Genetic tracing of market wildlife and viruses at the epicenter of the COVID-19 pandemic

Abstract

Zoonotic spillovers of viruses have occurred through the animal trade worldwide. The start of the COVID-19 pandemic was traced epidemiologically to the Huanan Seafood Wholesale Market. Here, we analyze environmental qPCR and sequencing data collected in the Huanan market in early 2020. We demonstrate that market-linked severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) genetic diversity is consistent with market emergence and find increased SARS-CoV-2 positivity near and within a wildlife stall. We identify wildlife DNA in all SARS-CoV-2-positive samples from this stall, including species such as civets, bamboo rats, and raccoon dogs, previously identified as possible intermediate hosts. We also detect animal viruses that infect raccoon dogs, civets, and bamboo rats. Combining metagenomic and phylogenetic approaches, we recover genotypes of market animals and compare them with those from farms and other markets. This analysis provides the genetic basis for a shortlist of potential intermediate hosts of SARS-CoV-2 to prioritize for serological and viral sampling.

Keywords: SARS-CoV-2; forensic genomics; metatranscriptomics; phylodynamics; phylogenetics; viral emergence; virus evolution; wildlife trade.

Figure 1:. Phylogenetic analysis of SARS-CoV-2 from early COVID-19 cases and virus sequences obtained from the Huanan market.

(A) Phylogenetic tree of SARS-CoV-2 viral genomes collected before Jan 20, 2020. Tip colors correspond to different samples (red: market environmental samples and directly linked cases; yellow: Wuhan cases, with those indirectly linked to the Huanan market labeled; blue: all global cases). Symptom onset dates for cases are shown when known. The branch leading to A20 is slightly longer than other sequences with two substitutions from the root due to its slightly greater proportion (2.2%) of undetermined nucleotides. (B) Schematic tree topologies and associated posterior probabilities of inferred ancestral haplotypes for the SARS-CoV-2 pandemic. The probability of the four most likely tree topologies are conceptually visualized along with two alternatively proposed topologies. Lineages are demarcated “A” and “B”, and the common ancestry of Huanan market linked viral sequences according to each topology is visualized in red. The probabilities shown on the panel correspond to the recCA-constrained analysis (see Methods and Table S1). (C) tMRCA distributions inferred for the global SARS-CoV-2 pandemic (constrained by recCA and unconstrained). The 95% highest posterior density interval of each distribution is highlighted.

Figure 2:. The spatial distribution of SARS-CoV-2 in the Huanan market and animal DNA/RNA in SARS-CoV-2 positive samples from wildlife stall A.

(A) Proportion of positive SARS-CoV-2 qPCR samples, across all stalls sampled on January 1^st and 12^th 2020. For A–C, the heatmaps represent the p-value surface distributions of a relative risk analysis, showing areas of significantly elevated positive test density. (B) Metatranscriptomic sequencing (mNGS)-based testing for SARS-CoV-2 in samples collected on January 12^th 2020. (C) SARS-CoV-2 qPCR testing of water drains tested throughout the market. Arrows indicate the direction of reported drainage flows. (D–H) Proportion of SARS-CoV-2 sequence reads and mammalian mtDNA reads in samples that tested positive for SARS-CoV-2 from one wildlife stall, relative to total mammalian mtDNA reads. The number after each bar is the raw number of reads. Only mammalian species reported in at least two samples or with greater than 300 total reads are shown.

Figure 3:. The spatial distribution of animal DNA/RNA in the Huanan market.

(A) Sequence read counts of the mammalian species with mtDNA detected in at least three samples across all wildlife stall samples sequenced on January 12^th (all located in the western part of the market). Samples are grouped by wildlife stall, horizontally ordered by detection of SARS-CoV-2 (red: positive by qPCR and sequencing; orange: positive by sequencing only; blue: negative), and species are those we detected whose genus was reported as sold live in Wuhan markets by Xiao et al., vertically ordered by abundance. The locations of the A–G wildlife stalls are provided in Table S2. (B–E) Spatial distribution of the mean proportions of mtDNA reads across sequenced samples collected in the market on January 1^st and 12^th for (B) humans (H. sapiens), (C) raccoon dogs (N. procyonoides), (D) hoary bamboo rat (R. pruinosus), (E) masked palm civet (P. larvata).

Figure 4:. Animal viruses in the Huanan market.

(A–C) The spatial distribution of detection of three wildlife viruses detected in the Huanan market in sequenced samples collected in the market on January 1^st and 12^th. Bubbles correspond to stalls, and the bubble color represents the mean proportion of reads mapping to the viral genome from samples in that stall. The heatmaps represent the p-values of relative risk analyses, showing spatial distribution of enriched positivity for each virus. (D–F) Maximum likelihood whole genome phylogenies of the market consensus sequence for each of the three viruses shown in (A–C). Each reference virus is colored by the species it was reported as isolated from.

Figure 5:. Mitochondrial phylogenetics of potential intermediate host species of SARS-CoV-2 in the Huanan market.

(A) Coverage of wildlife mitochondrial genomes in Huanan market samples. Covered bases are colored, and consensus SNPs from the reference genome for each species are shown as black lines. (B) Cytochrome B (cytB) phylogeny of raccoon dog reference sequences collected from mainland Asia. (C) Heatmap of raccoon dog cytB SNVs found in the Huanan market samples and samples collected from other markets to the north of Hubei province. The color of each square represents the read mapping frequency of that allele in the sample. White squares indicate no read coverage at the site. SNVs are grouped by the branch of the reference tree they fall on, corresponding to the colors from (B). Samples from the wildlife stall positive for SARS-CoV-2 (stall A) are in bold.