A fragile consensus that the global eradication of malaria is possible, and prospects for eventually achieving this audacious goal, are being threatened by the emergence in southeast Asia of parasite resistance to the drug artemisinin and its derivatives. On page 50 of this issue, Ariey et al.1 report welcome news: they have identified a molecular marker of artemisinin-resistant malaria that can be used to map resistance and guide efforts to eliminate it.

Artemisinin-based combination treatments have contributed to reductions in the global burden of malaria, prompting the Bill & Melinda Gates Foundation and the World Health Organization to issue a call in 2007 for an international push towards malaria eradication2. Artemisinin drugs normally clear malaria parasites from the blood of a patient within two days of starting treatment; now, however, increasing numbers of Plasmodium falciparum infections in western Cambodia, southern Vietnam, eastern Myanmar and western Thailand take up to five days to clear. In some areas, artemisinin-based combination therapies are starting to fail completely, with persistence of both infection and clinical illness after what should be curative treatment.

Efforts to contain artemisinin resistance in southeast Asia and to eliminate malaria would be aided enormously by the identification of a molecular marker for this drug resistance. Such markers are available for resistance to other antimalarial drugs for which the genetic determinants of resistance in the parasite are known. However, neither the mechanism of artemisinin action nor the mechanism(s) of resistance are understood. Examinations of the P. falciparum genome for regions of recent strong evolutionary selection, and targeted and genome-wide association studies, have implicated two adjacent regions on chromosome 13 as potential sites of a resistance-determining gene or genes3,4. Through dogged determination and a remarkable combination of approaches, Ariey and colleagues seem to have won the race to identify if not the gene, at the very least a key gene, responsible for artemisinin resistance.

In what seemed like a long shot, the researchers laboriously grew an artemisinin-sensitive parasite isolated from a Tanzanian individual in culture for five years, exposing it intermittently to artemisinin. The drug was removed when the parasites' growth faltered, and replaced after they bounced back. After 60 cycles of drug pressure, the proportion of parasites surviving a pulse of artemisinin had increased from less than 0.01% to more than 10%. Genome sequencing of this population revealed eight single-nucleotide mutations (single nucleotide polymorphisms, or SNPs) in seven genes that were present in the resistant parasites but not in their siblings grown in a parallel culture without drug exposure.

The suspects were cornered. The culprit was then identified when the authors looked for these candidate mutations in parasite lines from Cambodia that had varying susceptibilities to artemisinin drugs in vitro5. After ruling out candidate genes from the Tanzanian isolate that showed no sequence variation in the Cambodian isolates, and genes containing SNPs that were not associated with in vitro resistance, a single gene remained with resistance-associated SNPs. The gene is located on chromosome 13, within a candidate region identified in a recent genome-wide association study of clinical artemisinin resistance4.

The gene in question encodes a kelch protein called K13. Kelch proteins are involved in a variety of protein–protein interactions, and contain several regions of repeating amino-acid sequences, each forming a 'propeller blade' (Fig. 1). Further evidence for the central role of K13-propeller SNPs in resistance was provided by Ariey and colleagues' ecological survey of malaria parasites from several Cambodian provinces. K13-propeller mutations were rare or absent in samples from provinces with virtually no documented resistance, but were widespread in provinces where resistance has been reported. Moreover, their prevalence in these provinces has increased during the past decade, contemporaneously with increases in the prevalence of resistance.

Figure 1: Propeller mutations.
figure 1

Ariey et al.1 show that mutations in a Plasmodium falciparum gene that encodes the kelch protein K13 are associated with both in vitro and clinical measures of artemisinin-resistant P. falciparum malaria in Cambodia. The mutations (orange spheres) encode amino-acid changes in the 'propeller blades' of this protein, which resembles a child's pinwheel and is thought to be involved in various protein–protein interactions.

Source data

The authors went on to show that K13 SNPs were also strongly correlated with delayed parasite clearance following artemisinin treatment in clinical trials. And a final piece of evidence came from subpopulations of Cambodian parasites that can be segregated into sensitive and resistant groups6. The prevalence of K13 SNPs not only correlated with resistance among these subpopulations, they actually did a better job of explaining resistance than did population groupings: sensitive parasites assigned on the basis of their genomic profile to 'resistant' subpopulations had wild-type K13, and resistant parasites belonging to 'sensitive' subpopulations carried the SNPs.

Definitive proof that K13-propeller mutations confer artemisinin resistance will come from genetic transformation of drug-sensitive parasites into resistant ones by the replacement of their wild-type K13 gene with a mutated gene. It is of course possible — even probable — that other genes contribute to artemisinin resistance, but this study leaves little doubt that K13 is a major determinant of resistance to these drugs in P. falciparum malaria. Further study of gene variants in the chromosomal vicinity of the K13 gene will ascertain whether resistance arose once in western Cambodia and then spread — in which case the genomic regions containing the K13 SNPs will show extended surrounding sequence similarity, indicating common ancestry — or whether it emerged independently in different geographical locales. If resistance has arisen independently in many areas, local containment efforts will be futile and only regional elimination offers any hope of preventing its spread to Africa, where the arrival of drug-resistant Asian parasites has previously led to marked increases in malaria hospitalizations and deaths7.

Validation of this molecular marker of artemisinin resistance outside Cambodia will be easily achieved, and mapping of the marker throughout southeast Asia is already under way, thanks to early sharing of the results of this study with local malaria-control workers and researchers. With at least 17 SNPs residing in the propeller domains of K13, only one of which is found in any one parasite, sequencing of the K13 gene will initially be necessary to map resistance. But if a few specific SNPs emerge as being predictive of resistance in different settings, switching to rapid molecular assays using DNA extracted from dried blood spots will accelerate translation of this research finding into a practical tool for public-health surveillance.