|
|
Antimicrob Agents Chemother. 2009 April; 53(4): 1509–1515. Published online 2009 January 21. doi: 10.1128/AAC.00241-08. | PMCID: PMC2663078 |
Copyright © 2009, American Society for Microbiology Plasmodium falciparum pfmdr1 Amplification, Mefloquine Resistance, and Parasite Fitness ![[down-pointing small open triangle]](/corehtml/pmc/pmcents/x25BF.gif) † Piyanuch Preechapornkul,1 Mallika Imwong,1 Kesinee Chotivanich,1 Wirichada Pongtavornpinyo,1 Arjen M. Dondorp,1,2 Nicholas P. J. Day,1,2 Nicholas J. White,1,2* and Sasithon Pukrittayakamee1 Faculty of Tropical Medicine, Mahidol University, Bangkok, Thailand,1 Centre for Clinical Vaccinology and Tropical Medicine, Churchill Hospital, Oxford, United Kingdom2 Received February 21, 2008; Revised April 1, 2008; Accepted January 8, 2009. Mefloquine is widely used in combination with artemisinin derivatives for the treatment of falciparum malaria. Mefloquine resistance in Plasmodium falciparum has been related to increased copy numbers of multidrug-resistant gene 1 (pfmdr1). We studied the ex vivo dynamics of pfmdr1 gene amplification in culture-adapted P. falciparum in relation to mefloquine resistance and parasite fitness. A Thai P. falciparum isolate (isolate TM036) was assessed by the use of multiple genetic markers as a single genotype. Resistance was selected by exposure to stepwise increasing concentrations of mefloquine up to 30 ng/ml in continuous culture. The pfmdr1 gene copy numbers increased as susceptibility to mefloquine declined (P = 0.03). No codon mutations at positions 86, 184, 1034, 1042, and 1246 in the pfmdr1 gene were detected. Two subclones of selected parasites (average copy numbers, 2.3 and 3.1, respectively) showed a fitness disadvantage when they were grown together with the original parasites containing a single pfmdr1 gene copy in the absence of mefloquine; the multiplication rates were 6.3% and 8.7% lower, respectively (P < 0.01). Modeling of the dynamics of the pfmdr1 copy numbers over time in relation to the relative fitness of the parasites suggested that net pfmdr1 gene amplification from one to two copies occurs once in every 108 parasites and that amplification from two to three copies occurs once in every 103 parasites. pfmdr1 gene amplification in P. falciparum is a frequent event and confers mefloquine resistance. Parasites with multiple copies of the pfmdr1 gene have decreased survival fitness in the absence of drug pressure. Whereas a combination of antimalarial drug treatment and vector control were very successful tools for controlling malaria in the mid-20th century and eliminated malaria from most temperate zones in the world, the rates of morbidity and mortality from malaria have worsened over recent decades in many parts of the tropical world. The resurgence of malaria is largely related to the increased resistance of the Plasmodium falciparum parasite to many of the available antimalarial drugs, such as chloroquine, pyrimethamine-sulfadoxine, and more recently, mefloquine. Multidrug-resistant P. falciparum is widely prevalent in Southeast Asia. This has made the treatment of falciparum malaria increasingly difficult ( 27). The artemisinin combination treatment artesunate-mefloquine was first introduced in 1994 for the treatment of falciparum malaria on the Thailand-Myanmar (Burma) border and continues to be highly effective ( 29). Codon mutations in the gene encoding the P. falciparum P-glycoprotein transporter ( pfmdr1) have been incriminated in resistance to chloroquine and mefloquine ( 16, 18). A stronger association with mefloquine resistance is the increased expression of the amplified wild-type pfmdr1 gene ( 8, 16, 17, 19). In a large study conducted on the Thailand-Myanmar border, where multidrug resistant falciparum malaria is prevalent, an increase in the pfmdr1 copy number was estimated to account for about two-thirds of the variation in susceptibility to mefloquine ( 19). We evaluated the dynamic relationships between mefloquine resistance and pfmdr1 copy number and mutations in a long-term in vitro culture of a single strain of P. falciparum in which resistance was selected by stepwise increases in mefloquine pressure. The stability of the selected genotype containing multiple pfmdr1 copies was studied, and the relative fitness of parasites with multiple copies of pfmdr1 was compared to that of the original isolate with a single copy of pfmdr1. A mathematical model was constructed to estimate the frequency of the event resulting in pfmdr1 gene amplification. Parasite genotyping and culture. Mefloquine-sensitive Thai P. falciparum isolate TM036 was maintained in continuous culture ( 24). To confirm the presence of a single genotype, the isolate was genotyped with msp1, msp2, and glurp as genetic markers by nested PCR, as described previously ( 6). To further confirm this, the initial isolate was also genotyped by using 10 microsatellite markers distributed over different chromosomes, including TA1, TAA87, POLY, ARA2, PFPK2, 2490, TAA42, PFG377, TAA60, and TAA81, and fluorescent end-labeled oligonucleotides ( 3). The gene products were analyzed on an ABI 3100 capillary sequencer, and the alleles were scored by using Genescan and Genotyper software. Pull-up artifacts were prevented by dilution of the PCR product, and a minor allele was identified if its intensity amplitude was >33% of the height of the predominant allele ( 1). Parasites were cultured in human blood group O erythrocytes in a total volume of 40 ml to a maximum density of 5% parasitemia at 5% hematocrit. If the level of parasitemia exceeded 5%, the culture was routinely diluted to 0.5 to 1% parasitemia with a standard volume of fresh donor erythrocytes. The culture medium contained RPMI 1640 (ICN Biomedicals, United Kingdom) supplemented with 2 g/liter NaHCO3 (Life Technologies), 5.958 g/liter HEPES (Sigma), 1 ml/liter gentamicin sulfate (Government Pharmaceutical Organization, Thailand), 13.6 mg/liter hypoxanthine (Sigma), 4.5 g/liter d-glucose (Sigma), and 0.5% Albumax II (Gibco, New Zealand). Selection of mefloquine-resistant parasites. A stock solution of 1 mg/ml mefloquine hydrochloride (Pharmaceutical Technology Department, Institute de Technologia em Farmacos, Italy) dissolved in 0.01 M HCl was prepared in RPMI 1640 medium, stored protected from the light at 4°C, and used within 4 weeks of preparation. The stock solution was then further diluted with RPMI 1640 medium to the required concentrations shortly before each experiment. Resistance to mefloquine in the TM036 P. falciparum strain was selected by exposure of the cultured parasites to stepwise increases in mefloquine concentrations of 15, 20, 25, and 30 ng/ml, after an initial preliminary phase of 127 days of continuous culture with intermittent exposures to very low levels of mefloquine ranging from 1 to 10 ng/ml. The initial level of parasitemia was 4 to 5%, which corresponds to a total of approximately 10 9 parasites per flask, calculated as follows: parasite count per μl = parasite count per 1,000 erythrocytes × hematocrit (in percent) × 125.6 ( 23). The parasite culture serving as a control was routinely diluted to 0.5 to 1% parasitemia (1× 10 8 to 2 × 10 8 parasites per flask) in the absence of drug. After the addition of mefloquine, the level of parasitemia in a thin blood film was monitored every 24 h. If the level of parasitemia decreased to 2%, which corresponds to approximately 5 × 10 8 parasites per flask, the culture medium was replaced by a mefloquine-free medium. The parasites were then allowed to recover until the level of parasitemia was 5% (1.3 × 10 9 parasites per flask) and the growth rates were comparable to those in the control culture. The recovered parasites were then exposed to the same concentration of mefloquine for a second time. After that, the parasites were exposed to a higher dose of mefloquine and the same cycle of procedures was repeated. DNA samples were taken after removal of the mefloquine drug pressure and after the parasite population was allowed to recover until a level of parasitemia of 50 per 1,000 erythrocytes was reached. Mefloquine susceptibility assays. Mefloquine susceptibility assays were performed with samples taken from the continuous culture. Susceptibility was measured by assessment of the inhibition of schizont maturation, as described previously ( 30). In short, parasites were harvested when the level of synchronized ring-stage-infected red blood cell parasitemia reached 5%. The culture was then centrifuged at 800 × g at 4°C for 5 min, 100 μl of packed red blood cells was resuspended in culture medium to a 3% hematocrit, and the mixture of medium and cells was added to a microculture plate predosed with mefloquine at fivefold dilutions ranging from 0.064 ng/ml to 1 μg/ml. The parasites were incubated at 37°C with 5% CO 2 for 40 to 44 h, and the assay was terminated when 80% of the mature schizonts (more than six merozoites) presented in the control well containing no drug. Maturation inhibition was defined as the percentage of schizonts in the well containing drug compared to the percentage of schizonts in the control well. The drug concentration that gave 50% inhibition of schizont maturation (IC 50) was derived by fitting the dose-maturation inhibition curves ( n = 3) by using the statistical program Winnonlin (version 4.1; Pharsight, Mountain View, CA). Sample preparation and DNA extraction. Parasite DNA was extracted from the parasite-containing red blood cell pellet obtained by centrifugation at 800 × g for 5 min at 4°C. The pellet was either frozen at −20°C until further processing for template DNA purification with a QIAamp DNA kit (Qiagen, Hilden, Germany) or 5 μl of the parasite pellet (50% hematocrit) was transferred to filter paper (Whatman, Maidstone, United Kingdom) and the DNA was extracted with PrepMan Ultra reagent (Applied Biosystems, Foster City, CA), followed by precipitation of the supernatant and resuspension in Tris-EDTA buffer (50 μl). Assessment of pfmdr1 amplification by real-time PCR. Amplification of the pfmdr1 gene was assessed by real-time PCR (Rotor Gene 3000; Corbett Research, Australia). The primers and the fluorescence probe used have been described previously ( 19). The amplification reactions were done as multiplex PCRs in 7.5 μl of a reagent containing the buffer from the Biotool Quantimix easy probe kit (B&M Lab, Spain), MgCl 2 (5.5 mmol/liter), forward and reverse primers (each at 300 nmol/liter), a pfmdr1 probe and a β-tubulin probe (each at 100 nmol/liter), and 2.5 μl of template DNA. Fifty cycles were performed (holds at 95°C for 10 min, followed by cycling of 95°C for 15 min and 58°C for 1 min). Each copy number value was derived from three separate assessments per assay. Estimates were rejected if they did not conform to exponential kinetics. The assay was repeated if one of the following results was obtained: a spread threshold cycle ( CT) value greater than 1.5, a CT value greater than 35, or a copy number of more than 1.3. The assay was repeated at least twice for validation of consistent results. Both positive and negative control samples were added in each assay. Assessment of mutations in the pfmdr1 gene. The pfmdr1 gene was amplified from the DNA template obtained from the frozen parasite culture pellets by nested PCR. A PCR-restriction fragment length polymorphism assay was then used to assess whether there were point mutations in the pfmdr1 gene. Five polymorphic sites have been associated with mefloquine resistance. The polymorphic sites at positions 86, 184, 1034, and 1042 ( 20) were assessed by using ApoI, DraI, DdeI, and AseI (VspI), respectively, as described before ( 10). Mutations at another polymorphic site at position 1246 were assessed by gene sequencing (Genetic Analyzer 3100; Applied Biosystems). All restriction enzymes were purchased from New England Biolabs (Hitchen, United Kingdom). Digestion fragments were analyzed on a 3% Seakem LE agarose gel with TBE (Tris-borate-EDTA) buffer under constant power (120 V). Stability of pfmdr1 multiple-copy-number genotype in the absence of drug pressure. Three parasite cultures with multiple pfmdr1 copy numbers selected with stepwise increasing mefloquine concentrations (20, 25, and 30 ng/ml) were used to assess the backmutation frequency and the stability of the multiple-copy-number phenotype. For comparison, two cultures containing parasites with a single copy were also assessed. All five flasks were cultured in a drug-free complete medium for up to a maximum of 6 months. All parasite cultures were kept in 5% hematocrit in a total volume of 5 ml and were routinely diluted to 0.5 to 1% parasitemia to maintain the parasite population at 1 × 107 to 2 × 108 per flask. The pfmdr1 copy numbers were assessed once per month, as described above. The observations from this experiment were used to estimate an upper limit for the frequency of deamplification (switching down from three to two and from two to one copies of the pfmdr1 gene). Relative parasite fitness in relation to pfmdr1 copy number. The relative fitness of mefloquine-exposed Thai isolate TM036 containing multiple copies of pfmdr1 was compared with that of the wild-type isolate containing a single copy. For this, 1:1 mixtures of the wild-type culture and the mefloquine-exposed cultures were prepared. Equal volumes of the parasite cultures, both of which were adjusted with fresh erythrocytes to parasitemias of 0.1%, were mixed. It should be noted the accuracies of the levels of parasitemia in this range as assessed by light microscopy are inherently limited. The parasites with average copy numbers of 2.1 and 3.1 were each mixed with a single-copy-number parasite population. Each culture was at 0.1% parasitemia and 5% hematocrit, which corresponds to approximately 3 × 106 parasites per flask. The cultures were then expanded to a total volume of 40 ml, yielding approximately 109 parasites. All cultures were kept for up to 4 months. Every 2 to 3 days, parasites were sampled on filter paper for assessment of the pfmdr1 copy number. Modeling of parasite fitness. The changes in the pfmdr1 copy numbers over time in the mixtures of cultures containing single and multiple copies were used to model the relative fitness of the parasites with multiple-copy-number genotypes by using the SPLUS (version 6.0) Professional program (Insightful Corp.). Relative fitness was defined as the multiplication rate of the multiple-copy-number parasites compared to that of the single-copy-number parasites. A model was constructed by assuming that the observed copy number reflected a weighted average value of the copy number for an in vivo mixture of substrains with different copy numbers. It was also assumed that a pfmdr1 copy number of zero was not viable and that the accuracy of the pfmdr1 copy number assay was independent of the copy number ( 13). Another assumption of the model was that the pfmdr1 copy number of any isolate was stable in the absence of drug pressure, i.e., that there was a balance between the forward and reverse switching of pfmdr1 copy numbers. This assumption was supported by the results of the other experiments (see Fig. ). Therefore, the change in the average pfmdr1 copy number in a mixture of cultures over time represented differences in the fitness of the contributing parasite substrains with different pfmdr1 copy numbers. Details on the construction of the model can be found in the supplemental material. The number of parasites with x and y copy numbers at the beginning of the coculture experiment were calculated from the average pfmdr1 copy numbers before and after they were mixed. The total numbers of parasites with x copies at time t (parasite xt) or y copies at time t (parasite yt) can be calculated from. where parasite x0 and parasite y0 are the total numbers of parasites with x and y copy numbers at time zero, respectively, and where At and Bt are the multiplication rates of parasites with x and y copy numbers at time t, respectively. The following formula was used to describe the relationship between the average copy number at time t (av.copy t), parasite x0 and parasite y0, and Bt/ At, which is the ratio of the multiplication rates at time t: Since all parameters except Bt/ At are measured during the experiment, Bt/ At can be estimated by finding the value that best fits the curve represented by equation 3 to the observed data of the average copy numbers over time. Estimation of the event rate of pfmdr1 gene amplification. In order to derive a quantitative estimate of the net event rate of pfmdr1 gene amplification, a second mathematical model was constructed (see the supplemental material). The model inputs included the dynamics of the changes in pfmdr1 copy numbers over time, information on the relative fitness in the absence of drug pressure derived from the first model, and the relative fitness in the presence of drug pressure derived from the assessment of the IC50 for mefloquine in relation to the pfmdr1 copy number. Since the total number of parasites, the average copy number, and the relative multiplication rates of the parasites with different copy numbers are all known over time, the rates of net pfmdr1 gene amplification events, which denote the balance between the forward switch (from one to two copies and from two to three copies) and the reverse switch (from three to two copies and from two to one copies) can be derived by choosing the values that give the best fit to the observed data. The equations used can be summarized as follows for time t: where parasite 1t, parasite 2t, and parasite 3t are parasite populations with one, two, and three copies at time t (in cycles), respectively; sw12 and sw23 are the net rates of switching of the copy numbers from one to two and from two to three, respectively. μ 1c, μ 2c, and μ 3c are the multiplication factors for the parasite populations with one, two, and three copies numbers at a specific mefloquine concentration ( c), respectively. μ 1c, μ 2c, and μ 3c are derived from μ, which is the observed average multiplication rate of the parasite culture. For details, see the supplemental material. Statistical analysis. The repeatability of the assessment of copy numbers was assessed by the method of Bland and Altman ( 5). The repeatability coefficient was calculated as the 95% confidence interval (CI) of the differences between the measurements. Nonparametric data were analyzed by the Kruskal-Wallis, Mann-Whitney U, or Spearman rank test, as appropriate, by using the SPSS statistical program (version 10.0; SPSS Inc., Chicago, IL). Analysis of trends between two nonparametric variables was tested by the np trend test (Stata Inc.). The level of significance in these experiments was set at a P value of <0.05. Genotyping of P. falciparum isolate TM036. Genotyping of the starting isolate with three antigenic markers, msp1, msp2, and glurp, revealed the presence of only a single genotype. In addition, by use of a more sensitive typing system with 10 microsatellite markers, only a single allele was observed. Association between level of mefloquine resistance and pfmdr1 gene amplification and mutations. In P. falciparum mefloquine-sensitive isolate TM036, mefloquine resistance could be selected in vitro by use of a stepwise increase in the mefloquine concentration to which it was exposed to up to 30 ng/ml. The selection of tolerance was illustrated by the increased exposure time tolerated (defined as the ability to maintain a level of parasitemia above 2%) during the second exposure compared to that during the first exposure with the same mefloquine concentration. For instance, after 3 days of exposure to 15 ng/ml mefloquine, the level of parasitemia dropped from 5% to 2%; but after recovery of the parasite numbers, a second exposure to the same mefloquine concentration was tolerated for 7 days. This corresponded to an increase in the IC50 of mefloquine for the parasite population (see below). Over time, there was a stepwise increase in the median pfmdr1 copy number with exposure to increasing mefloquine concentrations, as shown in Fig. . The initial exposure for 127 days to intermittent low doses of mefloquine (1 to 10 ng/ml) in the early phase of the study did not affect the pfmdr1 copy number. The median pfmdr1 copy number of the original strain was 1.03 (range, 0.78 to 1.16). The first increase in the pfmdr1 copy number to a median value of 1.22 (range, 1.03 to 1.40; P < 0.05) was observed after the second exposure to mefloquine at a concentration of 15 ng/ml. After the mefloquine concentration was increased to 20 ng/ml, the copy number increased sharply to 1.92 (range, 1.71 to 2.26; P < 0.05), and there was a subsequent further significant increase in copy number after the second exposure to the same dose to 2.45 (range, 2.31 to 2.53). After this, exposure to higher mefloquine concentrations of 25 and 30 ng/ml did not significantly increase the median pfmdr1 copy number after either the first or the second exposure, with the exception of the pfmdr1 copy number after the first exposure to 30 ng/ml (median, 3.22; range, 3.03 to 3.22). Figure shows the relationship between the mefloquine drug exposure and the pfmdr1 copy number (values after the second exposures). There was a significant increase in the pfmdr1 copy number with increasing mefloquine concentrations (test for trend across ordered groups, P < 0.001). To confirm that tolerance to increasing concentrations of mefloquine corresponded to increased resistance to the drug, the IC 50s for parasites containing an average of 1.0, 2.1, and 3.0 copies of the pfmdr1 gene were assessed (in triplicate). The median IC 50s of mefloquine for parasites with 2.1 and 3.0 copies were 7.90 ng/ml (range, 3.99 to 9.04 ng/ml) and 9.29 ng/ml (range, 6.62 to 13.29 ng/ml), respectively, whereas the median IC 50 was 1.47 ng/ml (range, 0.62 to 1.88 ng/ml) for parasites with a single copy ( P = 0.03) (Table 1). The IC 50s of mefloquine correlated positively with the pfmdr1 copy number (Spearman's rho = 0.87; P = 0.001). | TABLE 1.Mefloquine IC50s for parasites with different copy numbers |
No codon mutations in the pfmdr1 gene at positions 84, 184, 1034, 1042, and 1246 were detected in any of the samples. Stability of parasite isolates containing multiple pfmdr1 copies in the absence of drug pressure. The stability of the genotypes of parasites containing different pfmdr1 copy numbers in the absence of mefloquine was evaluated with four samples of parasites, which were cultured for periods that varied from 2 to 6 months. Up to the 90th successive generation in a mefloquine-free culture medium, the increased pfmdr1 copy numbers in the selected parasite populations remained unchanged (t test for slope, P > 0.05) (Fig. ). This stability of the culture containing parasites with multiple copies was used to define a lower limit of the reverse switch rates (i.e., three to two and two to one pfmdr1 copies). From the total numbers of parasites in the experiment, a reverse switch rate from two to one copies to a more fit parasite subclone with a lower pfmdr1 copy number cannot be more frequent than the forward switch rate. Since the switch rate from one to two pfmdr1 copies occurs in approximately 1 in 108 parasites, it can be assumed that the switch-back rate has to be less frequent than 1 in 108, since any value lower than this would result in a measurable decrease in the mean copy numbers. Similarly, the switch-back rate from three to two copy numbers cannot exceed 1 in 5 × 103 asexual divisions. Relative fitness of parasites with multiple pfmdr1 copies compared to that of parasites with a single copy. Coculturing of parasite subclones containing a single pfmdr1 copy and multiple pfmdr1 copies in the absence of mefloquine drug pressure resulted in a gradual decrease in the weighted average copy number, indicating the preferential survival of parasites with single-copy-number genotypes over time (Fig. ). Parasite populations containing apparently single subclones with one or multiple copies remained unchanged over time. In contrast, in a coculture containing parasite populations with 1 and 2.3 pfmdr1 copies, the copy number decreased from an average of 2.0 to 1.1 in 21 days (10 generations); in the mix containing 1.0 and 3.1 copies, the number declined from 1.8 to 1.2 in 33 days (16 generations) (Fig. ). All assessments of pfmdr1 copy numbers were made in triplicate assays. The results of this experiment were used to model the relative fitness of the parasites containing multiple copies of the pfmdr1 gene compared to that of the wild-type isolate. Modeling of parasite fitness. The curve describing the reduction in the average copy number over time starting with an initial equal mix of parasites containing 1.0 and 2.3 copies numbers was fitted by using equation 3 in a nonlinear regression model. The best fit was achieved when the ratio of the multiplication rate of the isolates with multiple copies (2.3 copies) to the multiplication rate of the isolate with a single copy ( B/ A) was 0.937, with a residual sum of squares of 0.83 with 12 degrees of freedom (see Fig. SA1 in the supplemental material). Thus, compared to the fitness of parasites with a single-copy-number genotype, the relative fitness disadvantage of the parasites with an average copy number of 2.3 was estimated to be 6.3% (95% CI = 3.9 to 8.9%). Similarly starting with an equal mix of parasites with 1.0 and 3.1 pfmdr1 copies, the fitted model-derived ratio of multiplication rates (B/A) was 0.913, with a residual sum of squares of 1.03 with 16 degrees of freedom, which corresponds to a fitness disadvantage of 8.7% (95% CI = 5.9 to 11.5%) in parasites with an average of 3.1 pfmdr1 copies per parasite compared to the fitness of parasites with an average copy number of 1 (see the supplemental material). The incremental reduction in the fitness of parasites with 3.1 pfmdr1 copies (compared to the fitness of those with 2.3 copies) was thus smaller (2.4%) than the reduction in the fitness of parasites with 2.3 copies (6.3%) compared to the fitness of those with 1 copy. Modeling the rate of pfmdr1 amplification. By fitting the observed increase in the pfmdr1 copy numbers over time in the presence of mefloquine drug pressure by using the model described above and further detailed in the supplemental material, the rate of pfmdr1 amplification events was estimated. The relative fitness of parasites with one, two, or three copies in the presence of various mefloquine concentrations was derived from the in vitro dose-maturation inhibition curves. With increasing mefloquine concentrations in the culture medium, the relative fitness of parasites with lower numbers of copies decreased compared to the fitness of those with three copies (which was set equal to 100%): at a concentration of 15 ng/ml mefloquine, the relative levels of fitness of parasites with one and two copies were 24.0% and 93.9%, respectively; at a concentration of 20 ng/ml mefloquine, the relative levels of fitness of parasites with one and two copies were 22.2% and 92.9%, respectively; at a concentration of 25 ng/ml mefloquine, the relative levels of fitness of parasites with one and two copies were 21.1% and 92.3%, respectively; and at a concentration of 30 ng/ml mefloquine, the relative levels of fitness of parasites with one and two copies were 20.3% and 91.8%, respectively. The observed multiplication rates (μ) assessed from the growth curves and total parasite numbers varied greatly over time (range, 0.04 to 10.84), with the level of mefloquine exposure being the main determinant. By application of the model, net switch rates of 1 in 108 for amplification from one to two copies and 1 in 5 × 103 for amplification from two to three copies gave the best fit of the model to the observed data (see Fig. SA3 in the supplemental material). The residual sum of squares was 20.15. It should be noted that these values describe the net switch rate, which is the difference between the amplification rate and the reverse switch rate or deamplification (to a lower pfmdr1 copy number). When mefloquine was deployed in 1984 as treatment for uncomplicated falciparum malaria on the Thailand-Myanmar border, high-grade resistance developed rapidly over a 6-year period, despite the use of close controls that limited prescription of the drug for the treatment of symptomatic falciparum malaria ( 14). By mid-1994, 50% of infections treated with high-dose mefloquine monotherapy recrudesced and in 10% of infections parasites were not cleared at all from patients who lived in that region. After the drug was combined with a 3-day course of artesunate, the efficacy of the combination restored the cure rate to >90% and the in vitro susceptibility of P. falciparum to mefloquine initially improved ( 1). In several studies, mefloquine resistance has been associated with an increase in the parasite pfmdr1 copy number ( 9, 16, 19). Increased pfmdr1 copy numbers were found to be highly predictive of a subsequent recrudescence of the infection in both patients treated with mefloquine monotherapy and those treated with a combination of artesunate with mefloquine. The pfmdr1 copy number also correlated with in vitro susceptibility to mefloquine ( 19, 22). The results of in vivo studies suggested that changes in the pfmdr1 copy number were frequent and were certainly much more frequent than mutations in the gene ( 26) and could be observed within a single infection ( 25). To explore the relationships between pfmdr1 gene amplification and mefloquine resistance further, we studied the dynamics of the changes in pfmdr1 copy number under mefloquine drug pressure in relation to mefloquine sensitivity. The results showed that pfmdr1 gene amplification could be selected in vitro by exposing the parasites to stepwise increasing concentrations of mefloquine and that, indeed, amplification of the pfmdr1 gene was a frequent event. A weakness of the current study is that the data came from a single, albeit protracted and labor-intensive, experiment. The fresh parasite isolate used in the experiment was also not cloned, although extensive genotyping with both conventional genetic markers (msp1, msp2, and glurp) and 10 microsatellite markers revealed only a single genotype. The possibility that minute subpopulations of genotypically different clones were present cannot be ruled out entirely, but this is unlikely, since no new genotype was detected over several months of continuous culture. Any putative minority clone could not have contributed numerically to the estimates obtained in this study; otherwise, it would have been detected. Parasites with increased copy numbers exhibited increased tolerance to mefloquine in the culture medium, corresponding to a significant increase in the IC50 for mefloquine in the maturation inhibition assay. Other factors may also have contributed to the reduced mefloquine susceptibility, including transcriptional and translational modification of the pfmdr1 product. Net amplification of the pfmdr1 gene from one to two copies in this in vitro study occurred in approximately in 1:10 8 parasites per asexual cycle. Since the total number of parasites in an adult patient with symptomatic malaria varies between approximately 10 8 and 10 13 parasites, pfmdr1 gene amplification can be expected to occur in most clinical cases. This finding is concordant with clinical observations ( 25) and emphasizes the importance of using combination antimalarial treatment strategies to reduce the chances of selecting the resistant phenotype. The survival and expansion of parasites with multiple-copy-number genotypes depend on the presence of subtherapeutic and selective blood concentrations of mefloquine, which compensate for any intrinsic impairment of fitness and which provide a net fitness advantage for this genotype. These findings are in line with the rapid development of drug resistance observed when mefloquine, which has a long terminal half-life, was deployed as monotherapy. According to our model, the amplification from two to three copies was an even more frequent event, occurring in a genetic background of two copies per genome in every 5,000 parasites per cycle. Although this seems to be a very high rate, it is biologically plausible because the initial duplication event, which is the rate-limiting step, creates regions of homology in the genome that facilitate further amplification ( 13). A possible mechanism in P-glycoprotein gene amplification is a so-called breakage-fusion bridge cycle, as has been shown in a Chinese hamster cell line. Chromosome breakage at a major fragile site was shown to be associated with P-glycoprotein gene amplification in multidrug-resistant CHO cells ( 11). In the present study, a stepwise increase in copy number was observed. It is unclear whether gene deamplification also occurs in a stepwise pattern. Gene amplification is likely to be a random event, as is codon mutation, but the latter is much less frequent: for example, mutations in the mitochondrial DNA conferring high-grade resistance to atovaquone emerge in vivo at a frequency of approximately 1 in 10 12 parasites ( 28) and codon mutations at positions 46 and 108 of pyrimethamine-resistant parasites occur at less than 1 in 2.5 × 10 9 mutations/ dhfr gene/replication ( 15). Since codon mutations in the pfmdr1 gene occur at a much lower frequency (1:10 14) ( 28), this explains why codon mutations at any of the putative resistance loci (positions 84, 184, 1034, 1042, and 1246) were not detected in this study. This does not exclude the possibility that codon mutations in the pfmdr1 gene associated with mefloquine resistance happen in nature, but it shows that these are not essential for the development of mefloquine resistance. Indeed, epidemiological studies suggest that pfmdr1 amplification occurs more readily in parasites bearing the wild-type gene. In a field study at the Thailand-Myanmar border, no significant association between codon mutations at positions 86, 1034, and 1042 and the recrudescence of parasites was found ( 19). Other studies have shown an association between the deamplification of pfmdr1 and an increase in mefloquine sensitivity ( 4). However, some investigators claim that, on the basis of data from both in vitro studies ( 12) and in vivo studies ( 7), amplification of the pfmdr1 gene is not a prerequisite for mefloquine resistance. Our study directly assessed the fitness disadvantage of the parasites with multiple numbers of copies of the pfmdr1 gene compared to the fitness of parasites with a single copy. The decreases in fitness were 6.3% for parasites with two copies and 8.7% for parasites with three copies. Since the pfmdr1 gene product is an ATP-consuming P-glycoprotein pump, this could be explained by the fact that in the absence of drug pressure, an increase in the expression of the pfmdr1 gene product will give a less favorable metabolic balance. Moreover, given that amplified regions of chromosomes may exceed 100 kb ( 7) and contain multiple genes, it can be expected that there will be significant metabolic costs associated with replication. It should be noted that in our study, parasite populations with increases in the average pfmdr1 copy number were compared; the observed differences in fitness thus apply strictly to these particular populations containing mixtures of parasites with different copy numbers. Concurrent additional changes in the genome apart from the pfmdr1 gene also cannot be excluded to play a role in the observed differences in fitness. The magnitude of fitness loss associated with pfmdr1 amplification is on the same order as that observed for parasites with other mutations conferring resistance to antimalarial drugs, like atovaquone (9%) ( 21) and chloroquine (7 to 13%) ( 2). In the absence of drugs, no reversal of the pfmdr1 copy numbers in the genome of the parasite was observed over a period of up to 6 months, and we estimated from this experiment that the reverse switch rate from two to one pfmdr1 copies was less than 1 in 108 parasites. Since parasite populations with multiple copies are still likely to contain a small fraction of parasites with a single copy, which will outcompete the parasites with multiple copies in the absence of drug pressure, it is possible that a longer observation period would have shown a decrease in the pfmdr1 copy number. Competition between parasite clones with different pfmdr1 copy numbers is more likely to occur in areas in sub-Saharan Africa with high rates of transmission, where infections with multiple clones are the rule, than in areas like Southeast Asia, where the rates of transmission are low. This could be a factor contributing to the delay of antimalarial drug resistance in settings with high rates of transmission. In conclusion, the current study shows that pfmdr1 gene amplification is a frequent event and is associated directly with mefloquine susceptibility in an in vitro culture system. In the absence of drug, the presence of multiple pfmdr1 copies confers a significant fitness disadvantage. This study was supported by the University Development Committee Scholarship Thailand, the Thailand Research Fund, the Commission on Higher Education, and the Wellcome Trust of Great Britain. We thank Anne-Catrin Uhlemann, Sanjeev Krishna, and Sornchai Looareesuwan for their valuable support. We also thank Sue Lee and Kasia Stepniewska for their assistance. The funders had no role in the design of the study, data collection and analysis, the decision to publish, or the preparation of the manuscript. 1. Anderson, T. J., B. Haubold, J. T. Williams, J. G. Estrada-Franco, L. Richardson, R. Mollinedo, M. Bockarie, J. Mokili, S. Mharakurwa, N. French, J. Whitworth, I. D. Velez, A. H. Brockman, F. Nosten, M. U. Ferreira, and K. P. Day. 2000. Microsatellite markers reveal a spectrum of population structures in the malaria parasite Plasmodium falciparum. Mol. Biol. Evol. 17:1467-1482. [PubMed] 2. Anderson, T. J., and C. Roper. 2005. The origins and spread of antimalarial drug resistance: lessons for policy makers. Acta Trop. 94:269-280. [PubMed] 3. Anderson, T. J., X. Z. Su, M. Bockarie, M. Lagog, and K. P. Day. 1999. Twelve microsatellite markers for characterization of Plasmodium falciparum from finger-prick blood samples. Parasitology 119(Pt 2):113-125. [PubMed] 4. Barnes, D. A., S. J. Foote, D. Galatis, D. J. Kemp, and A. F. Cowman. 1992. Selection for high-level chloroquine resistance results in deamplification of the pfmdr1 gene and increased sensitivity to mefloquine in Plasmodium falciparum. EMBO J. 11:3067-3075. [PubMed] 5. Bland, J. M., and D. G. Altman. 1986. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet i:307-310. 6. Brockman, A., R. E. Paul, T. J. Anderson, I. Hackford, L. Phaiphun, S. Looareesuwan, F. Nosten, and K. P. Day. 1999. Application of genetic markers to the identification of recrudescent Plasmodium falciparum infections on the northwestern border of Thailand. Am. J. Trop. Med. Hyg. 60:14-21. [PubMed] 7. Chaiyaroj, S. C., A. Buranakiti, P. Angkasekwinai, S. Looressuwan, and A. F. Cowman. 1999. Analysis of mefloquine resistance and amplification of pfmdr1 in multidrug-resistant Plasmodium falciparum isolates from Thailand. Am. J. Trop. Med. Hyg. 61:780-783. [PubMed] 8. Cowman, A. F., D. Galatis, and J. K. Thompson. 1994. Selection for mefloquine resistance in Plasmodium falciparum is linked to amplification of the pfmdr1 gene and cross-resistance to halofantrine and quinine. Proc. Natl. Acad. Sci. USA 91:1143-1147. [PubMed] 9. Cowman, A. F., and S. Karcz. 1993. Drug resistance and the P-glycoprotein homologues of Plasmodium falciparum. Semin. Cell Biol. 4:29-35. [PubMed] 10. Duraisingh, M. T., P. Jones, I. Sambou, L. von Seidlein, M. Pinder, and D. C. Warhurst. 2000. The tyrosine-86 allele of the pfmdr1 gene of Plasmodium falciparum is associated with increased sensitivity to the anti-malarials mefloquine and artemisinin. Mol. Biochem. Parasitol. 108:13-23. [PubMed] 11. Kuo, M. T., R. C. Vyas, L. X. Jiang, and W. N. Hittelman. 1994. Chromosome breakage at a major fragile site associated with P-glycoprotein gene amplification in multidrug-resistant CHO cells. Mol. Cell. Biol. 14:5202-5211. [PubMed] 12. Lim, A. S., D. Galatis, and A. F. Cowman. 1996. Plasmodium falciparum: amplification and overexpression of pfmdr1 is not necessary for increased mefloquine resistance. Exp. Parasitol. 83:295-303. [PubMed] 13. Nair, S., D. Nash, D. Sudimack, A. Jaidee, M. Barends, A. C. Uhlemann, S. Krishna, F. Nosten, and T. J. Anderson. 2007. Recurrent gene amplification and soft selective sweeps during evolution of multidrug resistance in malaria parasites. Mol. Biol. Evol. 24:562-573. [PubMed] 14. Nosten, F., F. ter Kuile, T. Chongsuphajaisiddhi, C. Luxemburger, H. K. Webster, M. Edstein, L. Phaipun, K. L. Thew, and N. J. White. 1991. Mefloquine-resistant falciparum malaria on the Thai-Burmese border. Lancet 337:1140-1143. [PubMed] 15. Paget-McNicol, S., and A. Saul. 2001. Mutation rates in the dihydrofolate reductase gene of Plasmodium falciparum. Parasitology 122:497-505. [PubMed] 16. Peel, S. A., P. Bright, B. Yount, J. Handy, and R. S. Baric. 1994. A strong association between mefloquine and halofantrine resistance and amplification, overexpression, and mutation in the P-glycoprotein gene homolog (pfmdr) of Plasmodium falciparum in vitro. Am. J. Trop. Med. Hyg. 51:648-658. [PubMed] 17. Peel, S. A., S. C. Merritt, J. Handy, and R. S. Baric. 1993. Derivation of highly mefloquine-resistant lines from Plasmodium falciparum in vitro. Am. J. Trop. Med. Hyg. 48:385-397. [PubMed] 18. Price, R. N., C. Cassar, A. Brockman, M. Duraisingh, M. van Vugt, N. J. White, F. Nosten, and S. Krishna. 1999. The pfmdr1 gene is associated with a multidrug-resistant phenotype in Plasmodium falciparum from the western border of Thailand. Antimicrob. Agents Chemother. 43:2943-2949. [PubMed] 19. Price, R. N., A. C. Uhlemann, A. Brockman, R. McGready, E. Ashley, L. Phaipun, R. Patel, K. Laing, S. Looareesuwan, N. J. White, F. Nosten, and S. Krishna. 2004. Mefloquine resistance in Plasmodium falciparum and increased pfmdr1 gene copy number. Lancet 364:438-447. [PubMed] 20. Reed, M. B., K. J. Saliba, S. R. Caruana, K. Kirk, and A. F. Cowman. 2000. Pgh1 modulates sensitivity and resistance to multiple antimalarials in Plasmodium falciparum. Nature 403:906-909. [PubMed] 21. Schwobel, B., M. Alifrangis, A. Salanti, and T. Jelinek. 2003. Different mutation patterns of atovaquone resistance to Plasmodium falciparum in vitro and in vivo: rapid detection of codon 268 polymorphisms in the cytochrome b as potential in vivo resistance marker. Malar. J. 2:5. [PubMed] 22. Sidhu, A. B., A. C. Uhlemann, S. G. Valderramos, J. C. Valderramos, S. Krishna, and D. A. Fidock. 2006. Decreasing pfmdr1 copy number in Plasmodium falciparum malaria heightens susceptibility to mefloquine, lumefantrine, halofantrine, quinine, and artemisinin. J. Infect. Dis. 194:528-535. [PubMed] 23. Silamut, K., and N. J. White. 1993. Relation of the stage of parasite development in the peripheral blood to prognosis in severe falciparum malaria. Trans. R. Soc. Trop. Med. Hyg. 87:436-443. [PubMed] 24. Trager, W., and J. B. Jensen. 1976. Human malaria parasites in continuous culture. Science 193:673-675. [PubMed] 25. Uhlemann, A. C., R. McGready, E. A. Ashley, A. Brockman, P. Singhasivanon, S. Krishna, N. J. White, F. Nosten, and R. N. Price. 2007. Intrahost selection of Plasmodium falciparum pfmdr1 alleles after antimalarial treatment on the northwestern border of Thailand. J. Infect. Dis. 195:134-141. [PubMed] 26. Uhlemann, A. C., M. Ramharter, B. Lell, P. G. Kremsner, and S. Krishna. 2005. Amplification of Plasmodium falciparum multidrug resistance gene 1 in isolates from Gabon. J. Infect. Dis. 192:1830-1835. [PubMed] 27. White, N. J., F. Nosten, S. Looareesuwan, W. M. Watkins, K. Marsh, R. W. Snow, G. Kokwaro, J. Ouma, T. T. Hien, M. E. Molyneux, T. E. Taylor, C. I. Newbold, T. K. Ruebush II, M. Danis, B. M. Greenwood, R. M. Anderson, and P. Olliaro. 1999. Averting a malaria disaster. Lancet 353:1965-1967. [PubMed] 28. White, N. J., and W. Pongtavornpinyo. 2003. The de novo selection of drug-resistant malaria parasites. Proc. Biol. Sci. 270:545-554. [PubMed] 29. Wongsrichanalai, C., K. Lin, L. W. Pang, M. A. Faiz, H. Noedl, T. Wimonwattrawatee, A. Laoboonchai, and F. Kawamoto. 2001. In vitro susceptibility of Plasmodium falciparum isolates from Myanmar to antimalarial drugs. Am. J. Trop. Med. Hyg. 65:450-455. [PubMed] 30. World Health Organization. 1990. In vitro microtest (Mark II) for the assessment of the response of Plasmodium falciparum to chloroquine, mefloquine, quinine, sulfadoxine/pyrimethamine and amodiaquine. WHO document MAP/87.2, rev. 1. World Health Organization, Geneva, Switzerland.
|
This article has been 
![[filled square]](/corehtml/pmc/pmcents/x25AA.gif)
, copy number after exposure to mefloquine at 15 ng/ml;
