INTRODUCTION
The morbidity and mortality rates due to malaria have been declining gradually in recent years in Thailand, but multidrug resistant Plasmodium falciparum remains one of the major health problems in Thailand [1]. Artesunate-mefloquine combination was an artemisinin based combination therapy (ACT) that has been used as the first line treatment of uncomplicated falciparum malaria since 1995 in Thailand. Declining efficacy of this treatment regimen has been observed, especially along the Thai-Myanmar and Thai-Cambodia borders [1,2]. Artemether-lumefantrine (Coartem®) is an alternative fixed-dose ACT. Its efficacy was studied in 2 provinces along the Thai-Myanmar border, i.e., Tak and Ranong provinces in 2012 and 2013 [3]. Genetic characteristics of P. falciparum may play a role in the treatment outcome of malaria infection. The merozoite surface protein 1 (msp-1) and msp-2 are abundant surface proteins on the blood stage of P. falciparum. They are thought to play a role in erythrocyte invasion [4,5]. Four allelic families had been identified in block 2 of msp-1 gene; K1, MAD20, RO33, and MR [6,7] and 2 allelic families in msp-2 gene; FC27 and 3D7 [5]. These 2 markers and glutamate-rich protein (glurp) have been extensively used as markers to investigate the genetic diversity, multiplicity of infection, level of malaria transmission, as well as relationship with immunity against malaria [8,9,10]. Genetic diversity of msp-1 and msp-2 has been associated with clinical severity of malaria [11,12]. The 2 loci and the glurp loci have also been introduced as a discriminatory tool to distinguish new from recrudescent infections [13,14]. This approach has been used in the antimalarial drug efficacy monitoring program of Thailand to confirm the presence of drug resistant P. falciparum. No studies have been undertaken in Thailand to compare the genetic diversity of these markers among patients with different treatment outcomes.
The aim of this study was to examine whether the treatment outcome based on ACT was associated with multiplicity of infection and/or any particular allelic family of msp-1, msp-2, and glurp genotypes of P. falciparum. Allelic polymorphisms within these genes were analyzed in P. falciparum isolates collected from 3 groups of patients; day 0 of treatment failures, day 0 of recrudescences, and day 0 of successful treatments.
MATERIALS AND METHODS
Study site and samples
A total of 156 P. falciparum isolates were selected from the archival blood samples from uncomplicated falciparum malaria patients (adults) enrolled into the therapeutic efficacy monitoring of artesunate plus mefloquine combination (AM) during 2009 and 2013 and artemether-lumefantrine (Coartem®) during 2012 and 2013. The dosage of AM regimen was 12 mg/kg body weight of artesunate plus 25 mg/kg of mefloquine plus 30 mg of primaquine divided and given over 3 days. Coartem® was given as a total dose of artemether 2 mg/kg and lumefantrine 12 mg/kg twice a day for 3 days. Study patients were classified as treatment failure or successful treatment groups according to the WHO guideline. Treatment failure was patients having a parasite density of day 1>day 0 parasite density or day 3 parasite density >25% of day 0 parasite density or recurrent parasitemia after day 4 of follow-up examination. Patients with no recurrent parasitemia until day 42 of follow-up were classified as successful treatment group. The efficacies of the AM combination were 89.6%, 95.6%, and 78.4% in 2009, 2010, and 2013, respectively. The efficacy of Coartem® was 92% [3].
One hundred archival samples were selected from 50 treatment failure patients, 50 samples on day 0 (prior to treatment), and 50 samples on recrudescent day. Nineteen, 8, and 17 samples were from the AM treatment failure patients in 2009, 2010, and 2013, respectively. Six samples were from the Coartem® treatment failure patients. Fifty-six archival samples from successful treatment patients collected on day 0 prior to treatment were also selected for comparison. The inclusion criteria were archival samples with P. falciparum mono-infection and having parasitemia of 500-100,000 asexual forms per microliter. Ethical approval was obtained from the Research Ethics Committee of the Department of Disease Control, Ministry of Public Health, Thailand. The consent forms were obtained from patients on enrollment to the efficacy monitoring, including permission to use the blood for further analysis of parasite genotypes.
Genotyping of P. falciparum msp-1, msp-2, and glurp
Parasite DNA for PCR was extracted from dried blood spot on Whatmann 3 MM filter paper using the QiaAmp DNA mini kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. Polymorphic regions from P. falciparum
msp-1, msp-2, and glurp genes were used as genetic markers for the genotyping of parasite populations. For msp-1 and msp-2, the presence of unique sequences was used to divide the variants into distinct allelic families. The polymorphic repetitive regions block 2 of msp-1, block 3 of msp-2, and RII repeat region of glurp were amplified by nested PCR using the primers and methods as recommended by WHO [15]. In brief, the primary PCR, primer pairs corresponding to conserved sequences spanning the polymorphic regions of each gene were included in separate reactions. The product generated in the primary PCR was used as a template in 6 separate nested PCR, using in each case a specific primer pair in order to determine the presence of allelic variants from the K1, MAD20, and RO33 families of msp-1, the 3D7 and FC27 families of msp-2, and the RII blocks of glurp. Each polymorphic domain was amplified from 5 µl of DNA template in a 20 µl reaction mixture containing 0.12 µM of each primer, 2 mM MgCl2, 125 µM of each dNTP, 0.4 U Amplitaq Gold® 360 Master Mix (Invitrogen), and PCR buffer.
The cycling conditions in the thermocycler (Thermo Scientific Hybaid Px2 Thermal Cycler, Fisher Scientific, Middlesex, UK) for msp-1, msp-2, and glurp primary PCR and glurp nested PCR were as follows: 5 min at 95℃, followed by 30 cycles for 1 min at 94℃, 2 min at 58℃ and 2 min at 72℃, and final extension of 10 min at 72℃. For the msp-1 and msp-2 nested PCR, conditions were as follows: 5 min at 95℃, followed by 30 cycles for 1 min at 94℃, 2 min at 61℃ and 2 min at 72℃, and final extension of 5 min at 72℃. The amplified products were either stored at 4℃ or analyzed immediately by electrophoresis on a 2% molecular grade agarose gel and visualized by UV transilluminator after gel SYBR® safe staining. The sizes of the PCR products were estimated using Imagelab software version 3.0 (Biorad, Hercules, California, USA) with the size computed automatically by the software based on the 100 base pairs DNA ladder calibrator (Real Biotech Corporation, Taipei, Taiwan). Standard 3D7, Dd2, and RO33 clones were used as positive controls for 3D7 and K1, FC27 and MAD20, and RO33 alleles, respectively. The size polymorphism in each allelic family was analyzed. The detection of a single PCR fragment for each locus was classified as an infection with 1 parasite genotype (monoclonal infection) for that locus. Isolates with more than 1 genotype were considered as polyclonal infection [16]. Alleles in each family were considered the same if fragment size was within 20 bp interval for msp1 and msp2 genes [17] and 50 bp interval for glurp gene. In each isolate, the number of genotypes, size of corresponding PCR products, and allelic type (or family) of each gene were described.
Statistical analysis
The msp-1 and msp-2 allele frequencies were calculated as the proportion of alleles found for the allelic family out of the alleles detected in isolates. The proportions of alleles observed at each genetic locus within each treatment outcome group were compared using the Chi-square test statistics. Odds ratios and 95% confidence intervals were computed as measures of effect. Multiplicity of infection (MOI) was defined as the number of parasite genotypes per infection. It was calculated for each gene (msp1, msp2, and glurp) independently. Estimation of the overall MOI of the isolates was also calculated by combining the 3 markers, namely by using the highest number of bands detected in 1 marker. The maximum number of bands detected whatever the locus was considered as the MOI of that infection. Mean MOI was calculated by dividing the total number of fragments detected in msp1, msp2, or glurp by the number of samples positive for the same marker. The median multiplicity of infection was compared among successful treatment, treatment failure, and recrudescence groups using the non-parametric Kruskal Wallis H test. Statistical significance was defined at P≤0.05. All statistical analyses were performed using the SPSS statistical software, version 17.0.
RESULTS
The study population comprised of P. falciparum isolates collected from uncomplicated falciparum malaria patients before treatment; 50, 50, and 56 samples belonged to group 1 (day 0 samples of treatment failures), group 2 (day of recrudescent samples), and group 3 (day 0 samples of successful treatments).
Distribution and diversity of block 2 of msp-1
Allelic families of msp-1 were not evenly distributed among the isolates from patients with different treatment outcomes. Only 52%, 40%, and 36% of isolates from group 1 (day 0 samples of treatment failure group) contained K1, MAD20, and RO33 alleles, respectively. A similar proportion was found in isolates from group 2 (samples of recrudescence group); 50% of K1, 34% of MAD20, and 32% of RO33. In contrast, isolates from group 3 (day 0 samples of successful treatment group) had a higher proportion of MAD20 (50%) but less RO33 (16%) (Table 1).
The K1 and MAD20 allelic families had 6 and 5 different alleles ranging 150-320 bp and 150-290 bp, respectively. RO33 family was monomorphic with fragment size ranging 210-230 bp. The presence of 2 specific alleles of K1270-290 and RO33 were significantly associated with increased risk of treatment failure [OR: 19.3 (95% CI 2.4-154) and 2.9 (95% CI, 1.1-7.3)] (Table 2), when group 1 was compared with group 3. The OR was 15.5 (95% CI, 1.9-125) and 2.4 (95% CI, 0.9-6.2), when group 2 was compared with group 3. In contrast, the presence of K1150-170 was significantly associated with reduced risk of treatment failure [OR: 0.3 (95% CI 0.1-0.9)] (Table 2).
Distribution and diversity of block 3 of msp-2
FC27 and 3D7 had similar distribution patterns in all sample groups; 56% and 84% of isolates from group 1 contained FC27 and 3D7 alleles, respectively (Table 3). The similar proportion was found in isolates from group 2; 50% of FC27 and 84% of 3D7. Isolates from group 3 had slightly lower proportions of FC27 (46.4%) and 3D7 (75%), but there were no significant differences (Table 3).
FC27 and 3D7 had 9 and 16 different allelic variants with fragment sizes ranging from 250-510 bp, and 250-750 bp, respectively. The presence of a specific allele of 3D7610-630 was significantly associated with increased risk of treatment failure [OR: 4.9 (95% CI 1.3-19.0)] (Table 2), when group 1 was compared with group 3. The OR was 6.8 (95% CI, 1.8-25.6), when group 2 was compared with group 3. In contrast, the presence of 3D7670-690 was significantly associated with reduced risk of treatment failure [OR: 0.19 (95% CI 0.06-0.64)] (Table 2), when group 1 was compared with group 3. The OR was 0.25 (95% CI, 0.08-0.75), when group 2 was compared with group 3.
Nine different glurp allelic variants were detected with fragment sizes ranging 650-1,000 bp. The presence of glurp650-690 allele was significantly associated with an increased risk of treatment failure [OR: 4.9 (95% CI 1.3-19.0)] (Table 2), when group 1 or group 2 was compared with group 3.
Multiplicity of infection (MOI)
Table 4 shows MOI calculated by using data from each of the 3 marker genes for each of the treatment outcome groups. The mean MOI for msp-2 was the highest (2.93), followed by msp-1 (2.15), and glurp (1.26). The mean MOI of isolates from group 1 and group 2 was higher in all 3 marker genes when compared with group 3. In addition, 116 (74.3%) of 156 samples carried more than 1 allelic variants of msp-1 gene per isolate. The number of alleles per isolate ranged from 1 to 6 with a median of 2. Based on msp-2 gene, only 18 (11.5%) of 156 samples carried 1 genotype per isolate. The number of allele per isolate ranged from 1-7 with a median of 3. P. falciparum isolates from this study had less complexity when using glurp as a marker gene. The maximum number of glurp alleles per isolate was 5. But most isolates (84.6%) had 1 allele. However, significant difference among groups of samples was found only in the median MOI of msp-1 gene (P=0.042).
DISCUSSION
Artesunate-mefloquine combination has been used for the treatment of uncomplicated falciparum malaria in Thailand for 2 decades. The efficacy of the combination was initially over 95% despite having evidences of high mefloquine resistance in Thailand. The efficacy then has gradually declined. It was believed that the declining sensitivity attributed to mefloquine resistance. Until recently, artemisinin resistance was confirmed in western Cambodia close to Thai-Cambodia border [18,19]. Factors influencing the treatment failure outcome include antimalarial drug efficacy, pharmacokinetics of the drug in patients and genetic characters of the parasites. In the present study, genetic characteristics of P. falciparum isolates obtained from treatment failure and successful treatment patients were compared. Three polymorphic molecular markers, i.e., msp-1, msp-2, and glurp, were genotyped to characterize P. falciparum from both groups. The technique is simple, cheap, and routinely used in the antimalarial drug resistant monitoring program of Thailand.
Our study shows high diversity of msp-1 and msp-2 genes of P. falciparum isolates, derived from both polyclonal infection and allelic polymorphism. Most isolates (95.5%) had multiple genotype infections with an overall mean multiplicity of infection of 3.21. P. falciparum isolates in the eastern border of Thailand in the previous study also showed high diversity with 76% polyclonal infection [6]. Polyclonal infection has important implications for the epidemiology of drug-resistant P. falciparum and the outcome of drug treatment [8]. The number of parasite genotypes carried by the treatment failure subjects prior to treatment in this study was higher than successful treatment subjects similar to the finding in western Cambodia [19]. In addition, among the treatment failure subjects, P. falciparum isolates collected on day of parasite reappearance had a slightly lower number of parasite genotypes than those collected on day 0 before treatment. The initial presence of several parasite populations with different levels of drug sensitivity or resistance may result in the elimination of sensitive populations and selection of resistant populations. The reappeared parasites in this study were confirmed by the genotyping of msp-1, msp-2, and glurp showing that all were recrudescent parasites. The different number of pre-treatment parasite genotypes on day 0 before treatment and on day of recrudescence could be explained by the clearance of sensitive strain by the first treatment and leftover the resistant ones in the blood.
After a long period of continuous use of artesunate in Thailand, specific resistant variants of msp-1, msp-2, and glurp genes were possibly developed by the parasites, and variants beneficial to the parasites were selected. This resulted in the prevailing of some specific variants in different treatment outcome patients. In the present study, specific alleles and size variants were prevailed among isolates from treatment failure subjects, i.e., K1270-290 of msp-1 gene, 3D7610-630 of msp-2 gene, and glurp650-690, while K1150-170 and 3D7670-690 were prevailed among isolates from successful treatment subjects. However, patients in this study were treated with artesunate-mefloquine combination therapy so it could not elucidate artemisinin from mefloquine resistance.
High frequency of RO33 among treatment failure patients in this study supports the previous finding of association between RO33 and severity of the disease [20]. On the contrary, study in western Cambodia near the Thai-Cambodia border where artemisinin resistance was reported [19], RO33 was found more frequently in the fast parasite clearer. The inconsistent results may possibly due to the genetic difference of parasites from Thai-Myanmar and Thai-Cambodia borders. K1 alleles are dominant among P. falciparum isolates from the Thai-Myanmar border in the present study, followed by MAD20 (41%). A higher proportion (27.0%) of Thai-Myanmar border isolates carried RO33 compared to 4.9% of the isolates originated from the western Cambodia. Most isolates in the western Cambodia carried K1 (82.5%). Only limited numbers of MAD20 (4.9%) and RO33 (4.9%) were reported [18]. The proportion of the 3D7 and FC27 was similar in both Thai-Myanmar and western Cambodia isolates. Genetic characters of P. falciparum isolates in the eastern provinces of Thailand bordered to Cambodia deserve our attention to further study the role of RO33 in drug resistance malaria.
In conclusion, our study provides information on the profiles of allelic variants of msp1, msp2, and glurp genes of isolates from artemisinin derivatives treatment failure and successful treatment subjects. The approach used in this study could be used in addition to other molecular methods as a part of a surveillance program for monitoring the antimalarial drug resistant malaria. The same study in other areas of the country, especially malaria endemic areas in the eastern Thailand near Cambodia is needed to complete the allelic variant profiles of P. falciparum in Thailand.
