Accelerated evolution of SARS-CoV-2 in free-ranging white-tailed deer

Abstract

While SARS-CoV-2 has sporadically infected a wide range of animal species worldwide1, the virus has been repeatedly and frequently detected in white-tailed deer in North America2â€"7. The zoonotic origins of this pandemic virus highlight the need to fill the vast gaps in our knowledge of SARS-CoV-2 ecology and evolution in non-human hosts. Here, we detected SARS-CoV-2 was introduced from humans into white-tailed deer more than 30 times in Ohio, USA during November 2021-March 2022. Subsequently, deer-to-deer transmission persisted for 2-8 months, which disseminated across hundreds of kilometers. We discovered that alpha and delta variants evolved in white-tailed deer at three-times the rate observed in humans. Newly developed Bayesian phylogenetic methods quantified how SARS-CoV-2 evolution is not only faster in white-tailed deer but driven by different mutational biases and selection pressures. White-tailed deer are not just short-term recipients of human viral diversity but serve as reservoirs for alpha and other variants to evolve in new directions after going extinct in humans. The long-term effect of this accelerated evolutionary rate remains to be seen as no critical phenotypic changes were observed in our animal model experiments using viruses isolated from white-tailed deer. Still, SARS-CoV-2 viruses have transmitted in white-tailed deer populations for a relatively short duration, and the risk of future changes may have serious consequences for humans and livestock.

Figure 1. Geographic distribution of SARS-CoV-2 in Ohio by county.

Counties classified as urban are colored grey and rural counties are white. The size of circles plotted over the county centroids indicate the number of samples collected and the color scale indicates SARS-CoV-2 estimated prevalence in each county by rRT-PCR (A) and seroprevalence by surrogate virus neutralization (B). Counties that are outlined in bold borders indicate counties from which we obtained SARS-CoV-2 genomic sequences (Table S2). Counties marked with an asterisk indicate counties from which samples were collected from culled WTD as a part of population management programs (Table S1).

Figure 2. Human-to-deer transmission of SARS-CoV-2 in Ohio.

(A) MCC tree inferred for B.1.1.7 viruses collected from humans and WTD. Branches shaded by host species and location. The two Ohio WTD clusters are labeled. (B) AY.25 subtree (entire delta MCC tree shown in Figure S4). Ohio WTD virus transmission clusters are shaded similarly to Figure 3. (C) The number of bi-weekly COVID-19 cases in humans in Ohio from January 2021 to February 2022, shaded by the proportion of human SARS-CoV-2 sequences from Ohio that belong to one of four Pango lineages (or ‘other’). Red box delineates the B.1.1.7 wave in humans. Below, green bars show the estimated number of human-to-deer transmission events of B.1.1.7 viruses, per 20-week increments, based on “Markov jump” counts inferred on the MCC tree. Green circles indicate the collection dates of B.1.1.7 viruses in Ohio WTD. (D) Similar to panel (C), but for delta variants. (E) The detection lag (months) is the time difference between a human-to-deer transmission event (estimated) and the first observed sequence from a WTD transmission cluster, shown for 14 delta and 2 alpha WTD transmission clusters. (F) Estimated number of human-to-deer transmission events and long-distance deer-to-deer transmission events that span Ohio counties, shown for the alpha and delta variants. Data for North America does not include Ohio.

Figure 3. Map of SARS-CoV-2 transmission clusters in Ohio white-tailed deer.

Each shape represents a county in Ohio where SARS-CoV-2 virus was identified in WTD for this study (triangle = alpha variant; circle = delta variant). Large circles indicate WTD transmission clusters, as identified on the phylogenetic tree (black = clusters restricted to one county; shaded = clusters identified in more than one county). Large circles shaded the same color belong to the same transmission cluster. Small black circles indicate singleton WTD viruses. PANGO lineage provided for all clusters. Human population density is shown in the background (red = high; green = low) and major cities are labeled.

Figure 4. Evolutionary rate of SARS-CoV-2 in humans and white-tailed deer.

(A) The posterior distributions of evolutionary rates (substitutions per site per year) for five partitions of the SARS-CoV-2 genome (ORF1a, ORF1b, ORF3 – ORF8 plus envelope (E) and membrane (M), spike (S), and nucleocapsid (N) are presented for human (pink) and WTD (blue) for the delta variant. Alpha results (similar) are provided in Figure S11. (B) Mutations in spike protein that were found in delta WTD clusters (orange), alpha WTD clusters (green), and both alpha and delta WTD clusters (yellow). All recurrent mutations from WTD clusters are documented in Table S8. The log deviation (random-effect) from HKY model relative rates is presented for (C) alpha, humans, (D) alpha, WTD, (E) alpha, WTD-to-human ratio, (F) delta, humans, (G) delta, WTD, and (H) delta, WTD-to-human ratio. Box midlines indicate the median, the box limits show the upper and lower quartiles, and the whiskers extend to 1.5 times the interquartile range. Asterisks indicate transversions. WTD-to-human ratios that significantly differ from zero are highlighted.

Figure 5

Pathogenicity and replication of multiple strains of SARS-CoV-2 viruses in Golden Syrian hamsters. (A) Microneutralization titers of a-BNT162b2 or lineage specific serum against representative viruses from this study. For B-E, the mean for each group is plotted, and bars indicate standard deviation. Titers expressed as log₁₀ EC₅₀ were plotted and described as a fold change from the reference strains. (B) Body weight loss comparison between unvaccinated and BNT162b2 vaccinated animals at the peak of infection, day 7. Mean weights are displayed as a percentage of starting weight. Nasal wash was collected (unvaccinated groups only) (C) or lung and nasal turbinate were harvested (D and E) and used to quantify viral titers. Viral titers expressed as the log₁₀ TCID₅₀ were plotted. Statistical analysis was performed using one-way ANOVA (*, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001)