Attempted Transmission of Marburg Virus by Bat-Associated Fleas

Egyptian rousette bats (ERBs) are implicated as reservoir hosts for Marburg virus (MARV), but natural mechanisms involved in maintenance of MARV in ERB populations remain undefined. A number of hematophagous ectoparasites, including fleas, parasitize bats. Subcutaneous (SC) inoculation of ERBs with MARV consistently results in viremia, suggesting that infectious MARV could be ingested by blood-sucking ectoparasites during feeding. In our study, MARV RNA was detected in fleas that took a blood meal during feeding on viremic bats on days 3, 7, and 11 after SC inoculation. Virus concentration in individual ectoparasites was consistent with detectable levels of viremia in the blood of infected host bats. There was neither seroconversion nor viremia in control bats kept in close contact with MARV-infected bats infested with fleas for up to 40 days post-exposure. In fleas inoculated intracoelomically, MARV was detected up to 14 days after intracoelomic (IC) inoculation, but the virus concentration was lower than that delivered in the inoculum. All bats that had been infested with inoculated, viremic fleas remained virologically and serologically negative up to 38 days after infestation. Of 493 fleas collected from a wild ERB colony in Matlapitsi Cave, South Africa, where the enzootic transmission of MARV occurs, all tested negative for MARV RNA. While our findings seem to demonstrate that bat fleas lack vectorial capacity to transmit MARV biologically, their role in mechanical transmission should not be discounted. Regular blood-feeds, intra- and interhost mobility, direct feeding on blood vessels resulting in venous damage, and roosting behaviour of ERBs provide a potential physical bridge for MARV dissemination in densely populated cave-dwelling bats by fleas. The virus transfer might take place through inoculation of skin, mucosal membranes, and wounds when contaminated fleas are squashed during auto- and allogrooming, eating, biting, or fighting.