INTRODUCTION
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is a novel
Betacoronavirus and is the causative agent of the ongoing COVID-19 pandemic. SARS-CoV-2 was first identified in December of 2019 in Wuhan, China in a patient with severe respiratory disease who was a worker at a seafood and live animal wet market (
1). The virus spread rapidly worldwide following its discovery. Animal coronaviruses (CoVs) were first identified in the 1930s, while the first human coronavirus was identified in 1967 (
2). Since then, every human CoV has been zoonotic in origin, including Middle East respiratory syndrome coronavirus (MERS-CoV) and SARS-CoV-1. SARS-CoV-2 is suspected to have originated in a bat and has been shown to infect several animal species via reverse zoonotic spread of a pathogen from humans to animals (
3). This is in part attributable to the fact that the host receptor, angiotensin I converting enzyme II (ACE2), is highly conserved across species and abundantly expressed in target tissues in susceptible mammals (
4). There are at least 31 animal species that are known to be susceptible to SARS-CoV-2, including mink (
Neovison vison), white-tailed deer (
Odocoileus virginianus), domestic cats (
Felis catus domesticus), domestic dogs (
Canis lupus familiaris), nondomestic felid species (genus
Panthera), and golden hamsters (
Mesocricetus auratus) (
5,
6). These species have become infected via reverse zoonosis on multiple occasions, and all but domestic dogs seem to be capable of spreading SARS-CoV-2 between each other or potentially to humans (
5). There is concern that viral mutations may accumulate within these species, in particular, mink and white-tailed deer, which have been shown to spread SARS-CoV-2 rapidly within populations (
7,
8). The virus could potentially spread back into humans from these species, bringing with it new mutations that affect viral pathogenicity, virulence, and existing immunity in humans (
9). Because of the implications of zoonotic and reverse zoonotic spread of this virus, it is important to understand the social and environmental components which created the ideal environment that catalyzed the initial zoonotic spillover of SARS-CoV-2. The COVID-19 pandemic has highlighted the need for an improved One Health approach to disease surveillance—not only to continue to combat and recover from this pandemic but also to prevent future pandemics. One Health is defined by the World Health Organization as the “integrated, unifying approach that aims to sustainably balance and optimize the health of people, animals and ecosystems” (
10). In this review, we discuss spillover and spillback of SARS-CoV-2, the role animals have played in this pandemic, and the relationship of emerging infectious diseases with modern day human-animal-environment tripartite interactions (
Table 1).
ZOONOTIC SPILLOVER—HOW OUR INTERACTIONS WITH ANIMALS LED TO THE SPILLOVER OF SARS-CoV-2
Human coronaviruses characteristically have a zoonotic origin. Related to SARS-CoV-2 are the viruses MERS-CoV and SARS-CoV, which likely originated from bats before infecting an intermediate host and then spilling over into humans (
3). The origin of SARS-CoV-2 is still unclear, though it has been established that the virus does have a zoonotic origin (
11). Current hypotheses proposed about the origin of SARS-CoV-2 include the following: (i) direct spillover from bat species to humans after a random mutation event occurred, (ii) a less-virulent progenitor circulated and gained a number of mutations contributing to SARS-CoV-2’s current virulence; or (iii) an intermediate host became infected with the virus by a bat and then transmitted to a human (
12). A potential intermediate host has not been confirmed, though animals susceptible to sarbecoviruses (subgenus), such as palm civets, mink, racoon dogs, and red foxes, were all sold at wet markets including the Huanan Market, in Wuhan, China at the time of emergence of SARS-CoV-2 in 2019 (
13–15). Early COVID-19 infections were initially discovered in central Wuhan, with a high density of infected individuals residing near or working in the Huanan Market (
15). These early infections were classified into two lineages, A and B, which were genetically distinct at two key locations within the viral genome (
16). Lineage B emerged before lineage A and had a direct epidemiological link to the Huanan Market (
16). Early lineage A isolates did not have a direct link to the market, but were isolated in patients who had resided within 2.31 km from the market. Lineage A was later isolated from an environmental sample in the market (
15,
16). Lineage A was two mutations closer to known bat coronaviruses than lineage B (
16). Epidemiologic simulations did not support one spillover event into humans, rather, they supported multiple spillover events between late October and mid-November 2019 (
16). For multiple spillover events to occur, there had to have been a persistent source of introduction, and the Huanan Market has been suggested as the epicenter for these events (
15). Positive environmental samples from the market (cages, carts, and feather removers) were linked with both animal sales and human infections of SARS-CoV-2. This being said, no animals from the market tested positive for SARS-CoV-2, and the question stills stands as to whether the proximal origin of the virus lies within the Huanan Market, or if the associated cases are a result of an outbreak from a previously infected human bringing the virus to the market (
15,
17).
Interestingly, one coronavirus isolate, RaTG13, identified in 2013 in horseshoe bats (
Rhinolophus affinis), has a 96.3% genetic similarity to SARS-CoV-2 (
18,
19). This close genetic relationship supports that bats are a likely reservoir for SARS-CoV-like coronaviruses; however, the receptor binding domain (RBD) of the RaTG13 spike (S) gene is distinct from SARS-CoV-2 and is considered an unlikely SARS-CoV-2 progenitor (
20). Coronaviruses are known for their high rate of recombination, so it is not unreasonable to hypothesize that the emergence of SARS-CoV-2 is the result of one or more recombination events. Sequences from the RaTG13 genome, which were not synonymous with the SARS-CoV-2 genome, more closely matched those of pangolin coronavirus isolates, and thus it was hypothesized that SARS-CoV-2 is the result of recombination between existing pangolin and bat CoVs, with pangolins acting as an intermediate host (
21). Contradictory to this hypothesis, Boni et al. performed a phylogenetic analysis of lineage divergence as a result of recombination of RBD sequences of several related CoVs over time and found that sequences within the RBD of the S protein were present in a common progenitor for SARS-CoV-2, RaTG13, and two pangolin CoVs isolated in 2019 (
20). The analysis also showed that the RBD sequence underwent recombination in RaTG13 preceding the divergence of the RaTG13 lineage from the other two lineages (
20). This recombination event and subsequent divergence is hypothesized to have occurred at least 40 years ago, with RaTG13 and the SARS-CoV-2 progenitor evolving independently (
20). The two lineages remained genetically similar; however, human ACE2 binding affinity was maintained in the SARS-CoV-2 progenitor and not in RaTG13 (
20). This is evidence that SARS-CoV-2 may be a result of direct zoonotic spillover from bats into humans and may not be due to a recombination event (
20). SARS-CoV-2 has been shown to have incredible host adaptability once a species jump has occurred, with mutations occurring from positive selection within 10 days from experimental inoculation (
9). Therefore, it is possible that selective pressure within a human host may have caused the virus to quickly adapt to a human host after spillover from bats.
The hypothesis that SARS-CoV-2 originated in a lab is still widely discussed, though the possibility of such an occurrence is considered to be highly unlikely. Keusch et al. provides important insight into the possible origins of SARS-CoV-2 in this regard. These authors point out that ancestral bat CoVs (RaTG13, RmYN02, BANAL 52, etc.) with high sequence homology to SARS-Cov-2 do not bind to human ACE2 receptors nor do they have a furin cleavage site (FCS) within the S protein. FCSs are important for cell entry of SARS-CoV-2, and while some ancestral CoV’s possess FCS-like motifs, it is more suggestive that cleavage sites may have evolved over time (
22). Arguments that the FCS is the result of genetic engineering for gain-of-function experiments is a common argument for the theory of a “laboratory leak,” though FCSs are present in other naturally occurring viruses, including coronaviruses (
23). Additionally,
in vitro efforts failed to detect FCS evolution from related CoVs during passage in human cell cultures, indicating that it is unlikely that SARS-CoV-2 originated from passage within laboratory cell cultures (
22).
Bat-human interactions are not uncommon and have been shown to occur in rural areas and villages near bat caves in southern China (
24,
25). Li et al. found low seropositivity rates (0.6% of 1,497 residents) for known bat coronaviruses in residents living in the Yunnan, Guangxi, and Guangdong districts in China (
24). Factors that increased the likelihood that a resident would be seropositive for bat coronaviruses include eating raw or undercooked carnivores, interactions with poultry, socioeconomic status, and finding bats or rodents in the house (
24). While these spillover events are rare, it shows the importance of improving surveillance of novel SARS-CoV-like viruses in high-risk areas where humans may have increased interactions with bats and consumption of animals which may have come from wet markets (
24). This is of growing importance as populations grow and expand into the habitats of wild animals, which will increase the frequency of interactions between humans and animals known to carry coronaviruses and other zoonotic diseases.
CONFOUNDING ENVIRONMENTAL FACTORS: A RECIPE FOR INCREASED RISK OF NOVEL ZOONOTIC DISEASES SUCH AS SARS-CoV-2
A One Health perspective reveals potential environmental and anthropogenic factors which may have influenced the spillover of SARS-CoV-2 from its reservoir species to humans (
26). Proposed environmental disturbances and human behaviors that promote emergence of novel pathogens include land use change, intensified livestock production, and wildlife trade. Land use changes such as alteration of land for agricultural purposes, urban infrastructure expansion, and harvesting of natural resources result in enhanced human-wildlife contacts (
26). In particular, such changes often result from deforestation and habitat fragmentation, which causes alterations in wildlife ecology (
26,
27). Decreased biodiversity may result in decreased population-level immune function (e.g., when naive, genetically homogenous populations are exposed to novel pathogens) and increased probability of exposure to disease (
27). Habitat fragmentation also increases the likelihood of human-animal or wild- and domestic-animal interactions, increasing the potential for direct or indirect spillover (
27). For example, a recently published, long-term study described persistent changes in bat behaviors that mimic those witnessed during periodic food shortages. The authors posit that these behavioral changes increase contact between bats and domestic horses, a known intermediate host for Hendra virus (
28). Intensified livestock production also increases the proximity of domestic livestock to wild animals and may facilitate exchange of disease between these species and subsequently to humans (
26,
27). Further, livestock are being kept in increasingly dense numbers, a factor which promotes the spread of infectious disease and decreases host genetic diversity due to artificial selection and breeding, which in turn increases susceptibility to disease (
26,
27).
When coronaviruses are transmitted between wild and domestic animals, the chance of recombination between endemic and newly introduced viruses is increased, leading to emergence of novel viruses, which can infect multiple species (
26,
27). Coronavirus transmission may also occur in the illegal wildlife trade and in wet markets, locations where fresh meat, fish, and produce are sold in highly populated areas (
29). Wet markets are the destinations of some animal species caught in illegal wildlife trade that are valued for meat, fur, or as pets (
14). Nevertheless, they are important components of economies in east and southeast Asia, central and eastern Europe, and much of Africa (
30). These markets are mostly unregulated (despite laws in place to prevent illegal wildlife trade), and large numbers of live animals are sold or killed and processed onsite and sold as meat (
14). The convergence of many wild and domestic animals as well as humans creates an ideal environment for the spread of zoonotic disease, especially in sites where hygiene is overlooked (
30). Additionally, illegal wildlife trade displaces a diverse number of vertebrate species from differing habitats worldwide, namely, civets, racoon dogs, foxes, badgers, rodent species, primate species, and numerous bird and reptile species (
14,
30). Species that would not traditionally overlap in distribution are often closely housed and shipped together with little regard for biosecurity (
14). This introduces pathogens to naive animal species that may potentially amplify pathogens or become novel reservoir hosts. Because illegal wildlife trade and wet markets have been implicated in igniting viral epidemics and pandemics, it is vital that these establishments undergo regulatory reform in addition to pathogen surveillance to prevent spread of existing and emerging pathogens such as SARS-CoV-2 (
30).
ACE2 RECEPTOR ROLE IN SPECIES SUSCEPTIBILITY
SARS-CoV-2 is known to infect several species besides humans both naturally and experimentally (
4). This capability is largely attributable to the host cellular receptor, ACE2, which is highly conserved among mammals (
4). The SARS-CoV-2 S glycoprotein has two subunits, S1 and S2. S1 binds to specific amino acid sequences of ACE2 to initiate the process of cell infection, while S2 changes to a more stable configuration after S1 binding (
31).
In silico studies have predicted that mammal species which have homologous amino acid sequences within the ACE2 protein are more susceptible to SARS-CoV-2, though these predictions are not always reflective of true
in vivo susceptibility (
32–34). Wei et al. (
34) found that when specific critical binding sites within the ACE2 protein of mammals were compared to the same sites in the human ACE2 protein, predictions in species susceptibility were more accurate than analyzing the entire ACE2 sequence.
In addition to ACE2 binding affinity, expression levels of ACE2 influence susceptibility to SARS-CoV-2. Sun et al. performed a transcriptomic analysis of ACE2 mRNA expression in different tissues of humans and some mammal species (
35). They found that in humans, ACE2 is expressed in the testis, heart, small intestines, kidneys, corneal epithelium, and type II alveolar cells in the lungs (
35). They also reported that mice express ACE2 in the kidneys, heart, and intestine, but the protein is not expressed at high levels in the lungs (
35). This suggests that wild-type mice may not be a good model for respiratory infection of SARS-CoV-2 due to low ACE2 expression in the lungs (
35,
36).
Chinese rufous horseshoe bats and Malayan pangolins have high expression of ACE2 in lung tissues of both species (
35). In contrast, ACE2 in dogs and cats is localized to the kidneys, heart, and liver in addition to skin and retinas (
35). Additionally, cats and ferrets had high levels of ACE2 in the lungs, making them good potential models for infection with the virus, consistent with observations that these hosts are susceptible and may recapitulate human-like disease in natural settings (
35). Cattle, pigs, goats, and rabbits were found to have low expression of ACE2 in critical tissues such as the lungs and are, therefore, considered less susceptible to SARS-CoV-2 based on transcriptomic studies (
35). The findings of this study are consistent with current literature and reports of known animal infections. Using transcriptomics to analyze expression of the ACE2 protein in addition to protein structure can help to direct research and experimental infection in species suspected to be susceptible to SARS-CoV-2 infection and may help to direct biosecurity protocols to prevent spillover events.
SUSCEPTIBILITY OF DIFFERENT ANIMALS TO SARS-CoV-2
Understanding susceptibility of certain mammals to SARS-CoV-2 is a biological imperative for pandemic research. Such studies could reveal the original reservoir species of the COVID-19 pandemic or a potential intermediate host. Gaining detailed insight into the host range of SARS-CoV-2 will aid in understanding of the ecological factors that impact the COVID-19 pandemic. This understanding is crucial as human-animal interactions in domestic, educational, research, agricultural, conservational, and recreational industries facilitate opportunities for spillback (
11).
To comprehend the scope of animals susceptible to SARS-CoV-2, many current studies need to be assessed. A global open access database called SARS-ANI VIS reports on SARS-CoV-2 events in animals (
6). As of November 2022, 31 species have been reported in the database as testing positive for the presence of SARS-CoV-2 viral RNA or antibodies (
6). In addition to the notable species mentioned in the introduction, gorillas (
Gorilla gorilla), hippopotamus (
Hippopotamus amphibius), hyena (
crocuta crocuta), mule deer (
Odocoileus hemionus), Eurasian beaver (
Castor fiber), lynx (
Lynx canadensis/Lynx lynx), manatee (
Trichechus manatus), black-tailed marmoset (
Mico melanurus), giant anteater (
Myrmecophaga tridactyla), binturong (
Arctictis binturong), Asian small-clawed otter (
Aonyx cinereus), ring-tailed coati (
Nasua nasua), and mandrills (
Mandrillus sphinx) have been added to the list of reported species (
6). The number of mammals infected and/or exposed is ever-increasing (
Table 2).
Testing methods used to indicate active infection are reverse transcriptase PCR (RT-PCR) or PCR-based tests. Immune reactions are typically indicated by a serology (antibody detection) test. RT-PCR results dominate the extent to which prevalence has been characterized; however, additional studies are needed to accurately gauge quantitative pathogen burdens within each species affected.
Islam et al. reported that the Egyptian fruit bat, ferret, tree shrew, mouse, rabbit, hamster, and primates were susceptible by experimental infection (
11). Animals susceptible to natural infection include tiger, puma, lion, leopard, gorilla, mink, and domestic dog and cat. Chickens, ducks, guinea pigs, turkeys, and pigs have been resistant to experimental infection (
11).
A review by Rutherford et al. compiled SARS-CoV-2 susceptibility data from
in silico,
in vitro,
in vivo, and epidemiological studies performed since the start of the COVID-19 pandemic (
37). Their analysis found contrasting susceptibility of pigs to SARS-CoV-2, citing an
in vivo experimental infection in which two pigs tested positive for the virus via reverse transcriptase quantitative real-time PCR (RT-qPCR) (
37,
38). Cattle were also reported to have low susceptibility, but due to significant interactions with humans, spillback events remain a concern (
37). Other susceptible species include racoons, skunks, old world rodents (who are more susceptible to recent SARS-CoV-2 variants), otters, and hyenas (
37,
39).
REVERSE ZOONOSIS
In the broader context of SARS-CoV-2 transmission, it is important to contrast “reverse zoonosis” with “traditional zoonosis.” Reverse zoonosis is an area of concern for current and future pandemics, as viruses that are readily transmitted between animal and human hosts can accumulate mutations that lead to increased transmissibility and genetic variation (
9). Although it is known that SARS-CoV-2 has been found to infect a wide range of animals, including ferrets, mink, Syrian hamsters, bats, white-tailed deer, pangolins, felines, dogs, and rabbits, other suspected animals appear resistant to SARS-CoV-2 (such as most poultry) (
11). This discrepancy in predictive host susceptibility highlights the need for further investigation.
Social structure of different animal populations creates differences in transmission patterns during a reverse zoonosis event. Companion animals comprise communities where direct exposure often takes place with animal companions as well as their human counterparts. Likewise, domesticated livestock, wild animals in captivity, wild animals in conservation areas, and wild animals in rural landscapes are additional communities that can be directly or indirectly exposed to emerging infectious diseases.
Humans can have indirect contact with wild animals by the route of abiotic environments. These indirect forms of transmission can be sources of infection to animals in the wild. As an example, testing of local wastewater for SARS-CoV-2 is a way that researchers can track community infection rates. SARS-CoV-2 has been found in wastewater and urban runoff and could serve as a possible indirect form of exposure and transmission to wild animals (
40,
41). The disease ecology of SARS-CoV-2, transmission dynamics, and potential contributions of abiotic factors remain largely unidentified but are extremely important in understanding this pathogen.
KNOWN SARS-CoV-2 SPILLBACK
Since the beginning of the SARS-CoV-2 pandemic, several human-to-animal spillback events have been reported (
Fig. 1). Humans can infect domestic cats and felids in zoos with SARS-CoV-2 (
42,
43). Early in the pandemic, it was not clear what hosts were susceptible to infection as a result of mass human infection. Transmission could take place individually on the domestic level, in captive wild animals, or in wild animals in their natural environment, and these transmission events can come from direct or indirect human interaction.
Medkour et al. in France identified a domestic dog that presented with signs of infection while in close contact with the owners who tested positive for COVID-19 by RT-qPCR in a nasopharyngeal swab (
44). The dog also tested positive after nasal swab, and blood samples were collected and tested for SARS-CoV-2 by RT-qPCR after showing signs a week after the owners tested positive (
44). Antibodies were detected 12 days after the first diagnosis and up to 5 months later using enzyme-linked immunosorbent assay (ELISA) and Western blotting (WB) (
44). Genome sequences from the dog and the owners were identical, and the sequences also showed a match for the variant of concern in the region at that time, B.1.160, also known as the Marseille-4 variant (
44). Transmission of the virus between owner and pet may have a significant impact on the health of both the human and the animals involved; however, more research is needed on SARS-CoV-2 companion animal surveillance as well as transmission to susceptible hosts to understand overall risks of SARS-CoV-2 in domestic spaces.
HUMAN-TO-FELIDS
Felids were some of the first animals shown to be susceptible to SARS-CoV-2 after four tigers and three lions developed mild respiratory symptoms in the Bronx Zoo in New York (
45). PCR testing confirmed SARS-CoV-2 RNA in respiratory and fecal samples from which whole SARS-CoV-2 sequences were generated (
45). This example was a defining moment in the pandemic because it highlighted the initial intersection of transmission from humans to animals. After this time point in the pandemic, researchers turned to testing of other felids, namely, domestic cats.
Several experimental studies have demonstrated domestic cats are susceptible to SARS-CoV-2 and can transmit to other cats. Shi et al. studied the replication and transmission of SARS-CoV-2 in cats inoculated with the virus (
46). Viral RNA was detected in animals between 3 days postinfection (dpi) and 6 dpi, and the transmission study found viral RNA in the naturally exposed cat, suggesting horizontal transmission (
46). A supporting study investigated the nasal shedding of SARS-CoV-2 from cats and subsequent transmission of the virus by exposure of uninfected cats to virus-inoculated cats (
47). Nasal and fecal swabs were collected and assessed promptly for virus isolation on Vero cells (
47). Virus was detected in nasal swabs from all cats, as well as the cats that were cohoused that had no previous infection (
47). Viral titers peaked concomitant with viral shedding at 4 to 5 dpi (
47).
Transmissibility of SARS-CoV-2 among felids is concerning for a few reasons, and there is still little understanding of how it presents in our domestic and wild animals. Additionally, we are not fully aware of how this affects different species of felids. For example, we know that snow leopards (
Panthera uncia) can contract SARS-CoV-2 (
48). A snow leopard in Illinois and three in Nebraska have died of complications after contracting SARS-CoV-2 and more studies need to be conducted to be certain of their risk to COVID-19 (
49,
50). Snow leopards are considered vulnerable on the International Union for Conservation of Nature (IUCN) Red List, and populations are decreasing, which may make this species at extreme risk if wild populations are exposed (
51). This highlights the importance of using proper personal protective equipment (PPE) when handling at-risk animals such as snow leopards or other big cats, for both handler and animal protection.
HUMAN-TO-MINK
Mink are an example of spillback and secondary spillover but also an example of the transmission interface between humans and production animals. Early in the pandemic, mink were found to be susceptible to SARS-CoV-2 infections. Many countries, including the Netherlands, Denmark, the United States, Poland, and Sweden farm large numbers of mink for the international fur industry, and the susceptibility of mink to this virus is cause for concern. Indeed, it is probable that practices of housing high-density populations of farmed mink, coupled with low genetic heterogeneity, have contributed to outbreaks of SARS-CoV-2 on mink farms in many countries, including those mentioned above (
52). In the United States, mink farms typically range in size from a few hundred to approximately 38,000 individuals (
52). Among these farms, mortality rates of mink ranged from 1.8% to 42% during 18 outbreaks in the summer and fall of 2020 (
52).
A study by Oude Munnink et al. demonstrated sustained human-to-animal and back-to-human transmission of SARS-CoV-2 on mink farms in southeastern regions of the Netherlands, detailing the results of the first 16 SARS-CoV-2-positive mink farms (
8). The study combined whole-genome sequencing (WGS) data in combination with diagnostic testing, epidemiological information, and surveillance data (
8). As part of Danish public health measures and an awareness of animal transmission risk, mink farmers, veterinarians, and laboratories committed to report symptoms in mink to the Netherlands Food and Consumer Product Safety Authority (NFCPSA), resulting in the development of an extensive surveillance system (
8). Because of the surveillance efforts in the Netherlands, 1,750 SARS-CoV-2 isolates were sequenced from patients from parts of the country (
8). Owners and employees were tested by either serological assays or by RT-qPCR on mink farms with SARS-CoV-2-positive animals. In total, 66 of the 97 individuals demonstrated evidence of current or previous SARS-CoV-2 infection (
8). Forty-three of 88 upper respiratory samples tested positive by RT-qPCR and 38 of 75 serum samples tested positive for SARS-CoV-2 antibodies (
8). Some employees were symptomatic prior to testing positive, and when mink and humans tested positive, WGS was performed (
8). This study found that human sequences were almost identical to the mink sequences and that the virus was initially transmitted by humans and since then circulated among mink at the different farms (
8).
The susceptibility of mink to SARS-CoV-2 and transmissibility among mink has provided great insight and data. WGS analysis and epidemiological data helped to inform the public that SARS-CoV-2 was able to sustain transmission from human to nonhuman then back to human hosts. This led to researchers looking for the possibility of higher rates of mutation and types of mutations among mink populations. Therefore, understanding the risk of mink and other mustelids as a transmissible species needs to be further investigated (
53).
HUMAN-TO-WHITE-TAILED DEER
White-tailed deer have been a topic of recent interest in SARS-CoV-2 research because of the high numbers of infected individuals and because they migrate freely over distances and across state boundaries. A study in July of 2021 showed that 152 out of 385 (40%) free-ranging white-tailed deer sampled had antibodies for SARS-CoV-2 (
54). Samples were taken from deer in Michigan, New York, Pennsylvania, and Illinois, where the initial concern was for infected deer circulating virus within their populations (
54). A similar study by Hale et al. tested nasal swabs from 360 free-ranging white-tailed deer in northeast Ohio from January to March of 2021. This study reported 129 of 360 (35.8%) deer as positive via RT-qPCR and was the first study to declare PCR-confirmed results of naturally infected deer (
55). Three different SARS-CoV-2 lineages, B.1.2, B.1.596, and B.1.582 were identified with the use of WGS. Collections were taken between January and March of 2021, following a large outbreak in SARS-CoV-2 human cases in the same Ohio community where B.1.2 was a dominant variant in the population (
55). Collections for sequencing were done at 6 of the 9 sites, and B.1.2 was identified at four sites (
55). However, phylogenetic analysis found no evidence of transmission across deer populations, likely because the B.1.2 virus was genetically distinct (
55). These studies document four separate human-to-deer transmission events (
55). A study of 132 captive and 151 free-ranging deer in Iowa reported that 94 of the total 283 (33%) deer samples were positive for SARS-CoV-2 by RT-qPCR (
7). Subsequent WGS of the 94 positive samples along with geographic distribution data suggests not only human-to-deer transmission can occur but also deer-to-deer transmission (
7).
These studies suggest that transmission from human-to-deer and deer-to-deer likely persist. Deer that can transmit virus within a herd can also potentially transmit SARS-CoV-2 to other herds, or potentially other species, offering rationale in pursuing surveillance of white-tailed deer in future studies.
POTENTIAL SECONDARY SPILLOVER
Potential secondary spillover of SARS-CoV-2 from animals-to-humans is of concern for public health and One Health overall. As stated before, new variants can emerge from infected animal populations. Surveillance of infected or susceptible animal populations would inform agencies on next steps in prevention of outbreaks. There are other pathogens for which there have been documented reverse zoonosis events; however, Fagre et al. cite that “secondary spillover” is unique to the SARS-CoV-2 pandemic (
56). Mink, as an incidental host, provide an initial understanding of secondary spillover into human populations, and experimental and natural infections of felids have also provided insights into this phenomenon.
A study of mink farms in the Netherlands used WGS, phylodynamic analysis, and epidemiological information to show that 68 of 126 farms tested positive for SARS-CoV-2 between 24 April and 4 November 2020 (
57). Employees tested positive for SARS-CoV-2 by RT-qPCR at 41 of the 68 mink farms that had also tested positive (
57). Genome sequences were collected from 64 of the 68 infected mink farms (
57). Epidemiological data and phylogenetic sequencing showed that employees of mink farms were infected with mink strains of SARS-CoV-2 rather than other strains circulating among humans in that community at the time (
57). These genetic signatures also pointed to independent introduction of the virus from humans to the mink farms (
57). This in-depth One Health approach led the researchers to study farm-to-farm interactions and to be able to link infection clusters (
57). As this study points out, the emergence of novel variants may affect the viral host range as well as continuous transmission across species, possibly establishing a new reservoir for continuous spillback and secondary spillover (
57).
Other potential animal hosts capable of secondary spillover include companion felids. In Thailand, there has been a case of suspected cat-to-human transmission (
58). A previously healthy veterinarian visited a hospital reporting a history of fever, cough, and nasal discharge. Examination by a physician and a chest radiograph showed no notable signs of COVID-19 (
58). This patient became symptomatic 3 days after handling a cat in a veterinary setting (
58). The cat had been in close contact with its owners who were confirmed positive for SARS-CoV-2 by RT-PCR (
58). This patient, a veterinarian, was examining the cat, who was asymptomatic, and collected nasal and rectal swabs specimens (
58). During the nasal swabbing the sedated cat sneezed in the veterinarian’s face (
58). While the veterinarian and additional helpers were wearing PPE, and the sample collection lasted approximately 10 min. The veterinarian later tested positive for SARS-CoV-2 by RT-PCR with subsequent WGS to confirm the variant (
58). Contact tracing of those close to the veterinarian did not test positive for SARS-CoV-2 (
58). All parties, the veterinarian, the staff, the cat, and the cat’s owners, were all tested via RT-PCR prior to sequence determination (
58). The veterinarian, owners, and cat were all confirmed positive with the same viral genome, B.1.167.2 (
58). It is thought that because the veterinarian had no previous contact with the cat or its owners, the likely cause of infection was from the cat interaction (
58).
It is evident that reverse zoonosis is of high concern when studying the circulation of SARS-CoV-2 through human and animal populations. Mink, as reviewed, are potential sources for spillback and secondary spillover given their disposition as highly susceptible hosts. Although human-mink contact may be relatively rare, the context of exposure (high-density populations) may increase the likelihood of SARS-CoV-2 transmission. Conversely, domestic animals, such as cats and dogs, are of concern due to their close interactions with humans. Once we can understand the host range and disease variability of animal populations, then we can better protect vulnerable groups and possibly predict where the next variant of concern may come from. As we have shown, One Health approaches to disease surveillance efforts and an understanding of susceptible hosts can lead us to better pandemic preparedness measures.
ANIMALS AS A POTENTIAL SOURCE OF MUTATIONS AND VARIANTS
The COVID-19 pandemic has seen the emergence of several prominent SARS-CoV-2 variants as the virus continues to spread and evolve. The emergence of new variants is driven largely through mutations acquired from human infections; however, the possibility of variants emerging from infected animal species should not be overlooked given that a diverse number of species are susceptible to the virus (
11). One instance of a known variant to have likely emerged from an animal species is the Marseille-4 variant, B.1.160, which was previously discussed infecting a dog and is thought to have emerged from a French mink farm (
59). The first human infection of this variant occurred in March of 2020 and had spread outside of Marseille into other countries including New Zealand, Australia, and the United States by September of 2020 (
59). The variant was able to infect humans which had already been infected with SARS-CoV-2 as recently as 4 months earlier, and its success in nonnaive patients may be attributed to a mutation in the RBD of the S protein to confer weak cross-reactivity with antibodies generated against previous SARS-CoV-2 variants (
59). The variant appeared at a time when SARS-CoV-2 was limited in certain regions of France, one region with a large population of wild mink and another with a large mink farm (
59). The mink farm, located in Eure-et-Loire, had an outbreak of SARS-CoV-2 in May of 2020 and was forced to cull the population (
59). WGS data of a SARS-CoV-2 mink in Eure-et-Loire sampled in November 2020 was identical to the Marseille-4 variant, confirming the variant to be of mink origin (
59). Since it was discovered that SARS-CoV-2 could rapidly spread in mink and infect humans through secondary spillover, it has been shown that the virus has a faster mutation rate within mink populations than in human populations (
8). In humans, SARS-CoV-2 gains a mutation to fixation about once every 2 weeks (
8). As evidenced by the high genomic diversity of the virus within mink farms, though, the rate may be as little as 1 week after the virus was introduced to the mink (
8).
White-tailed deer is another species of concern for evolution of new mutations and variants (
7). Genome analysis of a group of white-tailed deer in ON, Canada showed a divergent lineage of SARS-CoV-2 and host adaptation following spillover into the population (
60). The lineage was shown to have mutations which had not been previously characterized (
60). This lineage has been detected in one sequenced human sample, which is suspected to have occurred via secondary spillover of SARS-CoV-2 from white-tailed deer (
60). It is unlikely any other secondary spillover events have occurred or that this lineage has any evolutionary advantage for spread among humans (
60). Continued surveillance and sequencing of SARS-CoV-2 in white-tailed deer populations is needed to detect new mutations and variant emergence.
There is additional concern regarding variant selection in companion animals susceptible to SARS-CoV-2 due to their close interactions with humans. Genome analysis of SARS-CoV-2 isolates from experimentally infected dogs, cats, ferrets, and hamsters was performed by Bashor et al. to identify variant emergence in these species during infection (
9). All animals were inoculated directly, except for one cat who was infected via contact with another cat in the study (
9). Fourteen mutations were found, with most located in the S protein and nonstructural genes. Variants of concern within the S protein arose in dogs, cats, and hamsters (
9). Mutations arising in dogs were largely within replicase genes, indicating selection for viral replication (
9). Positive selection occurred most significantly in hamsters and dogs (
9). Cats produced fewer synonymous variants and had higher viral titers than dogs, indicating that SARS-CoV-2 infects cats more similarly to humans than dogs, though more variants of concern arose from cats such as the D614G mutation in the S protein (
9). These findings showed rapid emergence of variants when the virus infects another host species and highlights accelerated host adaptation of SARS-CoV-2 (
9). Another recent study by Bashor et al. found on average 12 variants per cat when experimentally infecting 3 cohorts (
n = 23), indicating a 10-fold increase to what is seen in human patients (
61). From full viral genomes, they found 118 unique variants at ≥3% allele frequency, 59% of which were nonsynonymous single nucleotide variants (SNPs) (
61). These results highlight the potential capacity of variants that are coming from domestic cats alone. Therefore, it is in the interest of the scientific community to conduct surveillance of viral evolution using WGS within animal species in close contact with humans, especially cats, to prevent potential viral recombination and emergence of novel variants.
It is important to consider that while variants of concern may arise as a result of a host species jump, human-to-human infections also introduce many SARS-CoV-2 mutations, especially in those who are immunocompromised with persistent infection (
62). This is described in a retrospective cohort study by Hettle et al., in which WGS was performed on oropharyngeal samples from six patients who tested positive by RT-PCR for more than 56 days (
62). It was found that mutations quickly arose in these patients, including those within the spike protein that induce changes to B- and T-cell epitopes (
62). This presents the risk of evasion of existing population immunity, just as in spike mutations described in mammal infections (
9).
CONCLUSION
SARS-CoV-2 has proved itself as a generalist virus, which has complicated the COVID-19 pandemic. This is in part due to the fact that viral receptor ACE2 is conserved across many species of mammals (
4). Consequently, the capacity of SARS-CoV-2 to infect many species, including some companion and zoo animals in close contact to humans, has created concern for viral evolution and secondary spillover into humans. The virus has demonstrated host adaptive capabilities, mutating under selective pressure in new hosts, and in some cases, these mutations are maintained when spillover occurs back into humans as is the case shown in Dutch mink farms and in the spread of the Marseille-4 variant (
57,
59). As a One Health initiative, continued use of WGS for positive PCR samples from susceptible animals as a surveillance measure could be an important action to prevent the emergence of novel SARS-CoV-2 variants. This is of particular importance in domestic species such as dogs and cats, which are in close contact with humans, as well as secondary reservoir species such as mink and white-tailed deer, where spread occurs independently of humans once introduction occurs, and the likelihood of emergence of novel variants of concern may be increased (
7,
8). Vaccines in species of concern have been suggested as a One Health component of combating this pandemic, though feasibility of the implementation of widespread vaccination of animal populations may be difficult, and more research is needed to determine the effectiveness of implementing vaccination programs (
63).
Approximately 60% of existing human infectious diseases, and 75% of new emerging infectious diseases have a zoonotic origin (
64). Ebola, Hendra, Nipah, MERS-CoV, and SARS-CoV originated from bat-to-human transmission, and it is likely that the next novel virus of public health concern will be of bat origin. Because of the frequency of zoonotic spillover of coronaviruses from old world bats in the last 20 years, increased surveillance efforts and analysis of coronavirus evolution in known reservoir species is justified to better predict viruses that have the potential to spill over into humans and other animal species.
Due to occasional spillover of bat CoVs into human populations in rural China (
24), surveillance in these at-risk communities is important to detect which known bat CoVs are infecting humans. This surveillance can be extended into wildlife trade and wet markets, since the origin of the last three significant human coronaviruses have ties to that industry. Additionally, stricter regulations on species sold and trafficked must be implemented to prevent additional spillover events. These One Health strategies, if applied, can continue to mitigate spread of SARS-CoV-2 in animals and humans, detect future spillover events, and prevent spread of novel viruses into humans from animals.