Evidence of the absence of Human African Trypanosomiasis in northern Uganda: analyses of cattle, pigs and tsetse flies for the presence of Trypanosoma brucei gambiense

Background Large-scale control of sleeping sickness has led to a decline in the number of cases of Gambian human African trypanosomiasis (g-HAT) to <2000/year. However, achieving complete and lasting interruption of transmission may be difficult because animals may act as reservoir hosts for T. b. gambiense. Our study aims to update our understanding of T. b. gambiense in local vectors and domestic animals of N.W. Uganda. Methods We collected blood from 2896 cattle and 400 pigs and In addition, 6664 tsetse underwent microscopical examination for the presence of trypanosomes. Trypanosoma species were identified in tsetse from a subsample of 2184 using PCR. Primers specific for T. brucei s.l. and for T. brucei sub-species were used to screen cattle, pig and tsetse samples. Results In total, 39/2,088 (1.9%; 95% CI=1.9-2.5) cattle, 25/400 (6.3%; 95% CI=4.1-9.1) pigs and 40/2,184 (1.8%; 95% CI=1.3-2.5) tsetse, were positive for T. brucei s.l.. Of these samples 24 cattle (61.5%), 15 pig (60%) and 25 tsetse (62.5%) samples had sufficient DNA to be screened using the T. brucei sub-species PCR. Further analysis found no cattle or pigs positive for T. b. gambiense, however, 17/40 of the tsetse samples produced a band suggestive of T. b. gambiense. When three of these 17 PCR products were sequenced the sequences were markedly different to T. b. gambiense, indicating that these flies were not infected with T. b. gambiense. Conclusion The absence of T. b. gambiense in cattle, pigs and tsetse accords with the low prevalence of g-HAT in the human population. We found no evidence that livestock are acting as reservoir hosts. However, this study highlights the limitations of current methods of detecting and identifying T. b. gambiense which relies on a single copy-gene to discriminate between the different sub-species of T. brucei s.l. Author Summary The decline of annual cases of West-African sleeping sickness in Uganda raises the prospect that elimination of the disease is achievable for the country. However, with the decrease in incidence and the likely subsequent change in priorities there is a need to confirm that the disease is truly eliminated. One unanswered question is the role that domestic animals play in maintaining transmission of the disease. The potential of cryptic-animal reservoirs is a serious threat to successful and sustained elimination of the disease. It is with the intent of resolving this question that we have carried out this study whereby we examined 2088 cattle, 400 pigs and 2184 tsetse for Trypanosoma brucei gambiense, the parasite responsible for the disease. Our study found T. brucei s.l. in local cattle, pigs and tsetse flies, with their respective prevalences as follows, 1.9%, 6.3% and 1.8%. Further analysis to establish identity of these positives to the sub-species level found that no cattle, pigs or tsetse were carrying the pathogen responsible for Gambian sleeping sickness. Our work highlights the difficulty of establishing the absence of a disease, especially in an extremely low endemic setting, and the limitations of some of the most commonly used methods.

The absence of T. b. gambiense in cattle, pigs and tsetse accords with the low prevalence of g-HAT in the 30 human population. We found no evidence that livestock are acting as reservoir hosts. However, this study 31 highlights the limitations of current methods of detecting and identifying T. b. gambiense which relies on a 32 single copy-gene to discriminate between the different sub-species of T. brucei s.l. 33 Author Summary 34 The decline of annual cases of West-African sleeping sickness in Uganda raises the prospect that 35 elimination of the disease is achievable for the country. However, with the decrease in incidence and the 36 likely subsequent change in priorities there is a need to confirm that the disease is truly eliminated. One 37 unanswered question is the role that domestic animals play in maintaining transmission of the disease. The 38 potential of cryptic-animal reservoirs is a serious threat to successful and sustained elimination of the 39 disease. It is with the intent of resolving this question that we have carried out this study whereby we 40 examined 2088 cattle, 400 pigs and 2184 tsetse for Trypanosoma brucei gambiense, the parasite 41 responsible for the disease. Our study found T. brucei s.l. in local cattle, pigs and tsetse flies, with their 42 respective prevalences as follows, 1.9%, 6.3% and 1.8%. Further analysis to establish identity of these 43 positives to the sub-species level found that no cattle, pigs or tsetse were carrying the pathogen 44 responsible for Gambian sleeping sickness. Our work highlights the difficulty of establishing the absence of Introduction 48 The term "human African trypanosomiasis" (HAT) is used to describe two diseases that are clinically, geographically 49 and parasitological distinct. Both the cattle and pigs were sampled in the following manner, the animal was restrained, and a disposable lancet 152 was used to puncture a pineal (ear) vein. Blood was collected with three 50mm heparinised capillary tube which 153 collected 35µl of blood. Two tubes were centrifuged at 8,000 rpm for three minutes and the buffy coat layer 154 examined as a wet preparation at x400 magnification using a compound-microscope with a dark-field filter. The 155 contents of the third capillary tube was transferred to a Whatman FTA card (GE Health Care, Little Chalfont) and left 156 to air dry before it was heat sealed in a foil pouch with a packet of silica gel to ensure the sample remained 157 desiccated.

158
To extract the DNA from the FTA card, a modified version of the method described by Ahmed was carried out as 159 follows, 10 2mm hole-punches were taken from each bloodspot, using a Harris micro-punch. The punches were 160 washed three times in 1ml of distilled water and then 135µl of a 1% Proteinase K/10% Chelex TE suspension was 161 added to each batch of 10-hole punches. These were then incubated at 56ᵒC for an hour followed by 93ᵒC for 30 162 minutes. In total 14 sampling sites (   However as there is a difference in the copy number being targeted by the different primers not all those samples 213 initially identified as T. brucei s.l. will be identified down to a sub-species level due to insufficient DNA. Following the 214 detection of T. brucei s.l. positive samples using either the multiplex ITS primers or the FIND TBR-PCR primers (30) (Table 3).  In total 38/2,877 (1.3%; 95% CI=0.9-1.8) cattle and 25/400 (6.3%; 95% CI=4.1-9.1) pigs examined using the

300
Of the 17 bands, a subsample of three were sent for sequencing to determine the specific product size and 301 sequence. The samples were sent to SourceBioscience using both forward and reverse primers. The results 302 of the sequencing showed that the three bands sent were identical and that the product was 281 bp 303 (inclusive of primers) in length. The sequence when aligned against reference sequences for T. b. 304 gambiense using the NCBI database resulted in only a 90% identity and a query cover of 16% Fig 5. product. There was also significant variation in the 281bp sized sequences compared to T. b. gambiense sequence.

322
Based on the sequencing results these positive samples cannot be unequivocally identified as T. b. gambiense.

323
Three conclusions arise from the tsetse survey, the first is that despite screening 2,184 tsetse, no tsetse were found 324 to be positive for T. brucei gambiense, this could be due to T. b. gambiense no longer being transmitted in the area or that our sample size was too small to detect T. b. gambiense. Despite being understood as the sole vector of gHAT 326 (39) the prevalence of the disease amongst wild tsetse population is often extremely low (1,40,41)  The diagnostic methods used in this paper involved both microscopy and PCR, of which only PCR has the potential to 358 discriminate sub-species of T. b. gambiense (17,34). There are few diagnostic methods that are capable of 359 accurately distinguishing between the T. brucei sub-species (16,17,34). The molecular methods available for the 360 detection of T. b. gambiense are limited due to the practical aspect of conducting these assays in the field and the 361 limited diagnostic markers available. As mentioned previously the sensitivity of T. b. gambiense specific PCR is 362 limited to detecting a single copy gene. Some molecular assays attempt to overcome this problem by relying on the 363 human serums ability to lyse all salivarian trypanosomes (except for T. b. gambiense and T. b. rhodesiense) therefore 364 any T. brucei s.l. identified in a human sample would be one of the two HAT species (45). This allows for the targeting 365 of a higher copy region specific to the T. brucei species group. Using this method any positives would have to be one 366 of the two HAT species, however as the treatment of the two diseases differs and the only option to try and identify 367 if it was an East or West African sleeping sickness infection would be to try and determine the geographical location 368 of where the individual was infected. This approach would also only be limited to humans and could not be used in 369 either xenodiagnoses or screening animals, as all three T. brucei sub-species could be present in the vector or animal 370 populations. To put these differences of sensitivity in perspective, we can look at the limit of detection (LoD) of the 371 number of trypanosome per mL, between multiple diagnostic methods ( The lack of a highly specific, sensitive and field-friendly assay that is capable of screening for T. b. gambiense in both 375 the human, vector and local animal populations is sorely needed if the hope of eliminating sleeping sickness by 2020 376 is to be achieved.  (47). This study also highlights the lack of highly sensitive diagnostics that can 382 discriminate between the different sub-species of T. brucei s.l.. Despite not finding T. b. gambiense in the tsetse 383 population of Koboko vector control has been calculated to being essential to reach the elimination goal of 2030 (48) 384