Authors

Discussion

Tsetse flies (Glossina sp.) are vectors of subspecies of Trypanosoma brucei which cause human African trypanosomiasis (HAT). Catching tsetse and screening them for the presence of T. brucei is an important method of HAT surveillance. Classically, individual tsetse were dissected and subjected to microscopic analysis to identify trypanosomes if present. However in the ‘genomics age’ such techniques are being replaced by molecular xenomonitoring – screening vectors for genetic targets indicative of pathogen presence. Assays using a range of T. brucei genomic targets have been developed for this purpose, including the 10,000-copy T. brucei tandem repeat (TBR) region. The use of such highly sensitive targets in an end-point assay such as TBR-PCR can lead to difficulties in differentiating true biological infection from the mere presence of a DNA marker. These sensitive methods are also vulnerable to DNA cross-contamination. Such contamination may occur at capture, when live tsetse are held in a small trap cage for a day or more. Several xenomonitoring studies that have used TBR-PCR have reported higher-than-expected proportions (5% or more) of T. brucei-positive flies. Therefore, this study set out to investigate whether it is possible for T. brucei infected tsetse to contaminate uninfected tsetse with T. brucei DNA when housed in the same trap.

A total of 140 teneral G. morsitans morsitans were fed a bloodmeal spiked with T. brucei brucei TSW 196 and maintained in solitary cells. TBR-qPCR screening of tsetse faecal samples collected 9-14 days post inoculation (dpi) were used to determine individual fly infection status. At 19dpi, 48 infected flies (IFs) and 96 (not inoculated) uninfected flies (UFs) were placed in plastic bottles similar to the cages used with traps. The numbers of IFs and UFs in the bottles was varied according to four classes of treatment. The four treatments comprised IF:UF in the ratios: (T1) 9:3, (T2) 6:6 (T3) 1:11 and (T4) 0:12. Each treatment was replicated three times (A, B and C). After 24-hour incubation, all flies that had ingested an infectious bloodmeal (n=110) were dissected and microscopically analysed to determine infection status. All tsetse samples (n=206) then underwent DNA extraction and screening using TBR-PCR and TBR-qPCR.

Microscopy revealed that 100% (n=48) of IFs selected for experiments had developed mature midgut infection. Furthermore, screening of these IFs by TBR-qPCR revealed Cq values between 14.46-21.57 (mean=17.74, SD=0.7458) in 100% of flies. However UFs also contained TBR target DNA, the quantity of which varied according to the proportion of infected flies within the trap. For T1 and T2, 100% of UFs had detectable DNA with mean Cqs of 26.72 (±0.7498 SD) and 29.84 (±1.375 SD) respectively. For T3, 91% UFs had detectable DNA with a mean Cq of 33.70 (±1.352 SD).Low-level amplification was detected in 50% of UFs in T4-A and T4-B (placed within close proximity to traps containing infected flies) of mean Cq 35.23 ±1.396 SD. UFs in T4-C (placed in a separate room) recorded no amplification. Our results show that infected tsetse can contaminate uninfected flies with T. brucei DNA within a trap cage, and that the level of this contamination can be extensive. Whilst simple PCR may overestimate infection rates, quantitative PCR offers a means of identifying ‘true’ infections. We have also described a novel tsetse faecal screening method to determine T. brucei midgut infection status ante-mortem. This may facilitate future studies involving large-scale experimental infection of tsetse flies.