Background: Tsetse flies harbor at least three bacterial symbionts: Wigglesworthia glossinidia, Wolbachia pipientis and Sodalis glossinidius. Wigglesworthia and Sodalis reside in the gut in close association with trypanosomes and may influence establishment and development of midgut parasite infections. Wolbachia has been shown to induce reproductive effects in infected tsetse. This study was conducted to determine the prevalence of these endosymbionts in natural populations of G. austeni and G. pallidipes and to assess the degree of concurrent infections with trypanosomes.
Methods: Fly samples analyzed originated from Kenyan coastal forests (trapped in 2009-2011) and South African G. austeni collected in 2008. The age structure was estimated by standard methods. G. austeni (n=298) and G. pallidipes (n= 302) were analyzed for infection with Wolbachia and Sodalis using PCR. Trypanosome infection was determined either by microscopic examination of dissected organs or by PCR amplification.
Results: Overall we observed that G. pallidipes females had a longer lifespan (70 d) than G. austeni (54 d) in natural populations. Wolbachia infections were present in all G. austeni flies analysed, while in contrast, this symbiont was absent from G. pallidipes. The density of Wolbachia infections in the Kenyan G. austeni population was higher than that observed in South African flies. The infection prevalence of Sodalis ranged from 3.7% in G. austeni to about 16% in G. pallidipes. Microscopic examination of midguts revealed an overall trypanosome infection prevalence of 6% (n = 235) and 5% (n = 552), while evaluation with ITS1 primers indicated a prevalence of about 13% (n = 296) and 10% (n = 302) in G. austeni and G. pallidipes, respectively. The majority of infections (46%) were with T. congolense. Co-infection with all three organisms was observed at 1% and 3.3% in G. austeni and G. pallidipes, respectively. Eleven out of the thirteen (85%) co-infected flies harboured T. congolense and T. simiae parasites. While the association between trypanosomes and Sodalis infection was statistically significant in G. pallidipes (P = 0.0127), the number of co-infected flies was too few for a definite conclusion.
Conclusions: The tsetse populations analyzed differed in the prevalence of symbionts, despite being sympatric and therefore exposed to identical environmental factors. The density of infections with Wolbachia also differed between G. austeni populations. There were too few natural co-infections detected with the Sodalis and trypanosomes to suggest extensive inter-relations between these infections in natural populations. We discuss these findings in the context of potential symbiont-mediated control interventions to reduce parasite infections and/or fly populations.