Indel-informed bayesian analysis suggests cryptic divisions between Plasmodium knowlesi of humans and long-tailed macaques (Macaca fascicularis) in Malaysian Borneo.

Plasmodium knowlesi is a major causative agent of malaria in humans of Southeast Asia. Macaques are natural hosts for this parasite, which is considered to be zoonotic, but little is conclusively known about its patterns of transmission within and between these hosts. Here, we apply a novel phylogenetic approach to test for patterns of cryptic population genetic structure between P. knowlesi isolated from humans and long-tailed macaques from the state of Sarawak in Malaysian Borneo. Our approach differs from previous investigations through our exhaustive use of archival 18S Small Subunit rRNA (18S) gene sequences from Plasmodium and Hepatocystis species, our inclusion of insertion and deletion information during phylogenetic inference, and our application of Bayesian phylogenetic inference to this problem. We report distinct clades of P. knowlesi that predominantly contained sequences from either human or macaque hosts species for paralogous A-type and S-type 18S gene loci. We report significant partitioning of sequence distances between host species across both types of loci, and confirmed that sequences of the same locus type showed significantly biased assortment into different clades depending on their host species. Our results support the zoonotic potential of Plasmodium knowlesi, but also suggest that humans may be preferentially infected with certain strains of this parasite. Broadly, such patterns could arise through preferential zoonotic transmission of some parasite lineages or a disposition of parasites to transmit within, rather than between, human and macaque hosts. Available data are insufficient to address these hypotheses, although human-to-human transmission would provide a parsimonious explanation for the phylogenetic patterns that we observe. Regardless, P. knowlesi is often assumed to be a strict zoonosis. Our results suggest this may not be a safe assumption, and highlight the need for renewed and more vigorous explorations of transmission patterns in the fifth human malarial parasite.