Introduction

In the past two decades, two outbreaks of coronaviruses (CoVs) in Asian subcontinent and the Middle East increased the chances of such outbreak in the near future. The severe acute respiratory syndrome coronavirus (SARS-CoV) outbreak at the end of 2002 (November) in China in which 8,098 confirmed cases were reported with 774 total deaths (9.6%) (Ksiazek et al.2003; WHO 2002). Another coronavirus i.e. Middle East respiratory syndrome (MERS) occurred in 2012 with a fatality rate of around 35% (WHO 2019; Bermingham et al.2012). Recently, a new coronavirus called as novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) outbreak in December, 2019, in China. There is an emergence of global health threats and the situation became worse at the global level (Rodriguez-Morales et al.2020). The World Health Organization (WHO) termed the syndrome caused by the novel coronavirus as “COVID-19” on 11th, Feb 2020 (Gralinski and Menachery 2020).

Coronaviruses fall in Coronaviridae family and subfamily Coronavirinae, the members of which infect a broad range of hosts, producing symptoms and diseases of wide spectrum from a flu like to severe acute respiratory infection (SARI). The SARS-CoV-2 is considered to be one of the seven members of the CoV family that infect humans (Zhu et al.2020). Similar to other coronaviruses, the transmission of SARS-CoV-2 among humans occurs through droplets, direct contact, any physical objects, and aerosols (WHO 2020a; Liu et al.2020). Owing to the rapid transmission, SARS-CoV-2 epidemic has become a global public health challenge. On March 11, COVID-19 outbreak was declared as a pandemic by WHO (Cucinotta and Vanelli 2020). By June 10, more than 7 million people have been globally affected, with nearly 408,025 fatalities (WHO 2020b).

Due to the high morbidity and mortality rate of coronaviruses infection, there is an unmet need for therapeutic strategies to treat these diseases (Sarma et al.2020). Currently, there are no recommended therapeutic drugs or vaccine for coronaviruses. This emphasizes the surge for a suitable animal model to explore the pathogenesis and evaluation of countermeasures for the disease. With the outbreak of the coronaviruses, numerous animal models were developed in the past (Gralinski and Baric 2015). The current review highlighted the animal models of coronaviruses, summarized the strength and weaknesses of the existing animal models along with their key features.

Animal Models for SARS-CoV-2

Non-Human Primate Models

SARS-CoV-2 was recently recognized as a novel coronavirus causing the symptoms from mild to critical type (SARI). Munster et al. used SARS-CoV-2 infected rhesus macaques to study the pathogenesis of COVID-19. They observed high viral titers in nose swabs, throat swabs, and also lung lesions at varying degrees in all animals (Table 1). This model recapitulates COVID-19 symptoms and can be used further to elucidate the therapy for SARS-CoV-2 infection (Munster et al.2020).

Table 1 Animal models of SARS-CoV-2.
Full size table

Rockx et al. compared the pathogenesis of SARS and MERS with COVID-19 by inoculating cynomolgus macaques with virus infection (Rockx et al.2020). The virus samples were taken from the throat and nose. The diffuse alveolar damage was reported. The results of the study thus demonstrated that the infection with SARS-CoV-2 in non-human primates induced signs that resemble COVID-19 like disease, and the severity of virus infection lies in between the MERS and SARS (Mahase 2020).

Yu et al. studied the association of virus infection with age. The viral strain was inoculated in rhesus macaques of 3–5 years old and 15 years old through intra-tracheal route. The clinical signs, for instance, viral replication, and histopathological changes were analyzed. Replication of virus was more in lungs and nasopharyngeal swabs of old rhesus macaque as compared to young rhesus macaque after infection. Old rhesus macaques also observed to have diffuse severe interstitial pneumonia (Yu et al.2020).

In a comparative study, three species of non-human primates were infected with SARS-CoV-2 (rhesus macaque, Macaca fascicularis, and common marmoset). Severe gross lesions and histopathological changes in the vital organs of all animals were observed. The study found that the rhesus macaque model was more susceptible to SARS-CoV-2, as compared to Macaca fascicularis and common marmoset (Lu et al.2020).

Woolsey et al. determined that African green monkeys supported a high level of SARS-CoV-2 replication and developed a respiratory tract related illnesses that might be more substantial than reported for other non-human primate species like cynomolgus and rhesus macaques. The study also reported high viral loads in feces and in mucosal samples of all monkeys after 15 days post-infection (dpi) (Woolsey et al.2020) (Table 1).

Mouse Models

The inoculation of BALB/c mice with mouse-adapted SARS-CoV-2 at passage 6 (MACSp6) significantly affected all groups of mice irrespective of their ages. The infection resulted in moderate pneumonia and inflammation. A study reported higher viral load in lungs along with significant histopathological changes like denatured trachea, inflammation in pulmonary alveoli, detection of viral antigen in the trachea, bronchiole, in type II pneumocytes. There was a significant elevation of inflammatory chemokine and cytokines in sera and lung macrophages similar to clinical symptoms (Gu et al.2020).

Dinnon et al. constructed a recombinant virus (SARS-CoV-2 MA) that could replicate in both upper as well as lower airways of all groups of BALB/c mice irrespective of their ages. The results showed that there was more severe infection in aged mice in comparison to young mice (Dinnon et al.2020).

A study has been done based on the human ACE2 (hACE2) transgenic mice to investigate the pathogenesis of COVID-19. Infected mice reported weight loss and increased viral titer in lungs, and histopathological changes revealed the presence of interstitial pneumonia. The mice model seems to be promising for the evaluation of therapeutic measures for COVID-19 (Bao et al.2020).

One of the studies examined the infectivity and pathological changes in SARS-CoV-2 infected transgenic hACE2 mice, which was developed using lung ciliated epithelial cell-specific HFH4/FOXJ1 promoter 81. The model showed a partial simulation of the human COVID-19. The significant decrease in body weight, interstitial pneumonia, and fatality along with the involvement of other organs like heart, brain, and eye depicted its similarity to the human counterpart. The study also showed protection on reinfection in the survived mice (Jiang et al.2020) (Table 1).

Sun SH et al. developed the hACE2 mouse model using CRISPR/Cas9 knock in technology to study the SARS-CoV-2 infection and compared it with the wild type C57BL/6 mice model. There was high viral titer in brain, lungs and trachea of hACE2 mice than wild type mice. SARS-CoV-2 infection in hACE2 mice associated with increased cytokine levels and interstitial pneumonia with no mortality. Moreover, intragastric inoculation of the SARS-CoV-2 infection altered significant pathological changes in lungs when compared to wild type mice (Sun et al.2020b).

Sun J et al. developed a transgenic mice model by delivery of human ACE2 exogenously with a replication deficient adenovirus (Ad5-hACE2). The inoculation of mice with SARS-CoV-2 infection linked to high viral load in lungs along with weight loss and pneumonia. The role of type-1 interferon, signal transducer and activator of transcription 1 (STAT1) signaling is vital in viral clearance of virus and infection. Further, the study concluded the beneficial role of using human convalescent plasma and two antiviral agents (polyinosinic: polycytidylic acid and Remdesivir) (Sun J et al.2020a).

Hassan et al. developed hACE2 transduced mice and infected these mice with SARS-CoV-2. The study observed significant weight loss, high viral load in lungs and severe lung pathology in infected mice (Hassan et al.2020).

Hamster Model

Chan et al. observed significant binding of the novel coronavirus spike with the ACE2 receptor of the Syrian Hamster by in silico study (Table 1). The viral load significantly increased in the hamster which leads to diffuse alveolar damage in the initial stage and extensive apoptosis in the later phase of infection (Chan et al.2020).

The clinical and histopathological observations from SARS-CoV-2 hamster model closely resemble to the clinical condition. The airway involvement is evident from nasal turbinate to the trachea and pulmonary alveoli, associated with changes of inflammation, cellular viral N protein expression, and high viral load during the first week. The disease progressed with increasing respiratory rate, decreasing activity, and progressive weight loss, and was found to be most severe by 6 dpi, which is similar to the disease course of COVID-19 patients (Huang et al.2020).

Sia et al. evaluated the transmission and pathogenesis of SARS-CoV-2 infection in the golden Syrian hamster model. The animal groups infected with SARS-CoV-2 showed significant weight loss, and viral replication in the respiratory tract. Viral load was high on 2 dpi and rapid clearance was observed on 7 dpi, which could be attributed to the presence of CD3 positive T-lymphocytes. The involvement of olfactory sensory neurons simulates the anosmia in COVID-19 cases. The transmission of SARS-CoV-2 to naïve hamsters by aerosols can be helpful for further studies (Sia et al.2020) (Table 1).

Ferrets Model

Kim et al. used ferrets to study the SARS-CoV-2 infection and its transmission. The study reported that the ferrets were more susceptible to infection and transmission of the virus from one ferret to naive ferrets by direct or indirect physical contact. They showed that infected ferrets exhibited more replication of the virus in the upper respiratory tract and it recapitulates COVID-19 human conditions and can be further used to evaluate the countermeasures for the disease (Kim et al.2020) (Table 1).

Animal Models for SARS-CoV

Non-Human Primate Models

For the development of SARS-CoV infection model, many species, like old and new world primates, were used. There was significant infection observed in cynomolgus (Fouchier et al.2003) and rhesus macaques (Rowe et al.2004), common marmosets (Greenough et al.2005) and African green monkeys (McAuliffe et al.2004), when the virus was inoculated within respiratory tract in those primates. However, in the case of squirrel monkeys and mustached tamarins, investigators were not able to induce infection (Roberts et al.2008; Subbarao and Roberts 2006).

There was significantly high viral replication in cynomolgus, rhesus, and African green monkeys (McAuliffe et al.2004). While in marmosets, diarrhea, pneumonitis, fever, watery stool, and hepatitis were observed (Greenough et al.2005). Investigators have variable findings as a number of factors may affect the results, including age and animal source, dose, route of administration of infection and inoculation of the virus, and their history of the environment (Subbarao and Roberts 2006).

However, results in out-bred species are not consistent because of biological variability. Therefore, it is important to carry large sample sizes for the study to find out the meaningful conclusions (Table 2).

Table 2 Animal models of SARS-CoV.
Full size table

Mouse Models

There were different mice models used for the study of infection. When BALB/c mice were infected intranasally, there was an increase in viral titer in the lungs but no sign of morbidity or mortality was noticed. After 2 to 3 dpi, virus replication significantly increased in the respiratory tract but started decreasing at day 5 dpi (Subbarao et al.2004). This showed that young BALB/c mice failed to exhibit clinical signs of disease but permitted viral replication. On the other hand, a study showed that one-year-old BALB/c mice exhibited critical infection as compared to younger mice after SARS-CoV infection (Vogel et al.2007). Histopathological findings such as bronchiolitis, diffuse alveolar damage, and patchy interstitial pneumonitis were observed in BALB/c mice model of SARS-CoV. It was also observed that old age BALB/C mice were more prone to this disease as similar to the clinical symptoms observed in elderly people in the 2003 SARS outbreak.

One of the studies suggested that 129S mice infected with SARS-CoV exhibited virus replication and self-limited bronchiolitis. 129S mice strain also reported mild weight loss and pneumonitis after SARS-CoV infection. While on the other side, STAT 1 -/- mice in the 129S background observed sustained replication of the virus and histopathological changes, which were mimicking the human’s disease. However, vaccine studies are limited where targeted mice have immune defects (Hogan et al.2004; Frieman et al.2010).

Moreover, C57BL/6 mice were also used as a SARS-CoV model. Virus isolates administered intranasal into C57BL/6 mice showed high replication in the respiratory tract of mice with a peak on the 3rd day and clearance on the 9 dpi. The study observed the presence of transient systemic infection in the lungs which ultimately affects the brain (Glass et al.2004) (Table 2).

Hamster Model

Another model used to study the SARS-CoV infection is a golden Syrian hamster. There was a significantly high titer of viruses in the lungs along with interstitial pneumonitis (Roberts et al.2005). Because of the clinical signs like pulmonary histopathology and high virus load, the hamster can be used as a model for the study of therapeutics, immunotherapy, and immunoprophylaxis. It was observed that viruses got cleared off by 7dpi. (Table 2).

Ferrets and Domestic Cat Model

Various studies showed that domestic cats and ferrets are prone to SARS-CoV infection because virus replication, as well as specific antibodies, were found in both the species. After indicating drowsiness and epiphysitis 16 to 21 dpi, ferrets died (Martina et al.2003). However, with domestic cats, no further study has been conducted. After the inoculation of infection via intranasal route, the replication of the virus in the lungs and histological changes like pneumonitis were reported (ter Meulen et al.2004) (Table 2).

Animal Models of MERS-CoV

There is a difference in the homology of human and small animals DPP4 molecules which is the main virus receptor of MERS-CoV. Therefore, mice, guinea pig, and ferrets are not usually permissive to the MERS-CoV (Cockrell et al.2014; Coleman et al.2014; de Wit et al.2013a, b; Raj et al.2013). However, non-human primates, rabbits, pigs, and dromedary camels are naturally permissive for the virus.

Dromedary Camel Model

Adney et al. studied the pathology and transmission of MERS-CoV among dromedary camels. The animals displayed slight increases in body temperature and rhinorrhea. A large quantity of nasal shedding of the virus is suggestive of an animal-to-animal and animal-to-human transmission which may occur through droplets or direct contact. MERS-CoV mainly affects the upper respiratory tract and mimics mild clinical signs. Since in clinical settings, the disease shows severe pathology, so camel as a model to study therapeutics is not suitable (Adney et al.2014) (Table 3).

Table 3 Animal models of MERS-CoV.
Full size table

Alpaca Model

MERS-CoV is classified under Biosafety Level 3 which poses difficulty in using camels for infection study. Alpaca, belonging to the Camelidae family, provides an alternative to camels. Crameri et al. used alpaca as a prospective substitute to a camel for MERS-CoV vaccine testing (Crameri et al.2016). On exposure to MERS-CoV, the study observed dissimilarity between the immune response of two species. However, the pathogenesis of infection in alpaca is unknown (Table 3).

Non-Human Primate Models

de Wit et al. used rhesus macaques to develop the MERS model. There was high viral load of MERS-CoV in the lower respiratory tract. The viral strain mainly affected type I and II pneumocytes that formed the architecture of the alveolar space. Since high expression of MERS-CoV was present in the lungs, therefore, there was limited virus shedding. There was no viral replication in the kidney. The study concluded that renal failure in humans might be associated with hypoxia but not with viral dissemination. Due to its transitory nature, this model of rhesus macaques was not found to be exactly mimicking the critical clinical cases (de Wit et al.2013a). Similarly, Yao et al. observed a mild increase in temperature in rhesus monkeys. There was no virus detected in the upper respiratory tract. Significant pneumonia was observed with no systemic dissemination of the virus. The study concluded only lower respiratory infection with no extra-pulmonary effect (Yao et al.2014).

Human exposure to MERS-CoV leads to severe respiratory disease which might result in fatality and it is not easy to attain the same fatality rate in non-human primate models. Cockrell et al. used an infectious clone of novel MERS-CoV (icMERS-0) to induce MERS in the rhesus macaque. The clone strain was found to have higher pathogenicity in comparison to its wild type. MERS-CoV peaked at 3 to 5 dpi and resolved to 30 dpi. The respiratory disease was found to be placid and brief which resolves by day 30 dpi. The model mimics only mild infection of MERS in humans (Cockrell et al.2018).

In most of the cases, fatality is associated with comorbidities in MERS affected patients (Hui et al.2014; Saad et al.2014) and severity is linked to the immunocompromised state. In order to study the immune response and disease severity, Prescott et al. used the immunosuppressive drugs to down-regulate the immunity of rhesus macaques. The immunosuppressive state was found to be associated with a high titer of MERS-CoV but not supported by the pathology. In lung tissues, inflammatory cells significantly decreased in immunocompromised animals when evaluated histopathologically. This, in turn, shows the effect of the immune status of an individual on the shedding of the viral load. This study depicted the importance of considering the immune response along with using therapy to control infection in clinical settings (Prescott et al.2018) (Table 3).

Marmosets are more prone to MERS-CoV infection than the rhesus macaques and it might be suitable to screen therapeutic strategies for the disease in marmoset (Johnson et al.2015; Falzarano et al.2014). The study by Falzarano et al. displayed an increase in severity following exposure to MERS-CoV in marmoset than rhesus macaque. The titer was almost 1000 times greater in marmoset than in rhesus macaque. The infection persisted for a longer duration as compared in rhesus macaque and need euthanasia. In fact, there was a systemic dissemination of the virus. But there was no viral dissemination in the kidney. The study observed that the marmoset model recapitulates more closely to human MERS (Falzarano et al.2014).

Johnson et al. used two clones of MERS-CoV i.e. MERS-CoV-Jordan-n3/2012 and MERS-CoV-EMC/2012. The inoculation of MERS-CoV into common marmosets resulted in mild to moderate clinical presentation of the disease which might be due to the manipulations of the marmoset to a certain extent (Johnson et al.2015) (Table 3).

Yorkshire Landrace Pig Model

Pigs are naturally permissive to the MERS-CoV infection (de Wit et al.2017). Vergara-Alert et al. used HCoVEMC/2012 to infect Yorkshire Landrace pigs. The study observed no animal-to-animal transmission though RNA dissemination of the virus was significantly high. Thus, this model cannot be used to study therapeutic strategies as it does not depict the true clinical condition (Vergara-Alert et al.2017) (Table 3).

New Zealand White Rabbit Model

Houser et al. studied the effect of re-infection in the rabbits. The model supported the replication of MERS-CoV-EMC/2012 virus. The model displayed asymptomatic infection which is not associated with any significant clinical sign. The model is not suitable for preclinical screening of therapeutics or vaccines for MERS (Houser et al.2017) (Table 3).

Mouse Models

Mice are easily available, cost-effective, easy to handle, and suitable for initial screening of drug therapies and vaccines. Since, mice are not susceptible to MERS-CoV infection (de Wit et al.2013b; Coleman et al.2014), therefore Zhao et al. developed a mouse model for MERS using a recombinant adenovirus that expresses humanized dipeptidyl peptidase-4 (hDPP4) receptor (Zhao et al.2014). The study concluded significant weight loss and virus dissemination in the respiratory tract. Further interstitial pneumonia was observed in the mice. The transient hDPP4 expression, mild respiratory system involvement with no fatality are the drawbacks of the model (Zhao et al.2014).

Though the previous study used an adenovirus vector to develop the DPP4 receptor in mice vulnerable to MERS-CoV infection, it did not mimic more closely to the clinical conditions. Agrawal et al. used tissue-specific and inducible promoters, to derive humanized DPP4 transgenic mice. On exposure to MERS-CoV, the model showed high susceptibility to the virus (Agrawal et al.2015). Briefly, the transgenic model depicted progressive pneumonia with fatality.

Agrawal et al. developed a hCD26/DPP4 transgenic mice and the receptors were distributed throughout the body which was not physiologically relevant (Agrawal et al.2015). Pascal et al. used Veloci Gene technology to rapidly develop a new humanized model. The model showed active viral replication in the respiratory tract without cerebral infection (Pascal et al.2015). Tao et al. observed constant inflammatory infiltrates in both brain and lungs in humanized CD26 (hCD26)/DPP4 mice. The high viral titer was observed in the lungs and in the brain of mice on exposure to MERS-CoV infection. The pathological examination found gross and microscopic inflammation which was not restricted to lungs but a sign of systemic dissemination along with fatality in 4–6 dpi (Tao et al.2015).

Tao et al. showed acute severe infection which hampers the opportunity to fully understand the pathogenesis. Zhao et al. developed a mouse model that exhibits the codon-optimized- hDPP4 receptor which on exposure to viral strain showed severe disease pathology in lungs which further affected the brain and kidney. The severe infection and aberrant immune response lead to mortality. The exact mechanism of the severity of the disease model requires further evaluation (Zhao et al.2015).

Cockrell et al. used CRISPR–Cas9 technology to genetically alter a non-permissive host receptor. The model showed severe respiratory distress and fatality response to MERS-CoV. This was further supported by plethysmography that showed decreased pulmonary function which might be due to the inflammation in pneumocytes and airway epithelial cells. The viral titers were not detected in the brain thereby nullifying any confounding factors which could lead to fatality (Cockrell et al.2016).

Li et al. developed transgenic mice, and inoculated with mouse adapted MERS-CoV (MERSMA) viral strain which caused weight loss, infection in airway epithelia, pneumocytes, and macrophages. Though the model was rapid to develop, required no prior sensitization, the pathological changes were exclusively associated with the species-adapted mutations of the MERSMA clone. This hinders the model to fully mimic clinical conditions (Li et al.2017).

Fan et al. used CRISPR/Cas9 gene-editing technology to develop a knock-in-model of mice. The study displayed a high level of expression of hDPP4 in the respiratory tract and to a lesser extent in the brain. There is a clear site of infection along with acute respiratory distress syndrome (ARDS) in the lungs. There is also the dissemination of viruses into the central nervous system (CNS). The study observed high repeatability and the safety of the model (Fan et al.2018).

Iwata-Yoshikawa et al. developed a transgenic mouse model using the endogenous promoter. The hDPP4 mice model showed a high viral titer in the lower respiratory tract. Acute multifocal interstitial pneumonia was observed at 7 dpi. The viral titer was also observed in peripheral blood and lymphoid tissues. There was an age-associated immune response to exposure to the viral strain. The model expressed mild infection pathology. Moreover, after the infection, this model had not shown any signs of lesions in the brain and kidney (Iwata-Yoshikawa et al.2019) (Table 3).

Conclusions

  1. 1.

    The coronaviruses pathology is associated with the involvement of severe acute respiratory infection (SARI) and immune deregulation, thus, it is requisite to perform a study using an assay system that must involve all the cell signalling and thus, using animal model is an unparallel approach.

  2. 2.

    There are some animal models that recapitulate mild to moderate pathology, however, the clinical condition is related to high mortality and morbidity. Therefore, animal models which mimic critical pathologies are much closer to clinical conditions.

  3. 3.

    The novel coronavirus (COVID-19) pathology is linked to viral respiratory infection, hyper-immune response, and coagulopathy (Lin et al.2020; Connors and Levy 2020), therefore, to understand the mechanism or to evaluate therapeutic countermeasures, the animal models should involve all these interplays in a single model. Since with the advent of technology, transgenic mice are the significant tool but the transgenic mice lack the coagulopathy component which is indispensable in order to understand the complete mechanism and to evaluate the countermeasures.

  4. 4.

    Further, validation of the animal model is crucial. The error in the animal experimental study narrows the chances of the potential drugs or repurposing or repositioning drugs or vaccines to translate successfully to clinics and moreover, it is a wastage of resources. Thus, it is the need of the hour to validate the animal model using different criteria, for instance, face, construct, and predictive validity (Denayer et al.2014).