Survey of mosquito-borne flaviviruses in the Cuitzmala River Basin, Mexico: do they circulate in rodents and bats?

Background RNA viruses commonly infect bats and rodents, including mosquito-borne flaviviruses (MBFV) that affect human and animal health. Serological evidence suggests past interactions between these two mammalian orders with dengue viruses (DENV), West Nile virus (WNV), and yellow fever virus (YFV). Although in Mexico there are reports of these viruses in both host groups, we know little about their endemic cycles or persistence in time and space. Methods Rodents and bats were captured at the Cuitzmala River Basin on the Pacific coast of Jalisco state, Mexico, where MBFV, such as DENV, have been reported in both humans and bats. Samples were taken during January, June, and October 2014, at locations adjacent to the river. Tissue samples were collected from both bats and rodents and serum samples from rodents only. Highly sensitive serological and molecular assays were used to search for current and past evidence of viral circulation. Results One thousand nine hundred forty-eight individuals were captured belonging to 21 bat and 14 rodent species. Seven hundred sixty-nine liver and 764 spleen samples were analysed by means of a specific molecular protocol used to detect flaviviruses. Additionally, 708 serum samples from rodents were examined in order to demonstrate previous exposure to dengue virus serotype 2 (which circulates in the region). There were no positive results with any diagnostic test. Discussion To our knowledge, this is the first survey of rodents and only the second survey of bats from the Pacific Coast of Mexico in a search for MBFV. We obtained negative results from all samples. We validated our laboratory tests with negative and positive controls. Our findings are consistent with other empirical and experimental studies in which these mammalian hosts may not replicate mosquito-borne flaviviruses or present low prevalence. Conclusions True-negative results are essential for the construction of distribution models and are necessary to identify potential areas at risk. Negative results should not be interpreted as the local absence of MBFV in the region. On the contrary, we need to establish a long-term surveillance programme to find MBFV presence in the mosquito trophic networks, identifying the potential role of rodents and bats in viral dynamics.