Population genetic analysis of Plasmodium falciparum parasites using a customized Illumina GoldenGate genotyping assay

Abstract

The diversity in the Plasmodium falciparum genome can be used to explore parasite population dynamics, with practical applications to malaria control. The ability to identify the geographic origin and trace the migratory patterns of parasites with clinically important phenotypes such as drug resistance is particularly relevant. With increasing single-nucleotide polymorphism (SNP) discovery from ongoing Plasmodium genome sequencing projects, a demand for high SNP and sample throughput genotyping platforms for large-scale population genetic studies is required. Low parasitaemias and multiple clone infections present a number of challenges to genotyping P. falciparum. We addressed some of these issues using a custom 384-SNP Illumina GoldenGate assay on P. falciparum DNA from laboratory clones (long-term cultured adapted parasite clones), short-term cultured parasite isolates and clinical (non-cultured isolates) samples from East and West Africa, Southeast Asia and Oceania. Eighty percent of the SNPs (n = 306) produced reliable genotype calls on samples containing as little as 2 ng of total genomic DNA and on whole genome amplified DNA. Analysis of artificial mixtures of laboratory clones demonstrated high genotype calling specificity and moderate sensitivity to call minor frequency alleles. Clear resolution of geographically distinct populations was demonstrated using Principal Components Analysis (PCA), and global patterns of population genetic diversity were consistent with previous reports. These results validate the utility of the platform in performing population genetic studies of P. falciparum.

Figure 1. Distribution of theta values (proxy to genotype call) at 319 “reliable” SNPs in preparations from the laboratory clones 3D7 and HB3, and in artificial mixtures of the two clones.

Mixing proportions of the clones (HB3:3D7) are indicated in brackets. Red dots represent the SNPs that should have heterozygous genotypes in the artificial mixes (according to the genotypes from the haploid pure samples), black dots represent the SNPs for which both samples have the same genotype. Theta values approximating 0.5 correspond to having both alleles approximating 50% frequency. Values approximating 0 and 1 correspond to homozygous calls for allele a or b. In the pure clonal 3D7 and HB3 samples, the majority of genotype calls exhibit theta values <0.1 and >0.9.

Figure 2. Principal Components Analysis plots (x axis represents PC1, and y axis PC2) on the 143 clinical (non-cultured) and short-termed cultured samples using the 306 “globally reliable” SNPs.

Green = Mali Orange = Burkina Faso, Blue = Kenya, Black = Cambodia, Grey = Thailand, Red = Papua New Guinea. Clear continental differentiation of samples is observed (a). Within continental boundaries, moderate resolution of Thailand from Cambodia (b), and East Africa from West Africa (c) is observed.