TRENDS in Parasitology Vol.17 No.8 August 2001 http://parasites.trends.com 1471-4922/01/$ – see front matter © 2001 Elsevier Science Ltd. All rights reserved. PII: S1471-4922(01)02034-7 351 Research Update Research News Understanding how genetic variation in malaria parasites is structured across populations has benefits for researchers interested in finding loci that affect the ability of the parasite to resist host immunity or drugs. In a recent paper, Tim Anderson and colleagues have provided a broad-scale analysis of Plasmodium falciparum population structure. Their results provide an indication of the feasibility of using population-genetic and genetic-mapping approaches to pinpoint selected loci, in addition to reconciling some previously disparate views of population structure in malaria. Why are malaria parasites so genetically variable? Does this variability tell us anything useful about whether the parasite will continue to evade all control efforts successfully? Can we justify the effort of the past two decades in characterizing the genetic diversity of malaria parasite populations? Population geneticists do so on three grounds. First, it is argued that, if we know the basic population structure of the organism, we can predict the extent to which genes are exchanged among the genomes of parasites in the same population and between geographically separated populations. This information would help us to explain the dynamics of drug resistance 1–3 , how vaccine resistance might develop and to indicate which epidemiological process should be targeted for disease control 4 . Second, we need to know the basic level of genetic diversity in a population, as driven by processes in the absence of selection (‘neutral’ variation), before we can say whether a particular gene is under selection by host immunity or drugs. For example, an observation of a low estimate of diversity for a candidate antigen-coding gene might be interpreted as being caused by purifying selection (any mutations would be removed because they are deleterious) or advantageous mutations (the new beneficial variant would sweep through the population). However, if the population has a generally low level of neutral diversity, such interpretations cannot be made, and the original observation may be attributed to stochastic or deterministic dynamics affecting the genome as a whole. In other words, we need to know the background genetic noise in the neutral genes of the parasite before we can understand the signals of selection. Third, determining the extent of genetic rearrangement via recombination in natural populations is crucial if we are to use genome and map information to locate genes that underlie important Plasmodium phenotypes (e.g. parasite genes associated with virulence, transmission and immune evasion). Unfortunately, until now, almost all studies on population structure and evolutionary history at the nucleotide level have used loci that are either likely to be subject to selection 5 or ones that exhibit little diversity, thus making it difficult to assess population structure 6,7 . Inferences from population genetic parameters estimated using data such as WRIGHT’S F ST , LINKAGE DISEQUILIBRIUM (LD) and EFFECTIVE POPULATION SIZE (N e ) (see Glossary), all of which are assumed to be driven by neutral processes, could be misleading. For example, a highly clonal population structure has been proposed for Plasmodium falciparum on the basis of the rarity of synonymous changes in the genome relative to the observed frequency of nonsynonymous changes at several loci likely to be subject to some form of selection 5–8 . This conclusion flies in the face of observations of regular outcrossing and near panmixia observed in the field 9 , and high recombination as inferred by low levels of LD at the MSP-1 locus (msp1) in African populations 10 . Recently, Anderson et al. 11 provided us with the first large study of genetic diversity based on putatively neutral DNA The question of Plasmodium falciparum population structure Philip Awadalla, David Walliker, Hamza Babiker and Margaret Mackinnon Effective population size (N e ): The proportion of the population that contributes (is effective) to the evolution of genetic diversity. Not all individuals contribute equally to the next generation and, therefore, N e is generally lower than the census population size. Linkage disequilibrium (LD): A measure of association of alleles at pairs of loci. Two variable loci or sites in complete linkage disequilibrium have not experienced recombination between the loci. Linkage disequilibrium can arise through genetic drift, low recombination rates between loci, selection, and population structure. Used as an indicator of the frequency of outcrossing and recombination (i.e. linkage disequilibrium decays as outcrossing or recombination increases). MSP: Merozoite surface protein. ‘Neutral’ loci: A neutral locus is a locus or site that is not subject to selection. Thus, the observed level of diversity at such a locus is attributable only to mutation and stochastic processes (population size and structure). ‘Synonymous’ sites, nucleotide sites where a mutation does not change the amino acid encoded by that codon position, for the most part evolve neutrally (although selection for codon usage might be a significant force affecting synonymous variation in malaria). Rate of migration (m): The rate of gene flow between groups of individuals or populations. Small amounts of migration break down (homogenize) the population structure that develops through time via stochastic or deterministic processes. SSR: Simple-sequence repeat loci. Wright’s F ST : A coefficient that describes how genetic variation is partitioned among populations, or the amount of between-population diversity relative to within-population diversity. The F ST (S for subpopulation and T for total population) coefficient can be estimated from the ratio of the difference between total diversity (H T ) in all populations and average diversity estimated within populations (H S ), divided by H T [i.e. F ST = (H T -H S )/H T ]. When admixture between populations is high, F ST should be small and vice versa (the effects of mutation on this coefficient can also be included). Generally, F ST < 0.05 indicates little subdivision among populations (high migration) and F ST > 0.25 indicates extensive subdivision (low migration) a . An ANOVA approach a to estimate F ST is used by Anderson et al. b (Table 1). a Hartl, D.L. and Clark, A.G., eds (1997) Principles of Population Genetics (3rd edn), Sinauer Associates b Anderson, T.J.C. et al. (2000) Microsatellite markers reveal a spectrum of population structures in the malaria parasite Plasmodium falciparum. Mol. Biol. Evol. 17, 1467–1482 Glossary