Plasmodium knowlesi: reservoir hosts and tracking the emergence in humans and macaques

Abstract

Plasmodium knowlesi, a malaria parasite originally thought to be restricted to macaques in Southeast Asia, has recently been recognized as a significant cause of human malaria. Unlike the benign and morphologically similar P. malariae, these parasites can lead to fatal infections. Malaria parasites, including P. knowlesi, have not yet been detected in macaques of the Kapit Division of Malaysian Borneo, where the majority of human knowlesi malaria cases have been reported. In order to extend our understanding of the epidemiology and evolutionary history of P. knowlesi, we examined 108 wild macaques for malaria parasites and sequenced the circumsporozoite protein (csp) gene and mitochondrial (mt) DNA of P. knowlesi isolates derived from macaques and humans. We detected five species of Plasmodium (P. knowlesi, P. inui, P. cynomolgi, P. fieldi and P. coatneyi) in the long-tailed and pig-tailed macaques, and an extremely high prevalence of P. inui and P. knowlesi. Macaques had a higher number of P. knowlesi genotypes per infection than humans, and some diverse alleles of the P. knowlesi csp gene and certain mtDNA haplotypes were shared between both hosts. Analyses of DNA sequence data indicate that there are no mtDNA lineages associated exclusively with either host. Furthermore, our analyses of the mtDNA data reveal that P. knowlesi is derived from an ancestral parasite population that existed prior to human settlement in Southeast Asia, and underwent significant population expansion approximately 30,000-40,000 years ago. Our results indicate that human infections with P. knowlesi are not newly emergent in Southeast Asia and that knowlesi malaria is primarily a zoonosis with wild macaques as the reservoir hosts. However, ongoing ecological changes resulting from deforestation, with an associated increase in the human population, could enable this pathogenic species of Plasmodium to switch to humans as the preferred host.

Figure 1. Analyses of P. knowlesi csp gene sequences from infections of macaques and humans.

(A) Histogram showing proportion of human and macaque individuals with different numbers of full length csp alleles detected per infection. (B) Diversity of csp alleles in the P. knowlesi clade of the phylogenetic tree of Plasmodium spp. (Fig. S1), based on the non-repeat region of the gene. These intraspecific relationships clustered by the neighbor-joining method on a Kimura 2-parameter distance matrix represent observed pairwise sequence similarity (phylogeny cannot be determined within the species for a nuclear gene due to recombination). Figures on the branches are bootstrap percentages based on 1,000 replicates and only those above 70% are shown. The horizontal branch lengths indicate nucleotide differences per site compared with the scale bar. Parasite clones in the boxes represent sequences that are completely identical for the whole csp gene (including repeat sequences not analysed by alignment but given separately in Supplementary Table S1).

Figure 2. Diversity and haplotype network of P. knowlesi mtDNA genome.

(A) Histogram showing proportion of human and macaque individuals with different numbers of mtDNA haplotypes detected per infection. (B) Schematic diagram of genealogical network showing relationship among 37 mtDNA haplotypes of P. knowlesi. Numbers in larger circles represent number of haplotypes and unnumbered circles represent a single haplotype. Each line connecting the circles represents a mutational step and black dots represent hypothetical missing intermediates.

Figure 3. Time-calibrated maximum clade credibility phylogeny based on the 6 kb mtDNA of Plasmodium species of human and non-human primates.

Phylogenetic tree scaled to time generated using uncorrelated relaxed clock model and Bayesian skyline coalescent tree prior, with the divergence of Plasmodium spp. of Asian macaques and P. gonderi/Plasmodium sp. (Mandrill) as the calibration point (black circle). TMRCAs and HPDs for P. knowlesi and Plasmodium of Asian macaques are indicated. Numbers on branches are values of posterior probabilities. The accession numbers of sequence data of P. knowlesi were deposited in GenBank under the accession numbers EU880446–EU880499 and accession numbers of the other sequences are provided in the Methods section.

Figure 4. Demographic and evolutionary history of P. knowlesi.

(A) Pairwise mismatch distribution of the P. knowlesi mt genome. The bars represent observed frequency of the pairwise differences among mtDNA sequences and the line represents the expected curve for a population that has undergone a demographic expansion. (B) Bayesian skyline plot showing changes in effective population size (Ne) through time as estimated using uncorrelated log-normal relaxed molecular clock and Bayesian skyline coalescent model (10 coalescent-interval groups) with the substitution rate of 3.13×10⁻⁹ substitutions per site per year. The y-axis representing the effective population size is given on a logarithmic scale and the x-axis represents time in thousands of years ago. The thick solid black line is the median estimate and the blue shaded area represents the 95% highest probability density (HPD) intervals for effective population size.