In vitro susceptibility to quinine and microsatellite variations of the Plasmodium falciparum Na+/H+ exchanger (Pfnhe-1) gene: the absence of association in clinical isolates from the Republic of Congo

Background Quinine is still recommended as an effective therapy for severe cases of Plasmodium falciparum malaria, but the parasite has developed resistance to the drug in some cases. Investigations into the genetic basis for quinine resistance (QNR) suggest that QNR is complex and involves several genes, with either an additive or a pairwise effect. The results obtained when assessing one of these genes, the plasmodial Na+/H+ exchanger, Pfnhe-1, were found to depend upon the geographic origin of the parasite strain. Most of the associations identified have been made in Asian strains; in contrast, in African strains, the influence of Pfnhe on QNR is not apparent. However, a recent study carried out in Kenya did show a significant association between a Pfnhe polymorphism and QNR. As genetic differences may exist across the African continent, more field data are needed to determine if this association exists in other African regions. In the present study, association between Pfnhe and QNR is investigated in a series of isolates from central Africa. Methods The sequence analysis of the polymorphisms at the Pfnhe-1 ms4760 microsatellite and the evaluation of in vitro quinine susceptibility (by isotopic assay) were conducted in 74 P. falciparum isolates from the Republic of Congo. Results Polymorphisms in the number of DNNND or NHNDNHNNDDD repeats in the Pfnhe-1 ms4760 microsatellite were not associated with quinine susceptibility. Conclusions The polymorphism in the microsatellite ms4760 in Pfnhe-1 that cannot be used to monitor quinine response in the regions of the Republic of Congo, where the isolates came from. This finding suggests that there exists a genetic background associated with geographic area for the association that will prevent the use of Pfnhe as a molecular marker for QNR. The contribution of Pfnhe to the in vitro response to quinine remains to be assessed in other regions, including in countries with different levels of drug pressure.


Background
Quinine (QN) remains a first-line drug for the treatment of severe malaria that is still used as a second-line therapy for uncomplicated malaria in many countries [1]. Despite the efficacy of QN against chloroquine-resistant Plasmodium falciparum isolates, reports of QN resistance (QNR) have been increasing. In the 1980s, the frequency of clinical failures increased in Southeast Asia and Africa [2][3][4]. The mechanism of QNR is complex, multigenic, and not well understood. QNR has been associated with mutations in both the P. falciparum multidrug resistance gene mdr1 (Pfmdr1) [5] and the chloroquine resistance transporter gene Pfcrt [6]. Other genetic polymorphisms, such as mutations in the resistance protein gene Pfmrp [7] and variations of microsatellite length in the sodium/hydrogen exchanger gene Pfnhe-1 [8], may contribute to QNR. Using quantitative trait loci (QTL) on the genetic cross of HB3 and Dd2 strains, Ferdig et al identified genes on chromosome 5, encoding Pfmdr1, on chromosome 7, encoding Pfcrt and on chromosome 13, encoding Pfnhe-1, which were associated with QN reduced susceptibility [8]. Sequences of Pfnhe-1 showed multiple and complex variations such as point polymorphisms at three separate codons (790, 894 and 950) and as microsatellite variations in three different repeat sequences (msR1, ms3580 and ms4760). However, the three point polymorphisms and microsatellite polymorphisms msR1 and ms3580 showed no significant association with QN susceptibility. The investigations of the microsatellite ms4760 polymorphisms in culture-adapted isolates from around the world showed that an increased number of the amino acid motif DNNND was associated with a decreased QN susceptibility, and that an increased number of NHNDNHNNDDD motifs was associated with an increased QN susceptibility [8,9]. The association of two repeats with a high QN inhibitory concentration of 50% (IC 50 ) was found in a case of QN clinical failure in a traveller from Senegal [10]. In contrast, a recent multivariate analysis performed on 83 freshly collected clinical isolates from Madagascar and 13 African countries did not find an association between QN susceptibility and Pfnhe-1 microsatellite polymorphisms [11]. A similar absence of association was observed in 91 isolates from various countries on different continents (Pelleau S. et al. submitted). Given that the influence of Pfnhe on QN susceptibility has been shown to be strain-dependent, these apparently conflicting results may be explained, in part, by differences in the geographic origin of the parasites analysed, as their local selection history and genetic background varies.
Thus, further epidemiological investigation is required to determine the context in which Pfnhe can be used as a molecular marker of QNR. In this work, ms4760 polymorphism was analized and its association tested with in vitro susceptibility to QN in African isolates from a single geographical region, Pointe-Noire in the Republic of Congo.

Methods
Reference culture-adapted strains and clinical isolates of Plasmodium falciparum Between March 2005 and January 2006, 74 P. falciparum clinical isolates were collected from patients with uncomplicated malaria at the Medical Service of Total Exploration et Production, Pointe-Noire (Republic of Congo) [12]. Two cloned strains of P. falciparum were used for quality control (3D7 Africa and W2 Indochina). These clones were obtained from MR4-ATCC (Manassas, VA, USA). They were maintained in culture in RPMI 1640 (Invitrogen, Paisley, UK) that was supplemented with 10% human serum (Abcys S.A. Paris, France) and buffered with 25 mM HEPES and 25 mM NaHCO 3 . The parasites were grown in type A+ human blood under controlled atmospheric conditions (10% O 2 , 5% CO 2 , and 85% N 2 ) at 37°C with a humidity of 95%. They were synchronized twice with sorbitol before use [13]. Clonality was verified using PCR genotyping of the polymorphic genetic markers, msp1 and msp2, and the microsatellite loci [14,15].

In vitro assay
The QN was purchased from Sigma (Saint Louis, MO, USA). The QN was first dissolved in methanol and then diluted in water to obtain final concentrations ranging from 5 to 3200 nM.
For the in vitro isotopic microtest, 200 μl of the suspension of synchronous parasitized red blood cells (final parasitaemia, 0.5%; final haematocrit, 1.5%) per well were plated in 96-well plates that contained serial QN concentrations. Parasite growth was assessed by adding 1 μCi of tritiated hypoxanthine with a specific activity of 14.1 Ci/mmol (Perkin-Elmer, Courtaboeuf, France) to each well at time zero. The plates were then incubated for 48 h under controlled atmospheric conditions. Immediately after incubation, the plates were frozen and then thawed to lyse the erythrocytes. The contents of each well were collected on standard filter microplates (Unifilter GF/B; Perkin-Elmer) and washed using a cell harvester (Filter-Mate Cell Harvester; Perkin-Elmer). The filter microplates were dried and 25 μl of a scintillation cocktail (Microscint O; Perkin-Elmer) was added to each well. The radioactivity incorporated by the parasites was then measured using a scintillation counter (Top Count; Perkin-Elmer). The drug concentration that could inhibit 50% of the parasite growth (IC 50 ) was calculated assuming that it corresponded to the drug concentration at which the incorporation of tritiated hypoxanthine by the parasite was 50% of that incorporated in the drug-free, control wells. IC 50 values were determined using a non-linear regression analysis of log-based dose-response curves (Riasmart, Packard, Meriden, USA).

Statistical analysis
Data were analysed using R software (version 2.10.1) and GraphPad Prism (version 5.01). Differences between the median QN IC 50 values of isolates harbouring one, two or three DNNND repeats were tested using the Kruskal-Wallis test. The median QN IC 50 values of isolates with one or two NHNDNHNNDDD repeats were compared using the Mann Whitney test.

Results
The mean QN IC 50 of the 74 isolates was 381.9 nM, CI95% [330.7-433.1] (min-max: 36-1097 nM). Four isolates (5.4%) had a QN IC 50 > 800 nM ( Table 1) that corresponds to the in vitro QNR threshold. Twenty-seven different Pfnhe-1 genotypes were observed among the 74 isolates, and eighteen were identified as new genotypes (GenBank accession numbers FJ392810 to FJ392827). The DNNND repeats varied from 1 to 3. There was no statistically significant association (p = 0.84) between the median QN IC 50 value and the number of DNNND repeats (medians at 335, 316 and 330 nM, for isolates harbouring one, two or three repeats, respectively) ( Figure 1). No statistically significant association was observed (p = 0.92) between the median QN IC 50 value of isolates and the number of NHNDNHNNDDD repeats (medians at 328 and 329 nM, for isolates with one and two repeats, respectively).

Discussion
QN has been used to treat malaria for more than 350 years in Africa, with little emergence and spread of resistance. QN remains the first line anti-malarial drug for the treatment of complicated malaria in Europe and Africa. However, despite the efficacy of QN against chloroquine-resistant strains, the emergence of QN resistance has been observed. The first cases of QN clinical failure were observed in Brazil and Asia in the 1960s; then in the 1980s, clinical failures became more frequent in Southeast Asia, South America and Africa. However, QN resistance is not yet a significant problem in Africa, and QN remains the first-line drug to treat severe malaria and a second-line therapy for uncomplicated malaria in some areas of Africa.
Although some reports of QN treatment failure exist, it is difficult to confirm QN resistance because of the short elimination half-life of the drug, the requirement to administer it three times a day for at least five days, drug intolerance that often leads to poor compliance and a lack of reliable data on the correlation between QN IC 50 and clinical failure. Maximizing the efficacy and longevity of QN as a tool for the control of malaria will depend critically on the pursuit of intensive research toward the identification of in vitro markers of QNR and the implementation of in vitro and in vivo surveillance programs, such as those championed by the Worldwide Antimalarial Resistance Network [16,17]. Specifically, there is a need to identify molecular markers that effectively predict QN resistance and enable the active surveillance of temporal trends in parasite susceptibility [18]. The present study aimed to test the association between the Pfnhe polymorphism and QN susceptibility in clinical isolates from Pointe-Noire in Congo to assess the validity of Pfnhe as a molecular marker of QN susceptibility in this region.
The level of susceptibility to QN remained high, with a mean QN IC 50 of 382 nM and with 4/74 isolates (5.4%) that exceed the QNR threshold of 800 nM [19]. The threshold at which QN in vitro susceptibility is considered to be compromised is arbitrary, and different studies have set the following thresholds: 300 nM [20], 500 nM [21] or 800 nM. If a QNR threshold of 500 nM is applied here, 19/74 isolates exceed this value. These results suggest that QN treatment is still effective against the chloroquine-resistant parasites that are found in this area, where the prevalence of chloroquineresistant parasites has reached 75% [12].
One explanation for this could be variation in genetic background. A specific genetic background observed in Asia may allow the observed contribution of Pfnhe polymorphism on QN in vitro susceptibility. This explanation is consistent with: i) the first evidence of Pfnhe-QNR association from QTL analysis using Americano-Asian cross strains [8], ii) that most associations identified have been shown among Asian strains [8,9,22] and iii) that at least 5 genes spanning the P. falciparum genome influence the QNR in vitro phenotype with an additive effect or with pairwise interactions [8].
Additionally, these genetic dissimilarities between African and Asian plasmodial populations may be accentuated by different local selection histories. The best-documented genomic modifications by a local selection process relate to drug pressure. Several studies have shown that drug pressure may involve extended linkage disequilibrium around a drug resistance associated gene [23]. This is characterized by a strong genetic diversity loss, called a selective sweep. This process stretch may be modified by: i) drug use and ii) malaria transmission level. A reduced drug use and higher malaria transmission level in Africa would be consistent with the lower selective sweep and an absence of linkage disequilibrium between Pfnhe and other cooperative drug response genes or selected compensatory mutations. For example, QTL analysis on chromosome 13 located 60 genes of unknown function as being close to Pfnhe [8]. They would be more or less linked depending on drug pressure and malaria transmission level if QN pressure selects at least one of them. In Kenya there is an association between two DNNND repeats in the ms4760 Pfnhe microsatellite and a reduced susceptibility to quinine in 29 P. falciparum isolates [24]. This is consistent with the historic precedent of the spread of drug resistance around the world. The emergence of chloroquine resistance in Asia was followed by an initial introduction into East Africa and spreading across the African continent. Geographic proximity may explain plasmodial population migration. Moreover, East African plasmodial populations may exhibit genetic dissimilarities to other African populations [14].
As the QN response is controlled by multiple genes with complex interactions, one can expect: i) higher sensitivity to genetic background than if the response was controlled by only one gene, ii) higher sensitivity to parameters that may cause linkage disequilibrium between genes and iii) various combinations of gene polymorphisms may result in similar QNR phenotypes.

Conclusions
In summary, although studies have demonstrated that Pfnhe-1 contributes to QNR [24], the present study and another recent study [11] did not show an association between Pfnhe-1 polymorphism and QNR. Currently, Pfnhe-1 cannot be used as a QNR molecular marker when evaluating field isolates, especially in Africa in the context of a low level of quinine selection pressure. Further studies, using more parasite samples from South East Asia with reduced susceptibility to QN, are required to confirm or reject the use of the Pfnhe-1 gene as a QNR marker in this geographic region.