Monitoring of humoral immune responses against Marburg virus and evaluation of their role in protection against re-challenge in naturally immune Rousettus aegyptiacus fruit bats

Abstract

Marburg virus (MARV) is a zoonotic virus of significant potential public health concern in Africa. Together with Ebola virus (EBOV), MARV belongs to the family Filoviridae and causes a life-threatening haemorrhagic disease in humans and non-human primates. The occurrence of large outbreaks of MARV disease (MVD) within the past two decades, as well as the devastating EBOV outbreaks in West and Central Africa, indicates that filoviruses are a much more significant public health threat than previously anticipated and can emerge at any time without warning. These unprecedented outbreaks have emphasised the need for surveillance in reservoir host populations and for safe and reliable surveillance tools and diagnostic tests that may easily be performed in both laboratory and field settings. Egyptian rousette bats (ERB; Rousettus aegyptiacus) are reservoir hosts for MARV, and there is a need for understanding the dynamics of immune responses of these animals to MARV infection. This knowledge can assist in predicting periods of increased transmission within bat colonies and in turn, potential spillover events into human and other animal populations. In this thesis, the development of indirect enzyme-linked immunosorbent assays (I-ELISA) for the detection of specific anti-MARV immunoglobulin G (IgG) in bat sera is described. The IELISAs, based on two recombinant MARV protein antigens (nucleoprotein and glycoprotein), can be used without the need for high biocontainment facilities. Both I-ELISAs were found to be robust and repeatable, with good sensitivity and specificity. Applying the I-ELISAs in detecting IgG antibodies to MARV in sera collected from both wild-caught and experimentally infected bats indicated that the assays are suitable methods for MARV serosurveillance, with the MARV GP-based I-ELISA demonstrating higher diagnostic performance compared to the MARV NP-based assay. Based on this knowledge, the MARV GP I-ELISA was applied in monitoring and characterising the antibody responses of ERBs to MARV. Maternal antibodies to MARV were detected in juvenile bats up to approximately five months after birth. In bats experimentally infected with MARV, antibodies against the virus remained detectable in the majority of bats at 110 days post-infection. Furthermore, antibodies to MARV remained detectable in 84% of naturally exposed bats at least 11 months after capture, suggesting that bats develop long-term humoral immunity in response to active infection with MARV. To test whether pre-existing immunity in bats is protective against reinfection, 15 ERBs with differing levels of MARV-specific IgG antibodies were inoculated with the Watsa isolate of the virus. Levels of anti-MARV IgG antibodies increased swiftly from day 5 post inoculation. Viraemia was detected in 73% of reinfected bats, and the challenge virus was isolated from the serum of one reinfected bat. Viral ribonucleic acid was detected in the spleen (73% of bats), liver (47%) and lung (7%) at different days post inoculation. These results suggest that primary infection of ERBs with MARV does not induce sterilising humoral immunity; however, re-inoculation of previously infected bats produced only localised infection, with an absence of the virus in tissues potentially involved in viral transmission. Reinfection of previously infected bats is therefore not likely to be a key factor driving MARV maintenance in nature. The establishment of in-house capacity for the production of recombinant I-ELISA antigens as described in this thesis will assist in the biosurveillance programme in South Africa aimed at monitoring the presence and distribution of MARV infection in local bat populations. The assays based on these antigens will also assist in monitoring the immune status of reservoir host populations, predicting potential spillover events, implementing risk reduction strategies and improving virus-host modelling studies. These tools will further contribute to the characterisation of the antibody responses of ERBs to MARV, which may ultimately assist in elucidating the mechanisms by which bats are able to combat clinical MARV disease.