Prevalence of molecular markers associated with drug resistance of Plasmodium vivax isolates in Western Yunnan Province, China

Background Plasmodium vivax is the most widely distributed malaria parasite, and its drug resistance poses unique challenges to malaria elimination. The Greater Mekong Subregion (GMS) is known as the global epicenter of multidrug resistance. Surveillance of molecular markers associated with drug resistance in this area will help to inform drug policy. Methods Dry blood spots from 58 patients out of 109 with P. vivax infection between 2017, December and 2019, March were obtained from Yingjiang County, Yunnan Province, along the China–Myanmar border. Pvdhfr, Pvdhps, Pvmdr1 and Pvcrt-o were amplified and sequenced to assess gene mutations. The polymorphism and prevalence of these molecular markers were analyzed. Results Mutations in Pvdhfr at codons 57, 58, 61, 99 and 117 were detected in 27.59, 48.28, 27.59, 32.76 and 48.28% of the isolates, respectively. Single mutant haplotype (I13F57S58T61S99S117I173) was the most frequent (29.31%, 17/58), followed by double mutant haplotype (20.69%, 12/58). Of three types of tandem repeat variations of Pvdhfr, deletion type was the most common. Pvdhps showed a lower prevalence among mutation genotypes. Single mutant was dominant and accounted for 34.48% (20/58). Prevalence of Pvmdr1 mutations at codons 958 and 1076 were 100.00% and 84.48%, respectively. The proportion of double and single mutant types was 84.48% (49/58) and 15.52% (9/58), respectively. Eleven samples (18.97%, 11/58) showed K10 “AAG” insertion in chloroquine resistance transporter gene Pvcrt-o. Conclusions There was moderate diversity of molecular patterns of resistance markers of Pvdhfr, Pvdhps, Pvmdr1 and Pvcrt-o in imported P. vivax cases to Yingjiang county in Western Yunnan, along the China–Myanmar border. Prevalence and molecular pattern of candidate drug resistance markers Pvdhfr, Pvdhps, Pvmdr1 and Pvcrt-o were demonstrated in this current study, which would help to update drug policy.

(SP) in the 1970s, mefloquine in the late 1990s, and artemisinin in 2008, and then spread progressively throughout other malaria-endemic areas [3][4][5][6]. This has raised concern from the World Health Organization (WHO) and local health authorities [2,7,8]. Malaria transmission in international border areas is usually confounded by population mobility and distinct chemotherapy policies and antimalarial strategies. The China-Myanmar border, as part of the GMS, included 18 counties of Yunnan Province. Although no indigenous cases have been identified in Yunnan Province since 2017, P. vivax remains a challenge, with increasing evidence of abundant vector species richness and diversity, high malaria vulnerability resulting from mobile population, as well as drug resistance [9,10]. In Myanmar, the proportion of malaria cases caused by P. vivax has increased steadily since 2012 [11].
CQ was first produced in 1934 and quickly proved to be one of the most successful and important antimalarial agents [12]. Nevertheless, the heavy use of CQ throughout subsequent decades eventually led to drug resistance. P. falciparum developed resistance in various areas since the 1950s, but drug-resistant P. vivax was not reported until the 1980s in Indonesia and Papua New Guinea [12,13]. To date, CQ-resistant P. vivax has been confirmed in more than 10 countries, including Myanmar and China [14]. There has been a long history of successful application of SP in combating malaria due to its safety, good tolerance and long-lasting activity [15]. In China, as a component of the two combination regimens, pyrimethamine was widely used for malaria prophylaxis between the mid-1960s and early 1990s [16]. By now, SP is recommended by WHO as one of the partner drugs for treatment of P. falciparum in the GMS, as well as intermittent preventive treatment for infants, children and pregnant women [2,15]. Although SP is rarely used to treat P. vivax infection, the parasite is still under SP selection pressure, especially in endemic regions where co-infection with P. vivax and P. falciparum is common.
Compared with P. falciparum, it is more difficult to determine the underlying mechanisms of antimalarial drug resistance of P. vivax because there is no proper in vitro cultivation system for P. vivax. This means that the molecular mechanism of P. vivax resistance remains to be established. Several studies suggest that it involves multigenic loci, such as CQ resistance marker Pvcrt-o; multidrug resistance marker Pvmdr1; and antifolate resistance markers Pvdhps and Pvdhfr, which are conferred from homologous genes in P. falciparum [11,17]. Data for molecular markers associated with drug resistance would be beneficial in addressing the resistant parasite. Few studies to date have defined the molecular epidemiology of P. vivax resistance markers on the China-Myanmar Border [18,19]. Here, we report the prevalence of molecular markers of drug resistance in P. vivax to facilitate appropriate drug policy in this region.

Study site
Yingjiang (Longitude 97°31′~98°16′,Latitude 24°24′2 5°20′) is one of the 18 counties along the China-Myanmar border, located west of Yunnan Province. It was selected as the study site due to its long borderline with Kachin State, Myanmar and being well documented as an epidemic area of resistant P. falciparum [18].
Oligonucleotide primers and cycling conditions are listed in additional file (see Additional file: Table S1). A final 25-μL reaction volume was performed, of which 1 μL template genomic DNA was used in primary amplification reactions, and 1 μL primary reaction products in the second round of amplification in the case of nested PCRs. Amplification products were sequenced by Sangon Biotech Co. Ltd. (Shanghai, China).

Data analysis
Nucleotide and amino acid sequences of

General information
We collected data from 58 patients (  at position 57, were observed in 11(18.97%) and 5(8.62%)isolates, respectively, and the frequency for S117T and S117N was 16(27.59%) and 12(20.69%), respectively. Three types of tandem repeat variations were found in Pvdhfr. Type I was the same as the reference strain (accession number X98123), whereas type II showed mutant allele H99S, and type III exhibited a deletion of 18 nucleotides (ACACACGGTGGTGACAAC, translated into THGGDN) between amino acid positions 98 and 103 (Fig. 2). Type III was the most common, accounting for 37.93% (22/58), followed by Type II which was observed in 19 (32.76%) isolates. In addition, more than half of Type III isolates (12/22, 54.55%) also carried S58R and S117N mutations.
Among the mutant types, single mutant was dominant and accounted for 34.48% (20/58). Double mutant S 382 G 383 K 512 G 553 R 580 V 585 was less frequent (13.79%, 8/ 58). Quadruple mutant C 382 G 383 E 512 G 553 R 580 V 585 was only found in one P. vivax isolate. S382C and K512E were rarely observed in previous studies.

Discussion
Drug resistance is of great concern for malaria control and prevention, especially in GMS, necessitating monitoring resistance to antimalarial agents. However, since its first report in 1989, the burden of drug-resistant P. vivax is still unclear and its underlying mechanism, epidemiology and drug efficacy have not been well characterized [17]. Four main methods, in vivo therapeutic efficacy studies, in vitro assay, drug concentration measurement, as well as molecular markers analysis, are used to monitor antimalarial drug efficacy and resistance. Among these, molecular markers are widely preferred due to their practical advantages over in vivo and in vitro tests. Molecular markers allow population-level screening, and samples on filter paper are easily obtained, transported and stored, thus avoiding host confounding factors [22]. The epidemiology of drug resistance of P. vivax varies across the GMS, hence molecular marker surveillance is encouraged to inform local drug policy.
In the present study, the prevalence of Pvdhfr mutation type, including point mutation and mutant tandem repeat, was high (94.83%, 55/58), which was similar to the reports in southern Thailand and western Myanmar. However, one study in Xishuangbanna Prefecture of Yunnan Province (southern Yunnan, bordering Myanmar in the west and Laos and Vietnam in the south) between 2009 and 2010, and another in India between 2005 and 2007, found that prevalence of Pvdhfr was lower than in our study [16] [17,23,24]. Furthermore, point mutations at codons 57, 58, 61 and 117 in the Pvdhfr gene were detected in 27.59-48.28% of isolates in the current study. These results were lower than those in southern Thailand and western Myanmar, but higher than those in Xishuangbanna of Yunnan and India [16,17,23,24]. Regarding the patterns of mutation types, single and double mutants were the dominant genotypes in western Yunnan in our study, while quadruple mutation was the most common in Myanmar, Thailand and southern Yunnan [16,17,23,25]. Previous studies identified that mutations at residues 117 and 58 arose first under drug pressure, so they were more highly mutated than others [26]. These results confirmed that the mutation types at codons 117 and 58 were the most frequent. Triple and quadruple mutations were more associated with high level of SP resistance than double or single mutations were. Our study indicated that P. vivax in western Yunnan might be under stronger drug pressure than those in western Myanmar and southern Thailand [27].
Mutant tandem repeats are also suggested to be associated with P. vivax antifolate resistance, and the frequency of Type II (H99S type) and Type III (deletion type) was 32.75 and 37.93%, respectively. This was consistent with a previous studies that reported that most isolates in India and Cambodia were deletion type [24,28]. Nevertheless, the highest frequency of tandem repeat variants was for wild type in southern Thailand and Xishuangbanna Prefecture, Yunnan [16,23]. In Anhui Province (Central China), Type II (H99S type) was the most common [16].
Similar to Pvdhfr, the frequency of mutant Pvdhps, especially the highly mutant types (triple or quadruple types), was less than that in southern Thailand and southern Myanmar [21,23]. In addition, compared with another border region, Xishuangbannan of Yunnan, Pvdhps in our sampling region was more conserved, with higher proportions of wild type and fewer highly mutated types, although the isolates from Xishuangbannan were collected nearly 10 years ago [16]. Considering the similar drug policy in this study area and Xishuangbannan, it is still unclear whether the disagreement resulted from spatial heterogeneity or drug susceptibility to sulfadoxine.and as such, further study is required.
Several studies have provided evidence that Pvmdr1 mutations are associated with reduced susceptibility to CQ [17,29,30]. Therefore, Pvmdr1 is considered to be a strong candidate marker of drug resistance [23]. The prevalence of Pvmdr1 T958M and F1076L mutations in our study was consistent with previous studies, showing that T958F was harbored in all the isolates and F1076L was in most of them [13,21]. However, no Y976F was found in our study, while it was frequently reported with considerable prevalence in different endemic areas, including Indonesia, Thailand, Cambodia, India, Papua New Guinea and Ethiopia [29][30][31][32][33][34][35]. This is not surprising as 98.51% of patients were categorized as having an adequate clinical and parasitological response to CQ by an in vivo therapeutic efficacy study in Yingjiang and Tengchong, Yunnan [36]. Our study indicated that Pvmdr1 at codon 976 was conserved in this area, although this needs to be confirmed.
The possible role of Pvcrt-o in CQ resistance is controversial. Several studies have found a negative link between K10 insertion and reduced CQ IC 50 , while others have shown that Pvcrt-o expression decreased susceptibility to CQ by 2.2-fold [30,37]. The K10 Pvcrt-o gene insertion was found in 18.97% isolates in our study, which was less than in previous studies in Myanmar that reported 46.15% in Yangon in 1999, 72.73% in Shwegyin, 66.67% in Kawthaung and 48.33% in Buthidaung between 2009 and 2016 [17,35]. Conversely, K10 insertion was rarely observed in Thailand from 2012 to 2018, or from the Thailand-Myanmar border or Thailand-Cambodia border in 2008 or 2014 [21,23]. Given the geographical genetic differences among parasite populations from the GMS, the prevalence of K10 insertion in Pvcrt-o in the current and previous studies showed significant temporal and spatial heterogeneity [38]. The discrepancy may have resulted from differences in study sites or sample size.

Conclusions
In conclusion, the present study demonstrated the prevalence and molecular pattern of candidate drug resistance markers Pvdhfr, Pvdhps, Pvmdr1 and Pvcrt-o of imported P. vivax cases to Yingjiang county in Western Yunnan, along the China-Myanmar border. Diversity of molecular patterns of resistance markers Pvdhfr, Pvdhps, Pvmdr1 and Pvcrt-o was found. This study helped to provide evidence for drug policy update.
Additional file 1 Table S1 Primers and cycling conditions for Pvcrt-o, Pvmdr1, Pvdhps and Pvdhfr genotyping assay. Table S2 Combined analysis of all mutations from P. vivax isolates.