Animal Models of SARS-CoV-2: A Systematic Review

Background: The use of animal models for biomedical research provides us with a convenient and feasible route to establish causal relationships by recapitulating the temporal sequence of events in a controlled environment with a potential to manipulate the variables at multiple levels including genetic, protein, physiological or environmental. Objectives: The current review was conducted to gain insights into various animal models for the SARS-CoV-2 virus. Material and Methods: A literature review (PUBMED, PUBMED Central, PMC, Google Scholar, Google search engine) following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines performed in early April 2020 revealed 9 articles of interest. Search terms included covid 19, covid-19, novel corona virus, SARS-CoV-2, animal models, experimental models, laboratory models & covid 19 animal models. Two independent reviewers extracted the data; the third reviewer was involved in case of discrepancy. Results: SARS-CoV-2 shares an identical receptor binding domain with the SARS-CoV virus and has a superior binding affinity to the host ACE2. Based on this, the role of rhesus macaques, golden Syrian hamsters, transgenic hACE2 mice and ferrets as animal models have been studied. All four animals are susceptible to infection with SARS-CoV-2 with variable clinical presentation but universal recovery. The respiratory tract is primarily involved in all four models. Involvement of intestines was also seen in at least one study in each animal. Transfer to naïve animals in close contact has been documented in case of hamsters and ferrets. Seroconversion was documented in all although the role of convalescent sera was tested in hamsters only, with positive results though. Air-borne transmission was documented in ferrets and the possibility of feco-oral transmission was suggested for hamsters. The possibilities of recurrence and re-infection were ruled out by experiments upon the rhesus macaques. The fulfilment of Koch’s postulates has been highlighted. Discussion: The various studies available on animal models have been able to establish models of infection and transmission that recapitulate different aspects of disease in humans. Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 5 July 2020 doi:10.20944/preprints202007.0079.v1


Introduction
The virus has successfully invaded six continents, more than two hundred countries and 1.8 million patients resulting in a hundred-thousand deaths. The unpleasing and undesirable sequence of social distancing, travel restrictions and a universal lockdown, all resulting in severe disruption of global supply chain and economy started in late December 2019 [1]. A mysterious, SARS-like viral pneumonia was reported in few patients whom had a common history of exposure to the Huanan-Seafood-Wholesale-Market. It was on March 11, 2020 that the WHO escalated the disease-status to that of a pandemic. The etiological agent is a novel betacoronavirus, which has been nomenclated as the SARS-CoV-2 by the International Committee on Taxonomy of Viruses (ICTV) and the resulting disease is known as COVID-19 [2].
The unprecedented cross-border transmission of the disease has elicited unprecedented global public health response; a similar swiftness has been witnessed in the researchers' lobbies to gain insights into various disease mechanisms, novel vaccines and potential treatment options. The use of animal models for biological research facilitated by their remarkable anatomical and physiological similarity to humans provides us with a convenient and feasible route to establish causal relationships by recapitulating the temporal sequence of events in a controlled set-up with a potential to manipulate the variables at multiple levels including genetic, protein, physiological, behavioral or environmental.

Material and Methods
This systematic review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) statement [3]. The review methods had been established prior to conduct of the review including stating the review question, search strategy, inclusion criteria for studies and formulation of study protocol.
None of the authors' had any conflict of interest and the study did not receive any funding.

Inclusion and Exclusion Criteria for Studies
Studies (both randomized and non-randomized) discussing, describing or using an animal model of SARS-CoV-2 to highlight the clinical features, pathobiology of infection or modes of transmission following experimental inoculation with the virus were included in this review.
Non-randomized studies were included for several reasons. Firstly, the virus has been recently identified permitting little time for animal research. Secondly, there is scarcity of published literature on the subject. However, whatever information is available is indispensable due to the pandemic-spread of the infection and urgent need for treatment modalities and vaccines alike.
Studies reporting in language other than English were however, excluded.

Search Strategy
Search for a Previous Review: Preliminary search for published literature was done to make sure that there is no published systematic review on this topic nor there is one on the way to completion.
All searches, screening of titles & abstracts and study selection were done independently by two investigators (duplicate) and the results collated. Any discrepancy was resolved by verification with literature and consensus in the presence of third expert.

Data Extraction
Extraction of data from the studies included for this review was performed by two independent investigators. Discrepancies were resolved by re-review of the respective study and consensus in presence of a third expert.

Results
The literature on COVID-19 is expanding rapidly (with an average of 500 publications per month over the past five months) and a large amount of research is being showcased ahead of peer review. Authors' search retrieved 2 publications which were available while they were under peer-review [5,7].

Study Characteristics:
The characteristics of studies included in this study have been summarized in Table 1.

Choice of Animal Model for the SARS-CoV-2 virus
In an attempt to address important issues related to the causative agent of this new disease, experts compared the genome sequences of this virus with other coronaviruses known to infect humans, the SARS-CoV and the MERS-CoV in particular. The representative genome of the SARS-CoV-2 virus was shown to be more homologous to SARS-CoV than MERS-CoV [13].
The S1 unit of S-protein in the coronal virus is responsible for binding to the host receptors through direct interaction of its C-terminal Receptor Binding Domain (RBD) with the host receptors [14]. The new virus shared an identical 3-D structure in the RBD domain with the SARS-CoV, thereby retaining similar van der Walls and electrostatic properties in the interaction interface [13]. Although four-of-five critical (for cross-species and human-tohuman transmission) amino acid residues at the receptor-complex interface were replaced in the SARS-CoV-2, the virus had a superior (to SARS-CoV) binding affinity to the human ACE2 [6]. The interface between the SARS-CoV-2 spike-glycoprotein-RBD is conducive to development of more hydrogen bonds with the ACE2.
The ACE2 receptors are distributed predominantly in the epithelial cells of the lungs and small intestine thereby strongly suggesting these routes as the portal of virus-entry into human beings [15].

Monkeys as Animal Model for SARS-CoV-2
The experimental details along-with clinical and pathological findings in rhesus macques have been summarized in Table 2 [4][5]. There were no obvious clinical manifestations except reduced appetite (1 of 6) and a delayed weight loss (2 of 6; 7% between 7-14 dpi). Viral RNA was demonstrated in oro-pharyngeal, nasal and anal swabs but not in the blood samples. The oro-pharyngeal swabs depicted two peaks in viral RNA count on 1 and 5 dpi which could be explained by experimental virus inoculation and viral replication respectively. The chest X-ray depicted expanding ground glass opacities over first six days of infection which corroborated with diffuse interstitial pneumonia upon histopathology. Relapse or reinfection [17][18] was studied in the 2 rhesus monkeys by re-inoculating them with the same virus 28 days after first exposure once they were confirmed to be cured (disappearance of symptoms, absence of radiological abnormalities and two negative reports of RT-PCR) [4]. Although there was a transient rise in body temperature, the naso-pharyngeal and anal swabs tested negative for the virus and the findings corroborated with histopathology of the lungs and other tissues. The antibody titers escalated by 2.5 times in one monkey. A third monkey was not re-exposed to the virus but followed up longitudinally and no relapse was observed during the two week after 28 dpi.
Experimental studies ( Table 2) have shown that the golden Syrian hamsters simulate SARS-CoV-2 infection in humans with respect to clinical features, viral kinetics, histopathological changes and immune response. Clinical Features included progressive weight loss (more than 10%), lethargy, ruffled fur, hunch back posture and rapid breathing. Recovery was universal with no hamster mortality. Viral load was higher in the upper and lower respiratory tract and the intestine. Although low levels of viral RNA were detected in several organs, viral N protein was detected only in the intestine. Histopathology of the respiratory tract, intestines, spleen, heart and lymph nodes (bronchial and mesenteric) was studied ( Table 2). Serum neutralizing antibodies were detected at 7 dpi. Passive immunization with early convalescent serum reduced the viral loads in the nasal turbinate and the lungs.
Transmission to naïve animals in close contact was demonstrated; surface-spike gene sequencing in virus samples from contact animals was similar to that of index animals (exception: H6554-mutation in one hamster).
Similar results have been replicated in another study [7], although, the histopathological damage was restricted to the respiratory tract sparing the other organs ( Table 2). The fecal samples, however were positive for the viral RNA. Seroconversion happened at 14 dpi. The

disease was transmitted to all hamsters in close contact who manifested similar clinical course, shed comparable load of viruses and presented a comparable immunological
profile.
Clinical manifestations in mice are sparse ( Table 2). Viral RNA was demonstrated in the lungs (peak at 3 dpi) and intestines only. Infectious virus could also be isolated from the lungs by Vero E6 cell culture but not the wild type mice and mock-infected hACE2 mice. Mounting of immune response was concomitant with identification of MAC2+ macrophages, CD3+ T lymphocytes and CD19+ B lymphocytes upon immunohistochemistry (IHC). Sero-conversion with S-protein specific IgG antibodies was demonstrated at 21 dpi. There was no diseaserelated mice-mortality.
The theory propounding the need for hACE2 in host cells for virus entry was supported by the co-localization of the SARS-CoV-2 S-protein and hACE2 receptor in the alveolar epithelial cells of the hACE2 mice upon IHC.

Ferrets as Animal Model for the SARS-CoV-2
Kim et al. [10] inoculated ferrets (Infected Ferrets) with SARS-CoV-2 and placed them with naïve ferrets: either co-housed (Direct Contact) or housed in a cage with permeable partition (Indirect Contact) separating them from infected ferrets (designed to study air-borne transmission).
Clinical features included elevated body temperature, reduced activity and occasional cough. Viral RNA was isolated from blood, nasal washes, saliva, urine and feces and multiple routes of viral transmission documented. The highest viral titers were recorded in the nasal washes in both the Infected Ferrets and those in Direct Contact. Copies of viral RNA were also isolated from the nasal swab and fecal specimens of two ferrets in Indirect Contact although they were clinically asymptomatic (air-borne transmission). Infectious virus was isolated from nasal washes (highest yield) and saliva of Infected Ferrets and only nasal wash of those in Direct Contact. The highest viral titers were recorded in nasal turbinate and lungs upon necropsy; infectious virus recovery was directly related to the viral RNA copy numbers.
Naïve ferrets could sustain infection by coming in contact with infected urine and feces.
Seroconversion was consistent with disappearance of clinical symptoms and disappearance of viral RNA. One of six ferrets in indirect contact group also tested positive for serum neutralizing antibody (air-borne transmission).
Shi et al. [12] also conducted experiments with the primary objective to study the susceptibility of ferrets, cats, dogs and other animals to SARS-CoV-2. The authors have incorporated the results of ferret-based experiments in this review for completing the context while the other animals have been excluded ( Table 2). Viral replication in the upper respiratory tract of ferrets post intra-tracheal inoculation was documented for 8 days without causing significant disease or illness.

Discussion
The role of animals in biological research and medicine has been known for long. Although a matter of debate frequently, animals are indispensable to the understanding of disease pathobiology and in the pursuit for novel vaccines, drugs and therapies [22][23]. Basic research using animal models serves a multitude of functions such as, 1) mimic and assist in understanding the disease process in a controlled environment, 2) decipher the transmission routes of the infectious agents, 3) in vivo simulation of the immune response, 4) create an animal model to support active infection, shedding and transmission to sentinel animals, 5) development and assessment of diagnostic tools, 6) investigate newer vaccines and drugs for efficacy and safety prior to use in human beings, and 7) find out if the animals may serve as reservoirs of infection and perpetuate the virus over the years; does the pandemic reserve the potential to recur in a seasonal manner after the index event is contained.
The choice of animal models for the SARS-CoV-2 virus has been dictated by the presence of ACE2 receptors, similar to the SARS-CoV virus. Contrarily, the MERS-CoV virus attaches to the dipeptidyl peptidase-4 protein on the host cell through its surface spike (S) protein [24][25].
The binding affinity of ACE2 to the SARS-CoV-2, RBD is highest in rhesus macaque (100% similarity in amino acid residues' alignment at the interface region with human beings) followed by the hamsters among all the species [6].
The SARS-CoV-2 was shown to infect and replicate in humanized transgenic hACE2 mice, ferrets, hamsters and several species of nonhuman primates such as the rhesus monkeys [26].
The virus however, replicates poorly in dogs, pigs, chickens and ducks [12].
The non-human primates (NHPs) including the monkeys have played a very vital role in biomedical research such as deep brain stimulation in the management of Parkinson's disease, treatment of bronchial asthma, development of transplant-related drugs and the use of incubators for premature infants, cancer, AIDS, and obesity/ diabetes [27][28]. Although 90% of medical research is conducted on small animals such as mice, rats and other rodents and the NHPs account for less than 1%, the utility and vitality of the non-human primates may be traced back to the close phylogenetic relationship with the Homo sapiens. The two also share some apsects of physiology, neuro-anatomy, reproduction, development, cognition and social complexity [29]. The utility of NHPs, however, is limited by the expertise and the scarce availability of the Biosafety Level-3 facilities to handle non-human primates. Besides, the increased likelihood that the primates experience pain and suffering in ways similar to the human being raises a lot of ethical issues with their use for biomedical research.
Studies [4][5] have shown that they are susceptible to infection with the SARS-CoV-2 virus.
Although the clinical manifestations are scarce, a lot of similarities were documented between virus infection in monkeys and humans. The virus replication takes place not only in the upper and lower respiratory tract but in the gut as well similar to human beings. The virus isolated from the respiratory tract of the infected monkeys was identical (99.99%) to the original virus, thereby satisfying the Koch's postulates in-principle [21]. Seroconversion with rising antibody titers signifies resolution of infection with development of immunity. Longterm follow-up (14 days beyond 28 dpi) was not suggestive of any recurrence. Re-exposure of the monkeys with the same virus at 28 dpi could not re-establish infection; the immunity conferred by the antibodies was effective. There was no virus residue in the various body tissues after recovery from infection.
The study has offered the following suggestions, 1) the infection confers immunity postrecovery which is likely to be protective against future re-infections, 2) the convalescing patients are not expected to be contagious after mounting a strong antibody response, 3) the question as to how long will this immunity will be protective is yet to be answered since the study has a very brief follow-up and the numbers meagre. In case of the SARS virus, the immunity tends to wane off over a couple of years while it lasts for only a few weeks in case of some common cold viruses, 4) convalescent plasma therapy may be an option to rescue the critical cases, 5) development of antibody tests will be helpful in identifying individuals who have recovered from this infection such as in screening health-workers for immunity and to the epidemiologists in figuring out the hidden pattern of disease-spread, 7) the results are encouraging towards development of vaccines against this virus, and 8) the rhesus macaques may be useful in testing the safety and efficacy of vaccines, drugs and newer treatment strategies against this virus.
However, the 'as-is' applicability of these results to human being may be challenged by the observed rapid viral clearance mechanisms in the monkeys; human beings continue to be positive for the virus in the naso-pharyngeal swabs until upto 9-15 days after onset of symptoms (incubation period not be ignored) whereas the monkeys were cleared of infection on 15 th day post-infection.
Furthermore, the phenomenon of Antibody Dependent Enhancement which is a peculiar characteristic of the SARS-CoV virus was ruled out in context of the index virus [30]. Herein, the virus cannot re-infect the host cells by taking advantage of the host's humoral immune response to the initial infection.
Furthermore, the reason for repeat-positivity after discharge from the hospital cannot be reinfection, recurrence or relapse; the possibilities of false-negative results and a need to relook into the discharge criteria cannot be ignored.
The Syrian hamster (Mesocricetus auratus) has been used to study pathogenesis of human infections for more than six decades now. They are readily available, small in size, easy to handle, have a fast reproductive rate and mount an immune response similar to human beings [31]. The disease symptomatology and pathogenesis are also comparable. Hamsters have provided an effective platform in elucidation of various molecular-level developments in the immune system in response to viral (and other) infections [32][33][34]. Unlike the mouse models, they have fully functional human cytokines [35][36] Ferrets have a natural susceptibility to the human respiratory viruses and have been considered the ideal model for human influenza and proved to be useful in replicating the symptoms of several other viruses such as the respiratory syncytial virus and the SARS-CoV-1 [49][50][51][52]. This may partly be explained by their resemblance to human beings with regard to anatomic proportions of the upper and lower respiratory tracts, number of generations of the terminal bronchioles and the density of submucosal glands in the bronchial wall [51].
It has been shown that the SARS-CoV-2 recognizes ACE2 not only from humans but from ferrets and several other animal species (like pigs, cats, orangutans and monkeys) also, more so with similar efficiencies, because these ACE2 molecules are identical or similar in the critical virus-binding residues [53].
The ferret model which was initially developed to simulate the influenza virus infection [54] has been used in experiments with the SARS-CoV-2 virus to successfully demonstrate the Koch's postulates [21] and multiple aspects of SARS-CoV-2 infection including rapid viral replication, human-to-human transmission on close contact, air-borne transmission, multiple routes of virus shedding and transmissibility during the incubation period of the viral infection [10].  There is a crying need for a deeper dive in search of a quintessential animal model which can be studied for efficacy and safety of newer drugs and vaccines before a make-shift from the petri-dish to the human body can be contemplated.

Full-text articles not-qualifying for inclusion into study (n-4)
Sl. All experiments involving live SARS-CoV-2 followed the approved standard operating procedures of the HKU Biosafety Level-3 facility 4