Severity of COVID-19 is inversely correlated with increased number counts of non-synonymous mutations in Tokyo (preprint)

ABSTRACT

Background: SARS-CoV-2 genome accumulates point mutations in a constant manner. Whether the accumulation of point mutations is correlated with milder manifestations of COVID-19 remains unknown. Methods: We performed SARS-CoV-2 genome sequencing in 90 patients with COVID-19 infection treated at a tertiary medical center in Tokyo between March and August 2020. The possible association between disease severity and viral haplotype was then assessed by counting the number of mutations in addition to performing phylogenic tree analysis, comparative amino acid sequence analysis among {beta}-coronaviruses, and mathematical prediction of the functional relevance of amino acid substitutions. Results: The number of non-synonymous mutations was inversely correlated with COVID-19 severity, as defined by requiring oxygen supplementation. Phylogenic tree analysis identified two predominant groups which were separated by a set of 6 single nucleotide substitutions, including four leading to non-synonymous amino acid substitutions. Among those four, Pro108Ser in 3 chymotrypsin-like protease (3CLpro) and Pro151Leu in nucleocapsid protein occurred at conserved locations and were predicted to be deleterious. Patients with Pro108Ser in 3CLpro and Pro151Leu in nucleocapsid protein had a lower odds ratio for developing hypoxia requiring supplemental oxygen (odds ratio of 0.24 [95% confidence interval of 0.07-0.88, P-value = 0.032]) after adjustments for age and sex, compared with patients lacking this haplotype in Clade 20B. Conclusion: Viral genome sequencing in 90 patients treated in the Tokyo Metropolitan area showed that the accumulation of point mutations, including Pro108Ser in 3CLpro and Pro151Leu in nucleocapsid protein, was inversely correlated with COVID-19 severity. Further in vitro research is awaited.