Longitudinal tracking and quantification of individual Plasmodium falciparum clones in complex infections

Abstract

Longitudinal tracking of individual Plasmodium falciparum strains in multi-clonal infections is essential for investigating infection dynamics of malaria. The traditional genotyping techniques did not permit tracking changes in individual clone density during persistent natural infections. Amplicon deep sequencing (Amp-Seq) offers a tool to address this knowledge gap. The sensitivity of Amp-Seq for relative quantification of clones was investigated using three molecular markers, ama1-D2, ama1-D3, and cpmp. Amp-Seq and length-polymorphism based genotyping were compared for their performance in following minority clones in longitudinal samples from Papua New Guinea. Amp-Seq markers were superior to length-polymorphic marker msp2 in detecting minority clones (sensitivity Amp-Seq: 95%, msp2: 85%). Multiplicity of infection (MOI) by Amp-Seq was 2.32 versus 1.73 for msp2. The higher sensitivity had no effect on estimates of force of infection because missed minority clones were detected in preceding or succeeding bleeds. Individual clone densities were tracked longitudinally by Amp-Seq despite MOI > 1, thus providing an additional parameter for investigating malaria infection dynamics. Amp-Seq based genotyping of longitudinal samples improves detection of minority clones and estimates of MOI. Amp-Seq permits tracking of clone density over time to study clone competition or the dynamics of specific, i.e. resistance-associated genotypes.

Figure 1

Frequency of individual SNPs and haplotypes of three markers in 33 baseline samples from PNG. Non-reference allelic frequency (Non-Ref AF) of each SNP (left) and frequency of haplotypes in these baseline samples (right). n, number of observations per haplotype shown for 2 most prevalent haplotypes. Total number of different haplotypes: 30 for cpmp, 15 for ama1-D2 and 22 for ama1-D3. (Frequency of haplotypes for markers msp2-CE given in Supplementary Fig. S2).

Figure 2

Dynamics of multi-clone infections in 4 children. Multi-marker haplotypes could be generated in panels A, B and C. Inference of multi-locus haplotypes was not possible for the child in panel D; here the dynamics of individual clones tracked by marker ama1-D2 are shown. Each colour represents a clone. Individual markers are represented by different shapes: cpmp (diamonds), ama1-D2 (circles) and ama1-D3 (squares). Solid line connecting multi-locus haplotypes represents their median frequency. Grey dotted vertical lines represent sampling dates. Red dashed lines represent day of artemisinin combination therapy. Red dash-dotted line represents end of radical cure (artesunate-primaquine) at baseline.