Improved semiochemical trapping and population genetics of mosquito species vectoring Rift Valley fever in Kenya

Abstract

The East African region is a major hot bed for old and newly emerging arboviral diseases that are occurring with increasing frequency and magnitude. The lack of effective treatment or preventive vaccinations for most of these infections emphasizes the need for surveillance to monitor circulation, which is critical for informing public health decision for early warning and response. Monitoring mosquito populations and mosquito-borne virus activity are the cornerstones of surveillance programs. As a model, this project focussed on Rift Valley fever (RVF), a mosquito-borne zoonosis, which remains prevalent in most parts of Kenya. Improving mosquito-based arbovirus surveillance by increasing trap captures was identified as a priority for maximizing viral detection probability especially during the inter-epidemic period (IEP) which is characterized by low vector population density and sporadic transmission foci. Initially,mosquitoes’ response to color of artificial lights using light-emitting diodes (LEDs) was exploited for improved sampling of important RVF virus mosquito vectors by comparing efficiency of selected LED CDC light traps (red, green, blue, violet, combination of blue-greenred (BGR)) to sample RVF vectors relative to incandescent light (as control) in a CDC light trap in field trapping experiments in two RVF hotspots (Marigat and Ijara districts) in Kenya.Hotspots are defined as areas with RVF epidemic involving higher than normal occurrence of abortions or perinatal mortality in livestock (cattle, sheep, goats) herds, including disease and deaths in humans. Furthermore, the role of host skin odors in attraction of RVF vectors was investigated; the identification of key compounds involved for formulation into attractants (i.e.baiting system) in conjunction with CO2 to enhance trap captures as a strategy for improved vector surveillance especially during the IEP was established. These involved a series of bioassay-guided field trapping experiments, electrophysiology and chemical analyses.Additionally, as an adjunct to arbovirus surveillance and epidemiology, the population genetics of key RVF vectors (Aedes Neomelaniconion) mcintoshi Huang and Ae. (Aedimorphus)ochraceus (Theobald)) sampled from RVF-endemic / epidemic / virus-free areas of Kenya was conducted by analysing sequence variation in mitochondrial cytochrome oxidase subunit I (COI)gene and nuclear internal transcribed spacer (ITS) genome targets. Both genome targets have been used extensively in studies of molecular evolution and have resolved evolutionary relationships among closely related or cryptic mosquito species complexes. Reference data on public databases such as Genbank are therefore readily available, and the COI barcoding region, is well-represented. Although seasonal preference was observed for some species (Ae. mcintoshi and Ae. ochraceus) to certain coloured lights; generally, higher captures for all species examined were recorded in control traps (incandescent) compared to the other LED traps although this was only significantly different from red and violet. Initial field trapping assays showed that the addition of fur (skin volatile) from sheep, the most susceptible host for RVF virus, to the standard CO2-baited light trap improves captures of key RVF vectors. As an understanding of interspecific host preferences can reveal new semiochemicals that could be exploited to maximise development of better attractants, the attractiveness of different RVF virus hosts (cow,donkey, goat, human) in addition to sheep to RVF vectors was assessed further in field experiments. An analogous pattern was observed with an increase in mosquito captures recorded following the addition of skin odours from each of these animals to CO2 traps relative to control traps containing CO2 alone. Interestingly, a higher proportion of engorged mosquitoes (bloodfed + gravid) were recorded in CO2 traps containing skin odours from these animal hosts relative to control CO2 trap alone. Electrophysiology studies to find out which compounds RVFV vectors responded to, revealed a similarity in response profile to the aldehyde components; heptanal, octanal, nonanal and decanal, that were common to all the hosts evaluated. Following fieldtesting, it was shown that each of these compounds could be exploited as attractants singly and/or blends in a dose-dependent manner. A blend formulated from the optimal attractive dose of each of these compounds synergized with CO2 significantly increased trap captures over that of control traps baited with CO2 alone. The four-component blend attracted multiple mosquito vectors under field conditions suggesting that a trapping system based on this formulation offers the opportunity for its use as a tool for RVF mosquito vector surveillance. There was evidence of divergent lineages for Ae. mcintoshi that display geographic restriction coinciding with the magnitude of occurrence of RVF in Kenya; both gene loci indicated the presence of four genetic lineages with significant differentiation among them across the study areas as evident from phylogenetic, median-joining network analyses and from analysis of molecular variance (AMOVA). In contrast, a single, relatively homogenous population was evident for Ae.Ochraceus. Low mean evolutionary divergence estimates among and within sites and a single lineage was evident in the Neighbor-joining analysis, and in network and TCS parsimony analyses. Interestingly, significant negative neutrality tests of Tajima’s D and Fu’s Fs were evident for the COI locus only, and supported by a unimodal curve for the mismatch distribution