Assessing the evolution of SARS-CoV-2 lineages and the dynamic associations between nucleotide variations

Abstract

Despite seminal advances towards understanding the infection mechanism of SARS-CoV-2 (severe acute respiratory syndrome coronavirus 2), it continues to cause significant morbidity and mortality worldwide. Though mass immunization programmes have been implemented in several countries, the viral transmission cycle has shown a continuous progression in the form of multiple waves. A constant change in the frequencies of dominant viral lineages, arising from the accumulation of nucleotide variations (NVs) through favourable selection, is understandably expected to be a major determinant of disease severity and possible vaccine escape. Indeed, worldwide efforts have been initiated to identify specific virus lineage(s) and/or NVs that may cause a severe clinical presentation or facilitate vaccination breakthrough. Since host genetics is expected to play a major role in shaping virus evolution, it is imperative to study the role of genome-wide SARS-CoV-2 NVs across various populations. In the current study, we analysed the whole genome sequence of 3543 SARS-CoV-2-infected samples obtained from the state of Telangana, India (including 210 from our previous study), collected over an extended period from April 2020 to October 2021. We present a unique perspective on the evolution of prevalent virus lineages and NVs during this period. We also highlight the presence of specific NVs likely to be associated favourably with samples classified as vaccination breakthroughs. Finally, we report genome-wide intra-host variations at novel genomic positions. The results presented here provide critical insights into virus evolution over an extended period and pave the way to rigorously investigate the role of specific NVs in vaccination breakthroughs.

Keywords: COVID-19; SARS-CoV-2 genome evolution; mutational co-occurrence; pathogenesis; vaccination breakthrough.

Fig. 1.

Comparative timeline of major SARS-CoV-2 lineage distribution in Telangana, India. (a) Frequency distribution of the major lineages per month from March 2021 (a complete timeline starting from April 2020 is provided in Fig. S1a). The black and red lines show the frequency distribution trend of Delta and Delta plus its sub-lineages in India (excluding Telangana), respectively; white points show the corresponding monthly frequency. (b) Timeline of Delta variant distribution (blue) and total (grey) cases in India and rest of the world (all data were obtained from GISAID). (c) Timeline of changing frequency of Delta sub-lineages across the world. Only those sub-lineages present in >5 % of total sequences submitted from the region were included. In (b) and (c), while estimating frequencies for the Asian countries, Indian sequences were excluded.

Fig. 2.

Monthly timeline of SARS-CoV-2 NV frequencies; variants present in <3 % of samples were excluded. Key domains in the S protein (first panel) are indicated at the top: NTD, N-terminal domain (brown); RBD, receptor binding domain (green); S2 subunit containing SD1 (subdomain 1), SD2 (subdomain 2) and S1/S2 cleavage sites (orange); CTD, C-terminal domain (black). NV frequencies for N protein (second panel), nsp3 (third panel), and ORF7a and ORF7b (bottom panels) are also shown.

Fig. 3.

NV frequency in vaccinated cases and pairwise cross-correlation analysis. (a) Frequency of S protein amino acid alterations present in >3 % of completely (top, calculated for a total of 152 samples) and partially (bottom, total of 137 samples) vaccinated cases. (b) Odds ratio indicating the extent of association of specific variants with vaccination breakthrough cases. (c) Pairwise cross-correlation plot between all non-synonymous missense NVs present in >3 % of all the SARS-CoV-2 samples identified from Telangana, India, during April 2020 to October, 2021; the size of all coloured ‘squares’ is inversely proportional to the corresponding P-value of the correlation. Positions with P>0.05 appear as blank (or white). The colour key for positive (blue) and negative (red) correlation is given below the plot.

Fig. 4.

SARS-CoV-2 iSNVs identified from Telangana, India. (a) Distribution of number of iSNVs in samples. (b) Distribution of genomic loci with shared and unique iSNVs and NVs. (c) Timeline of allele frequency changes in the minor alleles in ORF1ab, ORF3a, N and S proteins. Box plots indicate allelic frequency distribution while points represent samples in which the allele was identified. Only those alternate alleles whose allele frequencies were either consistent or increased during the indicated timeline are shown.