MERS-CoV viral RNA could be detected in blood, urine and stool specimens of some MERS patients, suggesting that the virus dissemination may occur [
17]. We and others have shown that endothelial cells of blood vessel in human ex vivo lung tissues were permissive to MERS-CoV [
23,
33], which may provide the pathological basis of the potential virus dissemination. In addition, it has been demonstrated that MERS-CoV can infect human monocyte-derived dendritic cells and cause productive viral replication [
38]. Human primary T cells are also readily susceptible to MERS-CoV [
45]. Dendritic cells and T cells are migrating cells in the human body. Therefore, the MERS-CoV infected dendritic cells and T cells may allow the virus to disseminate systemically beyond the respiratory tract. Collectively, extrapulmonary organs and tissues are very likely to be involved in MERS-CoV infection in vivo. However, so far, no human autopsy study has been documented. To address the possible extrapulmonary involvement, we have to seek evidence from MERS-CoV infected experimental animals although none of these animals can fully recapitulate the human MERS disease.
MERS-CoV infection and pathogenesis in non-human primate model and small animal models
The receptor for MERS-CoV was identified to be an exopeptidase, dipeptidyl peptidase 4 (DPP4) [
47]. The role of DPP4 as the main determinant in the host tropism of MERS-CoV has been elucidated in several studies [
48‐
50]. Commonly used laboratory animal species such as Syrian hamster, mice and ferrets are not susceptible to MERS-CoV since DPP4 orthologs of these animal species are unable to bind MERS-CoV spike protein and mediate virus entry [
49,
51,
52]. The MERS-CoV inoculation in rabbits displayed an asymptomatic infection. Neither significant histopathological change nor clinical symptom was observed in these rabbits although the virus could be detected from lung tissues [
53]. Camels are susceptible to the MERS-CoV isolated from human. Although the infected camels can shed large amounts of virus from the upper respiratory tract, the disease signs were mild [
54]. Among all experimental animals tested for the development of MERS animal models, rhesus macaques developed a mild to moderate respiratory infection [
55,
56] whereas common marmosets displayed a moderate to severe respiratory disease after inoculation using a combination routes of intranasal, intra-tracheal, oral and ocular [
57]. In marmosets, the MERS-CoV infection was suggested to be a disseminated infection since viral RNA was detectable in nearly all tested tissues in all infected animals, including blood, kidney, intestine, liver and spleen etc. However, except for the samples from the respiratory tract, isolation of infectious virus in other organs was not successful. Interestingly, the susceptibility of macrophages to MERS-CoV as evidenced in the in vitro and ex vivo studies was verified in alveolar macrophages of MERS-CoV infected marmoset [
57].
The first MERS mouse model was generated by prior transduction of adenoviral vector expressing human DPP4 (hDPP4). These mice displayed a transient viral pneumonia which resolved within 1–2 weeks after infection [
44]. Several lines of human DPP4 transgenic mice have been subsequently reported. MERS-CoV infection and replication were invariably evidenced in these hDPP4 transgenic mice [
58‐
60]. However, disease sign and pathology in these mice differed, which appeared to depend on the promoters controlling the expression of hDPP4 gene. Pascal et al. replaced the mouse DPP4 ORF with human DPP4 so that the knocked-in hDPP4 is under the control of the endogenous mouse DPP4 promoter [
60]. The authors believed that this strategy may render hDPP4 to be expressed in a physiologically-relevant context. After virus inoculation in these mice, MERS-CoV robustly replicated in the mouse lung. However, the inoculated mice did not exhibit disease signs, without any extrapulmonary involvement. On the other hand, two lines of hDPP4 transgenic mice which had the transgene under the control of chicken β-actin promoter [
58] and cytokeratin 18 promoter [
59] respectively, were highly permissive to MERS-CoV infection. The mice developed progressive pneumonia with fatal outcome after intranasal inoculation. The infectious virus can constantly and exclusively be recovered from lung and brain tissues [
58,
59]. Similar to the marmoset study, while viral RNA can also be detected from extrapulmonary organs including the heart, spleen and intestines, virus isolation from these organs was unsuccessful. The hDPP4 transgenic mouse with cytokeratin 18 (CK18) promoter was generated in parallel with another transgenic mouse line with surfactant protein C (SPC) promoter. SPC promoter confers transgene expression in bronchiolar and alveolar epithelia while CK18 promoter can drive a more universal transgene expression in epithelia of liver, kidney, gastrointestinal tract and some cells in the nervous system, apart from respiratory tract. The hDPP4 mice with CK18 promoter developed lethal infection after intranasal inoculation of MERS-CoV. In contrast, the same inoculation in the hDPP4 mice with SPC promoter caused no mortality or body weight loss [
59]. Therefore, the morbidity and mortality in human DPP4 transgenic mice may correlate to the tissue/cellular distribution and/or the expression intensity of the transgene. Notably, a common discovery among these MERS-CoV susceptible animals was that the gene expression of antiviral cytokines, proinflammatory cytokines and chemokines was elevated [
55,
57,
58]. Collectively, the tissue tropisms of MERS-CoV in human hosts have not been fully elucidated although there has been accumulating evidence of possible extrapulmonary involvement in MERS patients. Undoubtedly, hDPP4 transgenic mouse studies are conducive for the development of antivirals or vaccines against MERS-CoV. However, in terms of implication for the pathogenesis in human MERS, findings from these hDPP4 transgenic mice must be interpreted with great caution. An in-depth investigation of the tissue tropisms of MERS-CoV in human hosts will facilitate our understanding towards the transmission route and pathogenesis of MERS.
The MERS-CoV mouse models have been utilized to test the efficacy of antiviral drug, neutralization antibody and vaccine. A Venezuelan equine encephalitis replication particle expressing MERS-CoV spike protein (VRP-S) was demonstrated to substantially reduce the virus titer in lung tissues of the immunized hDPP4 transduced mouse model [
44]. The same group subsequent examined the VRP-S in a more permissive hDPP4 transgenic mouse model [
59]. The VRP-S immunized hDPP4 transgenic mice were completely protected from lethal infection. Pretreatment with serum of the mouse immunized with VRP-S can also be protected from fatal infection [
59]. Two fully human neutralization antibodies binding to distinct epitopes of MERS spike protein, which were generated using the mouse expressing human antibody, displayed the pre- and post-exposure protection efficacy from MERS-CoV infection in hDPP4 transgenic mice [
61].