Computational construction of a glycoprotein multi-epitope subunit vaccine candidate for old and new South-African SARS-CoV-2 virus strains

Abstract

The discovery of a new SARS-CoV-2 virus strain in South Africa presents a major public health threat, therefore contributing to increased infections and transmission rates during the second wave of the global pandemic. This study lays the groundwork for the development of a novel subunit vaccine candidate from the circulating strains of South African SARS-CoV-2 and provides an understanding of the molecular epidemiological trend of the circulating strains. A total of 475 whole-genome nucleotide sequences from South Africa submitted between December 1, 2020 and February 15, 2021 available at the GISAID database were retrieved based on its size, coverage level and hosts. To obtain the distribution of the clades and lineages of South African SARS-CoV-2 circulating strains, the metadata of the sequence retrieved were subjected to an epidemiological analysis. There was a prediction of the cytotoxic T lymphocytes (CTL), Helper T cells (HTL) and B-cell epitopes. Furthermore, there was allergenicity, antigenicity and toxicity predictions on the epitopes. The analysis of the physicochemical properties of the vaccine construct was performed; the secondary structure, tertiary structure and B-cell 3D conformational structure of the vaccine construct were predicted. Also, molecular binding simulations and dynamics simulations were adopted in the prediction of the vaccine construct's stability and binding affinity with TLRs. Result obtained from the metadata analysis indicated lineage B.1.351 to be in higher circulation among various circulating strains of SARS-CoV-2 in South Africa and GH has the highest number of circulating clades. The construct of the novel vaccine was antigenic, non-allergenic and non-toxic. The Instability index (II) score and aliphatic index were estimated as 41.74 and 78.72 respectively. The computed half-life in mammalian reticulocytes was 4.4 h in vitro, for yeast and in E. coli was >20 h and >10 h in vivo respectively. The grand average of hydropathicity (GRAVY) score is estimated to be -0.129, signifying the hydrophilic nature of the protein. The molecular docking indicates that the vaccine construct has a high binding affinity towards the TLRs with TLR 3 having the highest binding energy (-1203.2 kcal/mol) and TLR 9 with the lowest (-1559.5 kcal/mol). These results show that the vaccine construct is promising and should be evaluated using animal model.

Keywords: Bioinformatics; COVID-19; Computational construction; Control; SARS-CoV-2 virus; South Africa; Vaccine candidate.

Fig. 1a

Result showing the Lineages of South African Isolates. This result shows the lineage distribution of the SARs-CoV-2 South African isolates. This is Fig.1a legend.

Fig. 1b

Results showing the Clades of South African Isolates. This result shows the clades of the SARs-CoV-2 South African isolates. This is Fig.1b legend.

Fig. 2

Vaccine construct. This figure shows the multi-epitope vaccine construction.

Fig. 3

Secondary structure Prediction of the vaccine construct; (a) indicating the score of alpha-helix, extended strand, random coil, and beta structure and (b) OMPL predictions of the secondary structure depicted by different colors. Blue color represents alpha-helix, and green color represents Beta strands, red color represents extended strand, and yellow color represents random coil.

Fig. 4a

Showing predicted distribution for each residue as arrow goes towards each residue.

Fig. 4b

The conformational B-cell epitopes of the final vaccine construct, grey sticks represent the bulk of the polyprotein, and a yellow surface represents the conformational B cell epitopes.

Fig. 5

Showing the refined tertiary structure showing predicted distribution for each residue as arrow goes towards each residue.

Fig. 6

Showing (a) Z-score graph and (b) Ramachandran plot.

Fig. 7

Disulfide engineering showing disulfide adherence by disulfide by design 2.

Fig. 8

Showing Molecular docking of the construct with TLR: (a) Docked complexes for vaccine-TLR2 complex with vaccine colored blue and TLR2 colored red.(b) Docked complexes for vaccine-TLR3 complex with vaccine colored blue and TLR3 colored red. (c) Docked complexes for vaccine-TLR4 complex with vaccine colored blue and TLR4 colored red. (d) Docked complexes for vaccine-TLR9 complex with vaccine colored blue and TLR9 colored red.

Fig. 9

Spin prediction result of the Molecular energy simulation: (a) Vaccine–TLR2 interaction spin prediction. (b) Vaccine–TLR3 interaction spin prediction. (c) Vaccine–TLR4 interaction spin prediction (d) Vaccine– TLR9 interaction spin prediction.

Fig. 10

Molecular energy simulation Eigenvalue result: (a) The Vaccine–TLR2 interaction's Eigenvalue. (b) The Vaccine – TLR3 interaction's Eigenvalue. (c) The Vaccine–TLR4 interaction's Eigenvalue. (d) TheVaccine–TLR9 interaction's Eigenvalue.

Fig. 11

Molecular energy simulation deformability B-factor result: (a) deformability B-factor region of the vaccine–TLR2 interaction. (b)Deformability B-factor region of the vaccine–TLR3 interaction. (c) Deformability B-factor region of the vaccine–TLR4 interaction. (d) Deformability B-factor region of the vaccine–TLR9 interaction.

Fig. 12

Molecular energy simulation variance result: (a) the vaccine–TLR2 interaction's variance. (b) the vaccine–TLR3 interaction's variance. (c) the vaccine–TLR4 interaction's variance. (d) the vaccine–TLR9 interaction's variance.

Fig. 13

Molecular dynamics simulation elastic network result: (a) the Vaccine–TLR2 interaction's elastic network. (b) the vaccine–TLR3 interaction's elastic network (c) the vaccine–TLR4 interaction's elastic network. (d) the vaccine– TLR9 interaction's elastic network.

Fig. 14

Molecular dynamics simulation mobility B-factor result: (a) the vaccine–TLR2 interaction's mobility B-factor's result. (b) the vaccine–TLR3 interaction's mobility B-factor's result. (c) the vaccine–TLR4 interaction's mobility B-factor's result. (d) the vaccine–TLR9 interaction's mobility B-factor's result.

Fig. 15

Molecular energy simulation residue index result: (a) the Vaccine–TLR2 interaction's residue index. (b) the vaccine–TLR3 interaction's residue index (c) the vaccine–TLR4 interaction's residue index (d) the vaccine– TLR9 interaction's residue index.

Fig. 16

C-ImmSim presentation of computational immune simulation of the projected vaccine peptide. (a) The production of Immunoglobulin in response to antigen injection; certain subclasses are depicted as colored peaks. (b) B-cell population evolution after the injection. (c) T-helper cells' population after injection. (d) T-helper cells evolution (e) Natural Killer cells' population after injection and depiction of TR (Regulatory) cell population per state (f) Depiction of TH cell population (g) Depiction of TH cell population per state (h) Depiction of TC cell population (cells per mm³) (i) depicts ng/nl versus days plot (j) depicts MA population cells per state (cells per mm³) (k) depicts NK cell population (cells per mm³) (l) depicts PLB cell population (cells per mm³) (m) depicts DC population per state (cells per mm³) (n) depicts EP population per state (cells per mm³).

Fig. 17

(a) Adaptation of codon and computational cloning: a graph depicting sequence adaptation (See Fig. 17a in the Supplementary material), (b) computational cloning for the adapted vaccine sequence into pET28A (+) vector to ensure the expression of the vaccine protein in an E. coli system.