Abortive vampire bat rabies infections in Peruvian peridomestic livestock

Abstract

Rabies virus infections normally cause universally lethal encephalitis across mammals. However, 'abortive infections' which are resolved prior to the onset of lethal disease have been described in bats and a variety of non-reservoir species. Here, we surveyed rabies virus neutralizing antibody titers in 332 unvaccinated livestock of 5 species from a vampire bat rabies endemic region of southern Peru where livestock are the main food source for bats. We detected rabies virus neutralizing antibody titers in 11, 5 and 3.6% of cows, goats and sheep respectively and seropositive animals did not die from rabies within two years after sampling. Seroprevalence was correlated with the number of local livestock rabies mortalities reported one year prior but also one year after sample collection. This suggests that serological status of livestock can indicate the past and future levels of rabies risk to non-reservoir hosts. To our knowledge, this is the first report of anti-rabies antibodies among goats and sheep, suggesting widespread abortive infections among livestock in vampire bat rabies endemic areas. Future research should resolve the within-host biology underlying clearance of rabies infections. Cost-effectiveness analyses are also needed to evaluate whether serological monitoring of livestock can be a viable complement to current monitoring of vampire bat rabies risk based on animal mortalities alone.

Fig 1. Spatiotemporal correlations between rabies mortality incidence and seroprevalence.

(A) Map showing the districts of Peru where the study was carried out, color coded by the number of livestock rabies outbreaks reported to SENASA during 12 months prior to serological sampling. Lower-case letters correspond to districts shown in panel C. (B) Scatterplot showing the relationship between seroprevalence and rabies cases in the 12 months prior to sampling, in the 6 months surrounding sampling, and in the 12 months following sampling at each district. The size of points is scaled to sample size per district (range = 7–51 samples). The correlation coefficient (r) and its associated p-value of a Pearson’s correlation test are shown for each relationship. (C) Time series of rabies incidence in each sampled district from 2003–2017 showing sporadic outbreaks. Pie charts below indicate the proportion of livestock in each district that were seropositive in 2016 (red = positive, white = negative).