During the last decade miltefosine (MIL) has been used as first-line treatment for visceral leishmaniasis in endemic areas with antimonial resistance, but a decline in clinical effectiveness is now being reported. While only two MIL-resistant Leishmania infantum strains from HIV co-infected patients have been documented, phenotypic MIL-resistance for L. donovani has not yet been identified in the laboratory. Hence, a better understanding of the factors contributing to increased MIL-treatment failure is necessary. Given the paucity of defined MIL-resistant L. donovani clinical isolates, this study used an experimental amastigote-selected MIL-resistant L. infantum isolate (LEM3323). In-depth exploration of the MIL-resistant phenotype was performed by coupling genomic with phenotypic data to gain insight into gene function and the mutant phenotype. A naturally MIL-resistant L. infantum clinical isolate (LEM5159) was included to compare both datasets. Phenotypically, resistance was evaluated by determining intracellular amastigote susceptibility in vitro and actual MIL-uptake. Genomic analysis provided supportive evidence that the resistance selection model on intracellular amastigotes can be a good proxy for the in vivo field situation since both resistant strains showed mutations in the same inward transporter system responsible for the acquired MIL-resistant phenotype. In line with previous literature findings in promastigotes, our data confirm a defective import machinery through inactivation of the LiMT/LiRos3 protein complex as the main mechanism for MIL-resistance also in intracellular amastigotes. Whole genome sequencing analysis of LEM3323 revealed a 2 base pair deletion in the LiMT gene that led to the formation an early stop codon and a truncation of the LiMT protein. Interestingly, LEM5159 revealed mutations in both the LiMT and LiRos3 genes, resulting in an aberrant expression of the LiMT protein. To verify that these mutations were indeed accountable for the acquired resistance, transfection experiments were performed to re-establish MIL-susceptibility. In LEM3323, susceptibility was restored upon expression of a LiMT wild-type gene, whereas the MIL-susceptibility of LEM5159 could be reversed after expression of the LiRos3 wild-type gene. The aberrant expression profile of the LiMT protein could be restored upon rescue of the LiRos3 gene both in the LEM5159 clinical isolate and a ΔLiRos3 strain, showing that expression of LdMT is dependent on LdRos3 expression. The present findings clearly corroborate the pivotal role of the LiMT/LiRos3 complex in resistance towards MIL.

Competing Interests: The authors have declared that no competing interests exist.

• Wellcome Trust/United Kingdom