Previously, we collected nasal samples from WTD in northeastern Ohio focusing on metropolitan area parks11. In the present study, we expanded geographical study area by approximately 1000-fold to target the entire state of Ohio. We collected 1522 nasal swabs from WTD across 83 of Ohio’s 88 counties from October 2021 to March 2022. Of those, 713 samples were opportunistically collected from hunter-harvested WTD during hunting season (November to December 2021) in 81 counties in Ohio, drawing primarily from rural areas. In addition to hunted WTD, we also collected 801 samples from WTD culled during deer population management programs from 9 counties in Ohio. We had 2 additional samples collected from roadkill WTD, and 6 nasal swabs that did not have sufficient metadata to identify the manner of death. We had a large sample size discrepancy between adult (n = 1199) and juvenile WTD (n = 297), which contributed to the reduction of power and may bias the estimates. This large difference in sample between ages is likely due to hunters’ common preference to take larger deer.
Members of the public are permitted to hunt WTD in Ohio during an explicit time frame each autumn (following WTD mating season). Hunter-harvested WTD samples were collected at locations where other cervid disease surveillance programs were being conducted concurrently. Samples from culled WTD were collected in partnership with each respective population management program at their individual deer processing sites. Culled WTD are typically baited at their respective reservations and managed land sites (where hunting by the general public is not permitted). Professional sharpshooters then harvest WTD from those sites to reduce WTD population in accordance with the deer management plan for that location. Nasal swabs were collected from free-ranging WTD during the six-month period from October 2021 to March 2022. Most samples were collected during November-December 2021, corresponding to the gun hunting season in Ohio (Fig. S1). Sample collectors from all partner organizations were trained on standard sample collection methods. Collectors wore a facemask and changed gloves between all samples. Two sterile polyester tipped swabs were used during nasal swab collection. The first swab cleared any debris from the exterior of the nostrils. A second sterile swab was inserted fully, scraped epithelial cells and nasal fluids of both nasal passageways and placed into a tube containing 3 mL viral transport media (BD UVT cat #220220). Post-collection, nasal swab samples were chilled temporarily in the field until transport to our laboratory at The Ohio State University and stored at −80 °C until diagnostic testing. We also collected blood samples to test for antibodies against SARS-CoV-2 in collaboration with USDA. We collected blood samples from 1164 of the 1522 WTD from which we collected nasal swabs. Blood was collected using 2 high purity cellulose fiber filter paper (Nobuto strips) dipped into pooled blood from WTD carcasses to saturate the paper with approximately 0.1 mL of blood. Nobuto strips were labeled and dried prior to transport. Any carcasses that did not appear fresh, had already been substantially processed, exterior blood appeared heavily contaminated, or did not have visible blood sources available for collection were excluded. Due to post-mortem sample collection, the study was exempt from a scientific permit from the Ohio Department of Natural Resources and beyond the scope of The Ohio State University Institutional Animal Care and Use Committee.
Viral RNA extraction was conducted using the Omega Bio-tek Mag-Bind Viral DNA/RNA 96 kit (cat# M6246·03), with 200 µl of sample11. Extracted viral RNA from samples was tested via real-time reverse transcription polymerase chain reaction (rRT-PCR) using the E gene primer/probe panel (Integrated DNA Technologies, Inc. cat #1006804) with Xeno VIC Internal Control Assay (Life Technologies cat #A29765)11. Any sample with a cycle threshold (Ct) value of 40 or below were considered E assay positive. All samples that screened E assay positive were confirmed using the Charité/Berlin RdRp confirmatory assay42 (Integrated DNA Technologies, Inc. cat #10006805) or the CDC N1/N2 kit (Integrated DNA Technologies, Inc. cat #10006606) diagnostic procedure. Samples with a Ct of ≤40 on either confirmatory assay were classified as positive for SARS-COV-2.
Blood samples were tested for evidence of prior exposure to SARS-CoV-2 as previously described with minor modifications10. Briefly, antibody elution from Nobuto strips was accomplished by incubating each strip in 1 mL BupH Tris-buffered saline (pH 7.2) (TBS, Thermo Fisher) containing 3% nonfat dried milk (Sigma) and 0.1% Tween 20 (Sigma) (TBSNT). Following incubation, samples were mixed by vortexing and debris then removed by centrifugation at 5000 × g for 10 min at ambient temperature. Supernatants were transferred to sterile microcentrifuge tubes and stored at −80 °C until use. The effective dilution of antibody in each eluate was estimated at 1:20. Sixty microliters of each eluted sample was directly (without further dilution) analyzed using the GenScript SARS-COV-2 Surrogate Virus Neutralization Test (sVNT, L00847-A) in accordance with the manufacturer’s instructions43. GenScript’s sVNT has a manufacturer’s provided specificity of 99.40% in humans, but no specificity analysis has been done in WTD. All samples were tested at least twice, with the average % inhibition of technical replicates used for the qualitative interpretation of SARS-COV-2 exposure.
The Ohio State Applied Microbiology Services Laboratory attempted genomic sequencing on all positive samples with a Ct value of 33 or lower (n = 86 samples). RNA was reverse transcribed into cDNA and PCR amplified using the ARTIC v 4.1 SARS-CoV-2 primer panel (Integrated DNA Technologies, Inc. cat #10011442) and NEBNext® FS Library Prep Kit for Illumina® (New England Biolabs, Ipswich MA) per manufacturers protocol instructions. Illumina sequencing libraries were prepared using RNA Prep with Enrichment (L) Tagmentation Kit (Illumina, San Diego, CA) per manufacturers protocol with unique dual indexes (Illumina). Libraries were pooled and quantified using ProNex NGS Library Quant Kit (NG1201, Promega Co. Madison, WI). Sequencing with NextSeq 2000 (Illumina) and assembly by DRAGEN COVID Lineage app v.3.5.6 (Illumina) were performed as previously described11.
Maps were generated using ArcMap (ESRI). For statistical analysis, location was coded using USDA Rural Urban Continuum Codes44. Codes range on a scale of 1–9 with increasing numbers corresponding to decreasing population size. For analysis purposes, an RUCC of 1 was considered urban – corresponding to counties surrounding Ohio’s three major metropolitan areas of Cincinnati, Cleveland, and Columbus. All other RUCC values (two through nine) were considered rural for analysis, but smaller metropolitan areas that fall into this category are indicated to aid in understanding spatial clustering. To evaluate risk factors associated with detection of SARS-CoV-2, we fit a mixed-effects logistic regression model (STATA 14.2, StataCorp LLC). We included random intercepts for county of sample collection to account for spatial clustering of sample collection. A highly significant (p value < 0.00005) likelihood ratio test compared to a standard logistic model supported that the mixed-effects model was better fit for our data. To evaluate the potential for unobserved heterogeneity at the county level, we fit a correlated random effects model incorporating cluster specific means45 and did not find significant evidence for differences in the individual level effects within clusters compared to their effects between clusters. Fixed effects were estimated for urban vs rural binary classification for the county, culled vs hunted WTD, WTD sex, WTD age, and a categorical effect for the month of sample collection to evaluate any changes over the course of the season (Table S3).
First, to determine how viruses obtained from WTD in Ohio (n = 80) were genetically related to SARS-CoV-2 viruses circulating in humans in Ohio and mink and WTD in other North American locations, a background dataset of complete genome sequences was compiled from GISAID (downloaded May 24, 2022). A total of 44,456 human SARS-CoV-2 viruses, collected in Ohio during February 2, 2020 – May 10, 2022, were downloaded from GISAID. This study benefited from the availability of a large number of SARS-CoV-2 sequences generated by the Ohio Department of Health, the US Centers for Disease Control, and other laboratories. County-level data was provided by the Ohio State Department of Health, which confirmed that the dataset was spatially representative, with the number of sequences available from an Ohio county proportional to the population size of the county (Fig. S23). Pangolin46 was used to assign a lineage to each human virus and ten viruses from each Pango lineage were randomly selected for the final background dataset using a customized Python script (n = 1592 human viruses). Outbreaks of genetically similar viruses from mink in Canada or a US state (Michigan, Oregon, Utah, and Wisconsin) were downsampled to 5 viruses per day, resulting in a total of 140 North American mink viruses in the final dataset. In addition, 145 viruses from WTD were included from Canada and 14 US states (Arkansas, Illinois, Iowa, Kansas, Maine, Massachusetts, Minnesota, New Jersey, New York, North Carolina, Oklahoma, Pennsylvania, Tennessee, and Virginia). The dataset was aligned using NextClade with Wuhan-Hu-1 as a reference. In-house python scripts were used to remove non-coding regions and mask sites that are known to be unreliable. A phylogenetic tree was inferred from this dataset using maximum-likelihood methods available in IQ-TREE 247 with a GTR + G model of nucleotide substitution and 1000 bootstrap replicates, using the high-performance computational capabilities of the Biowulf Linux cluster at the National Institutes of Health (http://biowulf.nih.gov). The inferred tree was visualized in FigTree v.1.4.4. White-tailed deer transmission clusters were defined by monophyletic groups of WTD viruses supported by high bootstrap values (>70) and confirmed or refined using UShER48 (Ultrafast Sample Placement on Existing tRees). One large cluster of 8 deer viruses in Miami, Columbiana, Harrison, Summit, and Tuscarawas counties turned out to be 3 independent human-to-deer transmission events (2 transmission clusters – one in northeast Ohio and one in western Ohio—and one singleton) upon further inspection in UShER, which created sub-trees with the 50 most closely related sequences out of a dataset of more than 12 million SARS-CoV-2 genome sequences from GISAID, GenBank, COG-UK, and CNCB. This exercise illustrated the importance of not overinterpreting apparent transmission clusters based on sequence similarity alone and the need to account for missing data in phylogenies.
Bayesian approaches were used to examine the evolutionary relationships between alpha variants and delta variants in humans and WTD in greater detail and compare their evolutionary rates. Separate datasets were generated for alpha (B.1.1.7) and delta (B.1.617.2 and AY lineages). In addition to the alpha and delta sequences obtained for the ML tree, above, sequences from more recently sampled alpha and delta viruses in humans globally were added to provide a longer period of data. The final alpha dataset (n = 786 sequences) included 9 viruses from WTD in Ohio collected for this study (November 8, 2021 – December 4, 2021); 31 viruses from WTD in Pennsylvania and New York (October 2, 2021 to December 4, 2021); 677 viruses from humans in Ohio (December 29, 2020 to August 23, 2021); and 69 additional human viruses sampled globally from November 1, 2021 to March 31, 2022. The final delta dataset (n = 1094 sequences) included 67 viruses from WTD in Ohio collected for this study (November 6, 2021 to January 20, 2022); 36 viruses from WTD in other North American locations (October 28, 2021 to January 30, 2022); 642 viruses from humans in Ohio (April 26, 2021 to April 18, 2022); 319 additional human viruses sampled globally from March 1, 2022 to July 19, 2022; plus 30 additional human viruses sampled from the United States during peak delta activity (July 10, 2021 to November 24, 2021) that helped resolve portions of the tree where human and WTD viruses were closely related.
We performed a time-scaled Bayesian analysis using the Markov chain Monte Carlo (MCMC) method available using the pre-release 1.10.5 version of the BEAST49 package available on GitHub (compiled on October 20, 2022), using GPUs available from the NIH Biowulf Linux cluster. A host-specific local clock32 was used to accommodate differences in the evolutionary rate between WTD and humans. Since WTD viruses were not monophyletic on the alpha or delta tree, owing to multiple independent human-to-deer transmission events, separate WTD transmission clusters identified on the ML tree were specified. The analysis was performed two ways, excluding WTD singleton viruses that are not positioned in a WTD transmission cluster and including WTD singleton viruses. A Bayesian non-parametric demographic model50 was used, with a general-time reversible (GTR) model of nucleotide substitution with gamma-distributed rate variation among sites. The MCMC chain was run separately 3–5 times for each dataset using the BEAGLE 351 library to improve computational performance, until all parameters reached convergence, as assessed visually using Tracer v.1.7.2. At least 10% of the chain was removed as burn-in, and runs for the same dataset were combined using LogCombiner v1.10.4. A MCC tree was summarized using TreeAnnotator v.1.10.4. To compare evolutionary rates across different regions of the SARS-CoV-2 genome, the analysis was repeated using five genome partitions: ORF1a, ORF1b, ORF3–ORF8, spike (S), and nucleoprotein (N). Several additional analyses were performed, including a phylogeographic discrete trait analysis52 to quantify rates of viral gene flow, particularly in the directions of human-to-deer and long-distance (across Ohio county lines) deer-to-deer transmission. A location state was specified for each viral sequence. All human viruses were categorized as “human,” whereas WTD viruses were also categorized by location of collection, with state information used for WTD viruses collected outside of Ohio and county information provided for all Ohio WTD viruses collected for this study. The expected number of location state transitions in the ancestral history conditional on the data observed at the tree tips was estimated using ‘Markov jump’ counts53,54, which provided a quantitative measure of asymmetry in gene flow between defined populations. To estimate absolute rates of synonymous and non-synonymous substitutions as well as dN/dS, we employ a ‘renaissance counting’ procedure that combines Markov jump counting with empirical Bayes modelling55. The outputs of these analyses (Markov jump counts, evolutionary rate distributions) were summarized and visualized using customized R scripts. Finally, we performed a flexible random effects analysis of the evolutionary substitution process21 to capture mutational bias differences along the human and WTD branches of evolutionary history. Mutational bias was measured by deviations from the Hasegawa-Kishino-Yano (HKY) substitution model that accommodates unequal base frequencies and different rates of transition and transversion substitutions. To facilitate efficient sampling of the additional random effects parameters, this analysis took advantage of gradient-based Hamiltonian Monte Carlo for phylogenetics within BEAST56. As a sensitivity analysis, we repeated our analysis to compare the early evolution of SARS-CoV-2 in white-tailed deer following a host-switch against the early evolution of SARS-CoV-2 in humans during December 2019 to February 2020, using the same dataset that was analyzed in Pekar et al. 21, (787 genomes, predominantly from China) except no rate prior was used.
The epidemiological curve of SARS-CoV-2 cases in humans in Ohio from January 1, 2021 to January 22, 2022 was generated using the number of daily reported COVID-19 cases in the state of Ohio (all age groups), available from the US Centers for Disease Control and Prevention (https://data.cdc.gov/Case-Surveillance/COVID-19-Case-Surveillance-Public-Use-Data-with-Ge/n8mc-b4w4). To estimate the proportion of COVID-19 cases belonging to different Pango lineages during each week of the epidemic, SARS-CoV-2 sequences collected from humans in Ohio during this time period were downloaded from GISAID. To account for inconsistencies in the intensity of viral surveillance, the number of viruses per lineage per week was normalized against the epidemiological curve derived from COVID-19 case counts and visualized using R. To further minimize biases only sequences categorized in the GISAID submission as obtained using a ‘baseline surveillance’ sampling strategy were included in the analysis. The dataset was further trimmed to include only submissions with complete collection dates and sufficient coverage to assign a Pango lineage, resulting in a final dataset of 27,187 sequences from Ohio. For simplicity, sub-lineages of B.1.617.2 (for example, AY.3) were consolidated into the Delta category, sub-lineages of B.1.1.7 (for example, Q.3) were consolidated into the Alpha category, and sub-lineages of B.1.1.529 (e.g., BA.1) were consolidated into the Omicron category.
We used root-to-tip regression to visualize the rate of substitution over time for the omicron variant, deer viruses, and all other SARS-CoV-2 lineages. To correctly assign mutation status during annotation to alignment in the phylogenetic analysis section, reference Wuhan genome (NC_045512.2) was added with MAFFT version 7.47557; the Wuhan genome was then removed to preserve the original set of sequences, while allowing for the correct alignment length representing all positions in the SARS-CoV-2 genome. This alignment together with the corresponding phylogenetic tree (see Phylogenetic analysis section) were used to reconstruct states at all tree nodes with TreeTime58 ancestral v 0.9.0-b.2 using default parameters. Mutations were extracted from the tree and reformatted into vcf format. Obtained vcf files were annotated with SnpEff59 v 4.5, and NC_045512.2 was utilized as a reference. Because mutations that happen along the tree do not always have the same nucleotide in REF field as position in genome, we corrected the annotation for the cases when those two did not match. The final annotated vcf and phylogenetic tree were used to count the number of mutations occurring from root to each leaf. We considered all mutations, synonymous and missense independently. The obtained results were visualized in R. To calculate the linear regression slope we excluded all WTD samples and human omicron data.
To accurately analyze mutations accumulating in WTD on delta and alpha backgrounds, datasets described in the Bayesian analysis section were utilized. First, we added reference to each alignment as described above, and then reconstructed phylogenetic trees with IQTree using the following parameters: –polytomy -m GTR + G –alrt 1000. Root in both cases was placed at the reference genome. States at nodes were reconstructed as described above. VCF files were produced independently for WTD clusters and WTD singletons. In alpha dataset we utilized not only the two clusters from Ohio, but all available alpha clusters (7 in total, Fig. S20). VCF annotation was produced as described above. Mutations in known problematic sites were filtered out using a list available at https://github.com/W-L/ProblematicSites_SARS-CoV2. Individual transmission clusters were visualized with ete3 python package60. Observed mutations on the spike trimer were visualized using the Protein Data Bank (PDB; rcsb.org), structure ID 7JJI. Structure visualization was performed with Open-Source PyMOL version 2.4.0 (Schrödinger, LLC). R-package MutationalPatterns61 was utilized to reconstruct mutational contexts. For input, we utilized mutations in WTD clusters and data on mutations in humans, the latter of which was extracted from the public version of the UShER (Ultrafast Sample placement on Existing tRee) tree downloaded on 2022-07-01 (http://hgdownload.soe.ucsc.edu/goldenPath/wuhCor1/UShER_SARS-CoV-2/2022/07/01/public-2022-07-01.all.masked.pb.gz), containing 5.7 million sequences. Variability in mutational contexts in humans was estimated by producing a series of subsamples containing the same number of mutations as observed in WTD delta clusters. The 10 subsamples were utilized to perform a permutation test to estimate significance of elevated C > T rate in WTD.
The HyPhy v.2.5.40 package was used to study positive and negative selection62. For this analysis we selected the four genes with highest number of homoplastic sites in clusters (N, S, ORF3a and nsp3). Samples that contained missing data (Ns) in the studied gene were removed, because HyPhy is unable to perform calculations on the missing data. Three different methods were run for each gene: aBSREL and BUSTED to check for positive selection in particular genes in WTD, and MEME to look for individual sites under positive selection. We independently tested selection for two sets of branches: all branches within WTD transmission clusters (called later on ‘clusters’); and clusters set plus singletons and the branches leading to transmission clusters (called ‘all’). While the clusters set represented the mutations happening only within the WTD population, the ‘all’ set was potentially contaminated by mutations that happened before the virus was transmitted to WTD, but it allowed incorporation of all available WTD samples. To search for sites under positive selection with MEME, we only looked for sites in foreground branches (e.g. included in ‘cluster’ or ‘all’ sets) with p < 0.001 in comparison with all other branches on the phylogenetic tree (background). The analyses for alpha and delta datasets were performed independently on the phylogenetic trees described in the previous section. Foreground branches were marked on the phylogenetic tree with phylotree.js (http://veg.github.io/phylotree.js/#). dN/dS value provided in Supplementary Table S6 were extracted from MEME output. Results of aBSREL and BUSTED outputs showed no signs of gene wise positive selection. As an alternative approach, we used codeml (from PAML version 4.9e63) to test whether there are signs of positive selection on branches leading for transmission clusters. Codeml was run in two modes with fix_omega =1 and fix_omega = 0. The LRT and p-values were calculated from obtained lnL values for each cluster. Again, no signs of gene wise positive selection were found. To study the frequencies of mutations in human populations for comparison to WTD, we utilized the outbreak.info package for R64 that utilizes GISAID data. Data was visualized with ggplot2 package for R 4.1.2.
In vitro and in vivo experiments
Animal studies were conducted in accordance with the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health and approved under St. Jude Children’s Research Hospital’s Animal Care and Use Committee protocol 442.
Positive swabs were received on dry ice and were transferred into the ABSL3+ at St. Jude Children’s Research Hospital. All virus isolation, characterization and animal experiments were performed under ABSL3+ conditions.
The Vero E6 cell line ectopically expressing both TMPRSS2 and ACE2 (Vero ACE2 T2) was a kind gift from Dr. Barney Graham at VRC, NIAID, NIH. Cells were maintained in DMEM (Sigma D6429) supplemented with 10% heat treated fetal bovine serum (HyClone SH30071.03) and 10 μg/ml Puromycin (Sigma P9620). Cultures were overlaid with 1 mL of inoculum consisting of 100 μL of swab suspension plus 900 μL of infection media (DMEM supplemented with 2% heat treated fetal bovine serum and 1× antibiotic solution (Gibco 15240-062)). After 1 h, the inoculum was aspirated off and fresh infection media was added to the cells. Cultures were checked daily, and media-cell suspension was harvested when greater than 90% cytopathic effect (CPE) was observed. The suspension was tested by BD Veritor System for rapid detection of SARS-CoV-2 (Catalog # 256082) for confirmation of virus isolation and streaked on blood agar plates for sterility. Select viral isolates were sequenced to confirm identity with the original sample sequences that produced the viral isolates using CLC Genomic Workbench v22.0.1 (Qiagen) and BioEdit v7.2.5 (open source).
Virus stocks or experimental samples were titered using a Vero E6 cell line ectopically expressing the TMPRSS2 gene (Vero E6 T2) sourced from JCRB Cell Bank in Japan (https://cellbank.nibiohn.go.jp/english/). To determine the 50% tissue culture infectious dose (TCID50), 96-well culture plates were inoculated with 100 μL of a 1:10 serially diluted sample in infection media. After 72 h, the plates were fixed and stained with 0.1% crystal violet in 10% formalin. Infectious dose titers were determined using the Reed and Muench method23.
Male LVG Golden Syrian Hamster, 4–5 weeks of age, were purchased from Charles River Laboratories (Wilmington, MA), assigned numbers sequentially upon arrival to the Animal Resource Center, and assigned to groups based on vaccine treatment and virus challenge on paper without investigators observing individual animals. Laboratory experiments were performed exclusively with male animals, as designated by the vendor prior to arrival. Male animals were chosen as it has been established male Syrian hamsters are more susceptible to disease caused by SARS-CoV-229. There were 10 total experimental groups with 120 total hamsters. Hamsters assigned to vaccine groups were vaccinated intramuscularly with 10 μg BNT162b2 vaccine (New York, NY), prepared as instructed with the exception that it had reached an expiration date and was no longer suitable for clinical use. Vaccine was administered in 50 μl at 2 injection sites (100 μl total volume) of the rear hind limb. Animals were boosted by the same procedure, on alternate limb, 21 days post vaccination (dpv). Sera was collected 21 days post boost (B + 21) and assessed for antibody response by microneutralization.
Approximately 3 weeks post boost, vaccinated or naïve control hamsters were inoculated intranasally with 104 TCID50 of SARS-Cov-2 virus. Each experimental group had 5 animals that were used for weight loss measurements, 4 animals that were sacrificed on 2-days post inoculation (dpi), and 3–4 animals that were sacrificed on 4 dpi for a total of 12-13 animals per group. Except the Human B.1.1.7 and Deer AY.25, which had 4, 3, and 3 animals for each of those time points respectively (10 total per group). These groups were limited by animal availability and included an unvaccinated control group for cross virus comparisons only (Fig. 5b). At 2 and 4 dpi, animals were sacrificed for lung and nasal turbinate. Alternatively, on 2, 4 and 6 dpi animals were anesthetized with 100 mg/kg Ketamine and nasal passages were rinsed with 0.5 mL phosphate buffered saline (PBS). Infectious viral load was determined by TCID50 as described. Longitudinal animals were scored for clinical signs and weighed daily for 2 weeks. All surviving animals were exsanguinated at 21 dpi and the serum was collected for microneutralization comparison. Animals were housed individually in standard filter top rat cages with day/night cycle from 6 a.m. to 6 p.m. Animal health observations were made at least 1/day or 2/day during peak infection. We did not observe any adverse events in these experiments other than the expected animal weight loss. Statistical analysis for all in vitro and in vivo experiments was performed using one-way ANOVA followed by a Tukey post hoc in Prism v.9 (GraphPad).
A viral microneutralization assay was performed to measure the neutralizing antibody activity of hamster sera of SARS-CoV-2/human/USA/WA-1/2020 (WA-1), hCoV-19/USA/CA_CDC 5574/2020 (Hu-B.1.1.7), SARS-CoV-2/human/USA/COR-21-192500/2021 (Hu-B.1.617.2), (hCoV-19/deer/USA/OH-OSU-2158/2021) (AY.103), hCoV-19/deer/USA/OH-OSU-1338/2021 (B.1.1.7-like), and BNT162b2 vaccine against representative WTD SARS-CoV-2 isolates. Fivefold serial dilutions were performed on heat inactivated sera (1 h at 56 °C) in infection medium, starting at a 1:40 dilution. A standardized amount of infectious SARS-CoV-2 virus (250 TCID50), diluted in infection medium, was added to the diluted serum at a 1:1 ratio and incubated for 1 h at 37 °C. A volume of 100 μL of the serum/virus mixture was added to Vero E6 T2 cells seeded in 96-well plates the previous day and incubated for 1 h at 37 °C under 5% CO2. Subsequently, an additional 100 μL of infection media was added and the cells incubated for a further 24–48 h. Following incubation, cells were fixed with 4% formaldehyde (Polysciences Cat #18814-20) for 30 min, washed with PBS (source) three times and then incubated with a block/permeabilization buffer (PBS supplemented with 3% Bovine Serum Albumin (BSA; Sigma-Aldrich Cat #A8327-500 ml) and 0.2% Triton-X-100 (ThermoFisherSurfact-Amps-X-100, 10% Solution Cat #28314)) for 30 min. Rabbit anti-SARS CoV-2 NP mAb (Sinobiologicals Cat # 40143-R040) at a 1:2000 dilution was added for 1 h. Cells were washed three times with PBS supplemented with 0.5% Tween (PBST; Thermofisher Cat #28314) before incubation with a secondary goat anti-rabbit IgG –HRP conjugated antibody (Cell Signaling Cat# 7074 S) at a 1:3000 dilution for 1 h. After washing the cells three times with PBST, 100 μL of TMB (Thermofisher Cat #N301) was added and color developed for 10 min before 1 N sulfuric acid (Fisher Scientific Cat #SA212-1) was added to stop the reaction. The optical density was measured at 450 nm on a Biotek Synergy plate microplate reader and the neutralization titers were calculated as the reciprocal serum dilution (IC50) causing 50% reduction of relative light units.
Vero E6 T2, Vero ACE2 T2, and Calu-3 cells were infected at a low multiplicity of infection (MOI 0.001 TCID50/cell) with representative WTD isolates of SARS-CoV-2 and parent viruses, washed, and maintained in infection medium. Supernatants were collected at 12, 24, 36, and 48 hpi, then titrated in Vero E6 T2 using TCID50. Titrations were calculated by the Reed and Muench method65. Data shown are mean titers ±SD of triplicate measures for each time point.
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.