As we know, an ideal animal model is important to the progress of SARS research. Previous studies indicate that the Golden Syrian hamster infected with SARS-CoV has a higher virus replication, a longer duration in the respiratory tract and is accompanied by significant pathological changes, but showed no evidence of the disease (27). Though primates have been confirmed to be optimal animal model for human SARS study, the expense and rarity of the animal obviates their use by most research institutes. The accompanying environmental requirements for the use of a biosafety level 4 laboratory is another obstacle to its widespread study, given the serious nature of SARS-CoV and the evidence of the person-to-person transmission pathway. Even so, many primates failed to mimic characters of SARS disease well. When rhesus and cynomolgus macaques were challenged with 107 PFU SARS-CoV, some of the animals developed a mild self-limited respiratory infection, very different from what observed in patients with SARS (28). McAuliffe and colleagues (19) also noticed that though moderate to high titers of SARS-CoV with associated interstitial pneumonitis could be detected in the lungs of African Green monkeys on day 2 and have been resolved by day 4 post-infection, virus replication was highly restricted and there was no evidence of enhanced disease 2 months later. Taken together, although these models will be useful for the evaluation of immunogenicity of candidate vaccines, the lack of apparent clinical illness and variability from animal to animal in the level of viral replication result in their limited usefulness in the study of SARS and the evaluation of therapeutic effects.
Given that small animal models are urgently needed, we established a murine SARS model, in which Balb/cJ mouse, a standard laboratory rodent was involved. Previous study has indicated that significant pathological features were absent, though SARS-CoV replication was evident in mice (10). Instead of SARS-CoV, we infected Balb/cJ mice with MHV-3, a member of the groupⅡ coronavirus. Significantly, we found that animals died from a multiple organ dysfunction disease in 5 days post MHV-3 infection. As described in our results, although many organs were involved, the predominant damage was restricted to lungs. Other atypical pathological changes, such as mild hydropic degeneration, interstitial cell proliferation and dot necrosis could occasionally be detected in spleen, liver and kidneys. The pathological features displayed here is in consistent with reports from the experience with SARS patients (5, 11, 13, 25, 33).
More recently in vitro evidence also suggests that SARS-CoV may have infected human cells derived from lungs, kidneys, liver, and intestine (9). Beside the respiratory distress, other clinical symptoms in SARS patients also supported the conclusion that virus could attack many organs in an infected individual. For example, liver impairment was commonly reported in up to 60%-66% of patients with SARS (2, 32-34). In addition, the virus replication found in renal tubular epithelial cells and the crypt epithelial of intestine suggested that these were two possible transmission pathways of SARS-CoV. In our study, we also detected virus in all collected organs from Balb/cJ mice infected with MHV-3 through trachea with a significantly higher level in lungs, spleen and liver when compared with those in kidneys, intestine, heart and brain.
Nucleocapsid protein of coronavirus is the key protein for the formation of the helical nucleocapsid during virion assembly. The nucleocapsid protein is also immunogenic and abundantly expressed during infection, and believed to be more conserved than other proteins of the virus, such as spike and membrane glycoproteins. Therefore, it has been widely used as an important viral Ag in immunological and immunohistochemical assays for clinical diagnosis and laboratory detection of coronavirus infection (15, 35). In order to determine the virus targeted cells post inoculation of MHV-3 through trachea in Balb/cJ mice, we detected the nucleocapsid gene transcripts in collected tissues by ISH, in which two oligonucleotide probes were designed according to nucleocapsid gene sequence. Both probes were combined to ensure sensitivity and specificity of the detection. The results revealed that MHV-3 could target a variety of cells post infection as described in the results, which was consistent with what has been reported in SARS patients (3, 30).
Of particular relevance to this report is the extensive studies of the pathogenesis of fulminant viral hepatitis using the same virus in Balb/cJ mice but infected through a different pathway, through the peritoneum instead of trachea (3, 30). Although the virus replication and distribution were shown similar patterns in both animal models, the pathological features in different organs especially lungs, liver and spleen were significantly different. In Balb/cJ mice infected with MHV-3 through peritoneum, the damage within liver is massive and fatal, characterized by widespread necrosis of hepatocytes within 48-72h and all animals have died by 72h with only mild or no disorder in other organs such as lungs and spleen (4). In Balb/cJ mice infected with MHV-3 through the trachea, animals died up to 120h post infection with widespread injury, especially in lungs, characterized by diffuse thickened alveoli septums with numerous lymphocytes and mononuclear cells inflammatory infiltrates, extensive hyaline membrane in pulmonary alveoli, engorgement of capillary or microthrombosis in accompanied vessels at 72h to 96h post infection. However, even by 96h post infection, the damage in liver was still mild, characterized by cloudy swelling, prominent ballooning degeneration with mild lymphocytic infiltration in the portal area and occasionally dot or zonal hepatocyte necrosis. We further measured the serum alanine aminotransferase (ALT) level at 48h post MHV-3 infection to show the liver function within both models. The results displayed that the ALT level (945.8 ± 32.2 IU/L) in Balb/cJ mice infected with MHV-3 through peritoneum is sevenfold as high as that(112.7±36.5 IU/L) in Balb/cJ mice infected with MHV-3 through trachea with the ALT level of 39±2.94IU/L in saline injected Balb/cJ mice (detail data not shown), indicating that instead of fulminant liver failure, severe lesion in lungs and disreputed host immune system were the fetal cause of the Balb/cJ mice when infected with MHV-3 through trachea. Since both the infectious pathway and the pathological features in this murine SARS model induced by MHV-3 through trachea in Balb/cJ mice mimic those in human with SARS-CoV infection, the murine SARS model therefore could provide an attractive and convenient system for the dynamic pathophysiology investigation in SARS-CoV infection.
Although the viral titers in lungs, spleen, liver where disease displayed were significantly higher than those in other organs such as kidneys and brain, there were still a large amount of virus harvested from tissues with only mild or no pathologic disorders, suggesting that in addition to cytopathic effect (CPE) mediated by local virus replication, host factors contributed greatly to the pathologic injury. To date, the details of the host response to SARS-CoV infection is still largely unknown and consequently the most appropriate treatment regime remains to be established. A human peripheral blood mononuclear cells model system was established by Genome Institute of Singapore to investigate genome-wide host responses to SARS-CoV infection (20), while detailed in vivo studies of the host response are now required. Data from our and other laboratories have suggested that host factors may be another important factor in the pathogenesis of MHV-3 disease (26), herein a murine SARS model was used to investigate the host response in vivo and as a proinflammatory gene. fgl2/fibroleukin is an immune coagulant, which can bypass the TF/factorⅦ extrinsic pathway and directly cleave prothrombin to thrombin, followed by fibrin deposition and thrombosis formation. Though having the characteristics of a serine protease, mfgl2 immunoactivity is not inhibited by antithrombin Ⅲ (4). Evidence showed that mfgl2 has an important role in many diseases, such as fulminant viral hepatitis (14, 18), syngeneic transplantation (22) and fetal loss syndrome, whose pathogenesis was dominated by thrombosis in the microvasculature. Since patients dying from SARS revealed the presence of vascular thrombosis in pulmonary vessels (8, 24, 25), we investigated the role of mfgl2 in the lung impairment of murine SARS by Balb/cJ mice infected with MHV-3 through trachea. In our study, specific mfgl2 transcripts and immunoreactive protein were seen in the cytoplasm of terminal and respiratory bronchioles, most of the alveolar epithelia and infiltrating cells in the stroma. Double immunochemistry staining revealed widespread fibrin deposition in association with mfgl2 expression in lungs, especially within the vessels and hyaline-membranes. Therefore fibrin induced by mfgl2 from alveolar epithelia is considered to be involved in this process. In a similar way, specific co-localization of mfgl2 and fibrin in micro-vasculars, in addition to accumulation of numerous erythrocytes indicated the activation of coagulation system induced by mfgl2. Hyaline-membranes appear to be a pivotal mediator in acute respiratory distress, while thrombin can target the vascular endothelial cells and induce secretion of proinflammatory mediators and adhesion of leukocytes, which together potentiate the inflammatory response. Therefore our data provides new insight into the importance of mfgl2 in the lung impairment in MHV-3 induced murine SARS. These findings warrant further investigation of human fibrinogen like protein 2 (hFgl2) expression in patients with SARS. If host factors such as Fgl2 play an important role in SARS progression, a new logical target for molecular manipulation can be investigated and will offer hope for development of new treatment approaches for patients with SARS.