Within-host diversity of SARS-CoV-2 lineages and effect of vaccination

Abstract

Viral and host factors can shape SARS-CoV-2 within-host viral diversity and virus evolution. However, little is known about lineage-specific and vaccination-specific mutations that occur within individuals. Here we analysed deep sequencing data from 2,146 SARS-CoV-2 samples with different viral lineages to describe the patterns of within-host diversity in different conditions, including vaccine-breakthrough infections. Variant of Concern (VOC) Alpha, Delta, and Omicron samples were found to have higher within-host nucleotide diversity while being under weaker purifying selection at full genome level compared to non-VOC SARS-CoV-2 viruses. Breakthrough Delta and Omicron infections in Comirnaty and CoronaVac vaccinated individuals appeared to have higher within-host purifying selection at the full-genome and/or Spike gene levels. Vaccine-induced antibody or T cell responses did not appear to have significant impact on within-host SARS-CoV-2 evolution. Our findings suggest that vaccination does not increase SARS-CoV-2 protein sequence space and may not facilitate emergence of more viral variants.

Figure 1.. Statistics of within-host mutations in SARS-CoV-2 samples.

(A) Distribution of number of iSNV site(s) in each sample, colored by ranges of Ct values. The dashed line shows the mean value of the distribution. (B) Distribution of number of sample(s) sharing iSNVs (e.g., if the iSNV identified in one sample was not shared with any other sample, then the number of samples sharing that iSNV equals to one (x = 1), and so on), colored by variant types. (C) Distribution of the frequency of iSNVs per sample per synonymous and per non-synonymous site (d_S and d_N) for different types of mutations, colored by variant types. The dashed lines show the average frequency of synonymous and non-synonymous iSNVs among all types of mutations. The points and error bars show mean and standard deviation values based on 10,000 bootstrap replicates at mutation level. (D) Distribution of the frequency of iSNVs per sample per Kb for synonymous and non-synonymous site (d_S per Kb and d_N per Kb) in different genomic regions of 1Kb length, colored by variant types. The points and error bars show mean and standard deviation values based on 10,000 bootstrap replicates at mutation level. (E) Distribution of high-frequency mutations shared by multiple samples, colored by variant types. Coding regions of the SARS-CoV-2 genome, based on the reference genome (GenBank: MN908947.3), are shown at the bottom of the figure.

Figure 2.. Comparison of within-host mutation profiles between unvaccinated VOC and unvaccinated non-VOC samples.

(A-C) Full genome incidence of iSNVs (adjusted number of iSNVs per Kb), abundance of iSNVs (minor allele frequencies, MAF), and nucleotide diversity (π) of different samples. For all box plots, the bold horizontal line inside the box shows the median, the upper and lower edges of the box indicate the first and the third quartiles, and whiskers extend to span a 1.5 interquartile range from the edges. Pairwise comparisons between groups were performed by two-sided two-sample Wilcoxon tests; the pairs with P-value ≤ 0.01 and ≤ 0.05 are labelled with “**” and “*” respectively. (D-E) Full-genome and gene-specific within-host nonsynonymous nucleotide diversity (π_N) and synonymous nucleotide diversity (π_S) in samples from different groups. The points and error bars show the mean and standard deviation values under 10,000 bootstrap replicates at codon level. Significance was evaluated using Z-tests of the null hypothesis that π_N – π_S = 0 (10,000 bootstrap replicates, codon unit); P-value ≤ 0.01, ≤ 0.05 and ≤ 0.10 are labelled with “**”, “*” and “^”, respectively.

Figure 3.. Comparison of within-host mutation profiles between vaccinated and unvaccinated Delta and Omicron samples.

(A-C) Full-genome incidence of iSNVs (adjusted number of iSNVs per Kb), abundance of iSNVs (minor allele frequencies, MAF) and nucleotide diversity (π) of different samples. For all box plots, the bold horizontal line inside the box shows the median, the upper and lower edges of the box indicate the first and the third quartiles, and whiskers extend to span a 1.5 interquartile range from the edges. Pairwise comparisons between groups were performed by the two-sided two-sample Wilcoxon test; the pairs with P-value ≤ 0.01 and ≤ 0.05 are labelled with “**” and “*” respectively. (D-E) Full-genome and gene-specific within-host nonsynonymous nucleotide diversity (π_N) and synonymous nucleotide diversity (π_S) in samples from different groups. The points and error bars showed the mean and standard deviation values under 10,000 bootstrap replicates at codon level. Significance was evaluated using Z-tests of the null hypothesis that π_N – π_S = 0 (10,000 bootstrap replicates, codon unit); P-value ≤ 0.01, ≤ 0.05 and ≤ 0.10 were labelled with “**”, “*” and “^”, respectively.

Figure 4.. Spike mutations identified in unvaccinated and vaccinated Delta and Omicron samples.

Each circle represents one mutation identified in one sample in this study. (A) Identified within-host mutations in Omicron samples; (B) Identified within-host mutations in Delta samples.

Figure 5.. Overlapping known SARS-CoV-2 CD8+ and CD4+ T cell epitopes per mutation in unvaccinated and vaccinated Delta and Omicron samples.

(A) Analysis based on all known SARS-CoV-2 T cell epitopes. (B) Analysis based on T cell epitopes associated with HLA alleles prevalent in the Hong Kong population. Pairwise comparisons within groups were performed by the two-sided two-sample Wilcoxon test. For all box plots, the bold horizontal line inside the box shows the median, the upper and lower edges of the box indicate the first and the third quartiles, and whiskers extend to span a 1.5 interquartile range from the edges.