Malaria due to Plasmodium falciparum affects 300 million people and claims an estimated 1.5 million lives every year. Our present inability to synthesise a fully protective vaccine means that chemotherapy stands as the only effective measure in the control of the disease. However, in many parts of the world the parasite P. falciparum has become resistant to most drugs presently used 1, seriously undermining efforts for controlling malaria.

Chloroquine (CQ) has long been the drug of choice for the treatment of malaria; however, CQ-resistant parasites are now present in most areas where malaria is endemic 2. Chloroquine resistance is especially well established in Thailand, after having been first described in that country in the late 1950's 3. The decline in the efficacy of chloroquine has led to the use of alternative antimalarials, such as antifolates, mefloquine and artemisinin derivatives, but parasite resistance to these drugs is also becoming a real problem 2. In this context, understanding the genetic basis of drug resistance is essential for implementing rational measures to overcome the problem.

Although significant progress has been made in trying to understand how resistance to CQ may occur, many aspects of it remain unclear, and the genetic mechanisms responsible for mefloquine and quinine resistance are largely unknown. Nevertheless, two main genes have been implicated in quinoline resistance; the pfmdr1 (P. falciparum multi-drug resistance1) and the Pfcrt (P. falciparum chloroquine resistance transporter). There is evidence from the analysis of a genetic cross which indicates that point polymorphisms in the pfmdr1 gene may modulate sensitivity to both mefloquine (MF) and artemisinin in P. falciparum4. Furthermore, recent genetic transfection work has suggested that single nucleotide polymorphisms in the pfmdr1 gene encoding changes in aminoacids 1034, 1042 and 1246 can influence parasite responses to mefloquine, quinine and halofantrine as well as to the structurally unrelated drug artemisinin, and modulate sensitivity to chloroquine depending on the genetic background of the parasites strains 5. However, chloroquine resistance was shown to segregate independently of the pfmdr1 gene, following a genetic cross between a CQ-sensitive parasite, P. falciparum HB3, and a CQ-resistant one, Dd2 6, and the absence of a clear association between pfmdr1 and chloroquine responses in natural parasite populations 789101112131415, strongly suggests the involvement of other gene(s). Recently, detailed linkage analysis and fine chromosome mapping of progeny clones of the HB3 × Dd2 cross has allowed the identification of another gene, pfcrt, in which a mutation at aminoacid 76 (pfcrt K76T) is highly correlated with increased CQ tolerance among field parasite isolates of P. falciparum161718192021222324. In addition, a causal relationship between pfcrt 76T and chloroquine resistance has been confirmed by genetic transfection experiments 16.

The study of the correlation between drug resistance in natural parasite populations and genetic polymorphisms may allow the development of molecular tools to help predict responses to drugs and, as mentioned above, the pfcrt and pfmdr1 genes have been identified as putative markers of quinoline resistance. In the present work we have investigated possible associations between four molecular markers in these genes and sensitivity to chloroquine, mefloquine, quinine (QUIN) and amodiaquine (AMQ) of P. falciparum parasites collected in Thailand.

The increasing failure rates of several antimalarial drugs in the majority of malaria-affected areas means that close monitoring of the epidemiology and dynamics of drug resistance are necessary if we are to implement measures to circumvent the problem. The identification and validation of easy, rapid molecular markers of drug resistance would greatly facilitate this process, and would allow us to overcome difficulties in the use of traditional methods for assaying drug sensitivity.

In Thailand, CQ resistance was first reported more than forty years ago 3, and after ten years, resistance to chloroquine had become so widespread that use of the drug against P. falciparum was abandoned. At present, even though the drug is used only against P. vivax, it is perhaps not surprising to find that most P. falciparum are largely unaffected by CQ. Our in vitro observations show a near total prevalence of CQR (96%) in the present study area, confirming what has been widely reported 1. Among all CQR isolates tested in our study, the presence of pfcrt 76T was universal (P = 0.001), indicating complete selection of this polymorphism by the drug, a result that is in agreement with recently published work not only from Thai-originated parasites 22, but also from parasites of different areas of the globe 32.

The correlation between pfmdr1 genotypes and quinoline resistance has often generated conflicting results; although it has been suggested that pfmdr1 86Y can be correlated with increased CQ resistance in parasites which originated from different areas of the world 33343536, other field studies have not corroborated these findings 789101112131415, and the results of a P. falciparum genetic cross indicated that CQR did not depend on the pfmdr1 gene 6. Our present findings do not implicate pfmdr1 86 in CQ resistance in Thailand, since the presence of both N and Y in our samples was largely independent of their CQ response, indicating that chloroquine does not appear to exert selective pressure on this area of the gene. Interestingly though, pfmdr1 1042D and pfmdr1 1246N, previously associated with increased CQ sensitivity following genetic transfection experiments 5, were largely absent in our samples, possibly suggesting a mechanism of chloroquine resistance that may in part depend on the presence of the alternative polymorphic alleles, pfmdr1 1042N and pfmdr1 1246D, respectively.

Mefloquine was introduced in Thailand in the form of Fansimef (mefloquine-sulfadoxine-pyrimethamine) during 1984 with an initial cure rate of 95%, but MF-resistant P. falciparum parasites have arisen and present a real threat to the control of malaria, especially in the Thai/Cambodia and Thai/Myanmar border areas 37. In the present work the prevalence of MF resistance was 62%, indicating a worrying trend. A correlation between pfcrt 76T and mefloquine sensitivity would always be difficult to establish, since the near complete presence of this polymorphism is likely to have been selected by chloroquine pressure, whose mechanisms of action and resistance are probably distinct from those of MF. Earlier work from Thailand and other areas of the world has indicated that an increase in the level of mefloquine sensitivity among field isolates of P. falciparum may be correlated with a mutation in codon 86 of the pfmdr1 gene (N86Y) 383940. In the present work, the occurrence of pfmdr1 86N was significantly associated with MF resistance as 31/32 resistant isolates carried this polymorphism (P < 0.001), strongly suggesting that 86N is an important event in the generation of MF resistance and may be a useful marker to monitor MF resistance in this country.

Interestingly, although the prevalence of AMQ resistance was high (58%), we did not detect a significant correlation between AMQ responses and any of the markers studied, contrary to what could be expected considering that AMQ is chemically very similar to CQ. These observations indicate that the mechanism of action and/or resistance differ between the two drugs, which may raise interesting questions about the design of new CQ-derivative compounds.

Quinine is one of the most effective drugs for the treatment of malaria in Thailand. Although the efficacy of this drug has been reduced, it has not yet become a serious problem, since the combination with tetracycline increases its cure rate. In fact, P. falciparum quinine sensitivity was total in all regions covered by our study, showing that despite its adverse side effects, quinine can still be used as a reliable resource of malaria therapy in Thailand. The reasons for the long-lasting efficacy of quinine may be manyfold, but its explanations lie outside the scope of this article. Most importantly, quinine may prove especially valuable in the treatment of multi-drug resistant Falciparum malaria, which we found to be largely present in our study as 41 out of 52 isolates were found to be resistant to more than one compound and 7 were unaffected by all drugs (CQ, MF and AMQ) except quinine.

Taken together, our results seem to be suggesting that CQ and MF are the major selective forces on the pfcrt and pfmdr1 genes, whereby the presence of pfcrt 76T, and possibly pfmdr1 1042N and pfmdr1 1246D in Thai-originated parasites has been selected by chloroquine pressure. The pfmdr1 86N mutation seems to be important only for mefloquine resistance, and may represent a useful marker for monitoring resistance in this country, although its validation may require in vivo correlates and the analysis of a larger number of samples.

DL carried out a proportion of the parasite phenotyping, performed the molecular analysis of the majority of the samples and contributed for the elaboration of the manuscript. KR performed the molecular typing of a number of the isolates. FN participated in the parasite's phenotyping and genotyping. AS performed parasite collection and did a proportion of the micro-tests. JPG, VR and PC, conceived the study, participated in its design and co-ordination and were involved in phases of the experimental work.

None declared.