The genetics of drug resistance in malaria parasites*

Bulletin of the World Health Organization, 58 (5): 799-804 (1980) The genetics of drug resistance in malaria parasites* G. H. BEALE 1 The available experimental data on the genetics ofdrug resistance in malaria parasites are reviewed. Seven possible mechanisms for the origin of drug resistance are considered, and it is pointed out that spontaneous gene mutation is probably the most important. Experiments on theproduction ofpyrimethamine-resistant and chloroquine-resistant strains ofrodent Plasmodium species, and on the inheritance ofsuch drug resistance, are reviewed. Relevant biochemical data are also considered in relation to the genetics ofdrug resistance. Studies on competition between drug-sensitive and drug-resistant parasites in mixed populations ofrodent plasmodia are described. The implications of thesefindingsfor drug resistance in P. falciparum are discussed. At the present time it is not possible to carry out genetic experiments directly with human malaria parasites, but the in vitro culture technique for Plasmodium fakciparum (1, 2), which at present is limited to the erythrocytic stages, may eventually be extended to the sexual and pre-erythrocytic stages, thus making genetic experiments by classical Mendelian methods technically feasible. In addition, molecular methods, involving recombinant DNA and gene cloning in plasmids, will probably soon be applied to P.falciparum, thus producing rapid advances in our knowledge of the genetics of this species. At present, however, we are limited to genetic studies on rodent plasmodia (P. berghei and related forms) by the classical methods of hybridization and progeny analysis. The results of this work, which are summarized here, can be used, with some caution, to interpret some characteristics of P.falciparum. A number of different mechanisms can be postu- lated to account for changes in organisms affecting their ability to grow in the presence of drugs, including: * This work was supported by grants from WHO/UNDP/World Bank, the Medical Research Council of Great Britain, and the Wellcome Trust. I Honorary Research Professor, Institute of Animal Genetics, West Mains Road, Edinburgh EH9 3JN, Scotland. (1) non-genetic and probably temporary changes causing physiological adaptations of the cells to the drugs; (2) selection of previously existing drug-resistant cells from mixed sensitive and resistant populations, under the influence of drug pressure; (3) spontaneous mutations in the nuclei of one or more cells and subsequent selection of the drug- resistant mutants; (4) induced gene mutation, as a result of the action of mutagenic drugs; (5) mutation of extranuclear genes, e.g., mitochon- drial genes; (6) changes in gene expression caused by alterations in cytoplasmic or environmental factors; and (7) introduction into cells of resistance-transfer factors or other plasmids. These mechanisms have been shown to operate in various other organisms, especially in bacteria and to some extent in free-living protozoa, e.g., Paramecium (3, 4, 5). Present indications are that, as regards drug resistance in Plasmodium, mechanism 3 (spontaneous gene mutation) is probably the most important, though others are also possible. Mechanism 2 (selec- tion from mixed populations) could operate, since it is known from studies of enzyme variation (6) that populations of parasites in individual vertebrate hosts sometimes comprise mixtures of genetically diverse 4006 -799 G. H. BEALE organisms. Mechanism 1 (physiological adaptation) also operates to some extent, as shown by the occurrence of unstable changes in laboratory selection experiments (7). However, this is unlikely to be important in the long run because of the temporary nature of the changes produced. Mechanism 5 (extra- nuclear mutations) can probably be neglected in plasmodia, in view of the minor metabolic role played by mitochondria in the blood stages of the parasites. Mechanism 7 (resistance-transfer factors) has so far not been demonstrated to occur in Plasmodium. If such a mechanism were to exist, as it does in bacteria, one would expect the dissemination of drug resistance about the world to be even more rapid than that already seen. EXPERIMENTAL METHODS Among the rodent Plasmodium species, drug resist- ance has been studied in P. berghei, P. yoelii, P. chabaudi, and P. vinckei, but genetic experiments have been carried out only with P. yoelii and P. chabaudi. Some preliminary work was also done on pyrimethamine resistance in P. gailinaceum (8, 9), but it has not been followed up and will not be further discussed here. Relevant work with P.falciparum has consisted only of comparisons of drug resistance, in vivo and in vitro, in different strains, without any genetic analysis. The methods used involve the production of lines of parasites with increased drug tolerance by exposure to a drug in one of two ways: (1) a 'single' treatment (usually lasting 4 successive days) with a high dose, which eliminates most of the parasites and allows only the rare resistant ones to persist; or (2) a long-lasting series of treatments, covering many passages, starting at a low concentration of the drug, with periodic increases. After parasites with increased drug tolerance have been obtained, the newly resistant line must be tested for its stability following successive blood and mosquito passages. If the resistance is found to be stable, the genetic basis is determined by making a series of crosses between resistant and sensitive lines. The method of carrying out such crosses has been described in detail elsewhere (6). In brief, mosquitos are fed on mouse blood containing a mixture of the two types of parasite, and then allowed to infect other mice with the sporozoites thus produced. To ensure that the progeny parasites tested result from cross- fertilization, genetic markers (enzyme variants) are included in the parental strains, and recombinants involving the markers are selected in the progeny. Cloning is an essential part of the technique, since the initial strains may contain several distinct types of parasite (possibly even different species), and drug pressure may result in the selection of one at the expense of another. The resistant lines might then belong to a species or strain which was a minority constituent of the original strain. Since the resistant parasites might differ from the sensitives in a number of characters unrelated to resistance, comparisons between the two groups could give very misleading results. Cloning must also be done with the progeny parasites derived from crosses. The technique of cloning of malaria parasites has been described by Demidowa (10), Bishop (11); and others. In current work with rodent plasmodia, a dilution technique has been adopted. The resistance of rodent malaria parasites to drugs is measured by inoculating samples of parasites into mice and treating the mice with known concentrations of drug (expressed as milligrams of drug per kilogram of mouse body weight), usually on 4 successive days. The drugs are administered either orally or by intra- peritoneal injection. Presence or absence of parasites in the blood is then recorded after various time intervals. These procedures are subject to several errors, and the effect of the drug cannot be measured precisely. In vitro methods (12, 13, 14) are now available for assessment of drug resistance of P.falciparum and should give more precise results, but so far they have not been widely used with rodent plasmodia. The genetic work done so far concerns pyrimeth- amine and chloroquine resistance, and to a lesser extent sulfadiazine and mefloquine resistance. EXPERIMENTAL FINDINGS Pyrimethamine resistance Pyrimethamine-resistant parasites have been readily obtained from initially sensitive strains of various Plasmodium species, by both low- and high- pressure selection techniques (15, 16). Moreover, it is well known that clinical treatment of human malaria patients with pyrimethamine leads to the rapid development of drug-resistant parasites (17). In this review, attention will be confined to cases where some genetic analysis has been done. With P. yoelii, Morgan (18) used a clone of sensitive parasites as starting material and found that when mice containing about 109 parasites were treated with a high dose (e.g., 50 mg/kg) of pyrimethamine, approximately 1 in 50 developed drug resistant para- sites. Similar results have been obtained with P. chabaudi (19, 20). The 'mutation rate' of sensitive parasites to pyrimethamine resistance has been esti- mated as being approximately 1 in 1010. This is a very 800 GENETICS OF DRUG RESISTANCE IN PLASMODIUM low frequency in comparison with other mutations in genetically well-studied organisms, and it seems likely that the procedures used with rodent malaria parasites succeeded in selecting only a small fraction of the mutations that actually occurred. In any case, only those mutants giving resistance to a high dose of pyrimethamine, in a single step, were selected in these experiments. It is likely that more mutations would be obtained if a lower concentration of drug were used. The pyrimethamine-resistant mutants that have been obtained are of various types in regard to cross- reaction with sulfadiazine and requirement for 4-aminobenzoic acid (PABA). Earlier workers (e.g., Jacobs) had found that pyrimethamine resistance was usually accompanied by an increased PABA requirement (21). Morgan (18) reported obtaining eight pyrimethamine-resistant mutants of P. yoelii with an increased PABA requirement, and one with a reduced PABA requirement. Macleod (22), working with P. chabaudi, isolated one mutant with a 20-fold increase in resistance to pyrimethamine accompanied by a 27-fold increase in sensitivity to sulfadiazine and associated with an increased PABA requirement. The same worker obtained another type of mutant with only a 3-fold increase in pyrimethamine resistance but a 5-fold increase in resistance to sulfadiazine, and reduced PABA requirement. It should also be mentioned that pyrimethamine resistance has often been found to be associated with resistance to other antifolate drugs, such as proguanil (23) and cycloguanil. To illustrate the stability of pyrimethamine-resist- ant mutants, it may be mentioned that Morgan (18) obtained a resistant clone of P. yoelii whose resistance remained unaltered after 55 blood passages, 18 mosquito transmissions, and 5 months' storage in liquid nitrogen. Following hybridization between pyrimethamine- resistant and pyrimethamine-sensitive parasite lines, and isolation of progeny clones, segregation and recombination have been shown to take place (19). Table 1 gives some typical results, which show that a substantial proportion of recombinant clones were found. Similar results have been found with P. chabaudi (22). These and many other similar results show that, following crosses between pyrimethamine-resistant and sensitive lines, segregation and recombination involving the genes concerned take place in mos- quitos, presumably at meiosis. Recombinants have not been found when mixtures of blood forms, or of sporozoites, are made, and the parasites cloned after a period of asexual growth only. The earlier claim of Yoeli et al. (24) that genetic exchange may occur among populations of trophozoites, has not been confirmed (25). Taking together the findings that abrupt changes Table 1. Crosses between pyrimethamine-resistant and sensitive lines of P. yoelii (after Walliker et al. (19)) Parents (1) pyr-r x (2) pyr-s GPI-1 GPI-2 Progeny Non-recombinants Recombinants clones pyr-r 21 clones pyr-r 13 clonesGPI-1 GPI-213coe pyr-s 30 clones pyr-s 7 clones pyr-r - pyrimethamine resistant; pyr-s - pyrimethamine sensitive. GPI-1, 2 - electrophoretic variants of glucose phosphate isomerase. from sensitivity to resistance occur in single clones of parasites, that the resistance thus obtained is stably inherited, and that segregation and recombination only occur during mosquito passage, it is concluded that nuclear gene mutation is the cause of pyrimeth- amine resistance. There is no evidence that pyrimeth- amine is mutagenic, though this point has not been studied specifically. It is therefore assumed that the mutations are spontaneous. In view of the phenotypic diversity of different pyri- methamine-resistant mutants, it is possible that mutation occurs at more than one chromosomal locus, but the detailed genetic studies needed to prove this have not been carried out. Chloroquine resistance In comparison with their apparently uniform sensi- tivity to pyrimethamine, different species of rodent malaria parasites display considerable variation in their response to chloroquine. P.yoelii is innately resistant to high doses (e.g., 50 mg per kg of body weight on four successive days) (26), while P. chabaudi, P. berghei, and P. vinckei normally tolerate only small doses (2-5 mg per kg) (16). Many workers have reported that it is more difficult to raise the resistance of malaria parasites to chloroquine, than to pyrimethamine, under labora- tory conditions (23, 16). However, stably resistant lines of P. chabaudi (27) and P. vinckei (28) have been produced by low-level drug treatments, usually lasting over many blood passages, with gradual increases in drug concentration. In this way resistance of P. vinckei to chloroquine was raised from 5 mg/kg to 200 mg/kg of body weight (the maximum dose toler- ated by mice) (28); the resistance of P. chabaudi, starting with cloned material, was raised from 2 mg/kg to 30 mg/kg of body weight (27). In both species the resistance finally obtained was stable after repeated passaging. Several reports of the development of lines of 801 G. H. BEALE P. berghei resistant to chloroquine have been summarized (16), and in one case, Peters (7) found that chloroquine resistance was unstable in the absence of the drug. In another study, the same worker (29) recorded the occurrence ofa dramatic rise in the chloroquine resistance of P. berghei (strain NS) following only a short course of chloroquine treatment; however this was from uncloned material which probably contained a mixture of sensitive P. berghei and resistant P. yoelii, the latter being selected by the drug treatment. It is possible that some of the earlier reports of chloroquine resistance in P. berghei have a similar explanation. Recent attempts to obtain mutations to a high level of chloroquine resistance in a single step, by treating cloned material with a single drug dose at a high concentration, have been unsuccessful. Genetic work on chloroquine resistance has so far been carried out only with P. chabaudi (27). Following crosses between parasites resistant to 3 mg/kg and others resistant to only 2 mg/kg, segregation of the two classes occurred, and recombinations with other genetic characters (pyrimethamine resistance and two types of enzyme variant) were also seen. Padua (personal communi- cation, 1979) made crosses between sensitive parasites and parasites from a highly resistant (30 mg/kg) line, and obtained in the progeny mainly sensitive and low- level resistant clones, although a small number of high-level resistant recombinants could also be detected by special selective techniques. These results are consistent with the hypothesis that high-level resistance is due to the combined action of several mutant genes at different chromosomal loci. As with pyrimethamine resistance, no recombination involv- ing chloroquine-resistant genes was found among asexual parasites. Resistance to other drugs Little or no knowledge is available concerning the genetics of resistance to other drugs. Ramakrishnan et al. (30, 31) obtained sulfadiazine-resistant strains of P. berghei by selection in the presence of the drug and also by selection in mice fed on a PABA-deficient diet. Sulfadiazine-resistant mutants of P. chabaudi have also been obtained (22), which, after crossing with sensitive strains, produced parasite clones in which segregation of resistant and sensitive genotypes had occurred. There was some evidence indicating inde- pendent segregation of sulfadiazine resistance and pyrimethamine resistance. Further study is needed to clarify the interrelations of the genes controlling these characters, however. Mefloquine-resistant lines of P. berghei (32) and P. chabaudi (Padua, personal communication, 1979) have also been obtained, but their genetics has not yet been investigated. Competition between drug-sensitive and drug- resistant parasites in mixedpopulations When vertebrate hosts are treated with a drug there is naturally a selection of drug-resistant parasites from mixtures of sensitive and resistant parasite popu- lations. It is, however, interesting to ask what are the selective advantages or disadvantages of resistant parasites in the absence of drugs. Some preliminary experiments have been carried out to examine this question by setting up populations containing mix- tures of sensitive and resistant parasites in various proportions, carrying out a series of passages in the absence of drugs, and determining the proportions of sensitive and resistant parasites that survive (33). With regard to low-level chloroquine resistance in P. chabaudi, a surprising result was obtained indicating that there was a selective advantage of resistant over sensitive parasites (Table 2). It should be stressed, however, that the numbers of clones tested were small, and further experiments of this type are needed. If chloroquine resistance is selectively favoured in P. chabaudi, it is difficult to understand why the wild strains of this species have not all become resistant. Presumably environmental conditions in nature somehow compensate for those producing a predomi- nance of resistant parasites in the laboratory. As regards pyrimethamine resistance, Rosario et al. set up a competition experiment starting with a population of P. chabaudi containing 50% resistant and 50% sensitive parasites. After 30 days in the blood of mice not exposed to pyrimethamine, 56 sensitive and 11 resistant clones were isolated. It is concluded that pyrimethamine resistance is neutral, or possibly selectively disadvantageous by comparison with sensitivity. Table 2. Drug responses of clones derived from mixed infections of chloroquine-resistant and sensitive lines of Plasmodium chabaudi (after Rosario et al. (33)) Blood induced infections at 30 days Initial inoculum Total (106 blood forms) clones Sensitive Resistant 100% resistant 23 0 23 50% resistant 9 0 9 50% sensitive 90% resistant 8 0 8 10% sensitive 10% resistant 4 0 4 90% sensitive 100% sensitive 13 13 0 802 GENETICS OF DRUG RESISTANCE IN PLASMODIUM DISCUSSION The results of genetic experiments with rodent Plasmodium species show that resistance to both pyrimethamine and chloroquine arises as a result of gene mutation and selection of resistant mutants under drug pressure. There are, however, some differences in the genetic mechanisms affecting resistance to the two drugs. With pyrimethamine, resistance to a high concentration of the drug may arise by a single step. Moreover, Bishop (15) reported that treatment even with a low dose could result in the appearance of mutants resistant to a high dose. With chloroquine, however, convincing evidence of single step mutations to a high level of resistance has so far not been presented, and a series of mutations, each having a small effect, is required. Hence, assuming that the situation in P.falciparum is similar to that in rodent and bird plasmodia, treatment with either low or high doses of pyrimethamine will result in selection of high-level resistant strains, but treatment with large doses of chloroquine should theoretically not yield highly resistant mutants. Some of the genetic characteristics of pyrimeth- amine and chloroquine resistance can be explained, at least in part, by reference to the biochemical steps affected. In the case of pyrimethamine, it has been found that the enzyme dihydrofolate reductase (EC. 1.5.1.4) is altered in drug-resistant mutants. Accord- ing to Ferone (34, 35) the enzyme in resistant mutants has reduced substrate- and drug-binding properties, and is produced in larger amounts than in sensitive parasites. This would account for the increased resis- tance of mutants to a given dose of pyrimethamine, and also for the accompanying increase in sensitivity to sulfadiazine and in requirement for PABA. How- ever, the class of mutants resistant to both pyrimetha- mine and sulfadiazine cannot be explained in this way. Further biochemical and genetic studies are required to clarify this situation. At present we are also not certain of the cause of the extraordinarily wide range of concentrations of pyrimethamine tolerated by different strains of P.falciparum. For example, the author, in collaboration with S. Thaithong, has tested a strain of P.falciparum from Thailand with a resistance to pyrimethamine some 10 000 times greater than that of some other strains. Conceivably these differences may be explained on the basis of amplification of the gene locus concerned with production of dihydrofolate reductase, similar to that demonstrated in the somewhat analogous case of methotrexate resistance in cultured mouse cells (36). This hypothesis however requires experimental proof. With regard to chloroquiine resistance, nothing is known with certainty about the biochemical steps that are altered in resistant parasites, though it is noteworthy that the latter are known to have a decreased capacity to concentrate the drug (37). Since chloroquine is known to react with many different cellular constituents, it is likely that drug resistance is controlled by a number of unrelated metabolic processes, which are themselves controlled by different genes. Finally, it is worth pointing out that the three classes of drug resistance in malaria parasites, exemplified by pyrimethamine, sulfadiazine, and chloroquine, illus- trate three different types of mechanism by which drugs act selectively on a parasite in a host cell (in this case both parasite and host being eukaryotic organisms): (1) pyrimethamine acts on an enzyme (dihydrofolate reductase) that is essential for both host and parasite, but binding of the drug to the parasite enzyme is much stronger than to the host enzyme; (2) sulfadiazine and other sulfa drugs act on another enzyme (dihydropteroate synthetase (EC. 2.5.1.15)) which is thought to be essential for Plasmodium, but is absent from vertebrate cells (35); and (3) chloroquine acts more powerfully on the parasites than on the host cells because of the selective uptake of the drug in parasitized cells. The genetic peculiarities of drug resistance no doubt reflect these biochemical features of drug action. Further bio- chemical and genetic studies are needed to give a more complete understanding of these mechanisms, which would then help in combating the growing problem of increasing drug resistance of malaria parasites. ACKNOWLEDGEMENTS The author thanks Dr D. Walliker and Dr A. Tait for their comments on the manuscript. 803 804 G. H. BEALE RLSUME GENETIQUE DE LA PHARMACORESISTANCE DES PARASITES DU PALUDISME Le present article passe en revue les donnees d'experience disponibles sur la g6netique de la resistance qu'opposent aux medicaments les parasites du paludisme. Plusieurs m6canismes pouvant etre A l'origine de l'apparition de la r6sistance sont 6voqu6s, parmi lesquels une mutation g6nique spontan6e-qui peut se faire en une seule etape A l'egard de la pyrimethamine-est sans doute le plus important. L'auteur d6crit les methodes utilistes pour la production, parmi des especes de Plasmodium des rongeurs, de souches resistantes A la pyrimethamine et A la chloro- quine, et expose les constatations faites sur l'acquisition par la descendance de cette resistance. Les caracteristiques biochimiques du mode d'action des medicaments, qui peuvent jouer un r8le dans l'acquisition genetique de la pharmacoresistance, sont aussi examinees. L'auteur rapporte egalement les resultats d'etudes sur la competition entre parasites sensibles aux medicaments et resistants dans des populations mixtes de plasmodies infectant les rongeurs, competition qui peut s'exercer aussi bien sous la pression des medicaments qu'en I'absence de traitement. Les conclusions a tirer des resultats de ces 6tudes en ce qui concerne la resistance aux medicaments de P.falciparum sont discut6es. REFERENCES 1. TRAGER, W. & JENSEN, J. B. Science, 193: 673-675 (1976). 2. HAYNES, J.D. ET AL. Nature (London), 263: 767-769 (1976). 3. BEALE, G. H. In: Porter, R. & Knight, J., ed., Parasites in the immunized host: mechanisms of survival. Amsterdam, Elsevier, 1974, pp. 21-32 (CIBA Sympo- sium 25). 4. BEALE, G. H. Genetics of Paramecium aurelia, London, Cambridge University Press, 1954. 5. BEALE, G. H. & KNOWLES, J. Extranuclear genetics, London, Arnold, 1958. 6. BEALE, G. H. ET AL. In: Killick-Kendrick, R. & Peters, W., ed., Rodent malaria, London and New York, Academic Press, 1978, pp. 213-245. 7. PETERS, W. Experimental parasitology, 17: 80-89 (1965). 8. GREENBERG, J. & TREMBLEY, H. L. Journal of parasitology, 40: 336-340 (1954). 9. GREENBERG, J. & TREMBLEY, H. L. Journal of parasitology, 40: 667-672 (1954). 10. DEMIDOWA, L. W. Giornale di batteriologia e immunologia, 13: 872-877 (1934). 11. BISHOP, A. Parasitology, 48: 210-234 (1958). 12. RIECKMANN, K. H. Journal of the American Medical Association, 217: 573-578 (1971). 13. NGUYEN-DINH, P. & TRAGER, W. Science, 200: 1397-1398 (1978). 14. RICHARDS, W. H. G. & MAPLES, B. K. Annals of tropical medicine andparasitology, 73: 99-107 (1979). 15. BISHOP, A. Parasitology, 52: 495-518 (1962). 16. PETERS, W. Chemotherapy and drug resistance in malaria. London and New York, Academic Press, 1970. 17. CLYDE, D. F. & SHUTE, G. T. Transactions oftheRoyal Society of TropicalMedicine andHygiene, 51: 505-513 (1957). 18. MORGAN, S. The genetics of malaria parasites: studies on pyrimethamine resistance. PhD Thesis, Edinburgh University, 1974. 19. WALLIKER, D. ETAL. Parasitology, 66: 309-320(1973). 20. WALLIKER, D. ET AL. Parasitology, 70: 19-24 (1975). 21. JACOBS, R. L. Experimentalparasitology, 15: 213-225 (1964). 22. MACLEOD, R. The genetics ofdrug resistance in malaria parasites. PhD Thesis, Edinburgh University, 1978. 23. BISHoP, A. Biological reviews, 34: 445-500 (1959). 24. YOELI, M. ET AL. Parasitology, 59: 429-447 (1969). 25. WALLIKER, D. ET AL. Parasitology, 72: 183-194 (1976). 26. WARHURST, D. C. & KILLICK-KENDRICK, R. Nature (London), 213: 1048-1049 (1967). 27. RoSARIO, V. Nature (London), 261: 585-586 (1976). 28. POWERS, K. G. ET AL. Experimental parasitology, 26: 193-202 (1969). 29. PETERS, W. ET AL. Annals of tropical medicine and parasitology, 72: 23-36 (1978). 30. RAMAKRISHNAN, S. P. ET AL. Indian journal of malariology, 10: 175-182 (1956). 31. RAMAKRISHNAN, S. P. ET AL. Indian journal of malariology, 15: 96-106 (1961). 32. PETERS, W. ET AL. Annals of tropical medicine and parasitology, 71: 419-427 (1977). 33. ROSARIO, V. ET AL. Lancet, 185-188 (1978). 34. FERONE, R. Journal of biological chemistry, 245: 850-854 (1970). 35. FERONE, R. Bulletin of the World Health Organization, 55: 291-298 (1977). 36. SCHIMKE, R. T. ETAL. Science, 202: 1051-1055 (1978). 37. FITCH, C. D. Proceedings of the National Academy of Sciences, 64: 1181-1187 (1969).