursor	2040	2040	100%	0	76.12%	ACZ72020.1
7	NC_004718.3	Betacoronavirus	Sarbecovirus	nucleocapsid protein	672	672	100%	0	90.52%	YP_009825061.1
				spike glycoprotein precursor	2038	2038	100%	0	75.96%	YP_009825051.1
8	MG987420.1	Betacoronavirus	Merbecovirus	N protein	288	288	88%	6E-90	50.26%	AWH65950.1
				S protein	561	636	85%	0	35.56%	AWH65943.1
9	MG021451.1	Betacoronavirus	Merbecovirus	N protein	284	284	82%	3E-88	50.97%	AVV62533.1
				S protein	556	618	83%	1E-179	34.70%	AVV62526.1
10	NC_006213.1	Betacoronavirus	Embecovirus	nucleocapsid protein	176	207	74%	6.00E-55	38.35%	YP_009555245.1
				spike surface glycoprotein	467	602	74%	2.00E-146	37.63%	YP_009555241.1
11	NC_006577.2	Betacoronavirus	Embecovirus	nucleocapsid phosphoprotein	197	212	82%	9.00E-63	36.74%	YP_173242.1
				spike glycoprotein	452	576	71%	3.00E-141	35.43%	YP_173238.1
12	KY983587.1	Alphacoronavirus	Duvinacovirus	nucleocapsid protein	89.4	89.4	65%	6.00E-24	29.43%	ARU07605.1
				spike protein	366	481	67%	2.00E-111	31.32%	ARU07601.1
13	NC_005831.2	Alphacoronavirus	Setracovirus	nucleocapsid protein	61.2	61.2	13%	9.00E-15	48.28%	YP_003771.1
				Spike protein	349	404	65%	3.00E-104	30.78%	YP_003767.1

^aNC_045512.2 is Wuhan seafood market pneumonia virus isolate Wuhan-Hu-1;

MT007544.1 is Severe acute respiratory syndrome coronavirus 2 isolate Australia/VIC01/2020;

MG772934.1 is Bat SARS-like coronavirus isolate bat-SL-CoVZXC21;

KY417144.1 is Bat SARS-like coronavirus isolate Rs4084;

KT444582.1 is SARS-like coronavirus WIV16;

FJ882944.1 is SARS coronavirus ExoN1 isolate P3pp23;

NC_004718.3 SARS coronavirus Tor2;

MG987420.1 is Middle East respiratory syndrome-related coronavirus isolate NL13892;

MG021451.1 is Middle East respiratory syndrome-related coronavirus isolate NL13845;

NC_006213.1 Human coronavirus OC43 strain ATCC VR-759;

NC_006577.2 Human coronavirus HKU1;

KY983587.1 Human coronavirus 229E strain HCoV_229E/Seattle/USA/SC3112/2015;

NC_005831.2 Human Coronavirus NL63;

^bQuery Coverage: coverage of the compared sequences

^cexpect value: the possibility of random matching. When the value of E is close to zero or zero, it is essentially a perfect match.

^dpercentage identity: the percentage of base number in the total sequence of the compared sequences

Epitope conservancy analysis

The identity proportion results of 16 epitopes within 13 N protein sequences above are shown in Table 6. The identity is 80%~100% for 16 epitopes within 7 N protein sequences belong to the same subgenus (Sarbecovirus, including SARS-CoV-2). The identity is 23.53%~80% for 16 epitopes within other 6 N protein sequences belong to the different genus (Alphacoronavirus) or subgenus (Merbecovirus). The identity proportion results of 34 epitopes within 13 S protein sequences above are shown in Table 7. The identity is higher (60%~100%) for NP13 (KRSFIEDLLF_814-823), NP17 (KWPWYIWLGF_1211-1220) and NP28 (YEQYIKWPWY_1214-1223) within 13 S protein sequences. For the remaining 31 epitopes, the identity is 33.33%~100% for the epitopes within 7 S protein sequences belong to the same subgenus (Sarbecovirus, including SARS-CoV-2); the identity is 28.57%~55.56% for the epitopes within other 6 N protein sequences belong to different genus (Alphacoronavirus) or subgenus (Merbecovirus).

Table 6. The epitope conservancy analysis results for N protein among the 13 coronaviruses

				Identity percentage (%)^d
Name^a	Epitope sequence	Length^b	Percent of protein sequence matches^c	YP_009724397.2	QHR84456.1	AVP78049.1	ALK02467.1	ATO98142.1	ACZ72030.1	YP_009825061.1	AWH65950.1	AVV62533.1	YP_009555245.1	YP_173242.1	ARU07605.1	YP_003771.1
NP 1	DGKMKDLSPRWYFYYL98-113^e	16	15.38% (2/13)	100.00	100.00	93.75	93.75	93.75	93.75	93.75	50.00	50.00	68.75	75.00	62.50	50.00
NP 2	IAQFAPSASAFF304-315	12	53.85% (7/13)	100.00	100.00	100.00	100.00	100.00	100.00	100.00	66.67	66.67	58.33	50.00	33.33	33.33
NP 3	MSRIGMEVTPSGTWLTY317-333	17	46.15% (6/13)	100.00	100.00	100.00	100.00	94.12	100.00	100.00	23.53	23.53	23.53	23.53	29.41	23.53
NP 4	NTNSSPDDQIGYY75-87	13	15.38% (2/13)	100.00	100.00	92.31	92.31	92.31	92.31	92.31	46.15	46.15	30.77	30.77	30.77	38.46
NP 5	FYYLGTGPEAGLPY110-123	14	23.08% (3/13)	100.00	100.00	100.00	92.86	92.86	92.86	92.86	78.57	78.57	71.43	71.43	57.14	64.29
NP 6	LPQGTTLPKGFY161-172	12	53.85% (7/13)	100.00	100.00	100.00	100.00	100.00	100.00	100.00	50.00	50.00	58.33	58.33	33.33	41.67
NP 7	ILLNKHIDAY351-360	10	53.85% (7/13)	100.00	100.00	100.00	100.00	100.00	100.00	100.00	60.00	60.00	40.00	40.00	30.00	30.00
NP 8	KFPRGQGVPI65-74	10	15.38% (2/13)	100.00	100.00	90.00	90.00	90.00	90.00	90.00	70.00	70.00	70.00	80.00	40.00	50.00
NP 9	NTASWFTAL48-56	9	53.85% (7/13)	100.00	100.00	100.00	100.00	100.00	100.00	100.00	66.67	44.44	44.44	33.33	33.33	33.33
NP 10	RQKRTATKAY259-268	10	15.38% (2/13)	100.00	100.00	90.00	90.00	90.00	90.00	90.00	60.00	60.00	50.00	60.00	40.00	40.00
NP 11	KAYNVTQAF266-274	9	15.38% (2/13)	100.00	100.00	88.89	88.89	77.78	88.89	88.89	55.56	55.56	44.44	55.56	55.56	44.44
NP 12	LPAADLDDF395-403	9	15.38% (2/13)	100.00	100.00	88.89	88.89	88.89	88.89	88.89	33.33	33.33	44.44	44.44	44.44	44.44
NP 13	LPNNTASWF45-53	9	53.85% (7/13)	100.00	100.00	100.00	100.00	100.00	100.00	100.00	66.67	44.44	44.44	33.33	44.44	44.44
NP 14	LLLDRLNQL222-230	9	53.85% (7/13)	100.00	100.00	100.00	100.00	100.00	100.00	100.00	55.56	55.56	33.33	33.33	33.33	33.33
NP 15	IGYYRRATR84-92	9	53.85% (7/13)	100.00	100.00	100.00	100.00	100.00	100.00	100.00	55.56	55.56	44.44	44.44	44.44	44.44
NP 16	NQRNAPRITF8-17	10	15.38% (2/13)	100.00	100.00	80.0	90.00	90.00	90.00	90.00	40.00	40.00	30.00	30.00	30.00	40.00

^aepitope name

^bepitope length

^cPercent of protein sequence matches at identity 100%

^dthe identity percentage results of each16 epitopes with 13 N protein sequences

^eAmino-acid position

Table 7. The epitope conservancy analysis results for S protein among the 13 coronaviruses

				Identity percentage (%)^d
Name^a	Epitope sequence	Length^b	Percent of protein sequence matches^c	YP_009724390.1	QHR84449.1	AVP78042.1	ALK02457.1	ATO98132.1	ACZ72020.1	YP_009825051.1	AWH65943.1	AVV62526.1	YP_009555241.1	YP_173238.1	ARU07601.1	YP_003767.1
NP 1	YSVLYNSASFSTFKCY365-380^e	16	15.38% (2/13)	100.00	100.00	81.25	93.75	87.50	87.50	87.50	31.25	31.25	31.25	37.50	31.25	31.25
NP 2	AYYVGYLQPRTFLLKY264-279	16	15.38% (2/13)	100.00	100.00	81.25	75.00	75.00	75.00	75.00	37.50	37.50	37.50	43.75	31.25	37.50
NP 3	RISNCVADYSVLY357-369	13	23.08% (3/13)	100.00	100.00	61.54	100.00	92.31	92.31	92.31	30.77	38.46	38.46	46.15	38.46	38.46
NP 4	LQIPFAMQMAYRF894-906	13	53.85% (7/13)	100.00	100.00	100.00	100.00	100.00	100.00	100.00	46.15	46.15	30.77	38.46	38.46	30.77
NP 5	RVYSSANNCTFEY158-170	13	15.38% (2/13)	100.00	100.00	69.23	53.85	53.85	53.85	53.85	38.46	38.46	30.77	38.46	30.77	38.46
NP 6	AYTMSLGAENSVAY694-707	14	15.38% (2/13)	100.00	100.00	92.86	78.57	78.57	78.57	78.57	42.86	42.86	35.71	28.57	35.71	35.71
NP 7	SWMESEFRVY151-160	10	15.38% (2/13)	100.00	100.00	50.00	40.00	40.00	40.00	40.00	30.00	40.00	40.00	40.00	40.00	30.00
NP 8	YTNSFTRGVYY28-38	11	15.38% (2/13)	100.00	100.00	81.82	54.55	54.55	54.55	54.55	36.36	36.36	36.36	36.36	36.36	36.36
NP 9	HWFVTQRNFY1101-1110	10	23.08% (3/13)	100.00	100.00	100.00	70.00	70.00	70.00	70.00	30.00	30.00	40.00	40.00	30.00	40.00
NP 10	FQFCNDPFLGVY133-144	12	15.38% (2/13)	100.00	100.00	58.33	41.67	41.67	41.67	41.67	33.33	33.33	33.33	33.33	33.33	33.33
NP 11	SVASQSIIAY686-695	10	15.38% (2/13)	100.00	100.00	40.00	50.00	50.00	50.00	50.00	40.00	40.00	40.00	40.00	50.00	50.00
NP 12	FLPFFSNVTW55-64	10	15.38% (2/13)	100.00	100.00	80.00	80.00	80.00	80.00	80.00	40.00	40.00	40.00	40.00	40.00	40.00
NP 13	KRSFIEDLLF814-823	10	53.85% (7/13)	100.00	100.00	100.00	100.00	100.00	100.00	100.00	80.00	90.00	80.00	80.00	70.00	70.00
NP 14	LLTDEMIAQY864-873	10	15.38% (2/13)	100.00	100.00	90.00	80.00	80.00	80.00	80.00	40.00	50.00	40.00	40.00	40.00	40.00
NP 15	RVYSTGSNVF634-643	10	15.38% (2/13)	100.00	100.00	50.00	90.00	90.00	80.00	80.00	40.00	50.00	30.00	40.00	40.00	40.00
NP 16	WTAGAAAYY258-266	9	15.38% (2/13)	100.00	100.00	66.67	55.56	55.56	55.56	55.56	44.44	44.44	44.44	33.33	44.44	33.33
NP 17	KWPWYIWLGF1211-1220	10	30.77% (4/13)	100.00	100.00	90.00	90.00	90.00	90.00	90.00	100.00	100.00	70.00	70.00	60.00	60.00
NP 18	KSNIIRGWIF97-106	10	15.38% (2/13)	100.00	100.00	80.00	70.00	80.00	70.00	70.00	40.00	30.00	40.00	40.00	40.00	40.00
NP 19	CYFPLQSYGF488-497	10	15.38% (2/13)	100.00	100.00	40.00	70.00	70.00	70.00	70.00	40.00	50.00	40.00	40.00	40.00	40.00
NP 20	FEYVSQPFL168-176	9	15.38% (2/13)	100.00	100.00	66.67	55.56	66.67	55.56	55.56	44.44	44.44	44.44	44.44	33.33	44.44
NP 21	FVFKNIDGY192-200	9	15.38% (2/13)	100.00	100.00	66.67	77.78	66.67	77.78	77.78	44.44	44.44	44.44	33.33	44.44	44.44
NP 22	LMSFPQSAPHGVVF1057-1070	14	23.08% (3/13)	100.00	100.00	100.00	92.86	92.86	92.86	92.86	42.86	42.86	42.86	42.86	42.86	42.86
NP 23	NATRFASVY151-159	9	15.38% (2/13)	100.00	100.00	88.89	77.78	77.78	77.78	77.78	55.56	55.56	44.44	44.44	44.44	44.44
NP 24	KSFTVEKGIY312-321	10	15.38% (2/13)	100.00	100.00	70.00	70.00	70.00	70.00	70.00	50.00	50.00	50.00	40.00	50.00	40.00
NP 25	LPFNDGVYF84-92	9	15.38% (2/13)	100.00	100.00	66.67	77.78	66.67	66.67	66.67	44.44	44.44	44.44	55.56	33.33	44.44
NP 26	TLLALHRSY248-256	9	7.69% (1/13)	100.00	88.89	55.56	33.33	33.33	33.33	33.33	44.44	44.44	44.44	55.56	33.33	44.44
NP 27	NDLCFTNVY396-404	9	15.38% (2/13)	100.00	100.00	77.78	88.89	88.89	88.89	88.89	44.44	55.56	44.44	44.44	44.44	55.56
NP 28	YEQYIKWPWY1214-1223	10	46.15% (6/13)	100.00	100.00	90.00	100.00	100.00	100.00	100.00	70.00	70.00	80.00	80.00	70.00	70.00
NP 29	KRFDNPVLPF77-86	10	15.38% (2/13)	100.00	100.00	70.00	80.00	60.00	60.00	60.00	40.00	40.00	30.00	30.00	40.00	40.00
NP 30	FPNITNLCPF337-346	10	46.15% (6/13)	100.00	100.00	90.00	100.00	100.00	100.00	100.00	40.00	40.00	50.00	40.00	40.00	40.00
NP 31	NVYADSFVIR402-411	10	15.38% (2/13)	100.00	100.00	70.00	80.00	80.00	80.00	80.00	40.00	40.00	40.00	40.00	50.00	40.00
NP 32	QLTPTWRVY636-644	9	15.38% (2/13)	100.00	100.00	77.78	88.89	88.89	77.78	77.78	44.44	33.33	44.44	44.44	44.44	44.44
NP 33	KVGGNYNYLY452-461	10	15.38% (2/13)	100.00	100.00	40.00	60.00	70.00	60.00	60.00	50.00	40.00	40.00	40.00	30.00	40.00
NP 34	WTFGAGAAL894-902	9	53.85% (7/13)	100.00	100.00	100.00	100.00	100.00	100.00	100.00	44.44	44.44	44.44	44.44	44.44	44.44

^aepitope name

^bepitope length

^cPercent of protein sequence matches at identity 100%

^dthe identity percentage results of each 34 epitopes with 13 S protein sequences

^eAmino-acid position

Among the 16 SARS-CoV-2 N protein epitopes, 7 epitopes showed 53.85% (7/13) match between 13 protein sequences (Table 6). For NP 3 (MSRIGMEVTPSGTWLTY_317-333), this match was 46.15% (6/13). The above 8 epitopes had 23 alleles (Table 8). HLA-B*35:01 (3/8) was most frequent in the epitopes, followed by B*15:01 (2/8), B*15:02 (2/8), B*15:25 (2/8), A*30:02 (2/8), and B*53:01 (2/8). The remaining 17 HLA alleles were present in only one epitope. No epitope was found with the N protein of other human coronaviruses except SARS-CoV-2 and SARS. Among the 34 SARS-CoV-2 S protein epitopes, 3 epitopes showed 53.85% (7/13) match between 13 protein sequences (Table 7). Two epitopes had a 46.15% (6/13) match. The NP17 epitope (KWPWYIWLGF_1211-1220) was found in the S protein of two SARS-CoV-2 and two MERS-related coronaviruses. The above 6 epitopes had 13 alleles (Table 8). HLA-A*23:01, B*35:01, B*58:01, C*03:04, and C*03:03 were present in 2 epitopes. The remaining 8 HLA alleles were present in only one epitope.

Table 8. The epitope sequences and its presented HLA alleles of the 14 combination epitopes

Protein	Epitope sequence	HLA Alleles
N protein	IAQFAPSASAFF304-315^a	A*23:01, B*15:01, B*15:02, B*15:25, B*35:01, C*03:02, C*03:03, C*03:04, C*12:03, C*14:02, C*16:01
	MSRIGMEVTPSGTWLTY317-333	A*29:02, A*30:02, B*18:01, B*35:01, B*44:02, B*44:03, B*53:01, B*58:01
	LPQGTTLPKGFY161-172	A*30:02, B*35:01
	ILLNKHIDAY351-360	B*15:25, B*15:01, B*15:02
	NTASWFTAL48-56	A*68:02
	LPNNTASWF45-53	B*35:01, B*53:01
	LLLDRLNQL222-230	A*02:01, A*02:06
	IGYYRRATR84-92	A*31:01
S protein	LQIPFAMQMAYRF894-906	A*23:01, A*29:02, B*15:01, B*15:25, B*35:01, B*53:01, B*58:01, C*03:02, C*03:03, C*03:04
	KRSFIEDLLF814-823	B*58:01, B*57:01
	YEQYIKWPWY1214-1223	B*18:01
	FPNITNLCPF337-346	B*35:01
	WTFGAGAAL894-902	C*03:04, C*03:03
	KWPWYIWLGF1211-1220	A*23:01, A*24:02

^aAmino-acid position

Based on the epitope conservancy analysis results, we selected epitopes with 30.77%~53.85% match percentage as epitope candidates, which means the selected epitopes completely matched at least two human coronaviruses. Finally, 14 total epitopes (8 epitopes for N protein, 6 epitopes for S protein) were selected, which cover 89.60% of the worldwide population (Table 9). The average number of epitope hits/HLA combinations recognized by the population was 3.22, and the minimum number recognized by 90% of the population was 0.96. For SARS-CoV (FJ882944.1) and SARS-like coronavirus (KT444582.1), 13 epitopes (8 clusters for N protein, 5 clusters for S protein) were covered, which occurred 84.94% of the worldwide population. For Bat SARS-like coronavirus (KY417144.1), 12 epitopes (7 clusters for N protein, 5 clusters for S protein) were covered, which occurred 81.91% of the worldwide population. For Bat SARS-like coronavirus (MG772934.1), 11 epitopes (8 clusters for N protein, 3 clusters for S protein) were covered, which occurred 84.94% of the worldwide population. For MERS (MG987420.1 and MG021451.1), only one epitope (S protein) was covered, which occurred 26.18% of the worldwide population (Table 9).

Table 9. The population coverage results (%) of the 14 combination epitopes

Description	Total Score	Population Coverage
		Coverage^a	Average hit^b	PC90^c
8 epitopes in N protein	0.50	0.8413	1.93	0.63
5 epitopes in S protein	1.07	0.5312	1.02	0.21
8 epitopes in N protein and 5 clusters in S protein	-	0.8494	2.95	0.66
14 epitopes for SARS-CoV-2 (NC_045512.2 and MT007544.1)	-	0.8960	3.22	0.96
13 epitopes for SARS-CoV (FJ882944.1 and NC_004718.3)	-	0.8494	2.95	0.66
13 epitopes for SARS-like coronavirus (KT444582.1)	-	0.8494	2.95	0.66
12 epitopes for Bat SARS-like coronavirus (KY417144.1)	-	0.8191	2.53	0.55
11 epitopes for Bat SARS-like coronavirus (MG772934.1)	-	0.8494	2.79	0.66
1 epitope (NP17 in S protein) for MERS (MG987420.1 and MG021451.1)	-	0.2618	0.27	0.14

^aprojected population coverage

^baverage number of epitope hits/HLA combinations recognized by the population

^cminimum number of epitope hits/HLA combinations recognized by 90% of the population

Discussion

Modern immunoinformatic methodologies provide new strategies for the design and synthesis of antigen-specific epitope-based vaccines against viral or pathogenic infections. In the current study, according to the SARS-CoV-2 S and N protein sequences, we predicted putative HLA-restricted CTL epitopes using immunoinformatic methods. We found 14-epitope combinations that have 30.77%~53.85% match percentage among the coronavirus sequences covering SARS-CoV-2, other 6 human coronaviruses and other coronavirus species (NC_045512.2, MT007544.1, MG772934.1, KY417144.1, KT444582.1, FJ882944.1, NC_004718.3, MG987420.1, MG021451.1, NC_006213.1, NC_006577.2, KY983587.1, and NC_005831.2). The worldwide population coverage is 89.60%, which indicates that these epitopes could serve as candidate epitopes for vaccines of SARS-CoV-2 among most of the global population.

Based on the antigenicity of the SARS-CoV-2 S and N proteins, they could be major targets for preventing and treating SARS-CoV-2 infection. In 2007 and 2008, Cheung et al. predicted the DNA vaccines encoding the N-protein peptides LLLDRLNQL_223-231 and LALLLLDRL_220-228 presented by HLA-A*02:01 could trigger the highest T-cell cytotoxicity toward N protein-expressing cells, which indicated that these two N-protein peptides could be valuable peptide candidates for SARS vaccine.^18,23 In the current study, we also predicted that LLLDRLNQL_223-231 could be presented by HLA-A*02:01 and HLA-A*02:06, which covered 39.08% and 1.95% individuals worldwide, respectively. Moreover, the sequence of the epitope is identical to those of SARS coronavirus (KT444582.1 and FJ882944.1) and nucleocapsid protein Bat SARS-like coronavirus (MG772934.1 and KY417144.1), but not MERS (MG987420.1 and MG021451.1), which indicated that LLLDRLNQL_223-231 could also be a valuable vaccine candidate peptide for SARS and Bat SARS-like coronavirus. In 2020, Ahmed et al. revisited previously tested and functional HLA-restricted SARS coronavirus epitopes and found many epitopes were also conserved in SARS-CoV-2, such as epitope LLLDRLNQL_223-231 presented by HLA-A*02:01 allele.³⁰ Thus, epitope LLLDRLNQL_223-231 could potentially offer protection against these two viruses. In addition, in 2020, Austin Nguyen et al. analyzed viral peptide-MHC class I binding affinity across HLA-A, -B, and -C genotypes for all SARS-CoV-2 peptides in silico, and found that different HLA alleles showed various capacities to present highly conserved SARS-CoV-2 peptides.³¹ Among the conserved sequences, PRWYFYYLGTGP_106-117 in N protein was highly conserved. In the current study, we also predicted that FYYLGTGPEAGLPY_110-123 was presented by HLA-A*29:02 and HLA-C*14:02 alleles. In 2006, Zhou et al. predicted the SARS S protein epitopes and identified KLPDDFMGCV_411–420 as a novel HLA-A*02:01-restricted S protein epitope.¹¹ Their results indicated that the epitope could be a novel SARS-associated coronavirus-specific CTL epitope and a potential target for characterizing virus control mechanisms and evaluating candidate SARS vaccines. Then, Liu et al. also predicted a SARS-CoV N protein-derived CTL epitope and identified QFKDNVILL_346–354, which was restricted by HLA-A*24:02, by a series of in vitro studies.¹⁹ In the current study, we did not predict these two epitopes in the SARS-CoV-2 N and S proteins. After aligning SARS-CoV-2 and SARS-CoV, we found only one mutation in the S protein epitope KLPDDFMGCV_411–420 and two different mutations in QFKDNVILL_346–354 between SARS-CoV-2 and SARS-CoV. Then, we replaced these mutations in SARS-CoV-2 and reconducted the predictions; KLPDDFMGCV_411–420 and QFKDNVILL_346–354 could be predicted, which indicated that these mutations could change the affinity between epitopes and HLA molecules.

Evasion of the host CTL response through the mutation of key epitopes is a major challenge to achieving natural or therapeutic vaccine-induced immune control of SARS-CoV-2. Therefore, we used two SARS-CoV-2 sequences (NC_045512.2 and MT007544.1) retrieved from NCBI to predict the epitopes and found that 14 epitopes shared identity with these two SARS-CoV-2 sequences, which include two major types of SARS-CoV-2 (Designated L and S). Thus, the 14-epitope combination could be feasible in vaccines for SARS-CoV-2, including SARS-CoV-2 types L and S.

Moreover, 13 of the 14 epitopes we selected in the current study could also serve as candidate epitopes for the SARS coronavirus and Bat SARS-like coronavirus (MG772934.1, KT444582.1, KT444582.1, and FJ882944.1) due to identical sequences between the epitopes and coronaviruses; additionally, these 13 epitopes could cover 84.94% of individuals worldwide. However, the 13 combination epitopes did not cover MERS (MG987420.1 and MG021451.1) due to the different subgenus (MERS belongs to Merbecovirus and SARS-CoV-2 belongs to Sarbecovirus) with significant differences in the sequences between SARS-CoV-2 (NC_045512.2) and MERS (MG987420.1 and MG021451.1). The identity of the N and S proteins was 50.0% and 35.0%, respectively, between SARS-CoV-2 (NC_045512.2) and MERS (MG987420.1 and MG021451.1). However, it was interesting that one epitope (KWPWYIWLGF_1211-1220) had the same sequences between SARS-CoV-2 and MERS (NC_045512.2, MT007544.1, and MG987420.1, MG021451.1). This epitope (KWPWYIWLGF_1211-1220) presented by HLA-A*23:01 and HLA-A*24:02 could cover 26.18% of individuals globally. In 2020, Ibrahim and Kafi also predicted the epitope KWPWYIWLGF_1211-1220 because of the high scores that indicate high efficiency due to the prediction of a quantity proportional to the amount of peptide presented by MHC molecules on the cell surface.²⁴ No epitope completely matched with other four human coronaviruses (HCoV-OC43, HCoV-HKU1, HCoV-229E, and HCoV-NL63), because they belong to the different subgenus or different genus from SARS-CoV-2. HCoV-OC43 (NC_006213.1) and HCoV-HKU1 (NC_006577.2) belong to Embecovirus. HCoV-229E (KY983587.1) and HCoV-NL63 (NC_005831.2) belong to Alphacoronavirus. Thus, the 14-epitope combination could serve as vaccine candidate epitopes for SARS-CoV-2, Bat SARS-like coronavirus, SARS-like coronavirus, and MERS.

Our results indicated the possibility of using candidate CTL epitopes to produce vaccines to effectively control SARS-CoV-2 infections and development. In the current study, we obtained the 14-epitope combination based on the distribution of HLA-A, -B, and -C could cover 89.60% of individuals globally and overcome the limitation of HLA specificity. However, the epitopes we selected were only predicted with respect to binding ability between epitopes and specific MHC molecules based on the total scores as the cut off value for epitope selection. Thus, there were epitopes being ignored because the total scores were below the cut off value (0.5 for N protein, 1.07 for S protein). For example, there are no HLA-A*30:01 predicted epitopes in the entire sequence of the SARS-CoV-2 S protein in our results because all total scores were ≤0.6 with all 8,841 peptides for HLA-A*30:01. Moreover, in 2004, Wang et al. demonstrated that SARS-CoV protein-derived peptide-1 (RLNEVAKNL_1167-1175) could induce peptide-specific CTLs both in vivo and in vitro.²¹ In the current study, we have predicted this epitope using epitope prediction analysis; however, its total score was ≤-0.54. So, this epitope was not selected in the current study. Thus, the candidate epitopes which were predicted to reduce infection effects in the current study should also be demonstrated using peptide-sensitized peripheral blood mononuclear cells or isolated CD8⁺ CTL responses in vivo or in vitro level and animal model in the future. Our results will aid in exploring the possible use of these epitopes for the vaccine against SARS-CoV-2 infections.

Author contributions

Conceived and designed the experiments: Li Shi and Yufeng Yao. Performed the HLA data analysis: Sun Ming, Shuying Dai, Le Sun. Performed the immunoinformatic analysis: Yina Cun, Chuanyin Li, Lei Shi. Wrote the paper: Li Shi and Yufeng Yao.

Additional information

Funding

This work was supported by grant from Special Funds for high-level health talents of Yunnan Province (D-201669 and L-201615) and the Yunnan Provincial Science and Technology Department (2019HC0060). The funders had no role in the design of the study, data collection and analysis, decision to publish, or preparation of the manuscript.