Immunoinformatic Responses of Host MHC I/II and CD4+ Cells Against Conserved Spike Fragments of SARS CoV-2 of 186 Countries With In-Silico Mutations: Implications in Universal Vaccination

Different types of quickly-developed vaccines have been introduced against Covid-19 with largely inconclusive results. On the course of time, somewhere Covid-19 declines its infection/mortality rate and in some countries it revived with some new mutant-variants. Considering the large number of global variants, in the current study several conserved (186 countries) sequences (checked by ClustalX2) epitopic regions (by SVMTriP and IEDB) and in-silico mutants of SARS CoV-2 spike protein fragments (Cut 1-4) were screened for their stability against proteases, antigenicity (VaxiJen V2.0 and for glycosylation effects NetOGlyc-NetNGlyc), MHCI/II reactivity (IEDB TOOLS) and CD4+ cellular responses utilizing Haddock 2.4 and PatchDock programme. The cut4 mutant and its peptide SRLFRKSNLKPFERD showed highest combined-score 48.23548 and Immunogenicity-Score of 92.0887. The core-sequence SRLFRKSNL showed highest Median Percentile Rank (7 HLA-allele) of 19. CD4+ immunogenicity also conrms representation of CUT4TM2 epitope SRLFRKSNL by MHC Class II. The epitope YNYKYRLFR from CUT4 showed IC50 value of ~30 nM with allele HLA-DRB1*11:01 and HLA-DRB5*01:01. Binding anity and RMSD score suggest that different epitopic regions of Cut4 mutants interacted MHC II with large number of H-bonds. Cut4 double mutants strongly interacted with exposed T-cell surface and facilitated by its receptors. According to the antigenicity analysis, epitopes 3, 4, and 5 were the potent antigens with antigenicity values of 0.6268, 1.2404, and 0.4639, respectively. Screening of conserved SARS CoV-2 spike fragments globally may help to nd most stable antigenic determinant and further their mutation-engendered counterparts have better immunogenic responses. Further studies are necessary to develop global vaccination strategies against Covid-19. ligand The stability of CUT4 and its different mutants in physiological level has been studied through proteasomal cleavage site analysis, N terminal and C-terminal glycosylation analysis and half life prediction in blood. There was no proteasomal cleavage sites found in CUT4 before and after mutation according to IEDB analysis. Whereas, according to neural network based program, NetCTL, predicted that there were some one Proteasomal processing site but the score of 0.1984 stands insignicant. So, the peptides may remain stable in physiological condition. All the un-mutated and mutated CUT4 showed no glycosylation at both end (N and C terminal), except TM2 with C terminal glycosylation. These indicated that, these peptides were effectively bound with T cell receptors, and enhance further immunological reactions as they were not glycosylated.


Introduction
At the end of 2020, at least 30 vaccines are in the process of clinical trials. Some of those like Oxford's AZD1222/AstraZeneca/Oxford's AZD1222/Moderna's mRNA-1273 and Sinovac's CoronaVac vaccines reported with inconclusive results [1]. Some of the countries have introduced self-generated vaccines to their populations. Covishield (Oxford-AstraZeneca) and Covaxin (produced by Bharat biotech) from India origin have been introduced to a signi cant number of populations (ref https://www.bbc.com/news/world-asia-india-55748124). Those have some diverse outcome, criticism and no de nite clinical-trial outcome. In natural way, the infection and the case fatality rate (CFR) are declining but ample evidence of multiple mutation and highly evolving SARS CoV-2 urges in the development of more effective vaccine effective against global variants. In some country with largest populations induction of herd immunity by mass vaccination has been effective in the prevention of these types of diseases [2]. The autoimmune suppressive effects of this virus prevent or delay in the immunotherapy processes. In multiple cases vaccination or other immmuno-therapy cannot effect at ease. Negative outcome was also noticed during drug-induced inhibition of some subsets of B-cell hinders the initiation of innate and CD8 + cell responses [3]. None of the COVID-19 vaccines is fully effective with safety to get priority at this point of time. Only rigorous clinical trial and post application follow-up can recognize serious health hazards [4].
Report suggests that ChAdOx1 nCoV-19 has an acceptable safety pro le against symptomatic COVID-19 [5]. But that statistical outcome generated from a small sample size. The candidate vaccine mRNA-1273 encodes the stabilized prefusion SARS-CoV-2 spike protein and generated some immune responses [6].
Not only the generation of speci c subsets of IgG but the MHC, T and B cell responses have also been important for the optimized immune responses.
Th1-targeted antibody, CD8 + and CD4 + T cell responses and neutralizing IgG antibody responses were noticed with S protein-speci c viral neutralization [7]. This suggested T and B cell mediated acquired immunity is one of the main natural immune therapy against Covid-19. However very few study detected MHC class recognition of spike protein epitope. Our earlier study has indicated some spike epitope capable in the generation of MHC responses [8]. In contrary, it is also evident that the antibodies may suppress viral replication through neutralization but might also augment SARS CoV-2 pathogenesis through a process termed antibody-dependent enhancement [9]. Many of the studies elicited vaccine e cacy in polyfunctional S-speci c (some time exposed amino acids of S proteins) CD4+ (mainly Th1) and CD8 + T-cell responses, with a T effector memory phenotype. In some other strategy, DNA/MVA immunizations elicited higher T-cell responses [10].
The current study was aimed to nd the role of small cut fragment of the globally conserved SARS Cov-2 Spike RBD and its in-silico mutants in the generation of the higher antigenicity. Further, antigen presentation by HLA and T-cell responses were tested to con rm the antibody-mediated immunogenic responses. Present results are signi cant to develop a potent and universal vaccination strategy against the Covid-19. Further studies are necessary.

Materials And Methods
Structure retrieval, analysis and prediction Structure of different CUT4 and mutated peptide sequences were taken from our previous analysis. The binding sites of these CUTs with ACE2 Receptor Binding Domain (RBD) were analyzed using PyMol [11].
For immunoinformatics study, T cell receptor, MHC I and MHC II structures were downloaded from RCSB PDB [12]. Different mutated CUTs were analyzed for T cell receptor interaction analysis using PDB ID 1kgc, 2ak4, 3kxf and 4jrx. For different mutated CUTs interaction with MHC class I, PDB ID 1i4f was used.
Different epitopes from mutated CUTs were also analyzed for MHC class I binding using the PDB structures of 6iex, 3x13, 4lnr, 6j1w and 6vb3. Different mutated CUTs interaction with MHC class II was analyzed with PDB ID 4p5m and 3c5j. Different mutated CUTs epitopes were analyzed with 1jk8, 3pdo, 3lqz, 1i4f, 5ni9 and 1BX2 MHC class II structures.

Docking studies
Protein-Protein docking studies were performed using Haddock 2.4 [13] and PatchDock [14] server to check the binding a nity with T cell receptor, MHC class I and MHC class II molecules. The results were analyzed using Pymol [11] and the CUTs which had the best binding a nity were chosen. The best mutated structures from the docking result were accepted. The interaction study of the accepted unmutated cut and the predicted epitopes with best interacting T cell receptor, MHC I and MHC II according to IEDB tools was done using HADDOCK 2.4 [13].

Epitope study
Epitope prediction using SVMTRIP [15] was done and the probable epitopes were taken for the nalized structure. All the probable epitopes were analyzed using all the IEDB Tools [16]. The T cell Epitope prediction and analysis were performed using IEDB tools [8]. Epitope analysis using SVMTRIP [15] and analyzing results using all IEDB tools were performed. Antigenicity was checked using VaxiJen V2.0 [17] keeping the threshold value as 0.4. Glycosylation study using NetOGlyc and NetNGlyc [18] was performed.

Global mutation study on SARS CoV-2 spike protein
Initially, 186 gene sequences of nCOV2 spike glycoproteins from different countries were retrieved from GISAID database (https://www.gisaid.org/). Then all the sequences were subjected to multiple sequence alignment for the identi cation of unique sequences using ClustalX2. From all the sequences, 103 sequences were selected with unique sequences and converted to protein sequences using SMS Sequence Manipulation Suite (bioinformatics.org/sms2/translate.html) online server. Among them 24 unique protein sequences were selected for further analysis. Among them two sequences were selected having sequential difference at the Receptor Binding Domain (RBD) in comparison to CUT4. Finally, sequence of CUT4, Model 1 and Model 15 were selected for epitope analysis. Model 1 represented the RBD sequences of Estonia, Latvia, Hong Kong, Costarica, Iran, Mexico, Mongolia, Japan, Italy, Egypt, Ireland, Denmark, Germany, France, India, DRC, Serbia, Pakistan, England and Wuhan (wild type). Whereas, the sequences of Model 15 represented Finland. As per the Epitope analysis and antigenicity of epitope, structure of selected sequences (Model 1 and Model15) was individually aligned with CUT4 and the respective epitope sequences and locations were analyzed using Pymol [11].

Results And Discussions
Different peptide CUT analysis According to our previous study a peptide with an amino acid length of 84 (cut 4), has been selected for mutation analysis (data submitted elsewhere). Among 11 different mutations (4 Single mutations, 5 double mutations and 2 triple mutations) three were selected for their antigenicity and other immunereactivity testing. The CUT4 peptide was shown to competitively inhibit the nCOV2-ACE2 attachment at the RB domain (data submitted elsewhere). These peptide CUT4 mutations were further studied for T-Cell reorganization, MHC class I and II attachment, protease cutting-sites, glycosylation effects stability for stable antigenic responses and that was compared with the corresponding mutants.

Class I and Class II immunogenicity analysis
According to the results, CUT4, CUT4SM3 and CUT4DM2 showed similar result in Class I immunogenicity (S Table 1). Though the length of the peptide remain same, score value differed due to single, double and triple mutations. Highest score of 0.27813 was found for CUT4 unmutated, CUT4SM3 also showed positive value of 0.07923. Whereas, score − 0.07827 was found for both CUT4DM2 and CUT4TM2 fragments.
For Class II analysis, single and double mutation did not affect the CD4 immunogenicity property.
Whereas, CUT4TM2 showed better result than others (S Table 2). The peptide sequence SRLFRKSNLKPFERD showed highest combined score of 48.23548, Immunogenicity Score of 92.0887. Interaction analysis of CUT4 and other mutants with MHC class I and II Different segments were analyzed for stabilized interaction with different alleles of MHC class II. All the data were selected on the basis of percentile rank below 10% and IC50 value below 50nM. Here stabilization matrix alignment method or SMM-align method has been used for IC50 value calculation (S Table 3). Only the lower IC50 values were represented in S Table 3, rest of the data not shown. The lower IC50 value indicated the higher chance to form stabilized structures with MHC II alleles. Here, epitope YNYKYRLFR from CUT4 showed IC50 value of 30 nM and 28 nM with allele HLA-DRB1*11:01 and HLA-DRB5*01:01 respectively. Similar results were found with CUT4DM2. For the same epitope, CUT4SM3 showed IC50 value of 28 nM using only allele HLA-DRB5*01:01. But CUT4TM2 showed higher value at the range of 44-47 nM with the same allele HLA-DRB5*01:01. The unique epitope ATSTGNYNY in CUT4SM3 showed lower IC50 value of 30 nM with HLA-DRB5*01:01. Whereas neural network based IC50 analysis, alignment represented that CUT4 and CUT4DM2 had the lowest and similar IC50 values with epitope sequences YRLFRKSNL, YNYKYRLFR and YRLFRKSNL. Sturniolo method is one of the best methods to predict the best t between epitope and MHC Class II alleles. Sturniolo score also represented CUT4 and CUT4DM2 with the similar values where CUT4SM3 showed lower values. According to as all of them were less than 50nM. So, from the above analysis it could be predicted that, just like CUT4 unmutated different epitopes were equally capable to form a stabilized structure with different alleles of MHC Class II. For MHC Class I analysis, no signi cant changes were observed (S Table 4). So, Class II type of interaction and representation was observed for our epitope e ciency analysis. Molecular docking analysis, CUT4SM3 and CUT4DM2 were found to interact with MHC Class I allele 1i4f with the docking score of -258.83 and − 267.99, and ligand RMSD of 120.06 and 103.43 respectively ( Fig. 1). At those positions CUTs were found to block the MHC Class I peptide presenting site, but no effective bond formation were observed. Whereas, CUT4TM2 was found to interact with 1i4f with the amino acids involvement in H-bond were, GLN498 of CUT4TM2 with GLY16 1i4f and GLN 493 of CUT4TM2 with GLU19 of 1i4f. The docking score and ligand RMSD of that interaction were − 254.36 and 169.46 respectively. Whenever, the CUTs were analyzed for epitope, 6 fragments were found to interact with different alleles of MHC Class I. Based on the IEDB results, all the fragments were docked with their respective alleles as represented in Fig. 1. Where, effective H-bonding was observed for epitope segment 4 with a sequence of TRNIDATSTG. It interacted with HLA-A*30:01 (PDB ID: 6J1W) with a score of 0.008784 and rank 9. SER445 of CUT4TM2 formed H-bond with TYR7 of 6J1W. Docking score was − 164.37 and Ligand RMSD was 193.75. Rest of the epitope segments were found exactly at the epitope presenting domain of respective MHC Class I alleles but no H-bonding observed.
Similarly, molecular docking with MHC class II were also performed CUT4SM3 and CUT4DM2 were found at the epitope presenting domain of MHC class II allele 4P5M without any H-bonding (Fig. 2). Whereas, CUT4TM2 formed H-bond through SER14 to ARG76 of MHCII (PDB ID: 3c5j). But more a nity was observed for different epitopes as presented in Fig. 1 T cell receptor binding analysis of CUT4 and other mutated CUTs.
All the CUTs, SM3, DM2 and TM2 showed remarkable a nity to T cell receptors. Several strong bonds as well as H-bonds were formed as presented in Fig. 3. SM3 showed interactions with two different T cell receptors like 1kgc and 4jrx. H-binding pattern and surface attachment pattern has been presented in Fig. 3a & b. Most of the surface area has been exposed for DM2 in the peptide presenting site of receptor 1kgc (Fig. 3c). TM2 showed highest representation by 4 different T cell receptors like 1kgc, 4jrx, 2ak4 and 3kxf. The representations of different CUTs were facilitated by T cell receptors.
Epitope pattern of Receptor Binding Domain (RBD) analysis within various nCOV2 spike proteins found in 186 countries worldwide (Fig. 4). According to CUT4, Model 1 represented the Estonia, Latvia, Hong Kong, Costarica, Iran, Mexico, Mongolia, Japan, Italy, Egypt, Ireland, Denmark, Germany, France, India, DRC, Serbia, Pakistan, England and Wuhan (wild type) and Model 15 represented the Finland. Though, sequential diversity was present among CUT4, Model1 and Model15, sequential similarity found among 4 epitopes out of 5. The structural alignment also showed similarity among all except epitope 4.
In silico analysis of CUT4 and other mutant stability at physiological level The stability of CUT4 and its different mutants in physiological level has been studied through proteasomal cleavage site analysis, N terminal and C-terminal glycosylation analysis and half life prediction in blood. There was no proteasomal cleavage sites found in CUT4 before and after mutation according to IEDB analysis. Whereas, according to neural network based program, NetCTL, predicted that there were some one Proteasomal processing site but the score of 0.1984 stands insigni cant. So, the peptides may remain stable in physiological condition. All the un-mutated and mutated CUT4 showed no glycosylation at both end (N and C terminal), except TM2 with C terminal glycosylation. These indicated that, these peptides were effectively bound with T cell receptors, and enhance further immunological reactions as they were not glycosylated.
Finally, the presence of CUT4 epitope was analyzed among different nCOV2 spike glycoproteins from 186 countries worldwide (Fig. 4). Though sequential diversity was present among different nCOV2 spike whole proteins, the Receptor Binding Domain (RBD) showed higher sequential similarity. As a result of which two sequences and their speci c predicted tertiary Models were selected as representatives of all 186 countries. Model 1 was representative of all countries except Finland. Model 15 was the representative of Finland. The 5 epitopes present within the CUT4 were also found within Model 1 and Model15. Epitope 1, 2, 3, and 5 were sequentially and structurally the same. The epitope 4 was found to be different. This result was also alike in 103 previously selected protein sequences (S Fig. 1.) According to the antigenicity analysis, epitope 3, 4, and 5 were the potent antigens with antigenicity values of 0.6268, 1.2404, and 0.4639. This epitope-based analysis could be applicable for worldwide varieties of nCOV2. The effect of mutation was re ected in antigenicity analysis (S Table 5). Comparative study among CUT4, SM3, DM2 and TM2 represented that, TM2 has the highest antigenicity of 0.6501. Whereas, CUT4, SM3 and DM2 have scores 0.6478, 0.6373 and 0.6197 respectively. On the other hand epitope mutation showed signi cant differences (S Table 5). In this study, both the wild and mutant type epitopes were found to be reactive with the CD4 + cells expected that cell mediated higher order of immune responses. SARS CoV-2 develops immunosuppression in autoimmunity, which leads to terminated and/or delayed immunotherapy. Although CD4 + T cells are the key player but it is ineffective until the naive Bcells upregulations. This is noticed in some severe patients to recover who have CD20-depleted state can overcome immunosuppression [19]. Virus mediated couple of vaccination strategies have been tested. Recombinant adenovirus vector has been used to carry SARS-CoV-2) spike glycoprotein and used as a heterologous vaccine that induced strong humoral and cellular immune responses [20]. Vaccine hypersensitivity reactions (VAH) are one of the demerits of this procedure. In some earlier cases of respiratory syndrome (SARS, MERS or respiratory syncytial virus) cases this adverse effect was noticed [21].
It is speculated that two human adenovirus-vector vaccines, three attenuated vaccines, and one peptide vaccine are in limited use and some more are under clinical trial but having no satisfactory results till date [22]. Reactivity against the global populations a vaccine is required which should comprise B-cell, CTL, HTL epitopes. The epitopes associated with maximum HLA alleles are the requirement of the time to ensure the viral neutralization. One report claims this outcome have been currently under trial [23]. CRISPR-mediated genome editing approach has also been used for anti-viral strategies but yet to take time to be decisive [24].
In the current study, the globally conserved spike fraction of SARS CoV-2 has been extensively studied for greater immunogenicity and further in silico several of its mutants were generated with more immunogenic epitopic fragments which can develop MHC multi HLA responses and T-cell CD4 + signals.
This is hypothesized that multi-epitopic combinations or cocktail can be used for global vaccination strategies. Further studies are necessary in this regard.

Declarations
Con ict of interests: none Funding: No speci c funding for this investigation

Consent for publication: Received
Availability of data and materials: All data are available upon reasonable requests. Figure 1 Interaction of different selected CUT4 mutant (SM3, DM2 and TM2) with MHC class I (1i4f). Interaction showed SM3 and DM2 at the nearby location of 1i4f Epitope presenting site but no interactions were observed (a, b). TM2 formed H-bond with 1i4f (c). Different Epitope from these selected mutants (d) showed that Epitope4 to interact with MHC class I (6j1w). It formed H-bond between TYR7 of Epitope4 and SER445 of MHC I.

Figure 2
Interaction of different selected CUT4 mutant (SM3, DM2 and TM2) with MHC class II (4P5M and 3C5J). Interaction showed SM3 and DM2 at the nearby location of 4P5M Epitope presenting site but no interactions were observed (a, b). TM2 formed H-bond with 3C5J (c). Different Epitope from these selected mutants (d) showed that Epitope1, 3 and 5 to interact with MHC class II (1jk8, 3lqz and 5ni9).  Epitope pattern of Receptor Binding Domain (RBD) analysis within various nCOV2 spike proteins found in 186 countries worldwide.