A global serosurvey of RNA virus-reactive antibodies in bats and rodents reveals the enzootic presence of flavi- and paramyxoviruses in African Pteropodidae

Abstract

Bats and rodents have been identified as reservoir hosts of diverse RNA viruses that pose potential risks to humans. However, overall virus prevalence in wildlife and underlying factors determining the zoonotic potential of reservoir-borne viruses remain poorly characterized. Virus detection often relies on the identification of viral nucleotide sequences and may therefore be limited by variable virus concentration, tissue tropism, sample quality, and duration of infection. In this study, we applied a multiplex immunofluorescence assay to examine bat and rodent seroprevalence against 14 different human pathogenic RNA viruses spanning seven virus families (Coronaviridae, Flaviviridae, Hantaviridae, Paramyxoviridae, Phenuiviridae, Pneumoviridae, Togaviridae). We retrospectively analyzed 1,135 bat (16 species) and 454 rodent (15 species) blood or transudate specimens collected at 63 sites in seven countries. Tropical bats exhibited notably high rates of seropositivity against certain virus families (Flaviviridae: 42.8%; Paramyxoviridae: 60.4%; Togaviridae: 7.5%). Multivariable logistic regression models were created to evaluate the association of tropical bat characteristics with seropositivity. Preliminary findings indicate that large colony size, found among species such as Eidolon helvum and Rousettus aegyptiacus, is a risk factor for orthoflavi- and orthorubulavirus infection, consistent with the communal spread of bat-adapted viruses. Author SummaryMany animal populations carry viruses that have the potential to spill over to humans and cause severe disease. Despite surveillance efforts, it is logistically challenging to detect virus infections in wildlife, as viruses may have a short duration of infection or unknown routes of transmission. In this study, we examined blood and transudate samples from rodents and bats around the world for antibodies against 14 different viruses known to cause disease in humans. We analyzed the samples via a mosaic chip-based immunofluorescence test, which has the advantage of detecting antibodies against whole-virus antigen. This method allows for broad surveillance, as we would also expect to uncover evidence of past infection with bat- and rodent-adapted viruses that are related, but not identical, to known human pathogens. We detected antibodies against all seven virus families we examined, with especially high detection rates of antibodies in tropical bats. We used the results to construct multivariable logistic regression models of tropical bat seroprevalence to examine potential associations between tropical bat characteristics and previous infection by certain virus families. Bats with large colony sizes, such as Eidolon helvum and Rousettus aegyptiacus, were apparently at increased risk of infection with orthoflavi- and orthorubulaviruses compared to other tropical bat species.