Drug resistance in Trypanosoma brucei : comparative genomics of melarsoprol-pentamidine cross-resistance and the role of aquaglyceroporin 2 in clinical resistance

Drug resistance in African trypanosomes had already been studied more than 100 years ago, by the pioneering work of Paul Ehrlich. The molecular mechanisms and most genes responsible for drug resistance, however, have not been discovered until recently. New technologies that allow genome-wide comparison were highly successful in identifying many new genes that were linked to drug resistance to all clinical trypanocides.
The overall aim of this thesis was to identify and validate candidate genes for drug resistance in Trypanosoma brucei. In Chapter 2, I applied next generation sequencing to find the mutations causing drug resistance in two bloodstream-form T. b. rhodesiense lines that had previously been selected in vitro for resistance against the clinical drugs melarsoprol and pentamidine, respectively. Both cell lines exhibited strong cross-resistance to either drug - a phenomenon first observed over 60 years ago and repeatedly many times - and nowadays the genes involved have been characterized. Comparative genomics revealed the deletion of the known melarsoprol-pentamdine cross-resistance (MPXR) determinants adenosine transporter 1 (TbAT1) in the melarsoprol-selected line and aquaglyceroporin 2 (AQP2) in both selected lines. The pentamidine-selected line had acquired a heterozygous point mutation (G430R) in TbAT1 that rendered the transporter non-functional. The gene TbAT1, encoding the adenosine/adenine permease P2 transporter, has been discovered more than 10 years ago. AQP2 has recently been discovered to play a role in MPXR in a genome-wide RNAi screen. Both transporters mediate the uptake of melarsoprol and pentamidine and, when functionally lost, lead to cross-resistance. AQP2 emerged as the main genetic determinant of MPXR and corresponds to the high-affinity pentamidine transporter. Mutations in AQP2 were found in all analyzed MPXR cell lines selected, either in vitro or in vivo, with arsenicals or pentamidine and from all three T. brucei ssp. (Chapter 3). An additional mutation became fixed in both resistant cell lines; the RNA-binding protein TbUBP1 carried the exact same coding point mutation (R131L). Overexpression of TbUBP1 in T. b. brucei led to a strong growth deficit whereas overexpression of the mutant did not, but intriguingly, those cells became about 2-fold hypersensitive to pentamidine. The physiological function of TbUBP1 and how it affects pentamidine sensitivity remains to be further investigated.
TbAT1 and AQP2 are well studied in laboratory cell lines, but knowledge from clinical isolates is scarce. Chapters 4 and 5 investigate drug resistance in clinical isolates. 16 T. brucei ssp. field isolates, 8 stemming from melarsoprol treatment-refractory cases, that had been adapted to axenic in vitro cultivation have been genotyped for TbAT1 and AQP2 to test whether they carry mutations in either transporter and the drug sensitivities have been determined for melarsoprol, pentamidine and diminazene. Indeed, five T. b. gambiense isolates from the Democratic Republic of Congo and one isolate form South Sudan carried a deletion in the AQP2 / AQP3 locus leading to the formation of a chimeric gene between AQP2 and AQP3 and loss of wild-type AQP2.
The identified mutant T. b. gambiense isolates were 3- to 5-times less sensitive to melarsoprol and 40- to 50-fold less sensitive to pentamidine compared to reference isolates. Functional expression of the chimeric AQP in a tbaqp2 null background did not restore the drug sensitivity but the introduction of the wild-type AQP2 in one of the resistant T. b. gambiense isolates rendered the cells sensitive to melarsoprol and pentamidine, comparable to fully drug susceptible isolates. This proves that the loss of 'wild-type' AQP2 is the cause of melarsoprol-pentamidine cross-resistance in the T. b. gambiense isolate. Thus AQP2 may serve as a molecular marker for drug resistance in the field.