Key research articles related to SARS-CoV-2 in non-human animals

Published June 12, 2020

This in-depth summary was prepared by AVMA staff and will be updated on a regular basis as new research of relevance is published. However, this is not meant to be an exhaustive list of all studies pertaining to SARS-CoV-2 research in animals. The information is tabulated and listed chronologically, with the earliest publication listed first. General descriptions of the primary lines of investigation used in these studies are provided in the SARS-CoV-2 in animals section of the AVMA COVID-19 online resources.

Because of the rapid rate of research on SARS-CoV-2, many studies are being posted on preprint websites (e.g., bioRxiv and medRxiv) before or without being submitted for potential publication in a peer-reviewed journal. These open-access sites allow for rapid dissemination of information and wide sharing of preliminary results, which in turn allows for greater collaboration among scientists from around the world. However, because these sites are open access, results posted there but not yet reaffirmed by additional studies or subject to peer-review may inadvertently be publicized in the general press or via social media as definitive statements and conclusive evidence. Readers are encouraged to pay attention to the source of new information regarding COVID-19 and SARS-CoV-2 and heed disclaimers on non–peer-reviewed platforms. For example, the disclaimer on the bioRxiv website states that papers posted are “preliminary reports and have not been peer-reviewed. They should not be regarded as conclusive, guide clinical practice/health-related behavior, or be reported in news media as established information.”
*Row number corresponds to section number in the summary text that follows this table


No.*
Publication or Posting Date
Journal or
Web Site
Title
Digital Object Identifier (DOI) Link
Notes

1

Feb 2, 2020

bioRxiv

Host and infectivity prediction of Wuhan 2019 novel coronavirus using deep learning algorithm

doi.org/10.1101/2020.01.21.914044

Comparative sequence & deep learning analyses to predict potential hosts of SARS-CoV-2

2

Feb 3, 2020

Nature

A new coronavirus associated with human respiratory disease in China

doi.org/10.1038/s41586-020-2008-3

One of first two papers describing the new human coronavirus, using comparative sequence and molecular modeling analyses

3

Feb 3, 2020

Nature

A pneumonia outbreak associated with a new coronavirus of probable bat origin

doi.org/10.1038/s41586-020-2012-7

One of first two papers describing the new human coronavirus, using comparative sequence and molecular modeling analyses

4

March 19, 2020

Biochemical & Biophysical Research Communications

Spike protein recognition of mammalian ACE2 predicts the host range and an optimized ACE2 for SARS-CoV-2 infection

doi./org/10.1016/j.bbrc.2020.03.047

Comparative sequence analyses and molecular modeling to predict potential host range of SARS-CoV-2

5

Mar 26, 2020

Clinical Infectious Diseases

Simulation of the clinical and pathological manifestations of Coronavirus Disease 2019 (COVID-19) in golden Syrian hamster model: implications for disease pathogenesis and transmissibility

doi.org/10.1093/cid/ciaa325

Molecular modeling & experimental infection/ transmission studies to develop an animal model of human disease

6

Apr 2, 2020

FLI Institute

Novel Coronavirus SARS-CoV-2: Fruit bats and ferrets are susceptible; pigs and chickens are not

Not applicable
(press release available online here)

Experimental infection/ transmission studies; press release of initial results only

7

Apr 3, 2020

bioRxiv

SARS-CoV-2 neutralizing serum antibodies in cats: a serological investigation

doi.org/10.1101/2020.04.01.021196

Serological survey of cats in Wuhan, China during first months of pandemic to determine whether cats had been infected

8

Apr 6, 2020

Cell Host & Microbe

Infection and Rapid Transmission of SARS-CoV-2 in Ferrets

doi.org/10.1016/j.chom.2020.03.023

Experimental infection/ transmission studies to develop an animal model of human disease

9

Apr 8, 2020

Science

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

doi.org/10.1126/science.abb7015

Experimental infection/ transmission studies to develop an animal model of human disease and identify potential intermediate hosts

10

Apr 9, 2020

bioRxiv

Absence of SARS-CoV-2 infection in cats and dogs in close contact with a cluster of COVID-19 patients in a veterinary campus

doi.org/10.1101/2020.04.07.029090

Serological survey to determine whether pets living with people with COVID-19 had been infected

11

Apr 11, 2020

bioRxiv

Potential host range of multiple SARS-like coronaviruses and an improved ACE2-Fc variant that is potent against both SARS-CoV-2 and SARS-CoV-1

 doi.org/10.1101/2020.04.10.032342

Molecular modeling and in vitro studies to identify potential intermediate hosts of SARS-CoV-2

12

Apr 14, 2020

Molecular Biology and Evolution

Extreme genomic CpG deficiency in SARS-CoV-2 and evasion of host antiviral defense

doi.org/10.1093/molbev/msaa094

Molecular modeling to better understand origin of SARS-CoV-2

13

Apr 17, 2020

bioRxiv

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

doi.org/10.1101/2020.04.16.044503

Experimental infection & transmission studies to develop an animal model of human disease

**

Apr 18, 2020

bioRxiv

Broad Host Range of SARS-CoV-2 Predicted by Comparative and Structural Analysis of ACE2 in Vertebrates

doi.org/10.1101/2020.04.16.045302

Comparative sequence analyses and molecular modeling to predict potential host range of SARS-CoV-2

**

Apr 20, 2020

bioRxiv

Broad and differential animal ACE2 receptor usage by SARS-CoV-2

doi.org/10.1101/2020.04.19.048710

Comparative sequence analyses and molecular modeling to predict potential host range of SARS-CoV-2

**

Apr 20, 2020

bioRxiv

Exceptional diversity and selection pressure on SARS-CoV and SARS-CoV-2 host receptor in bats compared to other mammals

doi.org/10.1101/2020.04.20.051656

Comparative sequence analyses and molecular modeling to better understand origin and evolution of SARS-CoV-2

**

Apr 23, 2020

bioRxiv

Coronavirus surveillance of wildlife in the Lao People’s Democratic Republic detects viral RNA in rodents

doi.org/10.1101/2020.04.22.056218

Surveillance of wildlife to determine potential intermediate and reservoir non-human hosts of SARS-CoV-2

**

Apr 24, 2020

bioRxiv

STAT2 signaling as double-edged sword restricting viral dissemination but driving severe pneumonia in SARS-CoV-2 infected hamsters

 doi.org/10.1101/2020.04.23.056838

Animal model of human infection with SARS-CoV-2 and pathogenic mechanisms of viral infection

**

Apr 30, 2020

bioRxiv

Susceptibility of tree shrew to SARS-CoV-2 infection

 doi.org/10.1101/2020.04.30.029736

Experimental infection study to develop an animal model of human disease

**

May 7, 2020

bioRxiv

The SARS-CoV-2-like virus found in captive pangolins from Guangdong should be better sequenced

doi.org/10.1101/2020.05.07.077016

Comparative sequence analyses to identify potential intermediate hosts of SARS-CoV-2

**

May 8, 2020

bioRxiv

Comparison of SARS-CoV-2 spike protein binding to human, pet, farm animals, and putative intermediate hosts ACE2 and ACE2 receptors

doi.org/10.1101/2020.05.08.084061

Comparative sequence analyses and molecular modeling to predict potential host range of SARS-CoV-2

**

May 2020

Lancet Microbe

Comparative tropism, replication kinetics, and cell damage profiling of SARS-CoV-2 and SARS-CoV with implications for clinical manifestations, transmissibility, and laboratory studies of COVID-19: an observational study

doi.org/10.1016/S2666-5247(20)30004-5

In vitro studies

**

May 10, 2020

Current Biology

A novel bat coronavirus closely related to SARS-CoV-2 contains natural insertions at the S1/S2 cleavage site of the spike protein

doi.org/10.1016/j.cub.2020.05.023

Comparative sequence analysis

**

May 10, 2020

bioRxiv

Prediction analysis of SARS-COV-2 entry in livestock and wild animals

doi.org/10.1101/2020.05.08.084327

Comparative sequence analysis

**

May 11, 2020

bioRxiv

SARS-CoV-2 spike protein predicted to form stable complexes with host receptor protein orthologues from mammals, but not fish, birds or reptiles

doi.org/10.1101/2020.05.01.072371

Comparative sequence analysis and molecular modeling

**

May 13, 2020

bioRxiv

Origin of Novel Coronavirus (COVID-19): A Computational Biology Study using Artificial Intelligence

 doi.org/10.1101/2020.05.12.091397

Comparative sequence analysis and artificial intelligence to better understand possible origin and evolution scenarios for SARS-CoV-2

**

May 13, 2020

New England Journal of Medicine

Transmission of SARS-CoV-2 in domestic cats

doi.org/10.1056/NEJMc2013400

Experimental infection/ transmission study in cats to identify potential intermediate hosts of SARS-COV-2; letter to editor (may not be peer reviewed)

**

May 14, 2020

Nature

Infection of dogs with SARS-CoV-2

doi.org/10.1038/s41586-020-2334-5

Course of natural infection in 2 dogs in Hong Kong exposed to SARS-CoV-2 from their owners, who were hospitalized with COVID-19; see also Section XXXX in AVMA in-depth summary of natural infections in animals.

**

May 18, 2020

bioRxiv

SARS-CoV2 infection in farmed mink, Netherlands, April 2020

doi.org/10.1101/2020.05.18.101493

Identification of SARS-CoV-2 in farmed mink in The Netherlands; see also Section 1 in AVMA in-depth summary of natural infections in animals.

**

May21, 2020

medRxiv

Detection of SARS-CoV-2 in pets living with COVID-19 owners diagnosed during the COVID-19 lockdown in Spain: A case of an asymptomatic cat with SARS-CoV-2 in Europe

doi.org/10.1101/2020.05.14.20101444

Screening of pets owned by people with COVID-19 for infection with SARS-CoV-2 in an area of Spain with high viral transmission rates.

**

May 22, 2020

bioRxiv

A previously uncharacterized gene in SARS-CoV-2 illuminates the functional dynamics and evolutionary origins of the COVID-19 pandemic

doi.org/10.1101/2020.05.21.109280

Comparative sequence analyses and evolutionary virology to identify potential intermediate hosts of SARS-CoV-2

**

May 23, 2020

bioRxiv

Predicting wildlife hosts of betacoronaviruses for SARS-CoV-2 sampling prioritization

doi.org/10.1101/2020.05.22.111344

Statistical and molecular modelling to develop priority recommendations for sampling potential wildlife hosts of SARS-CoV-2

**

May 25, 2020

medRxiv

Rising evidence of COVID-19 transmission potential to and between animals: do we need to be concerned?

doi.org/10.1101/2020.05.21.20109041

Retrospective case study report of naturally acquired SARS-CoV-2 infections in domestic animals

**

May 29, 2020

bioRxiv

Pathogenesis, transmission and response to re-exposure of SARS-CoV-2 in domestic cats

doi.org/10.1101/2020.05.28.120998

Experimental infection and transmission studies in cats and dogs

**Summary not yet available

Section 1 : Host and infectivity prediction of Wuhan 2019 novel coronavirus using deep learning algorithm

Posted February 2, 2020 to bioRxiv (non–peer reviewed). This paper describes computer-based studies, using deep learning algorithms to predict potential hosts of the 2019-nCoV (now known as SARS-CoV-2019) as a means to aid in determining which animals may have been involved with early transmission (spillover) to humans. All viral sequences (both DNA and RNA viruses) with host information available that were released before 2018 were used as the training set for the deep learning system; all released after 2019 were the testing set. Genomic sequences of the first six isolates of SARS-CoV-2 that had been submitted to GISAID (5 isolates) or GenBank (1 isolate) were then analyzed, using this deep learning system they developed. The SARS-CoV-2 isolates had been obtained from human patients living in Wuhan, China who developed severe respiratory disease in December 2019 or early January 2020. The summaries of the articles that describe these 6 isolates are found in Sections 2 and 3 below.

The authors conclude from the results of the analyses, using deep learning algorithms and comparison of genetic and putative protein sequences, that:

  • SARS-CoV-2 has a closer phylogenetic relationship with SARS-CoV and bat SARS-like coronaviruses than with other coronaviruses
  • The predicted genes and annotated proteins of SARS-CoV-2 match the genes and encoded proteins for other SARS-like coronaviruses.
  • The 6 isolates are more similar to two SARS-like coronaviruses of bats than to SARS-CoV or other SARS-like coronaviruses.
  • Canine, porcine, mink, tortoise and feline viruses, albeit not coronaviruses of those species, show a closer infectivity pattern to SARS-CoV-2, compared with other viruses in the database used to create the algorithms. Pattern of infectivity of the mink and bat viruses were the most similar to that of the predicted infectivity for the new coronavirus, suggesting that bat and mink may be two candidate reservoir hosts of SARS-CoV-2.

AVMA staff comments: This is a highly theoretical paper that analyzed the host range of viruses from many different viral families, genera, and species to predict potential hosts for the novel coronavirus (SARS-CoV-2). However, it is not at all clear whether predications made on the basis of the same deep learning algorithms are supported by real life. That is, if the same deep learning algorithms were used to predict the host range of another virus (not SARS-CoV-2) for which host range is already known, would the predicted host range match the known host range?

RETURN TO TABLE

SECTION 2: A new coronavirus associated with human respiratory disease in China

Published in Nature online on February 3, 2020 and in volume 589 of the journal on March 12 (peer reviewed). This paper describes the isolation of a new coronavirus from a 41-year-old man who was hospitalized in Wuhan, China on December 26, 2019, 6 days after onset of a severe respiratory syndrome that included fever, dizziness, and cough. Infection with influenza virus and other known viral and microbial pathogens, were ruled out, using various diagnostic tests. The patient worked at a local indoor seafood market that also sold a variety of live wild animals, including hedgehogs, snakes, and badgers but excluding bats. Metagenomic RNA sequencing of a sample of bronchoalveolar lavage fluid from the patient identified a new RNA virus from the family Coronaviridae, which was originally designated WH-Human 1 coronavirus, then 2019-nCoV, and is now known as SARS-CoV-2. Phylogenetic analysis of the complete viral genome (29,903 nucleotides) revealed that the virus was most closely related (89.1% nucleotide similarity) to a group of bat SARS-like coronaviruses (genus Betacoronavirus, subgenus Sarbecovirus) that had previously been found in bats in China. The authors submitted the complete genome sequence of this new coronavirus isolate to GenBank on January 17, 2020 (accession number MN908947). The genome organization of the new isolate was determined by sequence alignment to two representative members of the genus Betacoronavirus: one associated with humans (SARS-CoV) and one with bats (bat SL-CoVZC45). The order of genes (5′ to 3′) was as follows: replicase ORF1ab, spike (S), envelope (E), membrane (M), and nucleocapsid (N). Phylogenetic trees were estimated on the basis of comparisons of the nucleotide sequences of the whole genome, the non-structural protein genes ORF1a and ORF1b, and the main structural proteins encoded by the S, E, M and N genes of SARS-CoV-2 with those of known coronaviruses. The new coronavirus (SARS-CoV-2) clustered with members of the subgenus Sarbecovirus, including SARS-CoV, which is the causative agent of the global SARS pandemic of 2002–2003, and a number of bat SARS-like coronaviruses.

The receptor binding domain (RBD) on the spike protein of the new coronavirus was compared with those of SARS-CoV and bat SARS-like CoVs, which are known to use angiotensin converting enzyme II (ACE2) to enter host cells. Using molecular modeling, the authors predicted the three-dimensional protein structure of the RBD of the new coronavirus and compared it with that of two bat SARS-like CoVs, including one that does not use ACE2 to enter host cells. The predicted protein structures of the RBD domains of both the new human coronavirus and the bat SARS-like CoV that uses ACE2 to enter host cells were closely related to the RBD protein structure of SARS-CoV. These three similar RBD structures were different from that predicted for the bat SARS-like CoV that did not use host ACE2 to enter cells. In addition, the N terminus of the spike protein of the new human coronavirus is more similar to that of SARS-CoV than to two other human coronaviruses (HKU1 and OC43) that are known to bind to sialic acid.

The authors conclude that the high degree of similarity in amino acid sequence and predicted protein structure of the RBD domain between the new human coronavirus and SARS-CoV suggests that SARS-CoV-2 may efficiently use human ACE2 as a receptor for cellular entry, which could potentially facilitate human-to-human transmission without the need for an intermediate host.

AVMA staff comments: This and the following seminal article are included in this summary because they described, for the first time, the full genomic sequence of what is now known as SARS-CoV-2 isolated from one or more human patients with what is now known as COVID-19. Key points described in these two articles (sequence, structure, phylogenetic relationship with other coronaviruses [particularly SARS-CoV, the causative agent of the SARS outbreak from 2002-2003, and a bat SARS-like coronavirus isolated from an intermediate horseshoe bat], classification as a betacoronavirus, description of the receptor biding domain (RBD) on the virus’s spike protein, and the ability of antibodies in sera of recovered COVID-19 patients to neutralize the new coronavirus in vitro) also serve as the basis for many subsequent research studies and have aided in global collaboration toward better understanding of all aspects of the pandemic. As an example, as of early May 2020, more than 20,000 SARS-CoV-2 gene sequences have been submitted to GISAID, a database that is accessible to all researchers.

RETURN TO TABLE

Section 3: A pneumonia outbreak associated with a new coronavirus of probable bat origin

Published in Nature online on February 3, 2020 and in volume 579 of the journal on March 12 (peer reviewed). This paper describes the sequencing of a new coronavirus (2019-nCoV, now known as SARS-CoV-2) from people with severe pneumonia during the early stage of the outbreak in Wuhan, China. Samples from seven such patients, six of whom worked at or delivered to the local seafood market, were sent to the Wuhan Institute of Virology (WIV) for the diagnosis of the causative pathogen. The laboratory used pan-CoV PCR primers to test these samples and five were positive for RNA from coronaviruses. One sample collected from the bronchoalveolar lavage fluid of one of the patients was analyzed by metagenomics analysis, using next-generation sequencing to identify potential etiological agents. By use of de novo assembly and targeted PCR, a 29,891-base-pair coronavirus genome was identified that was 79.6% homologous to that of SARS-CoV BJ01. This sequence (2019-nCoV WIV01; now known as SARS-CoV-2) was submitted to GISAID (accession number EPI_ISL_402124). Full-length genome sequences of the new coronavirus were identified from the other four patients, using next generation sequencing and PCR. These four sequences (WIV02, WIV05, WIV06 and WIV07) were more than 99.9% identical to each other, and were also submitted to GISAID (accession numbers EPI_ISL_402127-402130).

Similar analyses as those outlined for the article described in Section 2 were conducted on the five new SARS-COV-2 sequences with similar results. In addition, the authors found that a short region of RNA-dependent RNA polymerase from a bat SARS-like coronavirus (BatCoV RaTG13) previously detected in an intermediate horseshoe bat (Rhinolophus affinis) from the Yunnan province of China, had high sequence homology to that of the new human coronavirus (ie, SARS-CoV-2). Comparative analysis of the full-length genomic sequences of this bat coronavirus and the new human coronavirus revealed an overall homology of 96.2%. Phylogenetic analyses of two of the proteins encoded by SARS-COV-2 and BatCoV RaTG13 indicate that these two viruses form a distinct lineage from other known SARS-related CoVs.

The authors then compared the sequence of the S gene that encodes the surface spike protein responsible for binding to the host cell surface receptor in SARS-CoV-2, BatCoV RaTG13, and other SARS-related CoVs. They found that the SARS-CoV-2 S gene was less than 75% homologous to the S genes of other known SARS-related CoVs, except for that of BatCoV RaTG13, in which there was 93.1% nucleotide identity. The authors conclude that the close phylogenetic relationship to BatCoV RaTG13 provides evidence that SARS-CoV-2 may have originated in bats.

The authors isolated SARS-CoV-2virus from the bronchoalveolar lavage fluid of a critically ill patient and found that sera from several patients could neutralize this virus in vitro. And finally, using molecular modeling, they confirmed that SARS-CoV-2 uses the same host cell entry receptor—angiotensin converting enzyme II (ACE2)—as does SARS-CoV.
AVMA staff comments: This and the preceding seminal article are included in this summary because they described, for the first time, the full genomic sequence of what is now known as SARS-CoV-2 isolated from one or more human patients with what is now known as COVID-19. Key points described in these two articles (sequence, structure, phylogenetic relationship with other coronaviruses [particularly SARS-CoV, the causative agent of the SARS outbreak from 2002-2003, and a bat SARS-like coronavirus isolated from an intermediate horseshoe bat], classification as a betacoronavirus, description of the receptor biding domain (RBD) on the virus’s spike protein, and the ability of antibodies in sera of recovered COVID-19 patients to neutralize the new coronavirus in vitro) also serve as the basis for many subsequent research studies and have aided in global collaboration toward better understanding of all aspects of the pandemic. As an example, as of early May 2020, more than 20,000 SARS-CoV-2 gene sequences have been submitted to GISAID, a database that is accessible to all researchers.

RETURN TO TABLE

SECTION 4 : Spike protein recognition of mammalian ACE2 predicts the host range and an optimized ACE2 for SARS-CoV-2 infection

Published in Biochemical & Biophysical Research Communications on March 19, 2020 (peer-reviewed). This paper describes results of studies that used comparative sequence analytical and molecular modeling techniques to predict other potential permissive hosts for SARS-CoV-2 as an aid in identifying the virus’s intermediate host and animals that might serve as experimental models for human infection.

The spike (S) protein found on the surface of SARS-CoV-2 virions is a type I glycoprotein that binds to angiotensin converting enzyme II (ACE2) on the surface of human cells to initiate viral entry into the cell, which is the first step leading to human infection with SARS-CoV-2. ACE2 is also the host receptor for SARS-CoV, the causative agent of the SARS outbreak of 2002-2003. The authors of this article used information known about the interactions between human ACE2 and SARS-CoV, results of recent studies that determined the crystal structure of SARS-CoV-2 bound to human ACE2, and comparative sequence analyses of human ACE2 vs the sequence of ACE2 from 42 mammals, both domestic and wild, to predict which species have potential to serve as hosts for SARS-CoV-2.

The five amino acids within the virus binding region (VBR) of human ACE2 that are key for interacting with the receptor binding motif on the S protein of SARS-CoV-2 are lysine (K) at position 31, glutamic acid (E) at position 35, aspartic acid (D) at position 38, methionine (M) at position 82, and lysine (K) at 353, with K31 and K353 of human ACE2 being the most critical residues for recognition of the viral receptor binding motif. The amino acid sequences of ACE2 from 43 mammalian species were aligned with that of human ACE2 and compared to determine the similarity in residues at these 5 key positions. Because the key amino acids of ACE2 from the golden snub-nosed monkey and rhesus macaque were identical to that in humans (K-E-D-M-K), the authors predicted that these two species of old-world monkeys would be permissive hosts for SARS-CoV-2. Three of five of these key amino acids in human ACE2 were the same in ACE2 from cats, dogs, ferrets, and ermine, or stoats (K-E-E-T-K). These latter two species belong to the Mustelidae family, as do mink. The sequences of the VBR of pangolin ACE2 (K-E-E-N-K) and Chinese hamster ACE2 (K-E-D-N-K) also shared three or four of five of the key amino acids, respectively, found in human ACE2. The molecular model used by these authors actually predicted that these latter two ACE2 sequences, each with N (asparagine) at or near position 82 rather than M (methionine; humans and some non-human primates) or T (threonine; cats, dogs, ferrets, & ermine) would bind with stronger affinity to the corresponding amino acid (F at 486) in the receptor binding motif of the SARS-CoV-2.

Based on the potential interaction between the viral S protein and mammalian ACE2 predicted by the molecular model these authors used, which focused on five key amino acids in the VBR of ACE2, the authors concluded that there is potential for SARS-CoV-2 to infect many mammals including cats, dogs, pangolins, and Chinese hamsters. Further, they speculate that ACE2 sequences with asparagine at position 82 would bind with greater affinity to the receptor binding motif of the SARS-CoV-2 S protein. Moreover, the authors state that: “…we should pay attention to monitoring whether domestic cats and dogs could be infected by SARS-CoV-2.”

AVMA staff comments: Molecular modelling and comparative sequence analyses provide theoretical evidence of permissive hosts for infection with SARS-CoV-2. However, it is theoretical, not definitive evidence. Additionally, this paper only looked at one specific region of the ACE2 protein in mammals, which is characterized by five amino acids, and modeled the interaction of that region with a very specific region of the SARS-CoV-2 spike protein. Although this interaction is essential for the virus to enter a host cell, it is not the only host-virus interaction necessary to allow the virus to replicate, amplify, and become infectious in each of these potential mammalian hosts. More information is needed from both studying natural and experimental infections in animals.

Since this paper was published, SARS-CoV-2 has been detected, albeit extremely rarely, in pet dogs and cats owned by or living with one or more persons with COVID-19 (see AVMA’s in-depth summary of reports of naturally acquired SARS-CoV-2 infections in domestic animals and farmed or captive wildlife). Most of these infected animals never developed clinical signs of illness attributable to infection with SARS-CoV-2. On the basis of sequence analysis of the virus infecting a few of these dogs and cats, it is likely that the infected owners transmitted the virus to their pets. To date, there is no evidence that dogs or cats are able to transmit SARS-CoV-2 to other animals, including humans, under natural conditions.

In late April/early May, the first reports of SARS-CoV-2 infected animals on Dutch mink farms provide evidence that species other than ferrets and stoats (identified as potentially susceptible in this paper) within the family Mustelidae are susceptible to infection with the COVID-19 virus. Early evidence from these farms suggests that it is plausible that infected mink may have transmitted the virus to two animal caretakers, with mink-to-human transmission occurring prior to federal enactment of requirements for farm workers to use personal protective equipment (PPE) when caring for the mink. The Dutch government is continuing to study the affected mink and farms, and as of May 20, initiated surveillance of all mink farms in The Netherlands to further elucidate details regarding transmission of SARS-CoV-2. On June 3, the Dutch government made the difficult decision to cull all mink on infected farms in the interest of animal and public health. Further details regarding SARS-CoV-2 in other species, including mink, are available on the AVMA’s COVID-19 web resources pages.

Cats, ferrets, and a number of hamster species, all of which were identified as potential hosts for SARS-CoV-2 in this article, have also been shown to become infected following experimental inoculation with SARS-CoV-2. In some cases, experimentally infected animals can also transmit the virus to other members of the same species under controlled laboratory conditions. Results of some of these studies are detailed in other sections of this in-depth summary of key research articles related to SARS-CoV-2 in non-human animals.

RETURN TO TABLE

Section 5: Simulation of the clinical and pathological manifestations of Coronavirus Disease 2019 (COVID-19) in golden Syrian hamster model: implications for disease pathogenesis and transmissibility

Published in Clinical Infectious Diseases on March 25, 2020 (peer reviewed). This article describes the use of Syrian hamsters as a potential animal model of human COVID-19. Researchers first used molecular modeling to demonstrate the theoretical affinity of the receptor binding domain (RBD) on the SARS-CoV-2 surface spike protein for angiotensin converting enzyme II (ACE2) on the surface of Syrian hamster cells. Binding of the virus via its spike protein to ACE2 on the surface of host cells is one of the necessary steps leading to replication and amplification of SARS-CoV-2 in permissive hosts. Young (6-10 weeks old) Syrian hamsters were then experimentally infected via intranasal inoculation with a high dose of tissue-cultured SARS-CoV-2 and used to study the time course of infection as well as transmission of virus from infected to naïve hamsters. On days 2, 4, 7, and 14 post inoculation in the time course of infection study or days 4 and 14 post inoculation or exposure in the transmission study, a subset of hamsters in each group were euthanized, and various respiratory and non-respiratory tissue specimens were assessed via quantitative RT-PCR and immunohistochemistry for presence of SARS-CoV-2; via quantitative RT-PCR for cytokine/chemokine profiles (lung tissue; time course of infection study only); and via histopathology for pathological changes induced by the virus. In the time course of infection study, serology was conducted to determine whether experimentally infected hamsters developed virus neutralizing antibodies. Although the overall number of hamsters assessed at each time-point post inoculation or exposure was relatively small (n = 5 or n = 3), results were consistent among all animals at each time point, suggesting that Syrian hamsters can be experimentally infected with SARS-CoV-2, with mild to moderate disease developing that is similar to that found in humans with respect to duration of clinical signs and subsequent recovery; virus localization, pathological lesions, and cytokine profiles in respiratory tissues; and neutralizing antibodies produced. In addition, experimentally infected Syrian hamsters effectively transmitted SARS-CoV-2 to naïve animals via direct contact, and those infected naïve animals also developed mild to moderate disease that is similar to that in the inoculated animals.

AVMA staff comments: This article provides initial evidence suggesting that Syrian hamsters may be a good animal model for human COVID-19. It does not, however, provide evidence that hamsters play any role in transmitting the virus to humans. For Syrian hamsters kept as pets, AVMA recommendations regarding keeping pets safe during the pandemic apply.

RETURN TO TABLE

SECTION 6 : Novel Coronavirus SARS-CoV-2: Fruit bats and ferrets are susceptible, pigs and chickens are not

Press release published on April 2 describing initial results of experimental infection/transmission studies conducted at the Friedrich-Loeffler-Institut (FLI), which is the Federal Research Institute for Animal Health in Germany (non–peer reviewed). Initial results reported by the FLI suggest pigs and chickens cannot be infected with SARS-CoV-2, but ferrets, which are a good model for other human respiratory infections, may also prove to be a good animal model of infection with the virus. However, details regarding number of animals used and experimental design were not provided. The press release indicates further studies are ongoing, with results anticipated for publication in May. Published results of such further studies will need to be assessed before more definitive statements can be made.

AVMA staff comments: As of June 3, 2020, results of the further studies mentioned in the April 2 FLI press release have not been published either as a second press release or an article in a peer-reviewed journal, nor have they been posted online as a preprint. Nonetheless, publication of this press release, together with publication of the April 8 article in Science, which is described in Section 9 of this document, raised public concern that cats and ferrets kept as pets might be able to be infected with SARS-CoV-2 and transmit the virus to other animals, including people. A 2003 Brief Communication that had been published in the journal Nature during the earlier SARS outbreak similarly provided results of experimental infection of cats and ferrets with SARS-CoV. Subsequent publication or posting of articles or reports describing additional SARS-CoV-2 experimental infection/transmission studies, particularly those using cats, exacerbates public concern. However, the AVMA emphasizes the need for caution in not overinterpreting results described in this press release or in published peer-reviewed articles or preprints posted online that describe experimental infection and transmission studies in animals. We also caution about extrapolating such results to the potential for SARS-CoV-2 to naturally infect or be transmitted by companion animals kept as pets. Our rationale is as follows:

  • Experimentally induced infection does not mirror naturally induced infection. That is, just because an animal can be experimentally infected via direct intranasal or intratracheal inoculation of high concentrations of purified tissue-cultured SARS-CoV-2 does not mean that it will easily be infected with that same virus under natural conditions.
  • Experimental transmission studies are typically done under ideal conditions that may include use of negative pressure test chambers and unidirectional flow of HEPA-filtered air from the infected to the naïve animal. Such highly controlled conditions do not reflect conditions found outside a laboratory setting. And, in the experiments described in the April 8, 2020 Science article, despite ideal transmission conditions, only two of six naïve cats (one in each age group) became infected via transmission of SARS-CoV-2 from experimentally infected cats. Results from so few animals should not be used as conclusive evidence that infected cats can readily transmit COVID-19, particularly under natural conditions.
  • The numbers of animals used in these experiments are typically small and the conclusions drawn may be based on data points collected from as few as two animals.
  • Only a small number of domestic animals (< 20) have been confirmed to be naturally infected with SARS-CoV-2 during the first 5 months of the COVID-19 outbreak. This despite the fact that as of June 1, 2020, the number of infected people exceeded 6.2 million globally and 1.8 million in the United States. In addition, there is no evidence as of yet that these relatively few naturally infected domestic animals play any role in transmitting COVID-19 to humans.

RETURN TO TABLE

SECTION 7 : SARS-CoV-2 neutralizing serum antibodies in cats: a serological investigation

Posted April 3, 2020 to bioRxiv (non–peer reviewed). This preprint describes a serological survey of cats conducted during the initial outbreak of COVID-19 in Wuhan, China to determine whether cats had antibodies against SARS-CoV-2, which would indicate they had been infected with the virus. Blood was collected from 39 cats prior to the onset of the outbreak (March-May 2019) and 102 cats after the onset (January-March 2020) and sera stored before testing. Results indicate:

  • Antibodies against SARS-CoV-2 were not detected in any samples collected prior to the outbreak, suggesting that virus was not circulating in Wuhan prior to the onset of the outbreak.
  • After the outbreak, SARS-CoV-2-specific antibodies were detected, using an ELISA, in 15 of 102 serum samples obtained from cats (14.7%).
    • These 15 cats either lived with an owner who had COVID-19 (n=3), at a veterinary clinic (n=6), or on the street as strays until they were moved to an animal shelter after the onset of the outbreak (n=7).
  • It was not reported how many of the 87 cats that were seronegative for SARS-CoV-2 lived with people who had COVID-19.
  • Eleven of the ELISA-positive samples were also positive via a tissue culture-based virus neutralizing test (VNT).
    • The highest titers of neutralizing antibodies (1:360 or 1:1080) were found in the three of 11 cats that lived with owners who had COVID-19.
    • Four of the 11 cats did not have detectable neutralizing antibodies,
    • The remaining four of the 11 cats had titers <1:40.

The authors conclude that SARS-CoV-2 infected the cat population in Wuhan during the initial outbreak, and recommend that “an immediate action should be implemented to keep in a suitable distance between humans and companion animals such as cats and dogs, and strict hygiene and quarantine measures should also be carried out for these animals.”

AVMA staff comments: The relatively low seroconversion rate that resulted in neutralizing antibodies (6.9% [7 cats] of the 102 cats tested had any level of virus neutralizing antibodies) and low to non-existent titers of neutralizing antibodies in all but the three cats who lived with people diagnosed with COVID-19 (these 3 cats represent 2.9% of total population of cats surveyed; they were the only cats that developed neutralizing antibody titers greater than just above background) suggests that cats may not be readily infected with SARS-CoV-2 under natural conditions. The significance of this seroconversion rate to development of virus-mediated disease in cats or transmission of the virus from cats to other animals, including people, is not known. As such, the AVMA does not necessarily concur with the recommendations of the authors of this study, and instead encourages veterinarians, other animal-care and public health professionals, and pet owners to review the AVMA recommendations regarding keeping pets safe during the pandemic and guidance provided by the CDC and OIE.

RETURN TO TABLE

SECTION 8: Infection and Rapid Transmission of SARS-CoV-2 in Ferrets

Published in Cell Host & Microbe on April 6, 2020 (peer reviewed). Ferrets are often used as an animal model of human respiratory infections, including those caused by the influenza virus, and this study provides evidence that ferrets may also serve as a good animal model of SARS-CoV-2 infection in people. Ferrets were readily infected with SARS-CoV-2 following intranasal inoculation with a high dose of tissue-cultured virus, and inoculated ferrets transmitted virus to naïve ferrets either placed in the same cage (direct contact) or an adjacent cage that shared airflow. Infection was detected via RT-PCR (viral RNA), virus isolation, and serology (neutralizing antibodies against SARS-CoV-2). Virus was detected via immunohistochemistry in tissue specimens collected following euthanasia of several of the ferrets, and infected ferrets developed mild, abnormal clinical signs (transient fever and cough).

The authors conclude that an animal model of human COVID-19 that uses ferrets “would be a useful tool to evaluate the efficacy prophylactic anti-virals and preventive vaccines.”

AVMA staff comments: Results reported in this article, which are still based on relatively small numbers of animals, provide additional evidence that ferrets can be experimentally infected with SARS-CoV-2 and transmit the virus to naïve ferrets in a laboratory setting, and as such, may prove to be a good animal model of COVID-19 in people. However, the results do not provide evidence regarding what role, if any, ferrets kept as pets might play in transmitting COVID-19 to humans. As such, the AVMA encourages owners of pet ferrets, as well as veterinarians and other animal-care and public health professionals, to review the AVMA recommendations regarding keeping pets safe during the pandemic and guidance provided by the CDC and OIE.

The AVMA notes that recent reports (late April, May, and June) of SARS-COV-2 infected animals on Dutch mink farms provide evidence that species other than ferrets within the family Mustelidae are susceptible to infection with the COVID-19 virus. Early evidence from these farms suggests that it is plausible that infected mink may have transmitted the virus to two animal caretakers, with mink-to-human transmission occurring prior to federal enactment of requirements for farm workers to use personal protective equipment (PPE) when caring for the mink. The Dutch government is continuing to study the affected mink and farms, and as of May 20, initiated surveillance of all mink farms in The Netherlands to further elucidate details regarding transmission of SARS-CoV-2. On June 3, the Dutch government made the difficult decision to cull all mink on infected farms in the interest of animal and public health. Further details regarding SARS-CoV-2 in other species, including mink, are available on the AVMA’s COVID-19 web resources pages.

RETURN TO TABLE

SECTION 9: Susceptibility of ferrets, cats, dogs, and other domesticated animals to SARS–coronavirus 2

Published in Science as a research report on April 8, 2020 (peer reviewed) after it had been posted online at bioRxiv on March 30. This article describes results of experimental infection and transmission studies of SARS-CoV-2, using ferrets, cats, dogs, pigs, chickens, and ducks. Pigs, chickens, and ducks did not become infected following experimental inoculation with cell-cultured SARS-CoV-2; however, ferrets, cats, and, to a lesser extent, dogs did. As such, this article primarily reports on results from these three species.

Ferrets were inoculated intranasally with a high dose (105 plaque forming units) of well characterized, tissue-cultured SARS-CoV-2 originally isolated from either a human COVID-19 patient or an environmental sample from the Wuhan seafood market that may have served as the source of at least some of the original human infections in China. These ferrets were shown to be infected as early as 4 days post inoculation by RT-PCR detection of SARS-CoV-2 RNA in upper respiratory tract tissue samples collected at necropsy. Additionally, virus in these samples was infectious, as determined by an in vitro infectivity assay. Other ferrets, inoculated in the same manner, were housed in separate cages within a negatively pressurized isolator unit and monitored over 10 days to assess the course of infection, development of clinical signs of disease, and viral shedding. Using RT-PCR, SARS-CoV-2 was detected in nasal washes from all ferrets on days 2, 4, 6, and 8 post inoculation, and also in some but not all rectal swabs. On the basis of results of in vitro infectivity assays, virus in the nasal washes was infective, whereas virus in rectal swabs was not. Only two inoculated ferrets developed a transient fever and loss of appetite on days 10-12 post inoculation. On day 13, blood was collected from these 2 ferrets and they were euthanized, with tissue samples collected during necropsy for histopathology and virus detection. Virus was detected by use of RT-PCR in only one of these two ferrets, and only in the nasal turbinate. Both had developed circulating antibodies against SARS-CoV-2, although neutralizing titers were relatively low, compared with titers in the remaining ferrets euthanized 7 days later. Histopathology revealed a lymphocytic perivasculitis/vasculitis in the lungs of both ferrets.

To investigate whether SARS-CoV-2 could replicate in the lungs of ferrets, additional animals were inoculated intratracheally with a high dose (105 plaque forming units) of SARS-CoV-2 originally isolated from a COVID-19 patient, and two animals euthanized at days 2, 4, 8, and 14 post inoculation to look for viral RNA in tissues and organs. Viral RNA was detected in the nasal turbinate and soft palate of all four ferrets euthanized on days 2 and 4; in the soft palate of 1 ferret and the nasal turbinate, soft palate, tonsil, and trachea of the second ferret euthanized on day 8; and was not detected in either ferret euthanized on day 14. Results were similar between ferrets infected with the human isolate of SARS-CoV-2 and those infected with the environmental isolate. Experiments to investigate transmission of SARS-CoV-2 from inoculated to naïve ferrets were either not conducted or not reported in this article.

Two differently aged groups of cats (seven 6-9 month old [subadults] and seven 70-100 day old [juvenile]) were inoculated intranasally with a high dose (105 plaque forming units) of the same human SARS-CoV-2 isolate used to inoculate ferrets. Two cats in each age group were euthanized on day 3 and two on day 6 post inoculation to evaluate viral replication in their organs by use of RT-PCR and in vitro infectivity assays. The subadult cats did not develop abnormal clinical signs, however, some of the juvenile cats did. Viral RNA was detected in the nasal turbinate of one subadult cat and in the soft palates, tonsils, tracheas, lungs, and small intestines of both subadult cats euthanized on day 3 post inoculation; whereas in the animals euthanized on day 6, viral RNA was detected in the nasal turbinates, soft palates, and tonsils of both cats and in the trachea of one and small intestine of the other. Viral RNA was not detected in the lungs of either cat euthanized on day 6. Infectious virus was detected in all RNA-positive samples collected from the nasal turbinates, soft palates, tonsils, tracheas, and lungs, regardless of day of euthanasia, but was not recovered from RNA-positive samples collected from the small intestines. Histopathologic studies performed on samples from the virus-inoculated juvenile cats that died or were euthanized on day 3 post inoculation revealed massive lesions in the nasal and tracheal mucosal epithelium and in the lungs. These lesions coincided with more severe abnormal respiratory signs in the younger cats. Viral RNA was detected in the nasal turbinate, soft palates, tracheas, and lungs of both juvenile cats on day 3 post infection, and in the tonsils and small intestine of one. Infectious virus was detected in all RNA positive samples from these two cats, except the lung from one of the cats. The pattern of viral detection in the two cats euthanized on day 6 was similar, except that virus was not detected in the small intestine of either of these cats.

The other three inoculated cats in each age group were used for a transmission study to determine whether SARS-CoV-2 could be transmitted from an experimentally infected cat to a naïve cat via droplet or aerosol spread. In these studies, each of the three infected cats within a given age group was housed in a negatively pressurized isolator cage with a double-layer net divider separating it from a naïve cat of the same age group that was placed on the day after inoculation. In each isolator, HEPA-filtered air flow was directed from the inoculated to the naïve cat with a HEPA-filtered airflow outlet preventing recirculation of infectious virus particles. In this way, transmission of virus from one cat to another could not be via direct contact, but only via droplet or aerosol spread.

Because the subadult cats were aggressive, multiple nasal washes could not be performed, so infection and transmission were monitored by detection of SARS-CoV-2 RNA in fecal samples, using RT-PCR. Viral RNA was detected in the feces of two of the three inoculated cats on day 3 post inoculation and in all three on day 5. Viral RNA was also detected in the feces of one naïve cat on day 3 post infection (day 2 post exposure). This naïve cat plus its inoculated pair were euthanized on day 11 post inoculation, and tissue samples collected at necropsy for RT-PCR. Viral RNA was detected in the soft palate and tonsils of the inoculated cat, and in the nasal turbinate, soft palate, tonsils, and trachea of the naïve, exposed animal. The other two pairs of cats were euthanized on day 12, and tissues collected at necropsy for RT-PCR. In these pairs, viral RNA was not detected in any of the naïve animals, but was detected in the tonsils of one of the inoculated cats and in the nasal turbinate, soft palate, tonsils, and trachea of the other inoculated animal. Virus neutralizing antibody was detected in the serum of all three inoculated cats and in the one naïve cat that became infected after exposure to an inoculated animal.

The transmission study on the juvenile cats was performed in the same manner, except infection status was assessed by detection of SARS-CoV-2 in nasal washes collected from each cat on days 2, 4, 6, 8, and 10 post inoculation. In the inoculated cats, viral RNA was detected in the nasal washes of all three on days 2, 4, and 6; in two of the three cats on day 8; and in none of the three cats on day 10 post inoculation. However, viral RNA was detected in the nasal wash of only one of the three naïve, exposed cats on days 6, 8, and 10 post inoculation (corresponding to days 5, 7, and 9 post exposure). Viral RNA was not detected in any of the three naïve cats on days 2 and 4 post inoculation (days 1 and 3 post exposure). Blood was collected from five of the six juvenile cats on day 20 post inoculation, and on day 10 from the one inoculated cat that died 13 days post inoculation. Serum antibodies against SARS-CoV-2 were detected, using an ELISA and virus neutralization test, in all inoculated cats; however, virus-specific antibodies were detected only in the naïve, exposed juvenile cat that was positive for viral RNA.

Dogs: Five young (3-month-old) beagles were intranasally inoculated with the same dose and isolate of purified SARS-CoV-2 as that used in the cat studies. Inoculated dogs were housed together along with two naïve beagles in the same room. Oropharyngeal and rectal swabs from each beagle were collected on days 2, 4, 6, 8, 10, 12, and 14 post inoculation for viral RNA detection by use of RT-PCR and virus titration by use of an in vitro infectivity assay. Viral RNA was detected in the rectal swabs of two inoculated dogs on day 2 post-inoculation and in the rectal swab of one dog on day 6. One of the five inoculated dogs in which viral RNA was detected in a rectal swab on day 2 was euthanized on day 4 post inoculation, but viral RNA was not detected in any organs or tissues collected from this dog. Infectious virus was not detected in any swabs collected from either the inoculated or naïve, exposed dogs. Sera was collected from the remaining four inoculated and two naïve, exposed dogs on day 14 post inoculation and assessed for SARS-CoV-2 specific antibodies by use of an ELISA. Two inoculated dogs seroconverted, but the other two inoculated dogs and the two naïve dogs did not.

Pigs, chickens, and ducks were also assessed in a similar manner to that described for ferrets, cats, and dogs. Viral RNA was not detected in any swabs collected from any of the inoculated animals or from naïve contact animals. None of the animals had antibodies against SARS-CoV-2 on day 14 post inoculation when tested by use of a virus-specific antibody ELISA. Because these animals could not be experimentally infected with SARS-CoV-2, transmission studies were not performed.

The authors concluded that:

  • SARS-CoV-2 can replicate in the upper respiratory tract of ferrets for up to 8 days after experimental infection without causing severe disease or death, which may make ferrets a good animal model of human COVID-19.
  • SARS-CoV-2 can replicate efficiently in cats, with younger cats being more permissive and, perhaps more importantly, the virus can transmit between cats via the airborne route.
  • Dogs have low susceptibility to SARS-CoV-2.
  • Pigs, chickens, and ducks are not susceptible to infection with SARS-CoV-2.

AVMA staff comments: Publication of this research report together with the press release from the FLI described in Section 6 of this document, raised public concern that cats and ferrets kept as pets might be able to be infected with SARS-CoV-2 and transmit the virus to other animals. A 2003 Brief Communication that had been published in the journal Nature during the earlier SARS outbreak similarly provided results of experimental infection of cats and ferrets with SARS-CoV. Subsequent publication or posting of reports describing additional SARS-CoV-2 experimental infection/transmission studies, particularly those using cats, exacerbates public concern. However, the AVMA emphasizes the need for caution in not overinterpreting results described in this or other published peer-reviewed articles or in preprints posted online that describe experimental infection and transmission studies in animals. We also caution about extrapolating such results to the potential for SARS-CoV-2 to naturally infect or be transmitted by companion animals kept as pets. Our rationale is as follows:

  • Experimentally induced infection does not mirror naturally induced infection. That is, just because an animal can be experimentally infected via direct intranasal or intratracheal inoculation of high concentrations of purified tissue-cultured SARS-CoV-2 does not mean that it will easily be infected with that same virus under natural conditions.
  • Experimental transmission studies are typically done under ideal conditions that may include use of negative pressure test chambers and unidirectional flow of HEPA-filtered air from the infected to the naïve animal. Such highly controlled conditions do not reflect conditions found outside a laboratory setting. And, in the experiments described in this April 8, 2020 Science article, despite ideal transmission conditions, only two of six naïve cats (one in each age group) became infected via transmission of SARS-CoV-2 from experimentally infected cats. Results from so few animals should not be used as conclusive evidence that infected cats can readily transmit COVID-19, particularly under natural conditions.
  • The numbers of animals used in these experiments are typically small and the conclusions drawn may be based on data points collected from as few as two animals.
  • Only a small number of domestic animals (< 20) have been confirmed to be naturally infected with SARS-CoV-2 during the first 5 months of the COVID-19 outbreak. This despite the fact that as of June 1, 2020, the number of infected people exceeded 6.2 million globally and 1.8 million in the United States. In addition, there is no evidence as of yet that these relatively few naturally infected domestic animals play any role in transmitting COVID-19 to humans.

The AVMA notes that results of several studies using comparative sequence analyses and molecular modeling techniques predict that cats and ferrets might be permissive for SARS-CoV-2 based on the sequence of the viral binding domain of the ACE2 receptor in these two species and the interactions between that region and the receptor binding domain of SARS-CoV-2. However, these same studies also predicted that dogs would be permissive hosts. Additionally, as of June 3, two dogs in Hong Kong, one in The Netherlands, and one in the USA have been confirmed to have been naturally infected with SARS-CoV-2, likely as a result of transmission from their infected owners, but only one of these pet dogs developed signs of illness, which were not confirmed to be a result of infection with the COVID-19 virus. The conclusion of molecular modelling studies and reports of naturally infected dogs appear to contradict results of the experimental infection/transmission study published in this Science article, which suggest that dogs are not easily infected via intranasal inoculation with purified SARS-CoV-2. These apparently contradictory findings reflect the hazards in overinterpreting results of experimental infection/transmission studies.

Finally the AVMA notes that recent reports (late April, May, and June) of SARS-COV-2 infected animals on Dutch mink farms provide evidence that species other than ferrets within the family Mustelidae are susceptible to infection with the COVID-19 virus. Early evidence from these farms suggests that it is plausible that infected mink may have transmitted the virus to two animal caretakers, with mink-to-human transmission occurring prior to federal enactment of requirements for farm workers to use personal protective equipment (PPE) when caring for the mink. The Dutch government is continuing to study the affected mink and farms, and as of May 20, initiated surveillance of all mink farms in The Netherlands to further elucidate details regarding transmission of SARS-CoV-2. On June 3, the Dutch government made the difficult decision to cull all mink on infected farms in the interest of animal and public health. Further details regarding SARS-CoV-2 in other species, including mink, are available on the AVMA’s COVID-19 web resources pages.

RETURN TO TABLE

SECTION 10: Absence of SARS-CoV-2 infection in cats and dogs in close contact with a cluster of COVID-19 patients in a veterinary campus

Posted April 9, 2020 to bioRxiv (non–peer reviewed).
This preprint, from the Institut Pasteur in Paris, France reported on 21 pets (12 dogs and 9 cats) living in close contact with their owners in a community of 20 French veterinary students. Two of the students had tested positive for COVID-19 and 11 others developed symptoms consistent with COVID-19 infection. Although three of the cats were reported to have developed mild abnormal respiratory or gastrointestinal signs, none of the animals tested positive for SARS-CoV-2 by RT-PCR and no antibodies specific for SARS-CoV-2 were detectable in serum, using an immunoprecipitation assay.

The authors concluded that the rate of SARS-CoV-2 transmission between humans and pets under natural conditions is probably extremely low. However, additional studies are needed to better understand the possible role pets might play in transmission of the virus. The authors recommend that such studies should include more animals and investigate the ages of the pets and the surrounding viral load on the ability of SARS-CoV-2 to infect pet dogs and cats.

AVMA staff comments: The results of this serological survey support other evidence suggesting that pets (dogs and cats) are only extremely rarely infected by SARS-CoV-2 via transmission from owners with confirmed or suspected COVID-19. However, it is possible that antibodies were not detected in these pets because of the timing of blood collection; that is, insufficient time had passed since infection to result in virus-specific antibody production. The preprint did not make it clear when blood was collected from the pets in relation to when their owners had been sick with COVID-19. The AVMA encourages veterinarians, other animal-care and public health professionals, and animal owners to review the AVMA recommendations regarding keeping pets safe during the pandemic and guidance provided by the CDC and OIE.

RETURN TO TABLE

SECTION 11: Potential host range of multiple SARS-like coronaviruses and an improved ACE2-Fc variant that is potent against both SARS-CoV-2 and SARS-CoV-1

Posted April 11, 2020 to bioRxiv (non–peer reviewed). This preprint provides details of molecular modeling and in vitro studies to assess binding properties between the receptor binding domain (RBD) of the spike (S) protein of four coronaviruses and angiotensin converting enzyme II (ACE2) found on the surface of a variety of cells in numerous animal species. The four coronaviruses the authors looked at were SARS-CoV-2 (the causative agent of the ongoing COVID-19 pandemic), SARS-CoV-1 (which also uses ACE2 for initial entry into host cells and has a genomic sequence that is 79.5% homologous with that of SARS-CoV-2), a bat coronavirus (Bat-CoV RaTG13; genomic sequence is 96.2% homologous with that of SARS-CoV-2), and a coronavirus recently isolated from Malayan pangolins (Pangolin-CoV). For simplicity, this summary will focus on interactions between SARS-CoV-2 and ACE2 only.

In humans, ACE2 is used by SARS-CoV-2 as a necessary receptor to gain entry into host cells. The receptor binding domain (RBD) of the SARS-CoV-2 S protein mainly relies on amino acid residues in its receptor binding motif (RBM) to bind ACE2. The amino acid sequence and 3-D structure of the RBD and RBM of the SARS-CoV-2 S protein are known, which allowed the authors to use molecular modeling to identify amino acid residues of ACE2 from a number of animal species that are within 51 Å of the key SARS-CoV-2 RBD residues. The authors identified 22 amino acids that are highly conserved across ACE2 from sixteen species, including humans and fifteen domestic or wild animals, and analyzed the ability of ACE2 to bind and permit cell entry of SARS-CoV-2, using a number of in vitro techniques.

To assess the ability of SARS-CoV-2 to bind to ACE2 from these animal species, the authors created expression plasmids for each of the ACE2 orthologs and transfected 293T cells (a derivative of human embryonic kidney cells that contains the SV40 [polyomavirus simian virus 40] T-antigen) to allow expression of the recombinant ACE2 on the cell surface. A vector plasmid was used as the negative transfection control in these experiments. The authors also expressed the S-protein RBD of SARS-CoV-2 fused with mouse IgG2 and used that to stain, or bind to, the transfected cells. Cells were then stained with a fluorophore-labelled goat anti-mouse IgG secondary antibody, and RBD-ACE2 binding was detected, using flow cytometry. The authors found that in vitro, the RBD of SARS-CoV-2 was able to bind to ACE2 from humans, camels, cattle, horses, goats, sheep, pigs, cats, rabbits, and pangolins, but did not bind ACE2 from guinea pigs, civets, bats, rats, mice, and chickens.

To assess the ability of SARS-CoV-2 to enter cells after it binds to ACE2 from various animal species, the authors used an in-vitro pseudovirus reporter assay to determine whether a retrovirus engineered to express the S protein of SARS-CoV-2 could bind to and enter 293T cells transfected with one of the 16 ACE2 ortholog expression plasmids. Results from the pseudovirus reporter assay indicated that the pseudovirus expressing the S protein of SARS-CoV-2 was able to enter cells transfected with ACE2 from humans, camels, cattle, horses, sheep, pigs, cats, rabbits, pangolins and bats, but not goats, guinea pigs, rats, mice, chicken, or civets.

A final experiment described in this preprint investigated the ability of molecules that inhibit ACE2 mediated cell entry could be used to inhibit binding of SARS-CoV-2 to ACE2 orthologs expressed on the surface of cells in vitro. Results indicated that SARS-CoV-2 entry could be blocked by use of two recombinant IgG Fc fusion proteins; that is, IgG Fc fused with the receptor-binding domain of the viral spike protein (RBD-Fc) and IgG Fc fused with soluble ACE2 (ACE2 Fc).

The authors conclude that camels, cattle, horses, goats, sheep, pigs, cats, and rabbits may serve as potential intermediate hosts for new human transmission of SARS-CoV-2, and rabbits may serve as a useful experimental model to study COVID-19. In addition, the authors speculate that RBD-Fc and ACE2-Fc could be used to treat and prevent infection with SARS-CoV-2.

AVMA staff comments: Although some of the potential SARS-CoV-2 hosts predicted in this study overlap those predicted in other studies, there are also some differences. For example, this study identified camels as a potential host, whereas the study published on March 19, 2020 in Biochemical & Biophysical Research Communications (summarized in Section 4 of this document) predicted camel ACE2 would not be able to bind SARS-CoV-2 so could not serve as an alternative host for this virus. Results described in both the March 19th article and this preprint suggest that cats, rabbits, pangolins, sheep, pigs, horses, and cattle could serve as alternative hosts for SARS-CoV-2. However, results of experimental infection studies described in Sections 6 and 9 of this document suggest that pigs do not become infected following intranasal inoculation with SARS-CoV-2.

Conflicting data can be explained in part because of the variations in experimental design between this preprint and the March 19th article. Interactions between the virus spike (S) protein and host ACE2 were modeled in this study, using a wider range of amino acids from the ACE2 (22 amino acids) than the March 19th article. In addition, in the studies described in this preprint, but not in those described in the March 19th article, recombinant constructs and in vitro procedures were used to support or refute predictions made on the basis of comparative sequence analyses and molecular modeling. However, even though these in vitro procedures supported the prediction made in the earlier molecular modelling study that pigs could serve as a host for SARS-CoV-2, at least two other studies have shown that domestic pigs do not become infected following experimental inoculation with SARS-CoV-2. We must emphasize that although binding of SARS-CoV-2 to ACE2 on host cells may be necessary for the virus to bind to host cells, multiple other virus-host interactions are also required for the virus to enter and replicate within those host cells while evading the host immune system in order to amplify and spread as infectious viral particles to other members of the same host species. And, unless all those interactions occur in a given host species, then infection with and/or transmission of SARS-CoV-2 may not be possible. Thus, although results of molecular modeling and in vitro studies such as those described here may provide clues that can help identify permissive hosts for SARS-CoV-2, they should not be used to make definitive statements regarding the ability of SARS-CoV-2 to infect or be transmitted by a given animal species under natural—or even experimental—conditions.

RETURN TO TABLE

SECTION 12: Extreme genomic CpG deficiency in SARS-CoV-2 and evasion of host antiviral defense

Published April 14, 2020 in Molecular Biology and Evolution (peer reviewed). In this paper, the author
used Data Analysis for Molecular Biology and Evolution (DAMBE) software to calculate the conventional index of CpG deficiency (ICpG) of beta- and alphacoronavirus sequences to assess the evolutionary pressure of host zinc finger antiviral protein (ZAP) on these viruses. Because ZAP binds to CpG dinucleotides in viral RNA genomes, which results in inhibition of viral replication and mediation of viral genome degradation, viruses that infect host tissues with high concentrations of ZAP face evolutionary pressure resulting in genetic point mutations that downregulate the expression of genomic CpG dinucleotides. This allows the virus to evade the antiviral effects of host ZAP. The conventional index of CpG deficiency (ICpG) can be used to determine whether a virus has downregulated CpG expression, and if so, to what extent. Viruses with an index < 1 are considered CpG deficient, and the lower the index, the greater the extent of downregulation. The author concluded that dogs may have served as an intermediate host in the evolution of SARS-CoV-2 from bats to humans. He based this conclusion on the fact that the canine betacoronavirus can cause respiratory illness in some dogs and, most importantly, the canine alphacoronavirus is the only coronavirus with an ICpG similar to that of SARS-CoV-2. This latter piece of evidence suggests that both viruses evolved in similar host environments with high concentrations of zinc finger antiviral protein (ZAP).

AVMA staff comments: Although the author presents an interesting theory to explain his conclusion, it is just that, an interesting theory. No evidence other than ICpG values and tissue tropism of various alpha- and betacoronaviruses was presented to indicate either that SARS-CoV-2 evolved from a canine coronavirus or that dogs play any role in the ongoing COVID-19 pandemic. In fact, the SARS-CoV-2 genome and protein sequences are not at all similar to those of canine alpha- or betacoronaviruses. Moreover, studies attempting to experimentally infect dogs with SARS-CoV-2 have not met with much success and to date, only four SARS-CoV-2 positive pet dogs from three countries (Hong Kong, The Netherlands, and the USA) have been reported, with the virus likely transmitted from human-to-dog via close contact with a COVID-19 positive owner. To date, there is no evidence, either from experimental or natural infections, that dogs can transmit SARS-CoV-2 to other animals, including humans.

RETURN TO TABLE

SECTION 13: SARS-CoV-2 is transmitted via contact and via the air between ferrets

Posted April 17, 2020 to bioRxiv (non–peer reviewed). Ferrets are often used as an animal model of human respiratory infections, and the experiments described in this preprint are similar to those reported in an article published on April 6, 2020 in Cell Host & Microbe (summarized in Section 8 of this document). Ferrets were readily infected with SARS-CoV-2 following intranasal inoculation with a high dose of tissue-cultured virus, and inoculated ferrets transmitted virus to naïve ferrets either placed in the same cage (direct contact) or an adjacent cage that shared airflow (indirect contact). Infection was detected via RT-PCR (viral RNA) and virus titration on throat, nasal, and rectal swab specimens collected at various times after inoculation or exposure. An ELISA was used to detect IgG antibodies against the receptor binding domain of the SARS-CoV-2 spike protein in serum collected 21 days post inoculation or exposure. Unlike the earlier published study, none of the infected or exposed ferrets described in this preprint developed any abnormal clinical signs.

Peak viral shedding, as detected by use of quantitative RT-PCR on throat, nasal, and rectal swabs collected from the four infected ferrets, occurred on day 3 post inoculation, with virus continuing to be detected in throat and nasal swabs through day 11 post inoculation in two ferrets, day 15 in the third ferret, and day 19 in the fourth. All four direct-contact ferrets began shedding virus between one and three days post exposure, as determined by detection of SARS-CoV-2 in throat, nasal, and rectal swabs. Viral RNA continued to be detected in samples from the direct-contact ferrets for an additional 13 to 15 days. Three of the four indirect-contact ferrets also became infected, with SARS-CoV-2 RNA first detected in samples collected between three and seven days post exposure. Viral shedding continued for an additional 13 to 19 days. Infectious virus was isolated from throat and nasal swabs collected from each SARS-CoV-2 positive ferret for at least two consecutive days. However, no infectious virus was isolated from the rectal swabs

Antibodies (IgG) against the receptor binding domain of the SARS-CoV-2 spike protein were detected 21 days post inoculation or exposure in all virus positive ferrets. The one indirect-contact ferret that was not infected did not seroconvert, which was expected.

The authors conclude that ferrets will likely serve as a good animal model of human COVID-19, because they can readily be infected via intranasal inoculation with SARS-CoV-2. Moreover, the virus can be transmitted efficiently between ferrets via airborne droplets or through aerosolization, with transmission resulting in a productive infection characterized by shedding of infectious virus.

AVMA staff comments: Results reported in this preprint, which are based on relatively small numbers of animals (4 per group), provide additional evidence that ferrets can be experimentally infected with SARS-CoV-2 and transmit the virus to naïve ferrets in a laboratory setting, and as such, may be a good animal model of COVID-19 in people. However, the results do not provide evidence regarding what role, if any, ferrets kept as pets might play in transmitting COVID-19 to humans. As such, the AVMA encourages owners of pet ferrets, as well as veterinarians and other animal-care and public health professionals to review the AVMA recommendations regarding keeping pets safe during the pandemic and guidance provided by the CDC and OIE.

Finally the AVMA notes that recent reports (late April, May, and June) of SARS-COV-2 infected animals on Dutch mink farms provide evidence that species other than ferrets within the family Mustelidae are susceptible to infection with the COVID-19 virus. Early evidence from these farms suggests that it is plausible that infected mink may have transmitted the virus to two animal caretakers, with mink-to-human transmission occurring prior to federal enactment of requirements for farm workers to use personal protective equipment (PPE) when caring for the mink. The Dutch government is continuing to study the affected mink and farms, and as of May 20, initiated surveillance of all mink farms in The Netherlands to further elucidate details regarding transmission of SARS-CoV-2. On June 3, the Dutch government made the difficult decision to cull all mink on infected farms in the interest of animal and public health. Further details regarding SARS-CoV-2 in other species, including mink, are available on the AVMA’s COVID-19 web resources pages.

RETURN TO TABLE

SECTIONS 14 and beyond: Summaries not yet available; will be added at a later date

RETURN TO TABLE