ABSTRACT
Quinoline resistance in malaria is frequently compared with P-glycoprotein-mediated multidrug resistance (mdr) in mammalian cells. We have previously reported that nonylphenolethoxylates, such as NP30, are potential Plasmodium falciparum P-glycoprotein substrates and drug efflux inhibitors. We used in vitro assays to compare the ability of verapamil and NP30 to sensitize two parasite isolates to four quinolines: chloroquine (CQ), mefloquine (MF), quinine (QN), and quinidine (QD). NP30 was able to sensitize (reversal, >80%) P. falciparum to MF, QN, QD, and, to a lesser extent, CQ. The presence of 2 μM verapamil had no effect on mefloquine resistance; however, the presence of verapamil modulated the activities of QN and QD in a manner parallel to that observed for CQ. Genetic analysis of putative quinoline resistance genes did not suggest an association between known point mutations in pfcrt and pfmdr1 and NP30 sensitization activity. We conclude that the sensitization action of NP30 is distinct both phenotypically and genotypically from that of verapamil.
The incidence of malaria is increasing, with an estimated 300 million to 500 million cases of malaria and 3 million deaths from malaria annually (30). Despite optimistic predictions of the eventual control and eradication of malaria made in the past, the spread of drug-resistant parasites, particularly those resistant to chloroquine (CQ), has led to a resurgence of the disease.
New antimalarials have been introduced in recent years (11, 22); however, they are significantly more expensive than CQ and pyrimethamine-sulfadoxine (Fansidar), the treatment options most commonly used in developing nations. The appearance of malaria resistant to CQ was first reported in the late 1950s in South America (33) and Southeast Asia (15) and has since spread to nearly all regions where malaria is endemic (18). The emergence of CQ-resistant falciparum malaria, particularly in Africa, has been a global health disaster (18) and is directly responsible for the increased number of deaths from Plasmodium falciparum infections (30).
The mechanism by which quinoline-based antimalarials inhibit the growth of blood-stage parasites remains controversial; however, they appear to interfere with the production of the malaria pigment hemazoin (for a review, see reference 28). Quinoline uptake into the food vacuole, where polymerization occurs, may be a passive process (27); and drug-resistant parasites appear to have developed or adapted mechanisms to remove quinolines from this environment (32). Quinoline resistance in P. falciparum is frequently compared to multidrug resistance in mammalian cells, a comparison that is supported by the observation that CQ resistance in P. falciparum can be reversed by known P-glycoprotein substrates such as verapamil (19), chlorpromazine (3), promethazine (21), chlorpheniramine (4), and citalopram (12). P. falciparum contains several multidrug resistance-like proteins, including Pgh1, which is encoded by pfmdr1 (14). Site-directed mutagenesis of Pgh1 has demonstrated that Pgh1 is associated with resistance to mefloquine (MF), quinine (QN), quinidine (QD), and halofantrine (26). On the other hand, mutations in Pgh1 were not directly responsible for CQ resistance. A gene on chromosome 7, pfcrt (9), which encodes a vacuolar transport protein, has been linked to CQ resistance. Parasite transfection studies have demonstrated that mutations in pfcrt are sufficient to confer CQ resistance in vitro (13). However, in clinical studies pfcrt mutations appear to be necessary, but not sufficient, to predict CQ treatment outcomes in vivo (10, 24). Other host and, possibly, parasite factors contribute to parasite clearance following CQ therapy. These include host immune response (premunition) and a possible role for pfmdr1 (26) and other gene products in mediating higher levels of CQ resistance. Furthermore, the observation that the reversal of CQ resistance by verapamil is not uniformly effective (31) is consistent with the hypothesis that multiple gene products modulate the responsiveness of malaria parasites to CQ.
NP30, an uncharged polyethoxylated nonylphenol surfactant containing an average of 30 ethoxylate units (Fig. 1), appears to belong to a novel class of resistance reversal agents since its structure does not contain a nitrogen atom and the surfactant has no ionizable groups (7, 8). We therefore wished to determine if NP30 had quinoline sensitization properties that were similar to or different from those of verapamil, the agent most frequently used to examine CQ resistance in vitro.
NP30. NP30 consists of a nonylphenol group onto which an average of 30 ethoxylate residues have been polymerized.
MATERIALS AND METHODS
P. falciparum cultures.Human malaria parasites were grown in type A-positive blood obtained by venipuncture of volunteers. Cultures were maintained by the method of Trager and Jensen (29) by using RPMI 1640 supplemented with 10% human serum (either type A or AB [Wisent Inc., St. Bruno, Quebec, Canada]) and 50 μM hypoxanthine (Gibco). Patient isolates were obtained from pretreatment blood samples from patients enrolled in ethically approved studies at the Tropical Disease Unit, University of Toronto (16, 34). The resistance genotypes (point mutations in pfcrt and pfmdr1) of these isolates were determined as described previously (1, 2, 6, 23, 24).
In vitro drug susceptibility testing.In vitro drug susceptibility testing was performed by using a lactate dehydrogenase assay modified to be specific for the presence of P. falciparum lactate dehydrogenase (5, 17, 25). The concentration of drug that resulted in a 50% inhibition of viability of the parasite cultures (IC50) was determined by a nonlinear regression analysis using the computer program Sigma Plot 2000 (Jandel Scientific). The IC50s represent the means of four determinations.
NP30.Samples of NP30 were the kind gift of Union Carbide and were made up as 1% (wt/vol) stock solutions in water after extensive drying by lyophilization to obtain accurate weight measurements. Checkerboard assays (two-dimensional arrays of the surfactant concentration versus the quinoline concentration) were initially undertaken to determine the IC50s of NP30 and the quinolines alone and in combination. A single concentration of NP30 (0.003% [wt/vol]; 21 μM) was used in assays for comparison of the activity of NP30 with that of verapamil on the basis of the fact that this concentration of NP30 offered a favorable degree of sensitization with minimal direct antiparasitic effect.
RESULTS
We have previously determined that NP30 both has anti-P. falciparum effects and can reverse CQ resistance in some isolates (8). To test the hypothesis that NP30 and verapamil interacted with a common site, we determined whether the presence of 2 μM verapamil enhanced the anti-P. falciparum effects of NP30. The presence of verapamil did not modify the anti-P. falciparum activity of NP30 (Fig. 2), suggesting that verapamil and NP30 have distinct sites of action. Examination of the capacity of verapamil to sensitize two P. falciparum isolates, designated isolate 1 (which originated in West Africa) and isolate 2 (which originated in the Indian subcontinent), to CQ indicated that a high degree of sensitization could be obtained in isolate 1 but that only a limited degree of sensitization could be obtained in isolate 2 (Fig. 3). When 21 μM NP30 was present, the IC50 of CQ for isolate 1 was nearly identical to that for the untreated sample; however, the IC50 of CQ for isolate 2 was reduced to a value less than that for the verapamil-treated sample. This finding suggests that fully verapamil-sensitive CQ resistance is independent of the site of action of NP30.
Effect of verapamil on the anti-Plasmodium action of NP30. The IC50 of NP30 for Plasmodium cultures was determined in the absence (closed bars) and presence (open bars) of 2 μM verapamil. The values represent the means ± standard errors of the means of three independent assays.
Effect of verapamil or NP30 on the IC50s of four quinolines in two P. falciparum patient isolates. The IC50s of CQ, MF, QN, and QD were determined in the presence of drug alone, 2 μM verapamil, or 21 μM NP30 for two isolates. The values represent the means ± standard errors of the means of four independent assays. Asterisks indicate that the values were significantly different (P < 0.05) from the control values.
Previous work has indicated that some forms of quinoline resistance may be unique to CQ and has suggested that resistance to MF, QN, and QD may result from mutations in a common efflux protein, such as Pgh1 (26). We therefore determined what effect the presence of verapamil and NP30 had on the IC50s of MF, QN, and QD in the two isolates. The IC50s of MF for isolates 1 and 2 decreased significantly in the presence of NP30; however, verapamil had no effect (Fig. 3). NP30 resulted in a marked reversal of QN and QD resistance; however, the effect of verapamil in the presence of QN and QD was variable and not as dramatic as that of NP30 (Fig. 3). We therefore conclude that (i) NP30 and verapamil have dissimilar sensitization properties, (ii) MF resistance results from a verapamil-insensitive but NP30-sensitive mechanism, and (iii) QN and QD resistance can be reversed by either verapamil or NP30.
The use of two isolates with different CQ resistance phenotypes suggested that isolate 1 and isolate 2 might not share common point mutations in putative quinoline resistance genes. We therefore determined if previously characterized point mutations in the pfcrt and pfmdr genes were present (Table 1). The point mutation patterns seen in the pfcrt and pfmdr genes were not able to differentiate between the two distinct CQ resistance phenotypes (NP30 sensitive versus NP30 insensitive) observed.
Genotype sequences of putative drug resistance genes in P. falciparum patient isolatesa
DISCUSSION
Quinoline resistance in the erythrocytic stages of P. falciparum is frequently compared to multidrug resistance in mammalian cells, in part because of the initial observation that chloroquine resistance could be reversed by verapamil. NP30 is also able to reverse some forms of quinoline resistance in P. falciparum cultures; however, the results obtained in this study suggest that its resistance reversal properties are different from those of verapamil. This is seen most clearly in the case of isolate 1, in which CQ resistance could be reversed only by verapamil, while MF resistance could be reversed only by NP30. While this suggests that MF and CQ resistance is due to the expression of independent NP30-inhibitable and verapamil-inhibitable mechanisms, the results observed for QN and QD do not support this hypothesis. The QN and QD IC50s for isolate 1 can be dramatically decreased by the presence of either NP30 or verapamil. This suggests that both the verapamil-sensitive and the NP30-sensitive resistance elements are a necessary part of the QN and QD resistance process, since disruption of either makes the parasites sensitive to the drug's presence. If the two resistance mechanisms ran in a parallel fashion, we would expect that the presence of either agent alone would have little effect because of redundancy. Both agents are effective; therefore, it appears that they have a serial arrangement (e.g., they may facilitate transport between different compartments within the parasite). The same relationship between QN and QD resistance and the MF and CQ resistance patterns was also seen in isolate 2, in which the degree of reversal seen for QN and QD in isolate 2 supports the hypothesis that QD and QN efflux requires both the CQ and MF efflux mechanisms to be functioning simultaneously.
While the results obtained with CQ and MF for isolate 1 strongly suggest that the actions of verapamil and NP30 can be independent of each other, the results obtained with CQ for isolate 2 suggest that there can also be some overlap in their actions. The presence of either verapamil or NP30 results in a moderate, but significant, drop in the CQ IC50 for isolate 2. When both verapamil and NP30 were added to the same assay, their effects did not appear to be additive (data not shown), suggesting that this result is not due to the presence of a mixed population of verapamil- and NP30-sensitive parasites. CQ resistance can be either fully or partially sensitive to the presence of verapamil, and recent reports suggest that this behavior may be associated with the presence of specific sequences in the pfcrt gene (20). It would be of interest to know if these pfcrt polymorphisms were related to or were independent of the NP30 CQ sensitization potential.
It would appear that in our original report of the sensitization properties of NP30 (8) we used a group of parasite isolates that were sensitized by the presence of NP30. Our present findings indicate that there is a subset of P. falciparum isolates that display CQ resistance that is insensitive to NP30. The difference between the NP30 sensitivities of these isolates cannot be explained by previously reported point mutations in pfcrt and pfmdr1 associated with quinoline resistance, since these mutations were present in the two isolates used in this study. Further work is required to characterize the interaction of both verapamil and NP30 with resistance elements in P. falciparum.
NP30 has both anti-P. falciparum and sensitization abilities. The concentration of NP30 chosen for the evaluation whose results are presented in Fig. 2 (21 μM) was chosen because it permitted us to observe a high degree of sensitization, despite the presence of the direct anti-P. falciparum action of the agent. Higher levels of NP30 in the assays may result in a higher degree of CQ resistance reversal in isolate 2 than that shown in Fig. 2; however, it is of interest that QN and QD resistance reversal was nearly complete at this concentration. This implies that CQ resistance in this isolate is sensitive to the presence of NP30; however, the concentration of NP30 required for CQ resistance reversal may be higher than that required for QN and QD resistance reversal.
The finding that NP30 can be used to reverse MF, QN, and QD resistance in vitro may prove to be clinically useful, since this agent does not have the pharmacological effects on the human host that previously described agents, such as verapamil and promethazine, have. Furthermore, the low cost of this material may make its use in developing countries an economically viable proposition.
ACKNOWLEDGMENTS
MF was a gift from Hoffmann-La Roche. This work was supported by an operating grant from CIHR (grant MT-13721) and a Career Scientist Award from the Ontario Ministry of Health (to K.C.K.).
FOOTNOTES
- Received 18 July 2002.
- Returned for modification 20 November 2002.
- Accepted 28 April 2003.
- American Society for Microbiology
