Plasmodium vivax and Plasmodium falciparum ex vivo susceptibility to anti-malarials and gene characterization in Rondônia, West Amazon, Brazil

Background Chloroquine (CQ), a cost effective antimalarial drug with a relatively good safety profile and therapeutic index, is no longer used by itself to treat patients with Plasmodium falciparum due to CQ-resistant strains. P. vivax, representing over 90% of malaria cases in Brazil, despite reported resistance, is treated with CQ as well as with primaquine to block malaria transmission and avoid late P. vivax malaria relapses. Resistance to CQ and other antimalarial drugs influences malaria control, thus monitoring resistance phenotype by parasite genotyping is helpful in endemic areas. Methods A total of 47 P. vivax and nine P. falciparum fresh isolates were genetically characterized and tested for CQ, mefloquine (MQ) and artesunate (ART) susceptibility in vitro. The genes mdr1 and pfcrt, likely related to CQ resistance, were analyzed in all isolates. Drug susceptibility was determined using short-term parasite cultures of ring stages for 48 to 72 hour and thick blood smears counts. Each parasite isolate was tested with the antimalarials to measure the geometric mean of 50% inhibitory concentration. Results The low numbers of P. falciparum isolates reflect the species prevalence in Brazil; most displayed low sensitivity to CQ (IC50 70 nM). However, CQ resistance was rare among P. vivax isolates (IC50 of 32 nM). The majority of P. vivax and P. falciparum isolates were sensitive to ART and MQ. One hundred percent of P. falciparum isolates carried non-synonymous mutations in the pfmdr1 gene in codons 184, 1042 and 1246, 84% in codons 1034 and none in codon 86, a well-known resistance mutation. For the pfcrt gene, mutations were observed in codons 72 and 76 in all P. falciparum isolates. One P. falciparum isolate from Angola, Africa, showing sensitivity to the antimalarials, presented no mutations. In P. vivax, mutations of pvmdr1 and the multidrug resistance gene 1 marker at codon F976 were absent. Conclusion All P. falciparum Brazilian isolates showed CQ resistance and presented non-synonymous mutations in pfmdr1 and pfcrt. CQ resistant genotypes were not present among P. vivax isolates and the IC50 values were low in all samples of the Brazilian West Amazon.


Background
Malaria, one of the most prevalent parasitic diseases in the world, still causes high morbidity and death, mainly in Plasmodium falciparum-infected, non-treated patients [1]. Plasmodium vivax causes intense morbidity and contributes to significant political, social and economic instability in developing countries of Latin America and Asia [2,3]. CQ is the drug of choice to treat vivax malaria in endemic areas of Brazil and primaquine (PQ) is used to avoid late malaria relapses [3]. The recommended dose for adults is 1500 mg of CQ (daily for three days) and 210 mg of PQ (daily for seven days) [4]. Plasmodium vivax resistance is now widespread and has rendered CQ ineffective in parts of Indonesia and Papua New Guinea [5][6][7]. Low levels of resistance have also been reported in Myanmar, South Korea, Vietnam, India, Turkey, Ethiopia, and in regions of Southern Africa and South America [3,8,9]. The occurrence of severe vivax malaria and patient's deaths has been reported in Brazil [10][11][12] raising the possibility of an association between malaria severity and drug resistance [13]. In areas of CQ resistance, treatment of uncomplicated P. falciparum malaria is carried out with artemisinin-based combination therapy (ACT) [3]. Drugs that complement ACT include lumefantrine, amodiaquine, AQ, MQ, sulphadoxine-pyrimethamine and antibiotics. In Brazil, the first option for falciparum malaria treatment is the combination of artemether (480 mg daily for four days) and lumefantrine (2880 mg daily for four days). PQ (45 mg) is administrated on day one to avoid malaria transmission. These doses are recommended for adults with 50 Kg weight or more [4]. A reduced susceptibility to artemisinin derivatives has been described in P. falciparum-treated patients [14,15].
Increasing evidence of a lower P. vivax susceptibility to CQ in malaria-endemic areas [16] includes the state of Amazonas [17] and is believed to be associated with malaria's clinical severity [18].
Molecular markers associated with CQ resistance are nonsynonymous mutations in the drug/metabolite transporter gene pfcrt (C72S, K76T) and in the multidrug resistance protein 1 gene pfmdr1 (N86Y; Y184F; S1034C; N1042D; D1246Y), described in P. falciparum [19][20][21]. One mutation of the multidrug resistance gene 1 (Y976F) of P. vivax is also associated with parasite susceptibility to CQ [8]. A nonsynonymous mutation of the pvdhps gene at codon 382 (S382C) was recently associated with in vitro susceptibility to CQ [18]. The present study aimed to examine the phenotypic and genotypic chemoresistance profile of P. falciparum and P. vivax to commonly used anti-malarial drugs in a Brazilian malaria-endemic area in the Amazon Region.

Subjects
All isolates were collected between August 2012 and March 2013 from patients recruited at the Centre of Malaria Control (CEPEM) in the city of Porto Velho, state of Rondônia, in the Brazilian Western Amazon, where P. vivax is highly prevalent. Only patients monoinfected with either P. falciparum or P. vivax and with high parasitaemia (between 2,000 and 80,000 parasites/ μl) were recruited. Patients who used any anti-malarial in the previous month and/or presented severe symptoms of malaria were excluded from this work. The study cohort encompassed 56 patients living in this highly endemic area which is close to Bolivia (Figure 1). Forty seven patients were diagnosed with P. vivax and eight with P. falciparum. In addition, one isolate of an individual infected with P. falciparum with imported malaria (from Africa) was also studied. One patient had mixed malaria (P. vivax and P. falciparum) and was not included. From each volunteer, a peripheral venous blood sample (5 ml) was collected by venipuncture in heparin-containing tubes and immediately used for the ex vivo drug susceptibility assay using pre-prepared plates with the diluted anti-malarials, as described below. DNA was also extracted from peripheral venous blood in EDTA containing tubes for parasite genomic analysis.

Ethical approval
This study was approved by the Ethics Committee, Centro de Pesquisas René Rachou-FIOCRUZ (CAAE −03209212. 7.0000.5091). All participants signed a written informed consent before blood collection.
Pre-dosed plates with test and control drugs CQ, MQ and ART were prepared as 10 μg/mL stock solution in dimethyl sulphoxide (DMSO) in 96-well plates (20 μL per well), then diluted two-fold in RPMI, with variable maximum drug concentration according to each previously determined activity (shown in parentheses), i.e. CQ (854 nM), MQ (724 nM), ART (738 nM), lyophilized and stored at 4°C until further use. Each lot of plates was assayed for quality control just before use, based on the profile response to CQ to the P. falciparum laboratory strains W2 and 3D7, for CQ-resistance and -sensitivity, respectively. All plates were prepared in our laboratory, in Belo-Horizonte, transported to the field in dry ice and kept at minus 70°C until use.

Ex vivo drug susceptibility assay
The drug susceptibility of the malaria parasites, from each patient, was measured in vitro. For P. falciparum the method used was described by Rieckmann and Antuñano [22] and for P. vivax by Renapurkar et al. [23] with modifications [24]. White blood cells were removed by filtration in a CF 11 cellulose column as described [25]. Immediately before the ex vivo drug susceptibility assays, packed red blood cells with the parasites (iRBC) were diluted for a 2% hematocrit, using either complete medium RPMI 1640 medium plus 10% AB human serum, with P. falciparum cultures; or the McCoy's 5A medium plus 20% AB human serum, with P. vivax samples. The iRBC (200 μl per well) were distributed in the pre-dosed drug plate. For the maturation of parasites, rings to schizonts, the plates were maintained in candle jars at 37°C as described [26], at different incubation times (48-72 hours). The control wells were iRBCs and cultured with drug free complete medium. The incubation parasite-drug was stopped when 40% of the ring stages reached the schizont stage (at least four distinct nuclei per parasite) in the drug-free control wells (n = 6 per plate). Thick blood films were then made from each well, dried, stained with 5% Giemsa solution for 30 min, and examined microscopically. The number of schizonts per 100 asexual stage parasites was determined for each drug concentration and then normalized by comparing with the schizont number in the drug-free control wells (considered as 100%). The half-maximal drug inhibitory response (IC 50 ) was estimated by curve fitting using software (OriginLab Corporation, Northampton, MA, USA) and comparing with parasite growth in the drug-free controls.

Real time PCR for Plasmodium detection
Genomic DNA of the parasites was extracted using QIAamp DNA kit (QIAGEN, Chatsworth, CA, USA) and then subjected to real time PCR (Applied Biosystems®). The 18S rRNA gene was chosen as target gene since it contains both highly conserved and variable regions (at least five copies of the gene are dispersed on separate chromosomes of the Plasmodium genome). Each 20 μl reaction mix contained 2 μl of sample DNA, 10 μl Fas-tSTART DNA SYBR Green reagent (Roche), 6.5 mM MgCl2 (final concentration), and 0.5 mM concentrations of each primer (5′-TAACGAACGAGATCTTAA-3′and 5′-GTTCCTCTAAGAAGCTTT-3′). The PCR conditions consisted of an initial denaturation at 95°C for 10 min, followed by amplification for 40 cycles of 10 sec at 95°C, 5 sec at 50°C, and 20 sec at 72°C, with fluorescence acquisition at the end of each extension step. Amplification was immediately followed by a melt programme, consisting of 2 min at 95°C, 2 min at 68°C, and a stepwise temperature increase of 0.2°C/sec until 90°C, with fluorescence acquisition at each temperature transition. Fluorescence was analysed using F1/F2 settings, which improved the detection of P. falciparum (a cutoff of 35 cycles was used to define Plasmodium-positive samples). A melting curve analysis was used to determine the species-specific mean melting temperature (Tm) based on values determined from the respective positive controls [27].

Analysis of the crt and mdr1 genes
The pfcrt and mdr1 genes were amplified by PCR with specific primers for each region (

Anti-malarial susceptibility
Ex-vivo drug susceptibility was assessed in field isolates from all 56 patients with a single species infection, either P. vivax (n = 47) or P. falciparum (n = 9); eight of the latter were Brazilian isolates freshly collected in the state of Rondônia, in the Brazilian Western Amazon ( Figure 1) and one was from an individual returning from Angola, Africa, diagnosed and studied in Belo Horizonte, Brazil, outside the endemic malaria region. Adequate parasite growth was achieved in 100% (9/9) of the P. falciparum and in 68% (32/47) of P. vivax isolates. The characteristics of these isolates are summarized in Table 2. The majority of P. falciparum Brazilian isolates studied were considered resistant to CQ (IC 50 70 nM; range, 0.19 to 223 nM); two susceptible showed IC 50 values of 0.19 and 37 nM; two isolates were considered resistant to MQ (IC 50 of 50 and 63 nM); but, all isolates were susceptible to ART, since the highest IC 50 was 5.8 nM.
Among the P. vivax isolates the results of drug susceptibility to CQ suggests no resistance; the median IC 50 was 32 nM; ranging between 3 to 69 nM (Table 3); four isolates (P010, P024, P026, P042) less susceptible to CQ showed IC 50 values higher than 51 nM. In further analysis 32 were examined in the thick smears for the stages of the parasite blood forms; three patients had approximately 30% of parasites in trophozoite stage, which is considered a high rate. And, they present high IC 50 values to ART  and MQ (Table 4). However this number is not enough to show a strong association between stage-specificity and drug activities and future studies are needed. The only African isolate studied was susceptible to the antimalarials tested ( Table 3).

Characterization of the CRT and MDR1 resistance
The parasite profiles of molecular resistance for P. vivax isolates, using the frequencies of SNPs (Single Nucleotide Polymorphism) in pvmdr1 codon 976, were examined in 47 samples ( Table 1). The product was compared with the genomic sequence of Salvador I as the reference wild type found in the Plasmo DB gene bank, and no SNPs were found. For P. falciparum isolates, frequencies of SNPs in pfmdr1 codons 86, 184, 1034, 1042 and 1246 are shown in Table 5. In addition, codons 72 and 76 of pfcrt gene were also evaluated (Table 5). Mutant alleles at positions 184, 1042 and 1246 on pfmdr1 gene were present in 100% of the samples. In codon 1034, the frequency of mutation was 84% and no mutations were found in codon 86.
The pfcrt gene carried SNPs in codons 72 and 76 in all Brazilian isolates. All isolates presented non-synonymous mutations ( Table 5). The P. falciparum isolate from outside Brazil (Angola, Africa) had no SNPs on the pfcrt or pfmdr1 genes.

Discussion
Due to the strong impact of chemo-resistance among the malaria parasites to most drugs used in the control of the disease, monitoring the development of resistant phenotypes and genotyping are priorities wherever endemic malaria is present. The in vitro methods used to examine anti-malarial drug sensitivity provide a profile of Plasmodium sensitivity to a variety of drugs, simultaneously assayed. However, a lower in vitro sensitivity of a parasite isolate does not imply drug-resistance in vivo, as other factors can interfere, which are not determined by in vitro tests. The ex vivo tests provide an outline of resistant-circulating phenotypes for each tested drug, provided that an adequate number of patients are examined in a given area. This was not the case for P. falciparum in our study: only nine patients were evaluated, reflecting the reduced transmission of this parasite species in Brazil [30]. Determining parasite genotype, performed in parallel, provides further information since mutations in the pfcrt gene alter the CQ flux and/or reduce drug efficacy [31]; and provides data for policy makers to decide the best drug to be prescribed in that area. Ideally, monitoring antimalarial chemo-resistance must be continuous since development and spreading of resistance are dynamic events, changing with time and according to human interventions and other factors such as population migration [32]. The antiplasmodial activity and the molecular profile of resistance by anti-malarial standards like CQ, ART and MQ, confirms P. falciparum resistance to CQ only, as demonstrated in ex-vivo tests against P. vivax and P. falciparum isolates from patients with naturally acquired malaria in the state of Rondônia in the Brazilian Western Amazon, which is substantiated by sequencing of genes related to resistance to anti-malarial drugs.
There is evidence of P. vivax resistance to CQ in the state of Amazonas, specifically in the city of Manaus where an increase in the proportion of P. vivax malaria parallels an increase of unusual clinical complications  TGT > TCT Cysteine > Serine related to this species [33]. These authors used an in vivo test to assess the efficacy of standard supervised CQ therapy. Among 109 volunteers with P. vivax who completed the in vivo test, 19 had positive blood smears within the 28-days follow-up (one on day 14, three on day 21, and 15 on day 28). All were then required to undergo an alternative therapy with MQ. In another study performed by the same group, a lower P. vivax susceptibility to CQ through the attainment of IC 50 by ELISA assay or using traditional methods [18,34] was examined; they demonstrated that drug-resistance was related to the presence of non-synonymous mutation at pvdhps, pvcrt and pvmdr1. All P. vivax isolates presently studied in the Brazilian Western Amazon were sensitive to CQ in ex vivo assays. Although the threshold of IC 50 to define a sample as resistant to CQ is not well established for P. vivax, it has been proposed that the same threshold used for P. falciparum should be used, with 100 nM threshold of CQ [35].
The IC 50 for ART and MQ in P. vivax were also examined and they were higher when compared with the values found in P. falciparum. It can be due to differences in the stage-specific activities of CQ, MQ and AQ in P. vivax, demonstrated here and by Marfut et al. [28]. It may be interesting to compare the susceptibility of P. vivax in strains from other regions of the world. This issue remains to be further studied using strains from different places.
Only one point mutation for CQ was studied for P. vivax in the pvmdr1 gene codon 976, but other genes were associated with CQ resistance in the Brazilian Amazon, e.g., in the pvdhps gene in codon 382 (S → C) [18].
Considering that the severity of P. vivax malaria in the state of Amazonas has been attributed to CQ resistance and to the increased levels of pvmdr1 and pvcrt-o compared to the levels expressed by parasites from patients with mild symptoms [36], these genes copy number could also be evaluated. In addition, the mdr1 copy number is strongly associated with recrudescence after artesunatemefloquine administration, and could be used as a surveillance tool for artesunate-mefloquine resistance, as reported in patients in Cambodia [37,38].
In conclusion, in West Amazon, most P. falciparum isolates were CQ resistance, a data confirmed by parasite genotyping. No mutations were found for P. vivax in the region supporting the lower prevalence of this strain in Brazil.