Population genetics

Malaria variorum

A genetic study of the malaria parasite finds that this species is unexpectedly diverse. A second study shows multiple independent origins of mutations in one parasite gene that confer resistance to a widely used drug.

Understanding genetic variation in the malarial parasite, Plasmodium falciparum, is of major importance to public health, especially as we contemplate widespread programmes of vaccination. If a vaccine controls only part of the P. falciparum population, then it might alter the genetic composition of the surviving parasites. The frightening prospect is that such an imperfect vaccine might select for variants of P. falciparum that are more virulent, potentially undoing any progress towards mitigating this scourge1. A thorough knowledge of the genetic diversity of the parasite population is the first defence against this possible disaster — hence the importance of papers by Mu et al.2 and Wootton et al.3 on pages 323 and 320 of this issue.

There is considerable disagreement in the scientific literature regarding the level of variability of existing P. falciparum isolates and, by inference, their depth of common ancestry (that is, the time at which they started to diverge). All agree that there is abundant variation in microsatellites — simple repetitive DNA sequences. And it cannot be disputed that different P. falciparum isolates vary in their ability to resist drugs and in molecules that trigger immune responses in humans. But studies of microsatellites and protein differences might not be optimal when making inferences about the long-term population size and time of common ancestry of different isolates. 'Silent' DNA changes, which do not alter encoded amino-acid sequences, provide a picture that is more clock-like and less distorted by natural selection. Some have argued that there is an exceptionally low level of silent variation in P. falciparum isolates, and a correspondingly recent time of common ancestry4,5, around 3,000–10,000 years ago. Others — including Mu et al. — have suggested a date of 50,000 years ago or more2,6.

Why do these estimates vary so wildly? Part of the answer is that the studies differed in several significant ways, including the criteria by which the isolates were identified, which genes were examined, and how the data were analysed. Rich et al.4 randomly sampled the protein-coding regions of 9 genes in 35 P. falciparum isolates, and found differences (none of which were in fact silent) at 51 positions in the sequences. The authors carefully calibrated the rate of evolution of the sequences, comparing them with data from rodent and chimpanzee Plasmodium parasites, yielding an expected time of common ancestry of 6,000–12,000 years ago. Volkman et al.5, meanwhile, sequenced 25 introns (non-protein-coding sequences in genes) from 8 P. falciparum isolates. They found 8 varying bases; 5 were either in or near microsatellites and so were excluded from further analysis, which gave a date of common ancestry of 3,200–7,700 years ago. These studies are consistent with the idea that malaria became a widespread human pathogen only after the spread of agricultural practices that expanded the range of the malaria-transmitting mosquito7. Another implication is that the estimated low level of variation should make control easier.

Hughes and Verra6, however, calculated a much higher degree of variation, consistent with a parasite population size of the order of 100,000 for the past 300,000–400,000 years. Extracting a larger data set from public databases, these authors had identified 23 genes that satisfied their criteria for 'neutral' evolution — there was no strong indication that natural selection had distorted the pattern of variation. They discovered 159 variable positions, including 11 silent changes, within these sequences. One weakness, not unique to this study, was in accepting the public databases as error-free, so any sequencing mistakes would have been incorporated into their model as variation. But even if error rates were at the high end of credibility, the level of variation in these data still exceeds that of the other studies4,5.

This sets the stage for the approach of Mu et al.2, who re-sequenced portions of 204 genes in the largest survey so far of variation on chromosome 3 of P. falciparum. The authors obtained more than 218,000 base pairs of coding and non-coding sequence from 5 widely divergent isolates, and found 238 single changes (62 of which were silent and 31 of which were in non-coding sequence) and 165 microsatellite variants. From these data Mu et al. estimate a relatively high level of variation at silent sites, consistent with a time of common ancestry of 150,000 years ago.

All these details are important because there have been several competing explanations for the many different estimates of P. falciparum ancestry. One proposed explanation is that P. falciparum is a recently evolved species, with mutation rates that vary wildly along its chromosomes, giving different pictures depending on the genes studied. Another is that P. falciparum is ancient but in the recent past suffered a massive restriction ('bottleneck') in the number of surviving lineages and consequently lost most variation — leading to the relatively young date of common ancestry calculated in studies that, by chance, failed to sample any divergent but rare alleles. A third is that natural selection favoured novel mutations, such as those for drug resistance, and that as these mutations increased in frequency, they dragged linked genes (in a process known as 'hitchhiking'; Fig. 1) with them. This too would create an impression of little variation in some genes. Although it might not settle the question, Mu and colleagues' study2 of DNA sequence variation is far larger than its predecessors, and provides strong evidence that P. falciparum is diverse enough to raise a greater challenge for public-health measures.

Figure 1: Genetic hitchhiking.

a, A single chromosome with 5 genes (ae), from 12 isolates, is shown. Each gene has considerable variation (it comes in different alleles), depicted by different colours. When a favourable mutation occurs (asterisk), it increases in frequency (b), taking not only the blue version of gene c to high frequency, but also the yellow form of gene b and the green form of gene d. These three genes end up losing most of their allelic variation after such 'hitchhiking'. Genes farther away from the favoured gene c have undergone exchanges with other chromosomes (recombination), so the loss of variation decreases with greater distance from the selected gene.

In a separate analysis3 the same group provide a beautiful example of genetic hitchhiking. Wootton et al.3 were looking at variation in the ability of P. falciparum to resist the widely used antimalarial drug chloroquine. They studied microsatellites in 87 worldwide isolates of the parasite, and found that a narrow region of chromosome 7 — including the pfcrt chloroquine-resistance gene and flanking sequences — shows little variability among chloroquine-resistant isolates but is more diverse among sensitive isolates. Further analysis supported the idea that there are at least four independent forms of pfcrt, which are rapidly becoming more common in different parts of the world. The inference that selective pressures (in this case, the use of chloroquine) cause such localized increases in gene frequency is receiving attention elsewhere8, but a technical advantage of Plasmodium is that it has just one copy, not two, of each chromosome. The results show that complex combinations of mutations arise readily. So, in principle, P. falciparum might rapidly develop resistance to multiple drugs. The genes also seem to have moved across continents with frightening speed, implying that there would be little time to contain the spread of new resistance genes.

How can this information help us to control malaria? First, although it is tempting to think that we now know the full extent of P. falciparum diversity, further surveys are still needed. A present-day snapshot of variation in many genes among more isolates, spanning broad geographical areas, will be important in determining the genetic impact of a vaccination or expanded drug programme. Had thorough genetic monitoring been in place, we might have detected the genetic variants causing the spread of chloroquine resistance early enough to take steps to prolong the usefulness of the drug. Second, we need a systematic collection of Plasmodium isolates to be shared among laboratories, common resources for identifying and confirming sequence variations on a large scale, and a common database of information. We should then rapidly reach consensus on the genetic variability and ancestral demographic history of this devastating parasite.


  1. 1

    Gandon, S., Mackinnon, M. J., Nee, S. & Read, A. F. Nature 414, 751–756 (2001).

    CAS  Article  ADS  Google Scholar 

  2. 2

    Mu, J. et al. Nature 418, 323–326 (2002).

    CAS  Article  ADS  Google Scholar 

  3. 3

    Wootton, J. C. et al. Nature 418, 320–323 (2002).

    CAS  Article  ADS  Google Scholar 

  4. 4

    Rich, S. M., Licht, M. C., Hudson, R. R. & Ayala, F. J. Proc. Natl Acad. Sci. USA 95, 4425–4430 (1998).

    CAS  Article  ADS  Google Scholar 

  5. 5

    Volkman, S. K. et al. Science 293, 482–484 (2001).

    CAS  Article  Google Scholar 

  6. 6

    Hughes, A. L. & Verra, F. Proc. R. Soc. Lond. B 268, 1855–1860 (2001).

    CAS  Article  Google Scholar 

  7. 7

    Coluzzi, M. Parassitologia 41, 277–283 (1999).

    CAS  PubMed  Google Scholar 

  8. 8

    Kim, Y. & Stephan, W. Genetics 160, 765–777 (2002).

    CAS  PubMed  PubMed Central  Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Andrew G. Clark.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Clark, A. Malaria variorum. Nature 418, 283–284 (2002). https://doi.org/10.1038/418283a

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.