Dihydrofolate-reductase mutations in Plasmodium knowlesi appear unrelated to selective drug pressure from putative human-to-human transmission in Sabah, Malaysia

• Main Authors: Yeo, Tsin Wen, Auburn, Sarah, Anstey, Nicholas M., Grigg, Matthew J., Barber, Bridget E., Marfurt, Jutta, Imwong, Mallika, William, Timothy, Bird, Elspeth, Piera, Kim A., Aziz, Ammar, Boonyuen, Usa, Drakeley, Christopher J., Cox, Jonathan, White, Nicholas J., Cheng, Qin
• Other Authors: Tsuboi, Takafumi
• Format: Article
• Language: English
• Published: 2018
• Subjects: Malaria; DRNTU::Science::Medicine; Plasmodium Knowlesi
• Online Access: https://hdl.handle.net/10356/87597; http://hdl.handle.net/10220/46773
• Institution: Nanyang Technological University

Background: Malaria caused by zoonotic Plasmodium knowlesi is an emerging threat in Eastern Malaysia. Despite demonstrated vector competency, it is unknown whether human-to-human (H-H) transmission is occurring naturally. We sought evidence of drug selection pressure from the antimalarial sulfadoxine-pyrimethamine (SP) as a potential marker of H-H transmission. Methods: The P. knowlesi dihdyrofolate-reductase (pkdhfr) gene was sequenced from 449 P. knowlesi malaria cases from Sabah (Malaysian Borneo) and genotypes evaluated for association with clinical and epidemiological factors. Homology modelling using the pvdhfr template was used to assess the effect of pkdhfr mutations on the pyrimethamine binding pocket. Results: Fourteen non-synonymous mutations were detected, with the most common being at codon T91P (10.2%) and R34L (10.0%), resulting in 21 different genotypes, including the wild-type, 14 single mutants, and six double mutants. One third of the P. knowlesi infections were with pkdhfr mutants; 145 (32%) patients had single mutants and 14 (3%) had double-mutants. In contrast, among the 47 P. falciparum isolates sequenced, three pfdhfr genotypes were found, with the double mutant 108N+59R being fixed and the triple mutants 108N+59R+51I and 108N+59R+164L occurring with frequencies of 4% and 8%, respectively. Two non-random spatio-temporal clusters were identified with pkdhfr genotypes. There was no association between pkdhfr mutations and hyperparasitaemia or malaria severity, both hypothesized to be indicators of H-H transmission. The orthologous loci associated with resistance in P. falciparum were not mutated in pkdhfr. Subsequent homology modelling of pkdhfr revealed gene loci 13, 53, 120, and 173 as being critical for pyrimethamine binding, however, there were no mutations at these sites among the 449 P. knowlesi isolates. Conclusion: Although moderate diversity was observed in pkdhfr in Sabah, there was no evidence this reflected selective antifolate drug pressure in humans.