Individuals with recent prior SARS-CoV-2 infection are at reduced risk of Omicron infection and associated hospitalization

Omicron sub-lineages such as BA2.12.1 and BA5 have breached prior infection-induced immunity and vaccine-induced immunity. This capacity of Omicron to reinfect patients calls for a characterization of vaccination, infection, and reinfection patterns. We analyzed de-identified longitudinal electronic health records for 389,746 individuals (88,679 fully-vaccinated, 184,205 boosted, 73,184 with prior infection) across multiple US states. Compared to individuals with only full vaccination, the rates of SARS-CoV-2 infections in the Omicron era were reduced for individuals with additional prior infection (1.4 to 1.8-fold reduced, depending on vaccine status) or booster vaccination (1.3 to 2.0-fold reduced). Although prior infection was associated with lower incidence of SARS-CoV-2 infection, we found that the relative risk (RR) of infections for individuals with prior infection has increased during Omicron. During October, 2021, RR was 0.11 [0.10-0.13, 95% CI] while during May, 2022, it increased to 0.57 [0.46-0.68, 95% CI], suggesting an increase in reinfections with Omicron. Furthermore, we found that time since prior infection is associated with risk of reinfection, providing evidence of waning immunity. Prior infections before June, 2021, were associated with marginal reduction in risk of infection (eg., RR = 0.80 [0.68-0.90] for prior infection during January, 2021), while recent prior infections were associated with significant reduction in risk (eg., RR = 0.24 [0.20-0.29, 95% CI] for prior infection during November, 2021). Despite an observed increase in reinfections and vaccine breakthrough infections, our findings emphasize the protective effect of natural and vaccine immunity, with prior infection providing [~]6 months of protection from reinfection.

Genetic alteration of human MYH6 is mimicked by SARS-CoV-2 polyprotein: mapping viral variants of cardiac interest

Acute cardiac injury has been observed in a subset of COVID-19 patients, but the molecular basis for this clinical phenotype is unknown. It has been hypothesized that molecular mimicry may play a role in triggering an autoimmune inflammatory reaction in some individuals after SARS-CoV-2 infection. Here we investigate if linear peptides contained in proteins that are primarily expressed in the heart also occur in the SARS-CoV-2 proteome. Specifically, we compared the library of 136,704 8-mer peptides from 144 human proteins (including splicing variants) to 9,926 8-mers from all 17 viral proteins in the reference SARS-CoV-2 proteome. No 8-mers were exactly identical between the reference human proteome and the reference SARS-CoV-2 proteome. However, there were 45 8-mers that differed by only one amino acid when compared to the reference SARS-CoV-2 proteome. Interestingly, analysis of protein-coding mutations from 141,456 individuals showed that one of these 8-mers from the SARS-CoV-2 Replicase polyprotein 1a/1ab (KIALKGGK) is identical to a MYH6 peptide encoded by the c.5410C>A (Q1804K) genetic variation, which has been observed at low prevalence in Africans/African Americans (0.08%), East Asians (0.3%), South Asians (0.06%) and Latino/Admixed Americans (0.003%). Furthermore, analysis of 4.85 million SARS-CoV-2 genomes from over 200 countries shows that viral evolution has already resulted in 20 additional 8-mer peptides that are identical to human heart-enriched proteins encoded by reference sequences or genetic variants. Whether such mimicry contributes to cardiac inflammation during or after COVID-19 illness warrants further experimental evaluation. We suggest that SARS-CoV-2 variants harboring peptides identical to human cardiac proteins should be investigated as viral variants of cardiac interest.

COVID-19 vaccines dampen genomic diversity of SARS-CoV-2: Unvaccinated patients exhibit more antigenic mutational variance

Variants of SARS-CoV-2 are evolving under a combination of immune selective pressure in infected hosts and natural genetic drift, raising a global alarm regarding the durability of COVID-19 vaccines. Here, we conducted longitudinal analysis over 1.8 million SARS-CoV-2 genomes from 183 countries or territories to capture vaccination-associated viral evolutionary patterns. To augment this macroscale analysis, we performed viral genome sequencing in 23 vaccine breakthrough COVID-19 patients and 30 unvaccinated COVID-19 patients for whom we also conducted machine-augmented curation of the electronic health records (EHRs). Strikingly, we find the diversity of the SARS-CoV-2 lineages is declining at the country-level with increased rate of mass vaccination (n = 25 countries, mean correlation coefficient = -0.72, S.D. = 0.20). Given that the COVID-19 vaccines leverage B-cell and T-cell epitopes, analysis of mutation rates shows neutralizing B-cell epitopes to be particularly more mutated than comparable amino acid clusters (4.3-fold, p < 0.001). Prospective validation of these macroscale evolutionary patterns using clinically annotated SARS-CoV-2 whole genome sequences confirms that vaccine breakthrough patients indeed harbor viruses with significantly lower diversity in known B cell epitopes compared to unvaccinated COVID-19 patients (2.3-fold, 95% C.I. 1.4-3.7). Incidentally, in these study cohorts, vaccinated breakthrough patients also displayed fewer COVID-associated complications and pre-existing conditions relative to unvaccinated COVID-19 patients. This study presents the first known evidence that COVID-19 vaccines are fundamentally restricting the evolutionary and antigenic escape pathways accessible to SARS-CoV-2. The societal benefit of mass vaccination may consequently go far beyond the widely reported mitigation of SARS-CoV-2 infection risk and amelioration of community transmission, to include stemming of rampant viral evolution.