Are pvcrt-o and pvmdr1 Gene Mutations Associated with Plasmodium vivax Chloroquine-Resistant Parasites?

(1) Background: Malaria remains a significant global public health issue. Since parasites quickly became resistant to most of the available antimalarial drugs, treatment effectiveness must be constantly monitored. In Brazil, up to 10% of cases of vivax malaria resistant to chloroquine (CQ) have been registered. Unlike P. falciparum, there are no definitive molecular markers for the chemoresistance of P. vivax to CQ. This work aimed to investigate whether polymorphisms in the pvcrt-o and pvmdr1 genes could be used as markers for assessing its resistance to CQ. (2) Methods: A total of 130 samples from P. vivax malaria cases with no clinical and/or parasitological evidence of CQ resistance were studied through polymerase chain reaction for gene amplification followed by target DNA sequencing. (3) Results: In the pvcrt-o exons, the K10 insert was present in 14% of the isolates. Regarding pvmdr1, T958M and F1076L haplotypes showed frequencies of 95% and 3%, respectively, while the SNP Y976F was not detected. (4) Conclusions: Since K10-pvcrt-o and F1076L/T958M-pvmdr1 polymorphisms were detected in samples from patients who responded well to CQ treatment, it can be concluded that mutations in these genes do not seem to have a potential for association with the phenotype of CQ resistance.


Introduction
Malaria is a major important public health problem worldwide.According to the World Malaria Report from the World Health Organization (WHO), there were 247 million new cases and 619 thousand malaria-related deaths in 2021 [1].Plasmodium vivax is the most widely distributed species causing most malaria cases in Asia and South America [2,3].Since 2007, Brazil has observed a stable incidence rate of malaria cases, with P. vivax accounting for about 80% of cases.This trend continued in 2022.The country recorded around 129,000 malaria cases, of which around 83% were attributable to P. vivax and 16%, to P. falciparum [4].In the last 60 years, the combination of chloroquine (CQ) and primaquine (PQ) has been used in Latin America to radically cure P. vivax, i.e., eradicate the blood forms and hepatic hypnozoites [5].However, emerging resistance to antimalarial drugs may threaten malaria control programs [1].
The first report of P. vivax resistance to chloroquine (CQR) in Papua New Guinea dates from 1989 [6], 30 years after resistance reports for P. falciparum [7,8].In Brazil, the first cases of P. vivax resistant to CQ were described in Manaus, Amazonas, in 1999 [9].This timeline reflects the emergence of chloroquine resistance in P. vivax occurring later than in P. falciparum.Since then, studies have reported a 10% prevalence of CQR [10] and a recurrence or reemergence of P. vivax in 5.2% of cases in the same region [11], threatening current international efforts to control and eliminate malaria [1].Given the emerging risk of drug resistance, drug efficacy monitoring studies using molecular markers represent an important tool for refining CQR surveillance and validating potential molecular markers associated with the P. vivax resistance phenotype through mutated single nucleotide polymorphisms (SNPs) in related genes [12].
Currently, the mechanisms of resistance of P. vivax to antimalarial drugs are still unclear due to the lack of continuous in vitro culture systems and the possible involvement of multigenic loci [13].The P. vivax multidrug resistance 1 gene (pvmdr1) and the chloroquine resistance transporter gene (pvcrt-o) are orthologous to two genes described in P. falciparum, the multidrug resistance 1 gene (pfmdr1) and the chloroquine resistance transporter gene (pfcrt), which have been identified as potential markers for CQR in P. vivax [14].
The pvcrt-o gene, described about 20 years ago [15], emerged as a candidate marker of drug resistance and in contrast to pfcrt, only a few SNPs (~10) have been described in pvcrt-o, and most of them were related but occurring at low frequencies [16].Although the sequence polymorphism in the pvcrt-o locus is relatively limited, a lysine (AAG) insertion in the first exon (amino acid 10), originally discovered in Southeast Asian strains, has been found to be associated with a significant reduction in chloroquine (CQ)'s half-maximal inhibitory concentration (IC50) [13,17,18].Thenceforth, studies have demonstrated the presence of the K10 insertion in parasites from Southeast Asia and South America [16,19,20].
In the context of pvmdr1, this gene was identified in 2005 and, due to its strong sequence similarity to pfmdr1, became one of the most important candidate genes investigated in drug susceptibility studies of P. vivax [16].Since then, several SNPs have been identified as potential molecular markers for CQ resistance in P. vivax, including T958M, Y976F and F1076L, which are non-synonymous amino acid mutations associated with resistance to CQ [13,14,20,21].Judging by their protein sequence, the 3 pvmdr1 mutations are located in 9 to 11 domains of the hydrophobic transmembrane [22].The presence of these polymorphisms at codons Y976F and F1076L was registered in malaria-endemic areas where CQ was being used as the first-line antimalarial drug [17,[23][24][25].However, the possibility has been raised that the presence of these two mutations together would be required for the re-emergence of P. vivax resistance to CQ [26].While the Y976F and F1076L polymorphisms are widely distributed in Latin America [27], in Brazil, the increased expression of pvcrt-o and pvmdr1 has been associated with P. vivax resistance to CQ [20].
Considering that knowledge of CQ-chemoresistant P. vivax circulating parasites in Brazilian endemic areas is crucial for predicting the spread of resistant phenotypes and the need for the introduction of alternative therapies, the aim of this study was to investigate the polymorphisms in the pvcrt-o and pvmdr1 genes to identify their potential predictive role of the CQR phenotype in P. vivax samples from Brazilian endemic areas.

Location of the Study and Samples Collected
Samples were collected from January 2018 to August 2022, from P. vivax-infected patients in six Amazonian states (Acre, Amazonas, Amapá, Pará, Rondônia and Roraima).Patients were treated and followed clinically and with laboratory testing to verify if they were cured after chemotherapy treatment at the Outpatient Clinic for Acute Febrile Syndromes/Instituto Nacional de Infectologia (INI), which is part of the Refer-ence Center for Malaria Research, Diagnosis and Training-CPD-Mal/Fundação Oswaldo Cruz (FIOCRUZ)/Rio de Janeiro of Extra-Amazonian (22 • 54 ′ S 43 • 12 ′ W).The treatment adopted for the patients was recommended by the National Program of Malaria Control (PNCM), which comprised the administration of a combination of CQ for 3 days (10 milligrams (mg)/kilogram (kg) on day 1 and 7.5 mg/kg on days 2 and 3) and PQ for 7 days (0.5 mg/kg/day).If patients returned, they were evaluated on days 0, 1, 2, 3, 7, 14, 28 and 42 and in case of symptoms, at any time during the follow-up period.
Besides at CPD-Mal, blood samples were also collected in Manaus (3.1190 • S, 60.0217 • W), the capital of the Amazon state, at the Fundação de Medicina Tropical Doutor Heitor Vieira Dourado (FMT-HVD) and in field conditions in the municipality of Guajará (bordering the Amazonas and Acre states; 02  1 and Figure 1).

Malaria Diagnosis
The malaria diagnoses were made by light microscopy (Giemsa-stained thick blood droplets) in situ.To ensure the presence of a mono-P.vivax infection, together with microscopic diagnosis, all samples were subjected to molecular diagnosis by polymerase

Malaria Diagnosis
The malaria diagnoses were made by light microscopy (Giemsa-stained thick blood droplets) in situ.To ensure the presence of a mono-P.vivax infection, together with microscopic diagnosis, all samples were subjected to molecular diagnosis by polymerase chain reaction (PCR).Firstly, conventional and real-time PCRs were performed using Plasmodium primers [28].Then, the positive samples were submitted to species-specific single or nested PCRs to detect P. vivax [29], P. falciparum [30] and/or P. malariae [31].The samples were stored at the Malaria Research Laboratory (LPM) at Instituto Oswaldo Cruz (IOC), headquarters of the Reference Center for Malaria Treatment and Diagnosis (CPD-Mal/Fiocruz).Only patients with P. vivax mono-infections were included in the study.

DNA Extraction, Amplification and Sequencing
The DNA from 1 mL blood samples was extracted using the QIAamp™ DNA Blood Midi Kit (QIAGEN, Hilden, Germany), according to the manufacturer's instructions.For this study, approximately 1186 base pair (bp) fragments of the pvcrt-o gene of P. vivax were amplified according to the protocol described by Cheong et al. (2020) [14] and a fragment of 800 bp was amplified for the analysis of the SNPs T958M, Y976F and F1076L in the pvmdr1 gene according to the protocol described by Brega et al. (2005) [32].PCR products were analyzed by electrophoresis on 2% agarose gel, visualized under a UV transilluminator (DigiDoc-It; UVP, Unpland, CA, USA) and purified using Wizard™ SV Gel and the PCR Clean-Up System (Promega, Madison, WI, USA), following the manufacturer's procedure.The purified DNA sequencing was carried out through Big Dye™ Terminator Cycle Sequencing Ready Reaction version 3.1 (Applied Biosystems, Carlsbad, CA, USA), with 3.2 µM of forward and reverse PCR primers.DNA sequences to investigate the SNPs in pvcrt-o and pvmdr1 genes were determined using the ABI Prism DNA Analyzer™ 3730 (Applied Biosystems, Carlsbad, CA, USA), at the Fiocruz Genomic Platform PDTIS/Fiocruz RPT01A.Nucleotide sequences were aligned using a ClustalW multiple sequence aligner in BioEdit version 7.7.1 software (North Carolina State University, Raleigh, USA), and the electropherograms were analyzed using NovoSNP ® version 3.0.1 software (University of Antwerp, Antwerpen, Belgium), using the quality cutoff set to 10, in order to avoid the lack of real variation, and using the Salvador 1 strain as a reference sequence (GenBank Accession No. AF314649.1 for pvcrt-o and GenBank Accession No. AY571984.1 for pvmdr1).DNA sequences were deposited in GenBank (the NIH's genetic sequence database; www.ncbi/nlm/nih.gov/GenBankaccessed on 21 August 2023) with the accession numbers OR461289-OR461401.

Prevalence of Polymorphisms in the pvcrt-o Gene
A total of 104 (80%) of the 130 samples were successfully sequenced for the pvcrt-o gene.The insertion of lysine (AAG codon) at position 10, named the K10 insertion and considered a candidate molecular marker of CQR, was detected in 15 (14%) samples: seven from Acre (n = 50; 14%), five from Amazonas (n = 30; 16%), one from Pará (n = 1; 100%) and two from Roraima (n = 19; 11%).Meanwhile, it was not detected in the Rondônia (n = 2) and Amapá (n = 2) (Table 2) samples.In the vast majority of pvcrt-o sequenced samples (n = 89; 86%), no mutations were detected and, therefore, these DNA sequences were identical to the Sal-1 strain used as the wild-type reference of CQ-sensitive parasites.
All patients who attended the Malaria Reference Center located in Rio de Janeiro (CPD-Mal) had uncomplicated malaria and were followed up clinically and through laboratory testing; clinical and parasitological cures were confirmed within the expected time frame (Supplementary Materials Table S1).Using the Malaria Epidemiological Surveillance Information System (SIVEP-Malaria data) and Malaria Control Nacional Program (PMCN) definitions, there was no recrudescence in the six patients from the endemic area who had parasites with the K10 insertion until 60 days after initiating treatment of the primary vivax infection.The three patients I, J and L, were considered cases of reinfections and not recrudescence by CQ-resistant parasites due to P. vivax notifications >60 days after initiating treatment of the primary vivax infection (Supplementary Materials Table S1). 1 NT: total number of samples investigated.

Prevalence of Polymorphisms in the pvmdr1 Gene
The pvmdr1 gene was satisfactorily amplified and sequenced in 113 (86%) of the 130 samples investigated.Three-point mutations (T958M, Y976F and F1076L) considered as potentially associated with the P. vivax resistance phenotype were examined.The SNP T958M was the most prevalent (93%) and was found in parasites from all states.The SNP F1076L was present in only three samples from Amazonas (n = 25; 12%), while the Y976F mutation was absent in all samples examined (Table 3).  1 The bold character represents a non-synonymous mutation detected. 2NT: total number of samples investigated.
The three P. vivax patients harboring double mutant MYL parasites in pvmdr1, whose samples were followed up clinically at the CPD-Mal laboratory, were cured within the expected time (Table 5).

Comparative Alignment between Here Presented pvcrt-o and pvmdr1 Sequences of CQ-Sensitive Parasites with Those Reported by Melo et al. (2014) [20]
When we analyzed our CQ-sensitive sequences with those that were CQ-resistant of Melo et al. [20], we found that 14% of the sensitive pvcrt-o gene sequences had the K10 insertion against only two resistant isolates presenting the K10 insertion (Figure 2).In relation to the pvmdr1 gene, 92% of CQ-sensitive isolates presented the single mutant MYF haplotype whereas 100% of CQ-resistant or -sensitive sequences [20] contained the MYF haplotype (Figure 3).

Discussion
P. vivax is the most geographically widespread pathogen of human malaria and is responsible for most cases outside the African continent [1].Vivax malaria is considered a public health problem in many parts of the world, particularly with regard to the morbidity of children and pregnant women [33], the rare but real possibility of fatal

Discussion
P. vivax is the most geographically widespread pathogen of human malaria and is responsible for most cases outside the African continent [1].Vivax malaria is considered a public health problem in many parts of the world, particularly with regard to the morbidity of children and pregnant women [33], the rare but real possibility of fatal  [20].The sequences named TYF correspond to the isolates with the wild-type haplotype sequenced in this study.Sequences named MYL and MYF correspond to the isolates with one and two non-synonymous mutations detected in this study, respectively.AC: Acre; AM: Amazonas; AP: Amapá; PA: Pará; RO: Rondônia; RR: Roraima.

Discussion
P. vivax is the most geographically widespread pathogen of human malaria and is responsible for most cases outside the African continent [1].Vivax malaria is considered a public health problem in many parts of the world, particularly with regard to the morbidity of children and pregnant women [33], the rare but real possibility of fatal infections [16] and the presence of CQ-resistant P. vivax parasites [5].Due to the biological characteristics of P. vivax that prevent a continuous in vitro culture system, molecular monitoring has become the tool of choice for the surveillance of resistance to antimalarials due to its practical and economic advantages over in vivo fieldwork and in vitro assays, recommended by the WHO [1,13].Previous work has shown a significant association between the allele pfcrt and pfmdr1 variants and P. falciparum drug resistance [25].However, putative mutations in these genes, considered as candidate markers of antimalarial drug resistance, are not clear with respect to P. vivax [34,35].In fact, regardless of the chemoresistance phenotype, the genetic polymorphism of P. vivax is notable [36] and this parasite is known to be more diverse than P. falciparum [37].As, so far, little is known about the genotype of P. vivax parasites circulating in Brazilian endemic areas, thus, it is quite pertinent to investigate the polymorphisms of pvmdr1 and pvcrt-o.
The main gene related to P. vivax CQ resistance is pvcrt-o.Indeed, the presence of the "AAG" lysine insert in exon I of pvcrt-o (referred as the "K10 insert") has been associated with a significant reduction in the IC50 of CQ [18].In our analysis, it was possible to identify the K10 insertion in 15 samples (14%) distributed among the states of Acre (7), Amazonas (5), Pará (1) and Roraima (2).This SNP was also reported in studies from French Guiana (57%) [38], China (32%) [39] and, to a lesser extent, India (5.6%) [40], Pakistan (16%) [41] and Thailand (18%) [42].Due to the scarcity of previous studies, we do not have knowledge about the expansion or retraction of this mutation in Brazil.One possibility would be that this mutation has been introduced to Brazil through the events of illegal mining in French Guiana, where this insertion is frequent, although not directly associated with CQR [38].In fact, when we analyzed the results of samples from patients who carried P. vivax with the K10 insertion, a good response to the CQ treatment occurred within the 28-day expected period, with no observed case of recrudescence.Thus, excluding the small possibility of primaquine acting against minority populations of CQR parasite clones, our data reinforce that the K10 insertion is not a predictor of the P. vivax resistance phenotype to treatment with CQ, as previously suggested [13,43,44].
Regarding the pvmdr1 gene, several studies describe that the substitution in the Y976F codon, which changes the amino acid tyrosine to phenylalanine, would be associated with a reduced susceptibility to CQ [39,[45][46][47][48].This SNP has even been identified in endemic areas of China, Cambodia and Ethiopia [19,24,47], where treatment failure has been reported in patients with vivax malaria treated with CQ.Although this SNP was not detected in the present study nor in 2018 in the triple border region involving Colombia, Peru and Brazil/Amazonas [49], previous studies by our group reported the Y976F polymorphism at a high frequency (85.7%) in 2009 [26] and at a low frequency (15%) from 2010 to 2014 [50] in the endemic areas of the Legal Amazon.These facts seem to indicate two possibilities of events: either the SNP Y976F was present in only a few locations in the endemic region of the Legal Amazon or this mutation was not becoming fixed in the Brazilian endemic region.On the other hand, since we did neither detect the Y976F polymorphism in any of the samples here examined, nor any therapeutic failure to CQ in this casuistic, we cannot draw a conclusion about the role of this polymorphism as a marker of P. vivax CQR.
In addition to Y976F, we also investigated SNPs at the T958M and F1076L codons in the pvmdr1 gene, both of which have also been proposed to be potentially associated with CQ resistance [20,34].The SNP at codon T958M was present in almost all analyzed samples (92%), including those of the 34 patients whose parasite genes could be sequenced and responded to CQ treatment within the expected period.This high frequency has been verified by previous studies carried out by our group in Brazilian endemic areas [26,50], as well as by other authors analyzing samples from Africa [51], South America [50] and Southeast Asia [19], indicating the fixation of T958M in parasitic populations of P. vivax circulating in Brazil and in the world.This fact indicates a low potential of this polymorphism as a CQR marker.
It has been mentioned that the T958M and F1076L polymorphisms alone may not be associated with CQR, unless if acting together, they could modify the protein's conformation and enable drug evasion [32,51].When analyzing these double mutants, we verified that the F1076L polymorphism was always present with T958M, originating the MYL double mutant haplotype.This dependency relationship between these alleles was already noted by our group in isolated parasites from Acre, Amazonas, Rondônia and in autochthonous cases from the Atlantic Forest of Rio de Janeiro [50], as well as by other authors in Southeast Asia [52][53][54], where there are cases of P. vivax resistant to CQ treatment.However, once again, the good therapeutic response in patients carrying MYL parasites in the present study, disregarding the small possibility of primaquine acting in minority clones of CQR parasites, indicates that the association of double mutants with P. vivax CQR is unlikely.
As the Y976F mutant allele was not detected in our series, it was not possible to establish a relationship between chemoresistance and CQ for the TYL double mutant haplotype (T958M + Y976F + F1076L), as previously suggested in the literature [39,55,56].
Finally, the tendency toward tolerance or resistance to the drugs recommended for the treatment of vivax malaria may be more complex, including other genes in addition to P. falciparum, where pvmdr1 SNPs located in the transmembrane domain are associated with CQ vacuole efflux.It is noteworthy that striking differences in the topologies and numbers of SNPs in these transporter genes between P. vivax and P. falciparum reinforce the idea that mechanisms other than mutations may explain this CQ-resistant phenotype in P. vivax.In fact, the P. vivax CQ resistance process may also differ by its combined CQ/PQ treatment regimen, so it could be only a matter of time before more CQ-resistant P. vivax cases appear.

Conclusions
The results presented in this study reinforce that the mutations here investigated on pvcrt-o and pvmdr1 may not be good markers of P. vivax chemoresistance to CQ.New approaches for the identification of robust genetic markers for monitoring chloroquine resistance in P. vivax populations are needed.

Supplementary Materials:
The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/biomedicines12010141/s1,Table S1: Patients followed up at CPD-Mal and SIVEP-Malaria Database Platform, carrying P. vivax parasites containing the insertion of K10 in pvcrt-o, according to the state of infection.

Figure 3 .
Figure 3. Alignment of representative CQ-resistant and -sensitive pvmdr1 sequences, with the SNPs investigated highlighted in red (T958M, Y976F and F1076L).Melo R1 to R12 correspond to CQ-resistant sequences of Melo et al.[20].The sequences named TYF correspond to the isolates with the wild-type haplotype sequenced in this study.Sequences named MYL and MYF correspond to the isolates with one and two non-synonymous mutations detected in this study, respectively.AC: Acre; AM: Amazonas; AP: Amapá; PA: Pará; RO: Rondônia; RR: Roraima.

Figure 3 .
Figure 3. Alignment of representative CQ-resistant and -sensitive pvmdr1 sequences, with the SNPs investigated highlighted in red (T958M, Y976F and F1076L).Melo R1 to R12 correspond to CQ-resistant sequences of Melo et al.[20].The sequences named TYF correspond to the isolates with the wild-type haplotype sequenced in this study.Sequences named MYL and MYF correspond to the isolates with one and two non-synonymous mutations detected in this study, respectively.AC: Acre; AM: Amazonas; AP: Amapá; PA: Pará; RO: Rondônia; RR: Roraima.

Figure 3 .
Figure 3. Alignment of representative CQ-resistant and -sensitive pvmdr1 sequences, with the SNPs investigated highlighted in red (T958M, Y976F and F1076L).Melo R1 to R12 correspond to CQresistant sequences of Melo et al.[20].The sequences named TYF correspond to the isolates with the wild-type haplotype sequenced in this study.Sequences named MYL and MYF correspond to the isolates with one and two non-synonymous mutations detected in this study, respectively.AC: Acre; AM: Amazonas; AP: Amapá; PA: Pará; RO: Rondônia; RR: Roraima.

Table 1 .
Localities of P. vivax parasite blood collection by Brazilian states.

Table 2 .
Distribution of the K10 insertion (AAG codon) in the pvcrt-o gene in 104 samples of P. vivax from the Legal Amazon.

Table 3 .
Frequency of alleles in the pvmdr1 gene in 113 samples of P. vivax, according to the malaria diagnosis localities.

Table 4 .
Distribution of haplotypes in the pvmdr1 gene in 113 samples of P. vivax from the Legal Amazon.

Table 5 .
Patients followed up at CPD-Mal, carrying P. vivax parasites containing the MYL double pvmdr1 mutant haplotype infected in the Amazonas state and the dates of diagnosis and of cure (parasitological and molecular negative assays).
1Cure: based on parasitological and molecular negative assays.