key: cord-0793741-wx1v0h0q
authors: Singh, Abhishek; Thakur, Mukesh; Sharma, Lalit Kumar; Chandra, Kailash
title: Designing a multi-epitope peptide-based vaccine against SARS-CoV-2
date: 2020-04-15
journal: bioRxiv
DOI: 10.1101/2020.04.15.040618
sha: 546acd5b1cd1d57f50340e9987e9fde2ada459fe
doc_id: 793741
cord_uid: wx1v0h0q

COVID-19 pandemic has resulted so far 14,395,16 confirmed cases with 85,711 deaths from the 212 countries, or territories. Due to multifacet issues and challenges in implementation of the safety & preventive measures, inconsistent coordination between societies-governments and most importanly lack of specific vaccine to SARS-CoV-2, the spread of Wuhan originated virus is still uprising after taking a heavy toll on human life. In the present study, we mapped several immunogenic epitopes (B-cell, T-cell, and IFN-gamma) over the entire structural proteins of SARS-CoV-2 and by applying various computational and immunoinformatics approaches, we designed a multi-epitope peptide based vaccine that predicted high immunogenic response in the largest proportion of world’s human population. To ensure high expression of the recombinant vaccine in E. coli, codon optimization and in-silico cloning were also carried out. The designed vaccine with high molecular affinity to TLR3 and TLR4, was found capable to initiate effective innate and adaptive immune response. The immune simulation also suggested uprising high levels of both B-cell and T-cell mediated immunity which on subsequent exposure cleared antigen from the system. The proposed vaccine found promising by yielding desired results and hence, should be tested by practical experimentations for its functioning and efficacy to neutralize SARS-CoV-2.

by consensus method of Immune Epitope Database (IEDB) tool [34] . The identified epitopes 123 were scrutinized based on the consensus score i.e. ≤2 and only those epitopes were considered as 124 strong binders and selected for further analysis which were predicted by more than one allele of 125 both the method used.

Helper T-cell (HTL) epitopes 127 Fifteen residue long HTL epitopes recognized by HLA Class II DRB1 alleles were predicted 128 using Net MHC II pan 3.2 server [35] . The thresholds for strong binder and weak binder were set 129 as default. Further another set of HTL epitopes of 15 residues length recognized by HLA DR 130 alleles were identified using the consensus method as implemented in the IEDB server [34] The 131 epitopes with a percentile rank of ≤2 and predicted by more than one allele of both the methods 132 were considered as a strong binder and scrutinized for further analysis. 134 Considering the fact that epitopes with affinity for multiple HLA alleles tend to induce relatively 135 more immune response in the host cell, we scrutinized overlapping epitopes with affinity to both 136 HLA class I and class II alleles for immunogenicity and allergen prediction using VaxiJen v2.0 were predicted using Ellipro server [38] . The prediction parameters like Minimum score and 146 Maximum distance (Anstrom) were set as 0.8 and 6, respectively. 148 Since, IFN gamma is the signature cytokine of both the innate and adaptive immune system and 149 epitopes with potency to induce IFN gamma could boost the immunogenic capacity of any 150 vaccine. Thus, IFN gamma epitopes were predicted for each structural protein by the IFNepitope 151 server [39] .

Predicted T-cell and B-cell epitopes were submitted to the IEDB conservancy analys is tool [40] 155 to identify the degree of conservancy in the structural protein sequences and the epitopes with 156 100% conservancy were selected for further analysis.

Population coverage and autoimmunity identification 158 Immunogenetic response of human population towards the selected overlapping CTL and HTL 159 epitopes against their respective HLA genotype frequencies was predicted using IEDB 160 population coverage analysis tool [41] . Those epitopes exhibiting 50% or more population 161 coverage were scrutinized and to red uce the probability of autoimmunity, all the selected 162 epitopes were subjected to BlastP search analysis against the Human proteome. Epitopes

showing similarity to any human protein were excluded from the further analysis.

Inte raction analysis of epitopes and their HLA alleles 165 The sequence of scrutinized epitopes was submitted to an online server PEPFOLD 3 [42] for 3D 166 structure modeling. The structure of the most common HLA alleles i.e. HLA-DRB1 *01:01 167 (HLA class II) and HLA-A* 02:01 (HLA class I) in human population were retrieved from 168 Protein data bank (PDB) with a PDB ID of 2g9h and 1QEW. Any ligand associated with the 169 HLA allele's structure was removed and energyminimization was carried out. Thereafter, the 170 modeled epitopes were docked with the corresponding HLA allele using a web server Patchdock

[43], to study the interaction pattern of receptor and ligand. The parameters like clustering 172 RMSD value and the complex type was kept as 0.4 and default. The result obtained from 173 Patchdock was forwarded to the Firedock server [44] for the refinement of the best 10 models.

Based on Global energy value, the complex with the lowest global energy was scrutinized and 175 the corresponding CTL and HTL epitopes were selected for the final vaccine construct.

Construction and quality control of final multi-epitope vaccine 177 For the construction of a multi-epitope vaccine against SARS-CoV-2, we followed Chauhan et 178 al.

[23] with some modifications. The epitopes were finally selected based on meeting the 179 following criteria: 1). epitopes must be immunogenic, non-allergic and overlapping with affinity 180 to both HLA class I and class II alleles; 2).epitopes must be capable to activate both CTL and 181 HTL cells and must have a minimum 50% of the population coverage; 3).epitopes should not be 182 overlapping with any human gene and the predicted B-Cell epitopes should overlap with 183 finalized CTL and HTL epitopes and should be present on the surface of the target protein.

construct of the multi-epitope vaccine. The HTL and IFN gamma epitopes were linked by 186 GPGPG linkers, whereas the CTL epitopes were linked by AAY linkers [23] . An adjuvant, 50S 187 ribosomal protein L7/L12 (Locus RL7_MYCTU) with an NCBI accessio n no: P9WHE3 was 188 added to the N terminal to enhance the immunogenicity of the constructed vaccine. The refined finalized 3D structures obtained from Raptor-X was better than the models generated 264 from I Tasser and Phyre2 as Raptor-X models showed least percentage of outliers and therefore 

The CTL epitopes overlapping with HTL epitopes were scrutinized and subjected to 275 immunogenicity and allergenicity prediction. In total, 16 CTL overlapped epitopes were 276 predicted in Envelop protein, 8 in Nucleocapsid protein, 25 in Surface protein and 9 epitopes in 277 Membrane protein (Table S4 -S11). All these selected epitopes were predicted to be non-allergic 278 with high antigenic scores. 314 We selected seven CTL, seven HTL and four IFN gamma epitopes for construction of the multi-315 epitope vaccine (figure S1). The adjuvant was coupled by EAAAK linker with CTL epitope and 316 subsequently, AAY linker was used to couple CTL epitopes and GPGPG linker was used to (Table S16) . with their corresponding cluster scores (Table S17-S18). Among these models, the model Figure. 6. In silico simulation of immune response using vaccine as an antigen after subsequent three injections. 1). Antigen and Immunoglobins. 2). B-cell population. 3). B-cell population per state. 4). Cytotoxic T-cell population. 5). Cytotoxic T-cell population per state. 6). Helper T-cell population. 7). Macrophages population per state. 8). Dendritic cell population per state. 9). Cytokine production. 

World Coronavirus disease 2019 (COVID-19) Situation Report -77, 9 th

A Genomic Perspective on the Origin and Emergence of 448 SARS-CoV-2

Organization Coronavirus disease (COVID-19) outbreak. Available

Stereochemistry and validation of macromolecular structures