Molecular Detection of Antimalarial Drug Resistance in Plasmodium vivax from Returned Travellers to NSW, Australia during 2008–2018

To monitor drug resistance in Plasmodium vivax, a multidrug resistance 1 (Pvmdr1) gene and a putative transporter protein (Pvcrt-o) gene were used as molecular markers for chloroquine resistance. The biomarkers, the dihydrofolate reductase (Pvdhfr) gene and the dihydropteroate synthetase (Pvdhps) gene, were also used for the detection of resistance to sulphadoxine-pyrimethamine (SP); this drug is often accidentally used to treat P. vivax infections. Clinical blood samples (n = 120) were collected from patients who had been to one of eight malaria-endemic countries and diagnosed with P. vivax infection. The chloroquine resistance marker, the Pvmdr1 gene, showed F976:L1076 mutations and L1076 mutation. A K10 insertion in the Pvcrt-o gene was also found among the samples successfully sequenced. A combination of L/I57:R58:M61:T117 mutations in the Pvdhfr gene and G383:G553 mutations in the Pvdhps gene were also observed. Mutations found in these genes indicate that drug resistance is present in these eight countries. Whether or not countries are using chloroquine to treat P. vivax, there appears to be an increase in mutation numbers in resistance gene markers. The detected changes in mutation rates of these genes do suggest that there is still a trend towards increasing P. vivax resistance to chloroquine. The presence of the mutations associated with SP resistance indicates that P. vivax has had exposure to SP and this may be a consequence of either misdiagnosis or coinfections with P. falciparum in the past.


Introduction
In Australia, about 500 malaria cases are reported each year [1] and just under half of them are diagnosed with Plasmodium vivax [2]. All cases are in travellers who have been to malaria-endemic countries. Plasmodium falciparum remains the predominant malaria species, but P. vivax has become the major cause of human malaria in endemic regions outside of Africa, such as Asia, Central and South America and Oceania. Globally, it is estimated that there are between 80 million and up to 400 million vivax malaria cases each year [3][4][5]. Although malaria control and elimination programs are promoted of our patients were male (89/120) and most were in the 21-30 years age group (27/88) (see Table 1). Only 23 females were present in this study with most in the 21-30 and over 61 years age groups. Nine samples had no gender recorded. Pvmdr1 amplicons (n = 109) were successfully sequenced and displayed mutations at codon 976 and 1076. As shown in Table 2, double F976:L1076 mutations were observed in all regions: Oceania (23/37), Southeast Asia (5/10) and South Asia (2/73). In Oceania, PNG showed the highest prevalence of mutations, with 22/29 of isolates (see Table S2). The single L1076 mutation was also detected in all three regions: South Asia (50/73), Oceania (8/37) and Southeast Asia (5/10). This mutation showed the highest prevalence in India (36/51) and Pakistan (14/17) (see Table S2). All single F976 mutations were found in India (7/51) (see Table S2) and nine isolates from the South Asia region were identified as wildtype. A K10 insertion in the Pvcrt-o gene was identified in nine isolates from India (South Asia) and two isolates from Southeast Asia. No other isolate from this study group had this insertion (see Table 2).

Sequence Analysis of The Dihydropteroate Synthetase (Pvdhps) Gene
Amplicons (n = 116) that were successfully sequenced had mutations at codons 382, 383, 512, 553 and 585. The wildtype genotype was found in all three regions, the majority of which were in South Asia (62/73), Oceania (35/37) and small numbers in Southeast Asia (5/10). Two mutations at G383:G553 loci were found in South Asia (6/73) (all six came from India), Southeast Asia (2/10) and Oceania (1/37). Isolates showing the single G383 mutation were identified in South Asia (2/73) and Southeast Asia (1/10). The single G553 mutation was only found in one isolate from South Asia and one from Southeast Asia.

The Prevalence of Tandem Repeat Variants in the Dihydrofolate Reductase (Pvdhfr) Gene
Tandem repeat variants found in the Pvdhfr gene locus have been described by different laboratories [26]; the typing of these variants in this study followed that of a recent Thai study from 2018 [7]. This group described three different tandem repeat variants found in the Pvdhfr gene sequence. The reference strain (Accession No. X98123.1) [27] contains three repeat sets of four amino acids (5'-GGDN-3') between codons 88 and 103 and was designated as Type 1 or wildtype. The tandem repeat sequence that lacks the six amino acid at codons 98-103 was designated as Type 2. Type 3 sequences have an additional six amino acids inserted between codons 103 and 104.

Countries
Pvdhfr gene Type 1 Type 2 Type 3 South Asia (SA): India and Pakistan; Southeast Asia (SEA): Indonesia, Thailand, Cambodia plus South Korea (includes one case with exposure to multiple countries of SEA (SEAN)); Oceania: Papua New Guinea and Solomon Islands.

Discussion
About 30 years ago, the first chloroquine resistance in P. vivax was found [8] and subsequently, it appeared in several other countries such as the Solomon Islands, Indonesia, Malaysia, Thailand, Ethiopia, Brazil and Peru [15]. Molecular markers of chloroquine resistance in P. vivax, Pvmdr1 and Pvcrt-o genes were used as they are orthologous to Pfmdr1 and Pfcrt genes, which are associated with chloroquine resistance in P. falciparum [6]. Among the studied samples herein, the presence of the F976 and L1076 mutations in the Pvmdr1 gene and the K10 insertion in the Pvcrt-o gene were shown previously to be related to chloroquine resistance in P. vivax by in vivo and in vitro studies [5,7,16,28,29].
This study obtained sequences from both Pvmdr1 (109/120) and Pvcrt-o (88/120) genes and mutations were found in all geographical regions of South Asia, Southeast Asia and Oceania. The F976:L1076 mutations were more frequent in Oceania (23/37) in comparison to the Southeast Asian (5/10) and South Asian (2/73) regions. A large number of isolates (50/73) from South Asia carried only the L1076 mutation and a few isolates from the Southeast Asia (5/10) and Oceania (8/37) regions. The single F976 mutation was detected only in seven isolates from the South Asian region.
The finding of the F976:L1076 mutations in Oceania is, however, different from earlier studies in PNG between 2006 and 2011; these studies showed the presence of the F976 mutation only [30][31][32]. This suggests that there was a rise in the number of chloroquine resistance markers since the first reported case of resistant P. vivax in the region, even though chloroquine has not been recommended in the antimalarial drug policy since 2009 [33]. In addition, the high prevalence of the L1076 mutation in the South Asian region was consistent with other recent studies in India and Pakistan with respective mutations of 54.5% and 98.0% [34,35]. Only two isolates carried the F976:L1076 mutations in the Indian study [34], whereas the previous study in Pakistan did not find either the F976 mutation or the F976:L1076 mutations [35]. Furthermore, if the hypothesis of a two-step mutation trajectory is true, then the L1076 mutation occurs first, followed by the F976 mutation before the emergence of chloroquine resistance [17,36,37]. The finding in the present study with the double F976:L1076 mutations is an indicator that fully developed chloroquine-resistant strains of P. vivax are present. A case report of a chloroquine-resistant P. vivax infection in a pregnant woman living in Pakistan [38] would further support the molecular finding in this study although mutations in the Pvmdr1 gene were not investigated. The increased rate of these mutations suggests that chloroquine resistance in P. vivax will become more common in the South Asia region where chloroquine has remained the first-line drug [39]. Our findings from clinical isolates from Southeast Asia were also similar to other studies in this location [5][6][7]16,40]. A clinical report of chloroquine-resistant P. vivax in a Thai pregnant woman indicated that recurrent parasites during Day 21 and 143 carried the double F976:L1076 mutation [40]. This double mutation was found in Thailand in 2008 (49.2%) and again in 2014 (18.4%) [7]. The same double mutations were found in 15.4% of samples in Myanmar between 2009 and 2016 [5]. This could indicate that chloroquine-resistant P. vivax is not only present in Thailand but also in neighbouring Myanmar.
Unlike other well-described mutations in the Pvmdr1 gene, the role of the K10 insertion in Pvcrt-o in chloroquine resistance in P. vivax is unclear because this mutation is present in some studies and absent in others [3,5,7,16,34,41]. In Thailand, the K10 insertion was found in 2008 [16], but was not detected again until recently [7,42]. In our study, Indian samples consistently showed the insertion from 2013 to 2017 (see Table S3). In addition to variations in finding the K10 insertion in the Pvcrt-o gene, a recent study suggests that the absence of this mutation (wildtype sequence) in isolates is associated with chloroquine resistance [43]. This would indicate that in our study, the 77 isolates from all three regions which carry the wildtype Pvcrt-o gene are of the chloroquine-resistant genotype.
The findings herein show that the chloroquine-resistant-associated genotypes are present in South Asia, Southeast Asia and Oceania, but chloroquine remains the drug of choice for treatment against P. vivax in India, Pakistan, Thailand and South Korea [21]. Reports of resistant genotypes [38,40,[44][45][46] and the increased mutation frequency found in this study do suggest that resistant phenotypes will continue to emerge in these regions.
A combination of sulfadoxine and pyrimethamine (SP) with other antimalarial drugs is not recommended for treating P. vivax [46]. Yet this species of malaria shows emerging mutant genotypes suggesting exposure to SP drug pressures [5,7]. In order to identify SP resistance in P. vivax, mutations in the Pvdhfr and Pvdhps genes were used. These genes are orthologues to Pfdhfr and Pfdhps genes in SP-resistant P. falciparum [5,7,10]. Furthermore, these gene targets have been correlated with SP resistance in P. vivax [47]. An in vitro study using yeast models which contained the Pvdhfr gene insertion showed that the IC 50 value of that yeast expressing the L/I57:R58 mutations was sevenfold higher than yeast expressing the wildtype gene [48]. Furthermore, genetically modified yeasts containing R58:N117 mutations and those with triple mutations of the Pvdhfr gene were 200-300-fold more resistant to SP than those with the wildtype gene [48]. The L/I57:R58:M61:T117 mutations in the Pvdhfr gene showed a high level of pyrimethamine resistance as the IC 50 value was over 500-fold in this yeast model [48]. This quadruple mutation was also linked to higher rates of therapeutic failure of SP in a volunteer group of patients than those isolates with wildtype, single, double and triple mutations in their gene sequences [49].
The present study found the L/I57:R58:M61:T117 mutations in a large number of isolates from Oceania (21/37), which was followed by the Southeast Asian (4/10) and South Asian (2/73) regions. Our findings were consistent with other studies from these regions [5,7,32,[50][51][52]. An increase in treatment failure by 12% occurred in P. vivax cases in PNG when the malaria treatment policy in 2000 changed to recommending the use of SP with either chloroquine or amodiaquine as the first-line drug for uncomplicated malaria [53]. India, the largest country in the South Asian region, continues to use SP with other drugs to treat uncomplicated P. falciparum malaria [23]. In the South Asian region, the R58:N117 mutations were present (18/73) in similar numbers found in past studies confirming the higher prevalence of the R58:N117 mutations (28.16%) than the L/I57:R58:M61:T117 mutations (19.71%) [51]. It was hypothesised that the N117 mutation occurs in parasites as an early development towards SP resistance [54]. This is followed by the R58 mutation, while the other two mutations might occur independently [55]. Therefore, a rise in early stages of SP-resistant mutations indicates ongoing SP drug pressure on P. vivax in the South Asian region.
Tandem repeat variants were found during the analyses of the Pvdhfr gene in the present study. The predominance of the tandem repeat Type 1 (wildtype sequence) occurs in three regions tested and the presence of low numbers of Type 2 was found in South Asia (9) and Southeast Asia (3). Type 3 variants, which contain four repeat sets of 5'-GGDN-3', were identified in Oceania (19), South Asia (4) and Southeast Asia (1). Other studies have found that all three types are present in India [54,55], whereas Indonesia, Thailand and South Korea have only Type 1 and Type 2 [7,52]. The role of the tandem repeat variants in SP resistance remains unclear [7,56] as these variants are not present in the Pfdhfr gene in P. falciparum [26,57]. However, Type 3 variants were exclusively linked to the L/I57:R58:M61:T117 mutations, which are associated with pyrimethamine resistance in P. vivax. There have been suggestions that these variants are involved in the development of mutations and higher levels of the resistance and therefore may be useful molecular markers [20,51]. Interestingly, 22/24 isolates carrying the Type 3 variant in this study had the L/I57:R58:M61:T117 mutations (see Table S2).
A large number of the studied isolates herein (103/120) from all three regions were found to be the wildtype Pvdhps genotype. However, nine isolates from South Asia (6/73), Southeast Asia (2/10) and Oceania (1/37) had the G383:G55 double mutations, which are linked to sulfadoxine resistance in P. vivax [5,7,58]. The current results correlate with the findings made in previous studies from these regions [7,31,32,51,52,55,59]. Some studies [47,60] showed that the mutations in the Pvdhps gene were more than twice as likely to emerge in isolates with multiple mutations in the Pvdhfr gene. However, only 3/120 of the isolates in the current study were found to carry the G383:G55 mutations when multiple mutations were present in the Pvdhfr gene (see Table S2).
Gene mutations will be influenced by the different drug policies of each country and these decisions will cause different drug pressures on the parasite genome. The presence of SP-resistant genotypes reflects the use of SP in the past and indicates a misguided use of SP for treating vivax malaria in those countries where SP is currently used. Nevertheless, the findings in our study are limited in scope because samples were collected over a 10-year period and malaria drug policies of each country may have changed during this period. The travel history associated with the clinical samples was not always complete and visits by these patients to multiple endemic countries at the time of infection cannot be excluded. In addition, patients' clinical profiles were not available, and drugs used for treatment could not be determined. Sample numbers from various countries such as Thailand, Solomon Islands, South Korea and Cambodia were not adequately represented.

Ethics Statement and Sample Collection
Clinical blood samples were collected during the time period of 2008-2018 from patients diagnosed with malaria at NSW Health Pathology-ICPMR, Westmead Hospital. At the time of diagnosis, all patients had recently been to malaria-endemic countries, such as India, PNG, Pakistan, Indonesia, Thailand, Solomon Islands, South Korea and Cambodia. There was one patient with no specific country of travel recorded because the patient had travelled to multiple countries within Southeast Asia (SEAN). Plasmodium vivax was diagnosed in these samples by an experienced microscopist and if the microscopic results were unclear, a nested Polymerase Chain Reaction (PCR) was used to determine the malaria species [61]. These clinical blood samples were used with the approval of the human ethics review committee for research in human subjects, research and development office, Western Sydney Local Health District (WSLHD) Research Governance Officer (HREC reference No. LNR/18/WMEAD/139 and SSA reference No. LNRSSA/18/WMEAD/140).
A total of 120 samples with P. vivax infections were included in this study and QIAamp mini DNA extraction kits were used to extract malaria DNA from the blood samples according to the manufacturer's instructions. Extracted DNA was stored at −20 • C until required.

Sequencing of The Multidrug Resistance 1(Pvmdr1) Gene, The Putative Transporter Protein (Pvcrt-o) Gene, The Dihydrofolate Reductase (Pvdhfr) Gene and The Dihydropteroate Synthetase (Pvdhps) Gene
Specific oligonucleotide primers listed in Table 5   A nested polymerase chain reaction (nested PCR) was used to amplify the Pvdhfr and Pvdhps genes with two pairs of oligonucleotide primers shown in Table 5. The outer PCR reactions were performed in a total volume of 25 µL per reaction containing: 0.2 mM of each dNTP, 1x of PCR buffer, 0.2 µM of each primer, 1.75 mM of MgCl 2 and 1 U of Taq polymerase. Extracted DNA (3 µL) was added to this PCR which was run under the following conditions: 94 • C for 5min, then 20 cycles of [94 • C for 30 s, 58 • C for 30 s, 68 • C for 60 s] and final annealing at 68 • C for 5 min. Products from the outer PCR reaction (2 µL) were added into the nested PCR reaction which contained 2 mM of dNTP mix, 1x of PCR buffer, 0.2 µM of each primer, 1.75 mM of MgCl 2 and 1 U of Taq polymerase. The following reaction condition was used: 94 • C for 5 min followed by 40 cycles of [94 • C for 30 s, 58 • C for 30 s, 68 • C for 45 s] with a final step at 68 • C for 5 min. All amplicons were analysed by electrophoresis on a 1.5% agarose gel.
The amplicons of Pvmdr1, Pvcrt-o, Pvdhfr and Pvdhps were sequenced by Macrogen in South Korea. All DNA sequences were assembled using the Sequencher 5.3 program. The assembled DNA sequences were aligned and translated to amino acid sequences using the MEGA 7 program and compared to Genbank reference strains (Reference No. Pvdhfr: XM001615032 [62] and X98123.1 [27], Pvdhps: XM001617159 [62], Pvcrt-o: AF314649 [12] and Pvmdr: AY618622 [18]) in order to identify the presence of amino acid changes in the expected protein. All nonsynonymous mutations in the targeted genes identified in this study are shown in Tables 6 and 7.

Conclusions
High prevalence of chloroquine-resistance-associated genotypes highlights the severe situation of chloroquine resistance in P. vivax in Oceania and South Asian and Southeast Asian regions. Genetic monitoring for-drug resistant P. vivax in travellers returning to Australia allows pre-empting rising treatment failures in these endemic countries. Furthermore, this data will serve as an early warning for imported cases of chloroquine-resistant strains of P. vivax so that alternative drug therapies can be instigated. The finding of mutations in Pvdhfr and Pvdhps genes suggests that SP drug pressure on P. vivax is already present, even though this drug is not recommended for use against P. vivax in these regions.

Supplementary Materials:
The following are available online at http://www.mdpi.com/2076-0817/9/2/101/s1, Table S1: The number of successful sequences obtained from 120 clinical samples, Table S2: Total of targeted sequences obtained from clinical samples, Table S3: The table of mutations in four genes found in Indian samples between 2008 and 2018, Table S4: The table of