The reservoir for zoonotic o’nyong-nyong virus (ONNV) has remained unknown since this virus was first recognized in Uganda in 1959. Building on existing evidence for mosquito blood-feeding on various frugivorous bat species in Uganda, and seroprevalence for arboviruses among bats in Uganda, we sought to assess if serum samples collected from bats in Uganda demonstrated evidence of exposure to ONNV or the closely related zoonotic chikungunya virus (CHIKV). In total, 652 serum samples collected from six bat species were tested by plaque reduction neutralization test (PRNT) for neutralizing antibodies against ONNV and CHIKV. Forty out of 303 (13.2%) Egyptian rousettes from Maramagambo Forest and 1/13 (8%) little free-tailed bats from Banga Nakiwogo, Entebbe contained neutralizing antibodies against ONNV. In addition, 2/303 (0.7%) of these Egyptian rousettes contained neutralizing antibodies to CHIKV, and 8/303 (2.6%) contained neutralizing antibodies that were nonspecifically reactive to alphaviruses. These data support the interepidemic circulation of ONNV and CHIKV in Uganda, although Egyptian rousette bats are unlikely to serve as reservoirs for these viruses given the inconsistent occurrence of antibody-positive bats.

Rift Valley fever phlebovirus (RVFV) is a zoonotic mosquito-borne pathogen endemic to sub-Saharan Africa and the Arabian Peninsula which causes Rift Valley fever in ruminant livestock and humans. Co-infection with divergent viral strains can produce reassortment among the L, S, and M segments of the RVFV genome. Reassortment events can produce novel genotypes with altered virulence, transmission dynamics, and/or mosquito host range. This can have severe implications in areas where RVFV is endemic and convolutes our ability to anticipate transmission and circulation in novel geographic regions. Previously, we evaluated the frequency of RVFV reassortment in a susceptible ruminant host and observed low rates of reassortment (0–1.7%). Here, we tested the hypothesis that reassortment occurs predominantly in the mosquito using a highly permissive vector, Culex tarsalis. Cells derived from Cx. tarsalis or adult mosquitoes were co-infected with either two virulent (Kenya-128B-15 and SA01-1322) or a virulent and attenuated (Kenya-128B-15 and MP-12) strain of RVFV. Our results showed approximately 2% of virus genotypes isolated from co-infected Cx. tarsalis-derived cells were reassortant. Co-infected mosquitoes infected via infectious bloodmeal resulted in a higher percentage of reassortant virus (2–60%) isolated from midgut and salivary tissues at 14 days post-infection. The percentage of reassortant genotypes isolated from the midguts of mosquitoes co-infected with Kenya-128B-15 and SA01-1322 was similar to that of mosquitoes co-infected with Kenya-128B-15 and MP-12- strains (60 vs. 47%). However, only 2% of virus isolated from the salivary glands of Kenya-128B-15 and SA01-1322 co-infected mosquitoes represented reassortant genotypes. This was contrasted by 54% reassortment in the salivary glands of mosquitoes co-infected with Kenya-128B-15 and MP-12 strains. Furthermore, we observed preferential inclusion of genomic segments from the three parental strains among the reassorted viruses. Replication curves of select reassorted genotypes were significantly higher in Vero cells but not in Culex—derived cells. These data imply that mosquitoes play a crucial role in the reassortment of RVFV and potentially contribute to driving evolution of the virus.