Case-Control Approach to Identify Plasmodium falciparum Polymorphisms Associated with Severe Malaria

Watcharee Chokejindachai,David J Conway

doi:10.1371/journal.pone.0005454

Abstract

BackgroundStudies to identify phenotypically-associated polymorphisms in the Plasmodium falciparum 23 Mb genome will require a dense array of marker loci. It was considered promising to undertake initial allelic association studies to prospect for virulence polymorphisms in Thailand, as the low endemicity would allow higher levels of linkage disequilibrium (LD) than would exist in more highly endemic areas.Methodology/Principal FindingsAssessment of LD was first made with 11 microsatellite loci widely dispersed in the parasite genome, and 16 microsatellite loci covering a ∼140 kb region of chromosome 2 (an arbitrarily representative non-telomeric part of the genome), in a sample of 100 P. falciparum isolates. The dispersed loci showed minimal LD (Index of Association, ISA = 0.013, P = 0.10), while those on chromosome 2 showed significant LD values mostly between loci <5 kb apart. A disease association study was then performed comparing parasites in 113 severe malaria cases and 245 mild malaria controls. Genotyping was performed on almost all polymorphisms in the binding domains of three erythrocyte binding antigens (eba175, eba140 and eba181), and repeat sequence polymorphisms ∼2 kb apart in each of three reticulocyte binding homologues (Rh1, Rh2a/b, and Rh4). Differences between cases and controls were seen for (i) codons 388-90 in eba175, and (ii) a repeat sequence centred on Rh1 codon 667.Conclusions/SignificanceAllelic association studies on P. falciparum require dense genotypic markers, even in a population of only moderate endemicity that has more extensive LD than highly endemic populations. Disease-associated polymorphisms in the eba175 and Rh1 genes encode differences in the middle of previously characterised erythrocyte binding domains, marking these for further investigation.

Highlights

A wide and unexplained spectrum of clinical disease caused by the malaria parasite Plasmodium falciparum is responsible for approximately one million deaths each year, mostly in African children in highly endemic populations [1] and in adults in areas of lower endemicity [2]
Fifty of these isolates had single alleles at each of the loci indicating unmixed haploid genotype infections, and these were suitable for analysis of multilocus linkage disequilibrium, as they exclude possibility of scoring a false haplotype from the presence of mixed genotypes
An earlier survey of unlinked microsatellites in parasites sampled over a single month in a small area on the Thai-Burma border (Shoklo) showed a higher index of association than seen here, but this difference disappeared when only unique haplotypes were analysed in that study, indicating a locally ‘epidemic’ population structure [21]

Summary

Introduction

A wide and unexplained spectrum of clinical disease caused by the malaria parasite Plasmodium falciparum is responsible for approximately one million deaths each year, mostly in African children in highly endemic populations [1] and in adults in areas of lower endemicity [2]. Given the imminent potential of large scale genotyping, there is a need to address whether a genome wide allelic association approach is feasible to identify loci in P. falciparum which affect the severity of malaria [20]. Studies to identify phenotypically-associated polymorphisms in the Plasmodium falciparum 23 Mb genome will require a dense array of marker loci. It was considered promising to undertake initial allelic association studies to prospect for virulence polymorphisms in Thailand, as the low endemicity would allow higher levels of linkage disequilibrium (LD) than would exist in more highly endemic areas

Methods

Results

Conclusion