Development and Characterization of a Mouse Model for Marburg Hemorrhagic Fever

The lack of a mouse model has hampered an understanding of the pathogenesis and immunity of Marburg hemorrhagic fever (MHF), the disease caused by marburgvirus (MARV), and has created a bottleneck in the development of antiviral therapeutics. Primary isolates of the filoviruses, i.e., ebolavirus (EBOV) and MARV, are not lethal to immunocompetent adult mice. Previously, pathological, virologic, and immunologic evaluation of a mouse-adapted EBOV, developed by sequential passages in suckling mice, identified many similarities between this model and EBOV infections in nonhuman primates. We recently demonstrated that serially passaging virus recovered from the liver homogenates of MARV-infected immunodeficient (SCID) mice was highly successful in reducing the time to death in these mice from 50 to 70 days to 7 to 10 days after challenge with the isolate MARV-Ci67, -Musoke, or -Ravn. In this study, we extended our findings to show that further sequential passages of MARV-Ravn in immunocompetent mice caused the MARV to kill BALB/c mice. Serial sampling studies to characterize the pathology of mouse-adapted MARV-Ravn revealed that this model is similar to the guinea pig and nonhuman primate MHF models. Infection of BALB/c mice with mouse-adapted MARV-Ravn caused uncontrolled viremia and high viral titers in the liver, spleen, lymph node, and other organs; profound lymphopenia; destruction of lymphocytes within the spleen and lymph nodes; and marked liver damage and thrombocytopenia. Sequencing the mouse-adapted MARV-Ravn strain revealed differences in 16 predicted amino acids from the progenitor virus, although the exact changes required for adaptation are unclear at this time. This mouse-adapted MARV strain can now be used to develop and evaluate novel vaccines and therapeutics and may also help to provide a better understanding of the virulence factors associated with MARV.The filoviruses, Marburgvirus and Ebolavirus (MARV and EBOV), cause severe hemorrhagic fevers in humans and nonhuman primates (27). The incubation time is estimated to be 3 to 21 days, with human case fatality rates reaching 90% in some outbreaks. Filoviral hemorrhagic fevers are characterized by a nonspecific viral prodrome in the early stage of infection, including fever, headaches, and myalgia (27). This is followed by a hemorrhagic phase that can include development of a maculopapular rash, petechiae, and bleeding from the gums, intestines, and other mucosal surfaces. Death usually occurs within a week of initial symptoms and is thought to be due to uncontrolled viral replication, hypotension-induced shock caused by increased vascular permeability, and multiorgan failure, likely caused by disseminated intravascular coagulation and extensive necroses in the liver, spleen, intestine, and many other major organ systems (27).Human-derived MARVs (isolates Angola, Musoke, Ravn, and Ci67) do not kill immunocompetent adult mice (23). Furthermore, there are no published reports of any lethal mouse-adapted MARV. The current mouse-adapted EBOV, strain Zaire (ZEBOV), was developed by performing nine sequential passages of ZEBOV 1976 virus in suckling mice, followed by two sequential plaque picks. The resulting virus was uniformly lethal to mice inoculated intraperitoneally (i.p.). Pathological evaluation of infected mice identified many similarities and only a few differences between this model (7, 22) and infections in nonhuman primates (21).In a previous study, we took a slightly different approach to mouse adaptation of MARV and found that serially passaging virus recovered from the liver homogenates of MARV-Ravn-infected adult mice with severe combined immunodeficiency (SCID mice) resulted in the generation of SCID-adapted MARV-Ravn (scid-MARV) that rapidly killed SCID mice but did not kill adult immunocompetent mice (51). In this study, we used scid-MARV as starting material for the first round of infection of adult immunocompetent BALB/c mice and serially passaged virus recovered from the liver homogenates of the BALB/c mice. MARV-Ravn was chosen over SCID-adapted MARV-Ci67 or -Musoke because it adapted more rapidly to SCID mice than the other isolates did. This produced a mouse-adapted MARV-Ravn strain (ma-MARV) that could kill adult BALB/c mice. Serial sampling studies to characterize the pathogenesis of ma-MARV revealed that this model was very similar to the guinea pig and nonhuman primate Marburg hemorrhagic fever (MHF) models, including rapid viremia, induction of D-dimers (fibrin degradation products), thrombocytopenia, profound loss of circulating and tissue lymphocytes, and marked liver damage. Additionally, we compared the immunological responses of mice after infection with either nonadapted wild-type MARV-Ravn (wt-MARV) or ma-MARV. This mouse model of MARV infection not only should advance our understanding of MARV pathogenesis and immunity but also may play a critical role in discovery of therapeutics for MARV infection.