Immunobiology, molecular virology and countermeasures of highly pathogenic viruses

National Institute Of Allergy And Infectious Diseases

Investigators

Linked publications, trials & patents

Abstract

Filovirus pathogenesis in the ferret model Ferrets have recently been established as a disease model for wildtype ebolavirus infections. Taï Forest virus (TAFV) is a lesser-known filovirus and has caused only a single human infection, but several disease outbreaks in chimpanzees have been linked to this virus. To investigate the potential of ferrets as a disease model for this virus, we infected ferrets with TAFV by the intranasal (IN), intramuscular (IM) or aerosol routes and observed disease progression including regular blood draws and radiographs. The IM group showed minimal signs of disease and 100% survival. In contrast, mucosal infection resulted in 83% (IN) and 50% (aerosol) survival. Ferrets developing severe disease were euthanized 9-10 days post-infection (dpi) presenting with fever, high-titer viremia, thrombocytopenia, viral shedding, and systemic viral spread. Ferrets surviving TAFV infection developed TAFV-specific IgG and were re-challenged either IM or IN with Ebola virus (EBOV) to assess the potential cross-protection since TAFV and EBOV are closely related filoviruses. Only ferrets infected IN with TAFV and IN with EBOV were uniformly protected from disease. However, all other ferrets succumbed to EBOV infection 5-7 dpi presenting with filovirus disease including fever, high-titer viremia, and thrombocytopenia (Fletcher et al. manuscript in revision with Plos Pathogens). This data shows that ferrets are a feasible model to assess TAFV pathogenicity by natural exposure routes and that cross-protection between filoviruses may be achieved depending on exposure routes. We performed additional ferret studies with infection by several filoviruses by the same 3 routes. In addition, we investigated if survivors of one filovirus infection were protected from another one exploring the cross-protection potential between filoviruses. The first isolate of the emerging filovirus, Lloviu virus (LLOV), was obtained in 2022, but no animal disease models have been established. We assessed its pathogenic potential in ferrets after IN, IM, or aerosol exposure. The observed lack of disease demonstrated that ferrets are not a disease model for LLOV and only a few of the ferrets seroconverted (Fletcher et al. 2024 Emerg Inf Dis). In contrast, Sudan virus (SUDV) caused highly lethal disease in the ferret model. We determined that the route of infection did not significantly impact overall SUDV pathogenicity; only subtle changes were detected in magnitude of viremia and oral viral shedding. Additionally, we sought to determine if preexisting LLOV immunity could protect ferrets from lethal SUDV infection. We found that the previous immunity elicited by LLOV infection was not sufficient to protect ferrets from lethal SUDV disease. In conclusion, our results indicate that the infection route has minimal effect on overall pathogenicity of SUDV in ferrets and that prior LLOV infection does not elicit a cross-protective immune response to SUDV (O’Donnell et al. 2025 Adv Virus Res). Next, the ferret was evaluated as a potential disease model for Bombali virus, another novel filovirus. This virus caused signs of disease after mucosal (IN and aerosol) infection in ferrets, but no lethality. After a resting period, surviving ferrets were challenged with EBOV by the IM or IN route resulting in uniform lethality without differences regarding infection route or pre-existing BOMV immunity (O’Donnell et al. manuscript in preparation). These models will be used for future preclinical countermeasure studies. Molecular determinants of filovirus pathogenesis Since its discovery in 1967, Marburg virus (MARV) has caused human disease outbreaks with an average case fatality of 62 %. The recent outbreaks of MARV disease (MVD) in Rwanda 2024 and Tanzania 2025 reiterate its present threat to global public health and emphasize the need to develop countermeasures. While ferrets recapitulate hallmarks of human disease when infected with Ebola virus (EBOV), they do not show any signs of disease when infected with MARV. Therefore, we aim to identify factors, likely interfering with the host innate immune response, determining the lack of pathogenicity of MARV in ferrets (Haase & Marzi 2025 Adv Virus Res). Primary ferret cells isolated from various tissues were immortalized and tested for their innate immune signaling ability. While wildtype MARV replicated poorly in these cells, EBOV and rodent-adapted strains of MARV replicated to similar levels. Most mutations acquired in the adaptation process are encoded in proteins antagonizing the innate immune response, therefore, we conducted a ferret pathogenesis study with hamster-adapted (HA-)MARV, which carries two mutations in the VP40 (known innate immune antagonist) compared to the wildtype MARV. HA-MARV mucosal (IN and aerosol) infection resulted in uniform lethality in ferrets. In contrast, 33% of the IM infected ferrets survived. This data confirms our hypothesis that the suppression of the host innate immune system plays a role in the development of lethal MVD in ferrets (Haase et al. manuscript in preparation). The findings provide valuable insight into the factors contributing to MARV virulence and pathogenesis and will result in the identification of novel targets for therapeutic intervention against MVD. In collaboration with Drs. Van Tol and Munster in the LV the importance of the innate immune response in the Jamaican fruit bat (JFB) was explored. EBOV-infected bats supported systemic virus replication and shed infectious virus orally. In contrast, MARV replicated only transiently and was not shed. Although the EBOV-infected bats experienced acute, mild hypothermia, they remained otherwise clinically healthy and seroconverted. In vitro, JFB immortalized kidney cells replicate EBOV 100 to 10,000 times more efficiently than MARV. These in vivo and in vitro data suggests that innate immune factors may determine bat-filovirus compatibility (van Tol et al. 2025 Nat Comm). In collaboration with Dr. Haigh in LNII we are investigating the ability of EBOV and MARV to infect cerebral brain organoids to investigate if reports of neurologic signs during acute disease but also as long-term sequalae may be modeled in this system. This project aims to identify factors that may serve as potential antiviral or host-directed therapeutic targets. Another study led by Dr. Furuyama (Asada) from Nagasaki University in collaboration with us highlighted the importance of the cellular endocytic pathway for MARV particle production (Furuyama et al. 2025 Microbiol Spectrum). These findings highlight potential cellular components that may be targets of future host-directed therapeutics. Countermeasure development Single-dose vaccines protect guinea pigs and nonhuman primates (NHPs) from lethal filovirus disease The most recent SUDV outbreak created a public health emergency in Uganda and the Eastern Africa region. Currently, there are licensed countermeasures for EBOV; however, there are no licensed vaccines or therapeutics against SUDV (Marzi & Feldmann 2025 npj vaccines; Marzi 2025 Plos Biol). We developed a VSV-based vaccine expressing the SUDV GP as the viral antigen and demonstrated protective efficacy with a single high-dose within 4 weeks (Anhalt & Marzi 2025 Methods Mol Biol; Marzi et al. 2023 Lancet Microbe; Provisional patent filed). We expanded upon this data by IM vaccinating NHPs with a single dose of VSV-SUDV either one month or one week prior to SUDV challenge. A third group was vaccinated with a single dose of VSV-EBOV one month prior to SUDV challenge to assess its cross-protective potential, and a control group received an unrelated VSV-based vaccine. All vaccinated NHPs developed antigen-specific IgG within 2 weeks of vaccination, including cross-reactive responses. After challenge with a lethal dose of SUDV, all VSV-SUDV-vaccinated NHPs were protected from disease. In contrast, the VSV-EBOV-vaccinated and control NHPs succumbed to disease between day 5 and 7 after challenge presenting with classical signs of SUDV disease associated with high titer viremia, high viral organ load, dysregulated cytokine profiles and typical pathological changes. The humoral immune response in the NHPs vaccinated with VSV-SUDV one month before challenge resulted in a profound and sustained antibody response with a diverse functionality profile which was not observed to the same extent in NHPs vaccinated one week before challenge. The fast-acting nature makes VSV-SUDV an ideal countermeasure for ring vaccination during outbreaks of SUDV disease (Fletcher et al. manuscript in revision with Lancet Microbe). In contrast, VSV-EBOV provided no relevant protection against SUDV infection in NHPs highlighting the need for species-specific filovirus vaccines (Marzi & Feldmann 2025 npj vaccines; Marzi 2025 Plos Biol). In collaboration with Dr. Messaoudi from the University of Kentucky we analyzed the whole blood transcriptome of these NHPs and found that NHPs vaccinated with VSV-EBOV (non-protective) generated a transcriptional response following SUDV challenge indicative of dysregulated inflammation. In contrast, NHPs that received the VSV-SUDV vaccine generated a transcriptional response indicative of a recall adaptive immune response. Finally, in-silico deconvolution methods indicated changes in immune cell frequency consistent with immune response and resolution in the VSV-SUDV-vaccinated NHPs that is not observed with VSV-EBOV-vaccinated NHPs (Doratt et al. manuscript submitted). We also used the SUDV guinea pig model for a comparative vaccine study. We compared the immunogenicity and protective efficacy of 2 versions of the LION-SUDV vaccine, one encoding only the SUDV glycoprotein (GP) and the other encoding both the SUDV and EBOV GPs (LION-Combination). The LION-EBOV vaccine alone was also included in our study to determine the potential for cross-reactivity of immune responses and cross-protection. We investigated the efficacy of these vaccines in a single-dose 3-week vaccination to challenge study in the SUDV guinea pig disease model. The study resulted in uniform protection for the LION-SUDV and the LION-Combination groups, while the LION-EBOV vaccine achieved only 50% protection. Antigen-specific humoral responses correlated with decreased virus replication and survival. This data supports the need for further studies in larger animal species to ensure that protective efficacy is maintained with this single-dose vaccine before translation into the clinical setting (O’Donnell et al. 2025 Nat Comm). Bundibugyo virus (BDBV) causes severe hemorrhagic disease in humans in Africa; outbreaks are associated with a mortality rate of approximately 20%–51%. We generated a BDBV-specific vaccine based on recombinant VSV expressing the BDBV glycoprotein (VSV-BDBV) and investigated its potential to provide rapid protection. Groups of NHPs were vaccinated with a single dose of VSV-BDBV either 7 or 3 days before lethal BDBV challenge. A control group received an unrelated vaccine 7 days before challenge. All VSV-BDBV-vaccinated macaques survived the challenge demonstrating uniform protection with limited to no viremia and viral shedding detected throughout the study. In contrast, 60% of the control group succumbed to the disease demonstrating with high viremia. The NHPs that recovered only reached mild viremia titers. All vaccinated macaques showed evidence of an early antigen-specific functional humoral response (3-9 days post-challenge (DPC)), while surviving NHPs demonstrated similar humoral responses only at 14 DPC. NHPs vaccinated 7 days prior to challenge had significantly higher levels of antigen-specific IgG 3 DPC than both the control NHPs and the NHPs vaccinated 3 days prior to challenge demonstrating the rapid speed of the formation of the humoral response post-vaccination. Overall, this study demonstrated that vaccination with VSV-BDBV provides a rapid survival benefit in macaques against BDBV infection, supporting the potential use of this vaccine during outbreaks in a ring vaccination approach (O’Donnell et al. manuscript in preparation). Vaccine efficacy in the context of underlying infections In collaboration with Dr. Maury at the University of Iowa we investigated the impact of malaria on VSV-EBOV (FDA-approved EBOV vaccine) protective efficacy. Regions in which EBOV outbreaks occur in Africa are also highly endemic for malaria, an infectious disease caused by Plasmodium parasites. Plasmodium infections persist in these regions due in part to the parasite’s ability to evade sterilizing immunity, a feature which can also dampen immune responses to heterologous pathogens and vaccines. We used murine Plasmodium infections to determine that an acute Plasmodium infection at the time of VSV-EBOV vaccination reduces vaccine-mediated protection against mouse-adapted EBOV challenge. Decreased protection was associated with a Plasmodium-induced IFN-γ-mediated decrease of VSV-EBOV replication in lymph node macrophages that resulted in reduced primary anti-EBOV glycoprotein antibody responses. Higher doses of VSV-EBOV partially overcame the Plasmodium-imposed antibody deficit and elicited protective responses. These data suggest a negative impact of Plasmodium on the efficacy of low dose VSV-EBOV vaccine protocols and support the use of high antigen loads to overcome Plasmodium-associated deficits in the efficient management of EBOV outbreaks (Elliff et al. manuscript submitted). HIV is endemic in areas of the world where MARV outbreaks occur. Antiretroviral therapy (ART) combats HIV/AIDS disease progression by inhibiting HIV replication. The lymphocytic dysfunctionality induced by HIV infection is of concern when it comes to providing immunity against other pathogens. The ability of vaccines to induce robust immune responses against common pathogens, such as influenza, is reduced in people living with HIV. Dr. Marzi shares a Rocky-Beth fellow with Dr. Brenchley in the LVD who is investigating the immunogenicity and efficacy of different MARV vaccines within the context of SIV infection and antiretroviral therapy (ART). Past studies in hamsters suggest that the VSV-MARV vaccine boosts the B-cell antibody response, while the LION-MARV skews immunity towards a more balanced B/T-cell response (O’Donnell et al. 2024 Int J Mol Sci). Both vaccines are currently being evaluated in SIV+ and SIV- NHPs to determine if one vaccine may induce greater protection against MARV infection in individuals living with HIV. We also included an ART regimen because it has been demonstrated that ART is crucial for maintaining robust B-cell immune responses to seasonal vaccination and other infections in HIV patients but has unknown influence on MARV replication. Therefore, it is imperative to examine the potential role of ART, especially since there is global access to these treatments. We are comparing MARV GP-specific humoral and cellular immune responses induced by the two different MARV vaccines in SIV+ and SIV- NHPs, with half of the SIV+ RMs receiving ART. In addition, protective efficacy will be evaluated with lethal MARV challenge. It is critical to establish which MARV vaccine may be the most protective in people living with HIV as it is endemic in areas that have a high potential for MARV outbreaks. Vaccines for emerging viruses In response to emerging infectious disease outbreaks, we developed VSV- and VSV-EBOV-based vaccines against a number of these pathogens and conducted preclinical NHP efficacy studies. We have developed VSV-based vaccines with preclinical efficacy for Crimean-Congo hemorrhagic fever virus (Tipih et al. 2025 npj vaccines) and influenza A virus (Furuyama et al. npj vaccines 2020; O’Donnell et al. manuscript in preparation). Currently, we are evaluating the protective efficacy of our VSV-HA vaccine after IN administration against recent H5N1 isolates from the USA in rodents (O’Donnell et al. manuscript in preparation). We are also supporting the development of filovirus vaccines by other groups including a promising approach in collaboration with Dr. Theresa Lambe based on the ChAdOx-platform for MARV which was successful in hamsters (O’Donnell et al. 2024 Int J Mol Sci; Smith & Marzi 2025 Methods Mol Biol).