Molecular surveillance of pvdhfr, pvdhps, and pvmdr-1 mutations in Plasmodium vivax isolates from Yunnan and Anhui provinces of China

Background Plasmodium vivax is the predominant species of human malaria parasites present in China. Although sulphadoxine-pyrimethamine (SP) and chloroquine (CQ) have been widely used for malaria treatment in China, the resistance profiles of these drugs are not available. Analysis of dihydrofolate reductase (dhfr), dihydropteroate synthase (dhps), and multidrug resistance (mdr-1) gene mutations in P. vivax isolates is a valuable molecular approach for mapping resistance to SP and CQ. This study investigates the prevalence of pvdhfr, pvdhps, and pvmdr-1 of P. vivax clinical isolates from China and provides baseline molecular epidemiologic data on SP- and CQ-associated resistance in P. vivax. Methods Plasmodium vivax clinical isolates were collected from two malaria-endemic regions of China, subtropical (Xishuangbanna, Yunnan province) and temperate (Bozhou, Anhui province), from 2009 to 2012. All isolates were analysed for single nucleotide polymorphism haplotypes in pvdhfr, pvdhps, and pvmdr-1 using direct DNA sequencing. Results In pvdhfr, 15% of Xishuangbanna isolates carried wild-type (WT) allele, whereas the majority of isolates carried mutant genes with substitutions at five codons. Eight mutant haplotypes of pvdhfr were detected, while limited polymorphism of pvdhfr was found in Bozhou isolates. A size polymorphism was present in pvdhfr, with the three-repeat type being the most predominate in both Xishuangbanna (79%) and Bozhou (97%) isolates. In pvdhps, mutations at four codons were detected in Xishuangbanna isolates leading to six haplotypes, including WT allele, single-mutation, double-mutation, and triple-mutation alleles. All Bozhou isolates carried WT pvhdps. In pvmdr-1, isolates from Xishuangbanna carried mutations at codons Y976F and F1076L, whereas all isolates from Bozhou had only a single mutation at codon F1076L. Conclusions Plasmodium vivax isolates from subtropical and temperate zones of China are shown to have dramatically different frequencies and patterns of mutations in pvdhfr, pvdhps, and pvmdr-1. Whereas P. vivax populations in subtropical China are highly resistant to SP and CQ, those in the temperate zone may still be susceptible to SP and CQ. This information is useful for establishing treatment policy and provides a baseline for molecular surveillance of drug-resistant P. vivax in these areas. Electronic supplementary material The online version of this article (doi:10.1186/1475-2875-13-346) contains supplementary material, which is available to authorized users.


Background
Plasmodium vivax infects approximately 100 million people annually and endangers 40% of the world's population [1]. Malaria has been a significant public health problem in China until recently, predominantly in two areas of China: subtropical zone (Yunnan and Hainan provinces) and temperate zone (such as Anhui, Henan, Hunan, Hubei, and Jiangsu provinces). Due to China's national malaria control programme launched in 1978, the intensity and scale of vivax malaria outbreaks have decreased dramatically through vector control and drug treatment of febrile individuals [2]; however, malaria incidence rose steadily between 2000 and 2006 [3]. In subtropical China, Yunnan is one of the two provinces with year-round local transmission of P. vivax and Plasmodium falciparum and has suffered one of the highest malaria morbidity and mortality rates in China. In temperate zone of Central China, outbreaks of malaria have been reported in Anhui and Henan provinces. In 2006, malaria prevalence increased considerably, with the highest numbers of malaria cases reported in Anhui province [4]. Although malaria incidence is low in China, the areas and populations at risk are large. P. vivax malaria accounted for > 95% of all malaria cases reported in China in 2007 [5], having become the sole parasite species and responsible for more than 90% of re-emerged malaria cases reported in temperate zone of China in 2009 [6]. In 2010, a countrywide malaria elimination policy was launched by the Ministry of Health of P. R. China with the goal of eliminating malaria by 2015 in China-with the exception of the border region in Yunnan province, and to completely eliminate malaria from China by 2020. Therefore, there is an urgent need to monitor vivax malaria transmission and drug resistance in subtropical and temperate zones of China.
Resistance to common anti-malarial drugs, such as chloroquine (CQ), has been reported for P. vivax around the world, including Indonesia [7], Myanmar [8], India [9], Vietnam [10], Turkey [11], and Ethiopia [12]. CQ and primaquine have been used as first-line therapies for radical cure of vivax malaria in China for the past 60 years, and clinical failures or reduced efficacy of CQ-primaquine drug combinations for treating vivax malaria had been reported at the Yunnan-Myanmar border [13], in Yunnan [14] and Anhui [15] provinces. Although sulphadoxinepyrimethamine (SP) is rarely used to treat patients with P. vivax malaria, pyrimethamine, a component of the two combination regimens [Maloprim® (plus dapsone) and Fansidar® (plus sulphadoxine)], has been widely used for malaria prophylaxis in China between the mid-1960s and early 1990s [16]. This may place P. vivax under the selection of SP drug stress in China, especially in the subtropical zone where P. vivax and P. falciparum mixed-species infections were common. P. vivax malaria re-emerged in many counties of China in recent years, possibly caused by social and environmental changes and intrinsic differences in parasites with long or short relapse patterns [17][18][19][20]. Even changes in mosquito strains or species could favour a subpopulation of vivax parasites and lead to changes in the parasite population structure [21], and the emergence and spread of drug-resistant malaria should, of course, be held accountable. Studies in other P. vivaxendemic regions such as Vietnam highlight the geographic heterogeneity and temporal dynamics of drug resistance in P. vivax [10], raising further concerns about the presence of clinical resistance to SP and CQ in these areas of China and highlighting the importance of continued surveillance of CQ drug efficacy.
Molecular genetic markers of resistance are useful for monitoring the emergence and spread of anti-malarial drug resistance; a better understanding of the mechanisms of drug action and resistance are essential for fulfilling the promise of eradicating malaria [22]. Currently, several genetic markers, including dihydrofolate reductase (dhfr), dihydropteroate synthase (dhps), and multidrug resistance (mdr-1), have been used to study the prevalence and spread of SP-and CQ-resistant P. vivax [23][24][25][26]; however, molecular epidemiologic information on P. vivax parasite resistance to CQ and primaquine in China is limited. This study investigated the frequencies and patterns of mutations in pvdhfr, pvdhps, and pvmdr-1 linked to SP and CQ resistance in P. vivax isolates from Yunnan and Anhui provinces of China, and the results provide important information for molecular surveillance of drug-resistant P. vivax in these areas.

Study areas
The study was conducted in areas in subtropical (Xishuangbanna prefecture, Yunnan province) and temperate (Bozhou city, Anhui province) zones of China ( Figure 1). Xishuangbanna prefecture had a population of around 113,000 in 2010. It is located in southern Yunnan (the south-west area of China), bordering Myanmar in the west and Laos and Vietnam in the south, in an area between 21°08' and 22°36' latitude and 99°56' and 101°50' longitude. Its climate is characterized by a cool, dry winter, and a warm, wet summer. It has a distinct seasonal climate characterized by rainy season from May to October and dry season from November to May, an average temperature of 18-22°C and average rainfall of 1100-2400 mm per year.
Xishuangbanna was a highly endemic area for both P. falciparum and P. vivax malaria. Five confirmed Anopheles species are widely distributed throughout Yunnan province, with the predominant malaria vectors being Anopheles minimus and Anopheles dirus [27]. Transmission in the study area is seasonal, and malaria incidence characterized by peaks of infection occurring from June to July and again in October to November [18]. Bozhou city's population was around 485,000 in 2010. It is located in the north-west of Anhui (temperate zone of China) bordering six provinces: Jiangsu in the south-east, Shandong in the north, Hubei in the west, Henan in north-west, Jiangxi in the south, and Zhejiang in the north-east, located in an area between 32°5 1' and 35°05' latitude and 115°53' and 116°49' longitude. The area has a distinct seasonal climate characterized by humid summer from June to September and dry winter from October to December, with an average temperature of 14-17°C and average rainfall of 500-750 mm per year [16]. Bozhou city was an endemic area for P. vivax malaria, and the main malaria vector was Anopheles sinensis and the secondary vector Anopheles anthropophagus.

Blood sample collection
Blood samples were collected from patients with symptomatic P. vivax malaria admitted to local hospitals for anti-malarial drug treatment. Inclusion criterion was patients infected with P. vivax but no other human malaria species as confirmed by peripheral smear examination. DNA extraction for pvdhfr, pvdhps, and pvmdr-1 Plasmodium vivax genomic DNA was extracted from 200 μL of each blood sample using a DNA blood kit following the manufacturer's instructions (Takara Bio Inc., Shiga, Japan) with minor modifications. The DNA was dissolved in TE buffer (10 mM Tris-HCl, 0.1 M EDTA, pH 8.0) and stored at −20°C until use. The quality of total DNA was analysed by running 5 μL of each DNA sample on a 1.0% agarose gel stained with ethidium bromide and visualized under ultraviolet illumination.
Polymerase chain reaction (PCR) amplification for pvdhfr, pvdhps, and pvmdr-1 To amplify pvdhfr, pvdhps, and pvmdr-1, a nested PCR amplification method was used following previously reported protocols with some minor modifications [15,16,28]. Oligonucleotide primers and cycling conditions are listed in Table 1  (10 mM each), and 2.5 μL of 10 × PCR buffer following the manufacturer's instructions (BioBasic Inc., Markham, ON, Canada). Primary amplification reactions were initiated with 1.0 μL of template genomic DNA prepared from the blood samples, and 1.0 μL of primary reaction amplification was used in secondary amplification reactions. The amplified nested PCR products were resolved on 1.0% agarose gel, and the sizes of the PCR products were determined using a 100-bp DNA ladder (NewProbe, Beijing, China). The nested PCR products were stored at −20°C until analysis.
Sequence analysis of pvdhfr, pvdhps, and pvmdr-1 The nested PCR products of pvdhfr, pvdhps, and pvmdr-1 from the above samples were directly sequenced in both directions, using an ABI PRISM3730 DNA sequencer from Sangon Biotech (Shanghai, China). Nucleotide and amino acid sequences of pvdhfr, pvdhps, and pvmdr-1 were aligned and compared with the following published sequences of pvdhfr (accession no. X98123), pvdhps (accession no. AY186730), and pvmdr-1 (accession no. AY618622) using Clustal W of the BioEdit 7.0 program.

Statistical analysis
Statistical significance was determined with SPSS software (version 13.0). U test was used to compare the prevalence of infection between collection sites. P < 0.05 was considered statistically significant.

Results
Prevalence and patterns of mutant pvdhfr in P. vivax isolates All blood samples (n = 114) were shown to be infected with P. vivax only by microscopy. The pvdhfr gene was successfully amplified from all 114 P. vivax isolates.
In addition to the above point mutations, variations at the central tandem repeat region between amino acid positions 88 and 105 of pvdhfr were also detected. Analysis    of repeat variant in pvdhfr by sequencing showed that most (89%, 101/114) of the isolates from both areas contained three copies of the GGDNXX repeat that could be divided into type A, B, and C based on substitutions in the middle repeat (Figure 2A). The rest of the isolates (11%, 13/114) had two repeats (type D). The majority of parasites from both Xishuangbanna and Bozhou isolates (79% and 97%, respectively) had the three-copy repeat type. However, the distribution and frequency of tandem repeats in pvdhfr in Xishuangbanna isolates were different from those of Bozhou isolates ( Figure 2B). Type A (60%) was the dominant allele in Xishuangbanna, whereas type C (56%) was the most common allele among the Bozhou isolates.

Discussion
Emergence and spread of CQ-or SP-resistant P. vivax strains has been reported in many malarial regions around the world, especially in Southeast Asia, which is responsible for the increasing morbidity and mortality of P. vivax [29]. The use of anti-malarial drugs exhibits enormous geographic heterogeneity in subtropical and temperate zones of China. With increasing movement of human populations, drug-resistant P. vivax populations and the parasite transmission pattern in the regions are also changing constantly. The emergence and spread of drug-resistant strains of human Plasmodium species may be partly responsible for re-emergence of malaria in China. SP was widely used for malaria prophylaxis between the mid-1960s and early 1990s, and CQ-primaquine is still the first-line antimalarial drug for treating P. vivax malaria in China [16]. Molecular markers have been validated as tools for surveillance of resistance. Analysis of point mutations in these marker genes thus serves as a valuable molecular approach for mapping drug resistance and monitoring malaria-control measures [30]. In this study, the profile of mutations in marker genes associated with CQ and antifolate drug resistance among the P. vivax parasites obtained from patients of the subtropical (Xishuangbanna in Yunnan) and temperate (Bozhou in Anhui) zones of China were determined to better understand the current and changing patterns of CQ-and SP-resistant P. vivax in different malaria-endemic areas. The P. vivax parasite populations in subtropical and temperate zones were shown to differ dramatically in pvdhfr, pvdhps, and pvmdr-1 allele frequencies, i.e., P. vivax populations in subtropical zone are mostly resistant to SP and are likely more tolerant to CQ, whereas the majority of P. vivax populations in the temperate zone are still effectively susceptible to SP and CQ. The long history of CQ use has exposed P. vivax to this drug pressure continually in China. Clinical failures after standard CQ treatment were reported in four cases of P. vivax malaria in Yunnan province in 1996 [14]. A trend of gradual decline of in vitro sensitivity to CQ had also been documented between 2005 and 2008 in some areas of China, especially at the Yunnan-Myanmar border and in central China [31], and reduced efficacy of CQ combinations for treating vivax malaria patients from 2007 to 2008 was reported at the Yunnan-Myanmar border [13]. The relatively high frequency (9%) of pvmdr-1 Y976F mutation among the Xishuangbanna isolates is consistent with the previous reports of declining sensitivity to CQ. The Y976F mutation had been reported in P. vivax isolates from many malaria-endemic regions around the world, including Indonesia [26], Thailand [32], Myanmar [32], and Mauritania [33, and is associated with an increase in CQ IC 50 values in vitro [15]. Cases of CQ-resistant P. vivax malaria have been reported in the areas of Yunnan bordering Myanmar, Laos, and Vietnam [8,10]. The long history of CQ use, as well as frequent population movement across the borders, may contribute to the CQ-resistant P. vivax detected in Xishuangbanna. However, the Y976F mutation was not detected in any Bozhou isolates, which is consistent with the results showing that CQ is still effective in killing P. vivax parasites from the temperate zone in an in vitro schizont maturation inhibition assay [15]. The presence of Y976F in pvmdr-1 in Xishuangbanna isolates suggests a trend toward decreased CQ sensitivity.
SP inhibits the activities of dhps and dhfr in the folate biosynthesis pathway of both P. falciparum and P. vivax parasites, resulting in a synergistic anti-malarial effect [35]. Resistance to antifolate drugs in both parasite species is found to be involved with point mutations in dhps and dhfr [23]. Although P. vivax infections are not generally treated with SP, the parasite is often exposed to the drug due to mixed infections and misdiagnosis [36]. In this study, five mutations at amino acid positions 57, 58, 61, 99, and 117 and seven mutated alleles in pvdhfr were detected in Xishuangbanna isolates, while only mutations at positions 99 and 117 of pvdhfr and three mutated alleles in pvdhfr were found in Bozhou isolates. It has been reported that double mutation S58R and S117N in pvdhfr may first arise under drug pressure and move toward development of resistance to SP [23]. Interestingly, approximately 62% of Xishuangbanna isolates was found to have double mutations at S58R/S117N, whereas no Bozhou isolates were detected with these mutations. Instead, 39% of Bozhou isolates had a single S117N mutation, indicating that the P. vivax isolate circulating in Xishuangbanna may be under stronger drug pressure than those in Bozhou. The most prevalent double mutations of S58R/S117N in Xishuangbanna isolates were similar to those in previous reports from East Timor [37], the Philippines [37], Vietnam [37], and Pakistan [38]. Clinical studies have shown that patients carrying triple and quadruple pvdhfr mutant vivax parasites were more likely associated with SP treatment failure than those with WT parasites [39,40]. The triple and quadruple mutations in pvdhfr have been detected in Yunnan [16], Hainan [16], and Guizhou [41] provinces of China's subtropical zone, and in some other countries including Indonesia [40], Papua New Guinea [40] and India [42]. Similarly, 61% of Xishuangbanna isolates had quadruple mutant alleles (57L/58R/61M/117T and 57I/58R/61M/117T), strongly suggesting that many Xishuangbanna P. vivax isolates are resistant to sulphadoxine drugs. The triple and quadruple mutations were not detected in isolates in Bozhou as well as in other temperate areas of China (such as Wuhe county in Anhui province) [41], Thailand [43], Vietnam [43], Korea [43], Afghanistan [44], Pakistan [45], and Iran [46], indicating that Bozhou P. vivax isolates may still be sensitive to antifolate drugs.
Among mutated codons in pvdhfr, the T61M mutation was mostly linked to the S117T mutation [23]. In the present study, two mutations at T61M/S117T with quadruple mutated alleles (57L/58R/61M/117T and 57I/58R/ 61M/117T) were detected in Xishuangbanna isolates. The observation is consistent with the early reports from Thailand [47], but is in contrast to other reports from the temperate zone of China (Wuhe county in Anhui province) that most T61M mutations were found as single mutations and arose independently [41]. In the current study, the three-copy repeat type was the predominate type in both Xishuangbanna (79%) and Bozhou (97%) isolates, which was similar to the findings from India [42], Afghanistan [44], Iran [46], and Wuhe county in Anhui province and Luodian county in Guizhou province of China [41]. The association between pvdhfr mutations and tandem repeat polymorphisms has been extensively investigated and used to predict the prevalence of drug-resistant malaria around the world [43]. Therefore, the observations from this study suggest increasing prevalence of antifolate-resistant parasites in the Xishuangbanna region. Additionally, no isolates were detected with the four-copy repeat type from either Xishuangbanna or Bozhou, in contrast to other reports showing that quadruple-mutant dhfr alleles were exclusively associated with the four-copy repeat type in Thailand [23], India [42], and Myanmar [48]. Whether the four-copy repeat type contributes to drug resistance requires further investigation.
Molecular epidemiologic studies in different areas have shown dramatically different mutation rates in pvdhfr and pvdhps, which may be attributed to different drug-selection pressures or the intrinsic differences among endemic strains of P. vivax. It is important to take into account the presence of A383G mutation of pvdhps along with double pvdhfr mutations, as SP drug-treatment failure was more frequently associated with multiple mutations in pvdhfr and pvdhps [50]; when the P. falciparum parasite carries mutant alleles of dhfr and dhps, clinical effectiveness of SP is compromised [51]. In the current study, disregarding the mutations at codon 99 of pvdhfr, which is not considered to be related to the anti-malarial drugs, 31.6% of 114 isolates (including Types 8, 14 to 19, and 21 to 28) contained the mutations in pvdhfr and pvdhps and 4.4% of 114 isolates (including Types 16, 19, and 26) mutated simultaneously in the three drug-resistant genes pvdhfr, pvdhps, and pvmdr-1 were found in the Xishuangbanna isolates but not in Bozhou isolates. Although SP was not recommended to treat P. vivax malaria in China due to the intensive use of antifolate drugs for treating P. falciparum infections, P. vivax has been under SP drugs pressure through mixed infections and/or incorrect diagnosis in Yunnan. In contrast, without the introduction of sulphadoxine in Anhui-where P. vivax is the only or most predominant malaria parasite, P. vivax is not under sulphadoxine but rather pyrimethamine drug pressure. Therefore, the different usage of antifolate drug in Yunnan and Anhui may lead to the high prevalence of pvdhfr and pvdhps mutated alleles in Xishuangbanna isolates and relatively low prevalence of these mutations in Bozhou isolates. In tropical area, in addition to longer transmission period, short-term relapses are more common (1-2 months), leading to increased expose of P. vivax parasites to sub-therapeutic concentrations of SP and CQ. Regardless, the results from this study indicates that the high prevalence of multiple mutations of pvdhfr, pvdhps, and pvmdr-1 genes may further reduce the sensitivity to SP and CQ in P. vivax populations in Xishuangbanna of Yunnan, while the P. vivax population may still be susceptible to SP and CQ in Bozhou of central China due to the absence of multiple mutations in pvdhfr, pvdhps, and pvmdr-1.

Conclusions
Results from the current study indicate that the prevalence and patterns of mutant pvdhfr, pvdhps, and pvmdr-1 exhibit enormous geographic heterogeneity in subtropical (Xishuangbanna, Yunnan) and temperate (Bozhou, Anhui) zones of China. P. vivax populations in subtropical zone of China are likely resistant to SP and CQ, whereas in temperate zone of China they are still relatively susceptible to SP and CQ, which provides valuable information for monitoring drug resistance in different malaria-endemic areas of China and for assessing the appropriateness of the current national anti-malarial drug policy. The increased resolution afforded by combination of molecular genetics and geographic information systems (GIS) tools has the potential to provide insights into the epidemiology, evolution, and ecology of these parasites in the future.