key: cord-0975938-x48xjedg
authors: Braun, Katarina M.; Moreno, Gage K.; Halfmann, Peter J.; Baker, David A.; Boehm, Emma C.; Weiler, Andrea M.; Haj, Amelia K.; Hatta, Masato; Chiba, Shiho; Maemura, Tadashi; Kawaoka, Yoshihiro; Koelle, Katia; O’Connor, David H.; Friedrich, Thomas C.
title: Transmission of SARS-CoV-2 in domestic cats imposes a narrow bottleneck
date: 2020-11-17
journal: bioRxiv
DOI: 10.1101/2020.11.16.384917
sha: 311c474171840b99cc9e2da3886481d334ef2fde
doc_id: 975938
cord_uid: x48xjedg

The evolutionary mechanisms by which SARS-CoV-2 viruses adapt to mammalian hosts and, potentially, escape human immunity depend on the ways genetic variation is generated and selected within and between individual hosts. Using domestic cats as a model, we show that SARS-CoV-2 consensus sequences remain largely unchanged over time within hosts, but dynamic sub-consensus diversity reveals processes of genetic drift and weak purifying selection. Transmission bottlenecks in this system appear narrow, with new infections being founded by fewer than ten viruses. We identify a notable variant at amino acid position 655 in Spike (H655Y) which arises rapidly in index cats and becomes fixed following transmission in two of three pairs, suggesting this site may be under positive selection in feline hosts. We speculate that narrow transmission bottlenecks and the lack of pervasive positive selection combine to constrain the pace of ongoing SARS-CoV-2 adaptive evolution in mammalian hosts.

Understanding the forces that shape genetic diversity of RNA viruses as they replicate within, 31 and are transmitted between, hosts may help us to forecast the future evolutionary trajectories 32 of viruses on larger scales. The level and duration of protection provided by vaccines, 33 therapeutics, and natural immunity against severe acute respiratory syndrome coronavirus 2 34 (SARS-CoV-2) will depend in part on the amount of circulating viral variation and the rate at 35 which adaptive mutations arise within hosts, persist between hosts, and become widespread. 36

Here, to model the evolutionary capacity of SARS-CoV-2 within and between hosts, we 37 characterize viral genetic diversity arising, persisting, and being transmitted in domestic cats. been recovered from lower respiratory tract lung parenchyma, where infection is most 48 commonly linked to severe disease in humans 1, 6, 9, 10 . 49

In a recent study, members of our team experimentally infected three index cats with a SARS-50

CoV-2 human isolate and introduced one naive direct contact cat per index one day following 51 index inoculation 11 . Each index cat transmitted SARS-CoV-2 to the corresponding contact cat, 52 and infectious virus was recovered from all six cats over multiple timepoints 11 . This study was 53 designed to address whether SARS-CoV-2 could infect and transmit between cats so viruses 54 collected from this study were not characterized beyond determining viral titers. Here, we use 55 deep sequencing to define patterns of SARS-CoV-2 genetic variation over time within the index 56 cats and following transmission to the contact cats. 57

Transmission bottlenecks -dramatic reductions in viral population size at the time of 58 transmission -play an essential role in the overall pace of evolution of respiratory viruses 12-21 . 59 For example, in humans airborne transmission of seasonal influenza viruses appears to involve 60 a narrow transmission bottleneck, with new infections founded by as few as 1-2 genetically 61 Genetic drift and purifying selection shape within-host diversity 108 To probe the evolutionary pressures shaping SARS-CoV-2 viruses within hosts, we first 109 evaluated the proportion of variants shared between cats. Eighty-six percent of variants (42 out 110 of 38 iSNVs and 11 indels) were found in a single cat (42/49), 8% of variants were found in 2-5 111 cats (4/49), and the remaining 6% of variants were found in all 6 cats (3/49). 112

Purifying selection, which acts to purge deleterious mutations from a population, is known to 113 result in an excess of low-frequency variants. In contrast, positive selection results in the 114 accumulation of intermediate-and high-frequency variation 36 . Importantly, especially in the 115 setting of an acute viral infection, exponential population growth is also expected to result in an 116 excess of low-frequency variants 37 . To determine the type of evolutionary pressure acting on 117 SARS-CoV-2 in cats, we plotted these distributions against a simple "neutral model" 118 (transparent grey bars in Fig 2B) , which assumes a constant population size and the absence 119 of selection 36 . This model predicted that ~43% of polymorphisms would fall in the 3-10% 120 frequency bin, ~25% into the 10-20% bin, ~14% into the 20-30% bin, ~10% into the 30-40% bin, 121 and ~8% into the 40-50% bin. The frequency distribution of variants detected in each index cat 122 across all available timepoints did not differ significantly from this "neutral" expectation 123 these iSNVs were detected above 3% frequency in the inoculum, but when we mined all 148 sequencing reads, S H655Y and E S67S can be detected at 0.85% and 0.34% in the inoculum, 149 respectively. S H655Y was the consensus sequence on days 2-5 and days 7-8 in index cat 1 as 150 well as on days 4 and 8 in index cat 2 and remained detectable above our 3% variant threshold 151 throughout infection (Fig 3) . Similarly, envelope S67S (E S67S) was the consensus sequence 152 on day 8 in index cat 1 and day 1 in index cat 2. S H655Y and E S67S) were detectable at all 153 timepoints in cat 3 on days that SARS-CoV-2 was detectable ≥ 10 4 copies/mL and day 8 but 154 stayed below consensus level. 155

Interestingly, S H655Y and E S67S became fixed together following transmission in two 156 transmission pairs (contact cats 4 and 6) and were lost together during transmission to contact 157 animal 5. In cat 5, however, two different variants in ORF1ab, G1756G and L3606F, became 158 fixed after transmission. ORF1ab G1756G was not detected above 3% and L3606F was found 159 at 17.2% in the day 5 sample from the index cat 2 (the cat transmitting to cat 5), and 160 interestingly was not found in the inoculum at any detectable frequency. The categorical loss or 161 fixation of these variants immediately following transmission, and in particular the fixation 162 following transmission of a variant that was undetectable before, are highly suggestive of a 163 narrow bottleneck 39 . 164

In addition, a synonymous variant in an alanine codon at amino acid position 1,222 in Spike 165 (nucleotide site 25,174) was found at >50% frequencies on days 4 and 8 in index cat 3, but was 166 not detected above 3% on any other days. All iSNVs over time are shown in Supplementary 167 To estimate the size of SARS-CoV-2 transmission bottlenecks, we investigated the amount of 173 genetic diversity lost following transmission in cats. We observed a reduction in the cumulative 174 number of variants detected in each contact cat compared to its associated index: 7 fewer variants in cat 4 (n=9) compared to cat 1 (n=16), 10 fewer in cat 5 (n=19) than cat 2 (n=10), and 176 10 fewer in cat 6 (n=16) than cat 3 (n=6). Likewise, the frequency distribution of variants in all 177 three contact cats following transmission differed from the distribution of variants in all three 178 index cats prior to transmission (p-value=0.052, Mann Whitney U test). Following transmission, 179 variant frequencies became more bimodally distributed than those observed in index cats, i.e., 180 in contacts most variants were either very low-frequency or near-fixed (Supplementary Fig 6) . 181

To quantitatively investigate the stringency of each transmission event, we compared the 182 genetic composition of viral populations immediately before and after viral transmission. We 183 chose to use the first timepoint when infectious virus was recovered in the contact cat coupled 184 with the timepoint immediately preceding this day in the index cat, as has been done previously 185 18 . We used days 2 (index) and 3 (contact) in pair 1, days 5 and 6 in pair 2, and days 4 and 5 in 186 pair 3 (these sampling days are outlined in red in Fig 1) . We applied the beta-binomial determined that a mean effective bottleneck size of 5 (99% CI: 1-10), 3 (99% CI: 1-7), and 2 190 (99% CI: 1-3) best described each of the three cat transmission events evaluated here (Fig 4) . immunity. Using domestic cats as a translational model, we show that genetic drift appears to 205 be a major force shaping SARS-CoV-2 evolution. Selection within hosts is weak, and 206 transmission bottlenecks, even with the potential for contact transmission, appear narrow. 207

These observations suggest that SARS-CoV-2 may already be well adapted to mammalian 208 hosts. The strong role of genetic drift may combine with the relatively slow mutation rate and 209 narrow transmission bottlenecks to slow the overall pace of viral evolution. 210

Here we use deep viral sequencing to carefully uncover within-host variants in 6 domestic cats 211 grouped into three defined transmission pairs. We find genetic drift and purifying selection 212 shape SARS-CoV-2 genetic diversity within feline hosts, and a stringent bottleneck defines viral 213 transmission. This latter finding is at odds with some recent studies in humans, which have 214 Humans are currently the primary reservoir for SARS-CoV-2, but the mink example shows that 283 SARS-CoV-2 is able to infect and transmit among other mammals with the potential for ongoing 284 zoonosis and anthroponosis. This exemplifies the need to understand the evolutionary 285 mechanisms and pace at which SARS-CoV-2 is able to adapt to, and transmit between a broad 286 range of host species. In our study we see variants arising early and being transmitted onward 287 in cats, a potential reservoir species. Our study and the mink example show that species-and 288 context-specific adaptations are inevitable as SARS-CoV-2 explores new hosts. While we do 289 not know the phenotypic impacts of these variants, the rapid rise of variants in potential 290 reservoir species may significantly impact humans if exposed to these new species-specific 291 SARS-CoV-2 adaptations. 292

As more than 300,000 new SARS-CoV-2 cases occur each day worldwide, we must have 293 models in place to recapitulate key evolutionary factors influencing SARS-CoV-2 transmission. 294

With the imminent release of SARS-CoV-2 vaccines and therapeutics and increasing incidence 295 of natural exposure-related immunity, these models can help us forecast the future of SARS-296

CoV-2 variation and population-level genetic changes. Here, we use six domestic cats to show 297 how SARS-CoV-2 genetic variation is predominantly influenced by genetic drift and purifying 298 selection within individual hosts. Additionally, we find a role for narrow transmission bottlenecks 299

shaping founding diversity in all three contact cats. Continued efforts to sequence SARS-CoV-2 300 across a wide variety of hosts, transmission routes, and spatiotemporal scales will be necessary 301 to determine the evolutionary and epidemiological forces responsible for shaping within-host 302 genetic diversity into global viral variation. Table 1 To generate SNP Frequency Spectrums (SFS), we binned all variants detected across 395 timepoints within each index cat into six bins -3-10%, 10-20%, 20-30%, 30-40%, 40-50%, 50-396 60%. We plotted the counts of variants falling into each frequency bin using Matplotlib 3.3.2 397 (https://matplotlib.org). We used code written by Dr. Louise Moncla to generate the distribution 398 of SNPs for a given population assuming no selection or change in population size, which is 399 expected to follow a 1/x distribution 36 . The code to replicate this can be found in the GitHub 400 accompanying this manuscript, specifically in the `code/SFS.ipynb` Jupyter Notebook. This 401 model predicts 42.8% of variants will fall within the 3-10% frequency range, 24.6% will fall within the 10-20% frequency range, 14.4% of variants will fall within the 20-30% frequency range, 

Susceptibility of ferrets, cats, dogs, and other domesticated animals to SARS-461 coronavirus 2

Comparative pathogenesis of COVID-19, MERS, and SARS in a 463 nonhuman primate model

Syrian hamsters as a small animal model for SARS-CoV-2 infection and 465 countermeasure development

Infection with novel coronavirus (SARS-CoV-2) causes pneumonia in 467

Rhesus macaques

SARS-CoV-2 infection in farmed minks, the Netherlands

The search for a COVID-19 animal model

SARS-CoV-2 is transmitted via contact and via the air between ferrets

Transmission of SARS-CoV-2 on mink farms between humans 475 and mink and back to humans

Epidemiology, virology, and clinical features of severe acute respiratory 477 syndrome -coronavirus-2 (SARS-CoV-2

Coronavirus Disease-19)

Clinical course and outcomes of critically ill patients with SARS-CoV-2 480 pneumonia in Wuhan, China: a single-centered, retrospective, observational study

Transmission of SARS-CoV-2 in Domestic Cats

Genetic bottlenecks in intraspecies virus transmission

Stochastic processes constrain the within and between host evolution 487 of influenza virus

Influenza B Viruses Exhibit Lower Within-Host Diversity than 489

Influenza A Viruses in Human Hosts

Selection on haemagglutinin imposes a bottleneck during mammalian 491 transmission of reassortant H5N1 influenza viruses

Selective Bottlenecks Shape Evolutionary Pathways Taken during 493

Mammalian Adaptation of a 1918-like Avian Influenza Virus

Influenza A virus transmission bottlenecks are defined by infection route 496 and recipient host

Mammalian adaptation of influenza A(H7N9) virus is limited by a narrow 498 genetic bottleneck

Reconciling disparate estimates of viral genetic diversity during 500 human influenza infections

Transmission bottlenecks as determinants of 502 virulence in rapidly evolving pathogens

Transmission bottlenecks and 504 the evolution of fitness in rapidly evolving RNA viruses

Deep Sequencing Reveals Potential Antigenic Variants at Low 506 Frequencies in Influenza A Virus-Infected Humans

Emergence of Oseltamivir-Resistant H7N9 Influenza Viruses in 508

Matters of Size: Genetic Bottlenecks in Virus Infection

Population Bottlenecks and Intra-host Evolution during Human-to-Human 512

Shared SARS-CoV-2 diversity suggests localised transmission of 514 minority variants

Mutational dynamics and transmission properties of SARS-CoV-2 516 superspreading events in Austria

Mutational 518 Signatures and Heterogeneous Host Response Revealed Via Large-Scale 519

Characterization of SARS-COV-2 Genomic Diversity

Hidden genomic diversity of SARS-CoV-2: implications for qRT-PCR 521 diagnostics and transmission

SARS-CoV-2 exhibits intra-host genomic plasticity and low-frequency 523 polymorphic quasispecies

Genomic diversity of SARS-CoV-2 in Coronavirus Disease

VERSO: a comprehensive framework for the inference of robust 527 phylogenies and the quantification of intra-host genomic diversity of viral samples

Structural and functional properties of 586

SARS-CoV-2 spike protein: potential antivirus drug development for COVID-19

SARS-CoV-2 mink-associated variant strain -Denmark

SNPGenie: estimating evolutionary 592 parameters to detect natural selection using pooled next-generation sequencing data

Simple methods for estimating the numbers of synonymous and 595 nonsynonymous nucleotide substitutions

Adaptive Evolution of Genes and Genomes

SARS_CoV-2_Zequencer: Zequencer to accompany 599 "Transmission of SARS-CoV-2 in domestic cats imposes a narrow bottleneck

Plot representing all 613 variants (iSNVs and indels) detected in any cat at any timepoint. Variant frequencies are plotted by 614 genome location and are colored by gene. Circles represent synonymous iSNVs, squares represent 615 nonsynonymous iSNVs, and stars represent indels. B) iSNV frequency spectrums with error bars showing 616 standard deviation for index cats plotted against a "neutral model

Variant frequencies in the 630 index cats (x-axis) compared with frequencies of the same variants in the corresponding contact cats (y-631 axis) that were used in the beta-binomial estimate are shown on the left. Estimates of SARS-CoV-2 632

10.2% of variants will fall within the 30-40% frequency range, and 7.9% of variants will fall within 404 the 40-50% frequency range. We used a Mann-Whitney U test to test the null hypothesis that 405 the distribution of variant frequencies for each index cat was equal to the neutral distribution. 406The code to replicate these results can be found in the `SFS.ipynb` Jupyter Notebook in the 407