Genetic Variations Associated with Drug Resistance Markers in Asymptomatic Plasmodium falciparum Infections in Myanmar

The emergence and spread of drug resistance is a problem hindering malaria elimination in Southeast Asia. In this study, genetic variations in drug resistance markers of Plasmodium falciparum were determined in parasites from asymptomatic populations located in three geographically dispersed townships of Myanmar by PCR and sequencing. Mutations in dihydrofolate reductase (pfdhfr), dihydropteroate synthase (pfdhps), chloroquine resistance transporter (pfcrt), multidrug resistance protein 1 (pfmdr1), multidrug resistance-associated protein 1 (pfmrp1), and Kelch protein 13 (k13) were present in 92.3%, 97.6%, 84.0%, 98.8%, and 68.3% of the parasites, respectively. The pfcrt K76T, pfmdr1 N86Y, pfmdr1 I185K, and pfmrp1 I876V mutations were present in 82.7%, 2.5%, 87.5%, and 59.8% isolates, respectively. The most prevalent haplotypes for pfdhfr, pfdhps, pfcrt and pfmdr1 were 51I/59R/108N/164L, 436A/437G/540E/581A, 74I/75E/76T/220S/271E/326N/356T/371I, and 86N/130E/184Y/185K/1225V, respectively. In addition, 57 isolates had three different point mutations (K191T, F446I, and P574L) and three types of N-terminal insertions (N, NN, NNN) in the k13 gene. In total, 43 distinct haplotypes potentially associated with multidrug resistance were identified. These findings demonstrate a high prevalence of multidrug-resistant P. falciparum in asymptomatic infections from diverse townships in Myanmar, emphasizing the importance of targeting asymptomatic infections to prevent the spread of drug-resistant P. falciparum.


Introduction
The emergence and spread of parasites resistant to antimalarial drugs and mosquitoes resistant to insecticides threaten the recent gains in malaria control and challenge the goal for malaria elimination in the Greater Mekong Subregion (GMS) of Southeast Asia [1]. Antimalarial drug resistance in P. falciparum tends to emerge in low-transmission settings, particularly in Southeast Asia and South America, before expanding to high-transmission settings in sub-Saharan Africa [2]. Resistance to chloroquine (CQ) and later to sulfadoxine-pyrimethamine (SP) has been responsible for the increased mortality in African children [3,4]. Artemisinin (ART)-resistant P. falciparum parasites were first reported in Cambodia and subsequently in all countries of the GMS [5]. This has accelerated resistance development in parasites to artemisinin combination therapy (ACT) partner drugs, resulting in increased treatment failure rates with dihydroartemisinin-piperaquine (DP) in Cambodia and with artesunate-mefloquine (AS-MQ) in Cambodia and on the Thai-Myanmar border [6,7]. The timely monitoring of antimalarial resistance spread is essential for maintaining the recent progress in malaria control and for achieving the goal of regional malaria elimination.
As a high malaria-burden country within the GMS, Myanmar uses several first-line treatments for P. falciparum infections, including artemether-lumefantrine, AS-MQ, and DP. However, delayed clearance has been observed in all the three first-line ACTs [8]. Similar to other malaria-endemic areas, Myanmar experiences a large proportion of asymptomatic malaria infections, as demonstrated by active case surveillance of malaria infections among healthy populations [9,10]. In the Shwegyin township of southern Myanmar, mutations in genes associated with drug resistance including those with ART resistance were identified in asymptomatic infections [11]. Asymptomatic or sub-clinical malaria infections seldom cause acute disease, but they are capable of infecting mosquitoes, thus serving as silent reservoirs for continued malaria transmission. Therefore, prevalent antimalarial drug resistant strains in asymptomatic infections may underlie the rapid spread of drug resistance. Given the limited data on drug resistance in asymptomatic infections, ongoing molecular epidemiological studies of drug resistance are needed in these parasite reservoirs.
Molecular epidemiology studies provide information for detecting the emergence and tracking the spread of antimalarial drug resistance. Several mutations in the P. falciparum dihydrofolate reductase (pfdhfr) and P. falciparum dihydropteroate synthase (pfdhps) genes (e.g., triple mutations at codons 51, 59 and 108 of pfdhfr and double mutations at codons 437 and 540 of pfdhps) are associated with SP treatment failures [12]. The P. falciparum chloroquine resistance transporter (pfcrt) K76T mutation and P. falciparum multidrug resistance protein 1 (pfmdr1) N86Y mutation have been linked to CQ and amodiaquine resistance [13]. Mutations in the P. falciparum multidrug resistance-associated protein 1 (pfmrp1), such as H191Y and S437A, were reported to be associated with resistance to CQ and quinine (QN) in vitro [14]. Point mutations in the propeller domain of Kelch protein 13 (k13) were collectively correlated with clinical ART resistance [15,16]. Molecular surveillance showed that k13 mutations associated with ART resistance were restricted to certain areas of the GMS, with C580Y being the predominant mutation in the Thai-Cambodian region and F446I in China and northern Myanmar [17][18][19][20].
Here, we profiled mutations in antimalarial drug resistance genes in asymptomatic P. falciparum infections identified from cross-sectional studies in three townships of Myanmar. The results will provide updated information on the drug resistance status of P. falciparum malaria in multiple sentinel sites in Myanmar, which is necessary for developing strategies to eliminate P. falciparum infections in this country.

Ethics Approval
Ethical approval for this project was obtained from the institutional review boards of China Medical University, China (2019086), University of South Florida, USA (Pro00036813) and the Ministry of Health and Sports, Myanmar (Ethics/DMR/2017/077AE/2018). Before conducting the study, all adult participants or legal guardians of children voluntarily signed the informed consent.

Study Sites and Samples
Three cross-sectional studies were conducted in Myanmar at the Laiza (Kachin State), Banmauk (Sagaing Region), and Paletwa (Chin State) townships in 2015, 2017-2018, and 2017, respectively ( Figure 1A). From 2015 to 2017, a country profile of malaria showed that Paletwa had the highest annual incidence of P. falciparum, followed by Banmauk and Laiza [21][22][23]. These cross-sectional surveys recruited a total of 3,495 residents, who provided finger-prick blood to prepare blood smears and dried blood spots on filter paper ( Figure 1B).
( Figure 1A). From 2015 to 2017, a country profile of malaria showed that Paletwa had the highest annual incidence of P. falciparum, followed by Banmauk and Laiza [21][22][23]. These cross-sectional surveys recruited a total of 3,495 residents, who provided finger-prick blood to prepare blood smears and dried blood spots on filter paper ( Figure 1B).

Diagnosis of P. falciparum Asymptomatic Infections
Smears were stained with Giemsa and examined by two experienced microscopists. For molecular diagnosis, parasite genomic DNA was isolated from blood spots on filter paper according to the protocol of QIAamp ® DNA Mini kit (Qiagen). Detection of P. falciparum isolates was performed by nested PCR using an established protocol targeting the 18S rRNA gene as described previously [24]. This method has a detection limit of ~2 parasites/μL. For PCR assessment, one positive and one negative control (from a symptomatic P. falciparum isolate and sterile water, respectively) were used in each of the amplification plates. The genomic DNA from P. falciparum infections was used for the amplification of genes associated with drug resistance.

Diagnosis of P. falciparum Asymptomatic Infections
Smears were stained with Giemsa and examined by two experienced microscopists. For molecular diagnosis, parasite genomic DNA was isolated from blood spots on filter paper according to the protocol of QIAamp ® DNA Mini kit (Qiagen). Detection of P. falciparum isolates was performed by nested PCR using an established protocol targeting the 18S rRNA gene as described previously [24]. This method has a detection limit of~2 parasites/µL. For PCR assessment, one positive and one negative control (from a symptomatic P. falciparum isolate and sterile water, respectively) were used in each of the amplification plates. The genomic DNA from P. falciparum infections was used for the amplification of genes associated with drug resistance.

Sequence Analysis and Statistics
We evaluated the quality of sequences by examining the chromatograms. Those showing ambiguity were re-amplified and re-sequenced. If the re-sequencing did not improve the quality, these samples were excluded from the analysis. The sequences were aligned using ClustalW in MEGA7.0.26. The samples showing mixed chromatograms were excluded and only single-peak sequences were considered to be monoclonal P. falciparum infections and were used for single nucleotide polymorphism (SNP) and haplotype analysis ( Figure S2). Nucleotide and amino acid positions were numbered according to the 3D7 reference sequences. The SNP data from different genes were analyzed using Microsoft Soft Excel 2007 and SPSS 22.0. Pearson's Chi-square test or Fisher's exact test were used to determine statistical significance (p < 0.05). The haplotype network was constructed by DnaSP 6 and Network software 5.01.0 using the median joining algorithm [29].

Availability of Data and Material
The datasets used and/or analyzed during the current study are available from the corresponding authors upon request.

Asymptomatic Infections in Blood Samples from Cross-Sectional Surveys
Asymptomatic P. falciparum infections were prevalent in malaria-endemic areas of Myanmar [24,30]. Dried blood samples on filter papers from 3495 healthy residents collected in cross-sectional surveys in three regions of Myanmar were screened by nested PCR. A total of 107 asymptomatic P. falciparum infections were identified for an overall prevalence of 3.1%. After excluding the samples with double peaks suggestive of mixed infections, 80 pfdhfr, 82 pfdhps, 75 pfcrt, 80 pfmdr1, 82 pfmrp1, and 57 k13 sequences were used for SNP and haplotype analysis ( Figure 1B).

K13 Mutations Associated with ART Resistance
The k13 gene was successfully sequenced in 57 samples. Three different point mutations were found, including two previously reported mutations (F446I and P574L) and one mutation (K191T) observed for the first time in these samples ( Table 6). Of these mutations, F446I was most frequent (14.0%; 8/57). Besides point mutations, N, NN, and NNN insertions between amino acids 136 and 137 were also identified in 1 (1.8%), 26 (45.6%) and 4 (7.0%) samples, respectively. All P. falciparum isolates from Laiza had the NN insert, whereas in Paletwa and Banmauk, it was present in 40.0% and 33.3% of the samples, respectively (p < 0.001). The F446I mutation always occurred together with the NN insert and was present exclusively in samples from Laiza along the China-Myanmar border. Of the 57 asymptomatic P. falciparum isolates, 22 samples were WT ( Figure 2F and Table 6).

Haplotype Network
SNPs in the six genotyped drug-resistance genes gave rise to 43 distinct haplotypes in 52 P. falciparum isolates that were successfully sequenced at all six target genes ( Figure S3), demonstrating high-level genetic diversity of asymptomatic P. falciparum infections in these regions. Among the samples, 11.6% (5/52), 7.0% (3/52), 7.0% (3/52), and 4.7% (2/52) shared haplotypes No. 2,9,30, and 26, respectively, whereas each of the remaining isolates was unique. To examine the phylogenetic relationship of the asymptomatic P. falciparum parasites, a haplotype network was generated based on the SNPs observed in the six genes ( Figure 3). The network had three main branches. Isolates of the three study areas appeared to be polyphyletic and were distributed across the entire network, suggesting independent origins of resistant mutations in different genes from diverse genetic backgrounds. Three haplotypes with the k13 K191T mutation originated from Banmauk. Four haplotypes with the k13 F446I mutation and one haplotype with the P574L mutation associated with ART resistance appeared independently in Laiza. Significantly, there were 20 haplotypes with the NN insert in k13 gene, including eight haplotypes in Laiza, six in Banmauk and six in Paletwa. 26, respectively, whereas each of the remaining isolates was unique. To examine the phylogenetic relationship of the asymptomatic P. falciparum parasites, a haplotype network was generated based on the SNPs observed in the six genes ( Figure 3). The network had three main branches. Isolates of the three study areas appeared to be polyphyletic and were distributed across the entire network, suggesting independent origins of resistant mutations in different genes from diverse genetic backgrounds. Three haplotypes with the k13 K191T mutation originated from Banmauk. Four haplotypes with the k13 F446I mutation and one haplotype with the P574L mutation associated with ART resistance appeared independently in Laiza. Significantly, there were 20 haplotypes with the NN insert in k13 gene, including eight haplotypes in Laiza, six in Banmauk and six in Paletwa. Figure 3. Median-joining haplotype network of asymptomatic P. falciparum isolates harboring mutations in six drug resistance-associated genes. The haplotype network was constructed for asymptomatic P. falciparum isolates using 43 haplotypes obtained from amino acid changes observed in pfdhfr, pfdhps, pfcrt, pfmdr1, pfmrp1, and k13. The size of each circle shows the same haplotype prevalence, with a color corresponding to a population of origin: Laiza (red), Banmauk (blue), or Paletwa (yellow). The length of an edge is proportional to the number of variations between two haplotypes, and enlarged is the torso of the tree. K13 individual mutations are labeled by amino acid position. . Median-joining haplotype network of asymptomatic P. falciparum isolates harboring mutations in six drug resistance-associated genes. The haplotype network was constructed for asymptomatic P. falciparum isolates using 43 haplotypes obtained from amino acid changes observed in pfdhfr, pfdhps, pfcrt, pfmdr1, pfmrp1, and k13. The size of each circle shows the same haplotype prevalence, with a color corresponding to a population of origin: Laiza (red), Banmauk (blue), or Paletwa (yellow). The length of an edge is proportional to the number of variations between two haplotypes, and enlarged is the torso of the tree. K13 individual mutations are labeled by amino acid position.

Discussion
As the GMS is moving towards malaria elimination, Myanmar requires immediate attention because of the high malaria burden and its geographical connection with South Asia, through which drug resistance could spread rapidly. According to WHO reports, the latest malaria drug resistance map reveals that both treatment failure and resistant genotypes are prevalent in Myanmar, including our three study sites [31]. Irrespective of the ACTs used, from 2010 to 2018, treatment failure among P. falciparum patients in Kachin State was higher than that in Chin State. Our study found a higher prevalence of k13 and pfmdr1 mutations in Kachin State (east) than in Chin State (west) which is possibly contributing to the treatment failure in Kachin State. Asymptomatic Plasmodium infections, as silent reservoirs of malaria parasites, play an important role in continued malaria transmission. Critically, these asymptomatic infections may carry genetic variations conferring drug resistance, thus facilitating the spread of drug-resistant parasites [11]. Therefore, this study aimed to provide molecular epidemiology information about drug-resistant P. falciparum in asymptomatic infections at three sentinel sites located in southwestern, northern and northeastern Myanmar, each with a different malaria epidemiology.
The drug combination SP, commercially known as Fansidar, has been used as an antimalarial chemotherapy in Southeast Asia and Africa. The mechanism of resistance to sulfadoxine has been associated with five point mutations at the S436A/F, G437A, K540E, A581G, and A613T/S codons of the pfdhps gene. Mutations at 437 and 540 were strongly associated with SP treatment failure, whereas the 436, 581, and 613 mutations confer some degrees of resistance [32]. In this study, A613T/S was not observed, but K540N was prevalent, consistent with a previous study from northeast Myanmar [33]. A581G, G437A, and K540E were the predominant mutations in Laiza, Banmauk, and Paletwa, respectively. In population studies, mutations at codon 59 of the pfdhfr gene and codon 540 of the pfdhps gene are strongly predictive of SP treatment failure [34,35]. The prevalence of this mutation combination in Laiza and Paletwa was 100% and 83.3%, respectively, and was slightly lower in Banmauk (34.1%), suggesting that strong selection pressure against SP was emerging in western and northeastern Myanmar. Similarly, triple mutations at codons 108, 51, and 59 of pfdhfr and double mutations at codons 437 and 540 of pfdhps are associated with SP treatment failure [12]. Here, one isolate was found to have this quintuple mutation combination in Banmauk, and 12 isolates (nine in Laiza, two in Banmauk and one in Paletwa) harbored the sextuple mutations with pfdhfr (N51I, C59R, S108N, and I164L) and pfdhps (K540N/E and A581G), suggesting that P. falciparum from these asymptomatic carriers would be highly resistant to SP.
Mutations in the pfcrt gene, especially the K76T mutation, are major determinants of CQ resistance [36,37]. In addition, pfcrt mutations also influence P. falciparum susceptibility to mefloquine (MQ), halofantrine (HF), and ART [38]. The K76T mutation is associated with different sets of mutations at different codons, most commonly C72S, M74I, N75E, A220S, Q271E, N326S, I356T, and R371I [39]. Consistent with a previous study conducted in Thailand [25], C72S was not found. Except for N326S, other mutations reached high prevalence in all three sites. Notably, almost all parasite isolates had the CIVIET haplotype around codons 72-76, which is the typical CQ-resistant haplotype in Southeast Asia. The stable and highly prevalent of pfcrt mutations may be the result of continued use of CQ as the first-line treatment for P. vivax infections in this region [40,41]. In addition to CQ resistance, pfcrt has gained recognition as a multidrug resistance transporter, which influences parasites' susceptibilities to multiple first-line antimalarial drugs [42,43]. Emerging mutations in pfcrt, H97Y, F145I, M343L, and G353V, also have been linked to piperaquine (PPQ) resistance [44,45], but these mutations were not identified in this study. Thus, the high prevalence of pfcrt mutations in asymptomatic P. falciparum parasites may be linked to failures of multiple ACTs in Myanmar.
The ATP-binding cassette (ABC) transporters, including pfmdr1 and pfmrp1, are potentially involved in resistance to multiple antimalarial drugs [46,47]. Several point mutations in pfmdr1 are associated with changed sensitivities to CQ, MQ, QN, HF, and ART [38,[48][49][50]. In the presence of pfcrt K76T, point mutations in pfmdr1, primarily at codon 86 [13,14] and additionally at positions 184, 1034, 1042, and 1246 [51], can increase resistance of the parasites to CQ. Decreased susceptibility to lumefantrine (LMF) has also been linked to polymorphism in pfcrt and pfmdr1 [52,53]. In this study, only two P. falciparum isolates from the western township Paletwa had the pfcrt K76T and pfmdr1 N86Y mutations, but the combination of pfcrt K76T and pfmdr1 Y184F had much higher prevalence in Laiza (58.3%) than Banmauk (18.0%) and Paletwa (12.5%). The pfmdr1 Y184F was not detected in an earlier study [54], where I185K was reported as a novel mutation. The authors speculated that the decrease in parasite's susceptibility to the drug was associated with substitution of mutation at codon Y184F for I185K. In our study, the pfmdr1 I185K mutation was highly prevalent (~90% in all samples) and reached fixation in Banmauk and Paletwa. It would be worthwhile to establish whether this mutation is linked to altered drug sensitivity through genetic studies given the high prevalence of this mutation. Since disruption of pfmrp1 in a CQ-resistant P. falciparum isolate rendered this parasite more sensitive to CQ, QN, PPQ, primaquine, and ART, PfMRP1 is proposed to mediate resistance of the parasite to multiple antimalarial drugs [55]. Association studies suggest that mutations at codons Y191H, A437S, H785N, I876V and F1390I in pfmrp1 contribute to decreased sensitivity to CQ, PPQ, LMF, artesunate (ATS) and dihydroartemisinin (DHA) as well as reduce both in vivo and in vitro susceptibility to ACT [28,33,56]. Our results suggest that P. falciparum isolates from asymptomatic infections in these areas might be resistant to CQ, PPQ, LMF, ATS, DHA, and ACT.
ART resistance in P. falciparum has emerged in western Cambodia and its potential spread to Africa threatens global malaria control [57]. ART resistance is characterized by delayed parasite clearance, which corresponds to the decreased susceptibility of ring-stage parasites [58]. Non-synonymous SNPs in K13 were found to be strongly associated with resistance to ART [16]. To date, more than 200 non-synonymous mutations in the k13 gene have been reported. N458Y, Y493H, R539T, I543T, and C580Y are validated ART resistance mutations, and a number of mutations are candidates for ART resistance, including P574L [8]. In Africa, non-synonymous k13 mutations are relatively rare [59]. Recently, k13 mutations have emerged independently in multiple locations in Southeast Asia, including Myanmar [17]. C580Y, Y493H, R539T, and I543T k13 mutations were prevalent in the Thai-Myanmar border, whereas F446I was the most prevalent mutant allele of k13 in the northern Myanmar and China-Myanmar border, and P574L next to C580Y were reported among isolates collected from migrant goldmine workers in southern Myanmar [19,[60][61][62]. In our study, the validated mutations of k13 resistance were not found, while F446I, P574L, and K191T mutations in the k13 propeller domain were observed among parasites from asymptomatic infections. The K191T has not been reported in previous studies. F446I mutation was observed with high prevalence only at Laiza along China-Myanmar border, which is consistent with previous studies. F446I mutation in k13 has been found to be associated with delayed parasite clearance [61], and a recent study showed that F446I was associated with the increased ring survival rates by genetically introducing the mutation into the k13 gene [63]. Consistent with previous studies of P. falciparum from a symptomatic population, the absence of k13 propeller mutations in Paletwa suggest that ART resistant parasites may not have spread to the western Myanmar region [18]. Moreover, we found the N, NN, and NNN insertion in the N terminus of k13 protein through analysis of the full-length k13 sequences. A previous study reported that an NN insertion in the N terminus of k13 protein was associated with increased in vitro ring-stage survival in the presence of DHA, but the possibility of the NN insert acting cooperatively with F446I in conferring ART resistance was not ruled out [33]. In view of the high prevalence (45.6%) of the NN insertion among P. falciparum from asymptomatic populations in Myanmar, it would be valuable to determine whether it independently affects ART resistance.
The generation of the haplotype network can be used to study the phylogenetic relationship of parasites, to determine the origins of drug resistance gene mutations from different geographical regions, and to assess the diversity of parasite populations. Consistent with the results from clinical P. falciparum isolates, parasites from asymptomatic individuals also displayed extraordinarily high diversity in Myanmar. The genotypes of the six genes associated with drug resistance indicated that the asymptomatic P. falciparum infections in Myanmar are resistant to multiple drugs. More importantly, 58.8% (20/34) haplotypes carried k13 mutations (N458Y, R539T, P574L, and F446I) associated with ART resistance in clinical isolates, whereas 62.5% (5/8) haplotypes of asymptomatic P. falciparum isolates carried F446I and P574L. Although only two k13 mutations related to ART resistance were detected in our study, the presence of such high haplotype diversity in the limited isolates of asymptomatic infection should be concerning. Despite reduced P. falciparum prevalence in recent years, the evidence suggests that ART resistant P. falciparum parasites appear independently and spread widely along the China-Myanmar border, which is a reminder that surveillance of P. falciparum should be strengthened both in Myanmar and China. Although k13 mutation associated with ART resistance was not found in the other two areas, the diversity of multiple drug resistance reflected by haplotype analysis should not be ignored.

Conclusions
In this study, analysis of six molecular markers of drug resistance among asymptomatic P. falciparum infections revealed a high prevalence of mutations linked to drug resistance in Myanmar. High mutation frequency in pfdhfr, pfdhps, and pfcrt means a serious resistance to SP and CQ in these areas, which supports the ACT being used for P. falciparum treatment in Myanmar. Fortunately, ART resistance has not yet spread to western Myanmar. Given that pfcrt also is a multidrug resistance transporter with high frequency of mutations, it is necessary to continuously monitor for the emergence of ART resistance in western Myanmar. However, the drug resistance pattern of P. falciparum from asymptomatic carriers along the China-Myanmar border is not optimistic; perhaps the best option to prevent the spread of drug-resistant parasites in border areas is rapid elimination through mass drug administration. Haplotype analysis also revealed extremely high diversity in these areas. These asymptomatic P. falciparum infections carrying mutations associated with drug resistance may be an important reservoir for the transmission and spread of resistance in this region. Molecular surveillance of antimalarial resistance might be helpful for developing and updating guidance for the use of antimalarials in Myanmar.

Supplementary Materials:
The following are available online at http://www.mdpi.com/2073-4425/10/9/692/s1, Table S1: PCR primer sequences for the amplification of sequences containing P. falciparum pfk13, pfcrt and pfmdr1 genes, Table S2: PCR primer sequences for the amplification of sequences containing P. falciparum pfdhfr, pfdhps, and pfmrp1 genes, Figure S1: Schematic diagram showing successfully sequenced regions from different genes, Figure S2: Sequencing chromatograms showing single and mixed infections, Figure S3: Phylogenetic clustering of the haplotypes based on the mutations in six genes associated with drug resistance of the asymptomatic P. falciparum isolates.