Introduction

Plasmodium vivax has replaced Plasmodium falciparum as the primary human malaria species in Thailand since 2008 (WHO 2016). In Thailand, the high prevalence of malaria is found near the international borders of Thai-Myanmar, Thai-Cambodian, Thai-Malaysian, and Thai-Laos. The control of P. vivax infection is limited by the relapsing nature of this malaria species due to its hypnozoite stage in the liver. Although resistance of P.vivax to the first-line drug chloroquine has been reported in several parts of the world particularly South East Asia (Marlar et al. 1995; Thanh et al. 2015), there has been no confirmed case of chloroquine-resistant P. vivax in Thailand (Phyo et al. 2011). This may suggest the revolution of P. vivax in adapting its biology to overcome antimalarial action. Genetic analysis of P. vivax essential proteins is important as background information to guide emergence and development of drug resistance.

Studies on the genetic diversity of P. vivax have been limited due to the lack of optimal procedure for continuous culture of the clinical isolates (Panichakul et al. 2007). Analysis of genetic markers is, therefore, a preferable approach to determine antimalarial drug resistance, as well as to distinguish reinfection, relapse, and multiple infections (Kim et al. 2006; Kirchgatter and del Portillo 1998; Veron et al. 2009). Besides, these genetic analyses constitute the essential elements for drug and vaccine development for malaria control. A number of malaria genetic markers have recently been identified including apical membrane antigen 1 (AMA1), which encodes a protein of all malaria species for red cell invasion (Kim et al. 2006; Triglia et al. 2000), and Duffy binding protein (DBP) which encodes a protein on human red cells carrying Duffy-negative blood group (Hamblin and Di Rienzo 2000). However, both genes are not highly polymorphic and, thus, are not suitable for use as genetic markers for parasite diversity. The circumsporozoite surface protein (CSP) which encodes a protein located on the surface of malaria in sporozoite stage (Imwong et al. 2005) is also not suitable to use as a genetic marker due to high variation in the immunodominant T cell epitopes of CSP (Imwong et al. 2005; Souza-Neiras et al. 2010).

Several merozoite surface proteins (MSPs) which encode proteins located on the surface of P. vivax merozoites have been identified including PvMSP1, PvMSP3, PvMSP4, PvMSP5, PvMSP7, and PvMSP10 (Garzon-Ospina et al. 2011; Gomez et al. 2006; Imwong et al. 2005; Putaporntip et al. 2009; Rungsihirunrat et al. 2011). Among these, PvMSP1 and PvMSP3 are highly polymorphic and may be reliably genetic markers for P. vivax diversity (Imwong et al. 2005; Rungsihirunrat et al. 2011). The protein is essential for parasite invasion into human red blood cells. The PvMSP1 gene is divided into 13 blocks; the highly variable blocks are located in the three regions, i.e., block 2 (F1 region), block 6–8 (F2 region), and block 10 (F3 region) (Imwong et al. 2005). It has been reported that antibody against PvMSP1 can inhibit P. vivax invasion to red cells, and this gene is proposed as a potential target for vaccine candidate (Ak et al. 1998; Espinosa et al. 2003). The PvMSP3 is also crucial for parasite red cell invasion; however, the inhibitory effect on red cell invasion was not found using an antibody against PvMSP3. Ten of PvMSP3 multigene families have been identified (Jiang et al. 2013); among these, three original members of PvMSP3 have been demonstrated, i.e., PvMSP3-α, PvMSP3-β, and PvMSP3-γ (Galinski et al. 2001). Genetic analysis of PvMSP3-α and PvMSP3-β has been performed in different P. vivax endemic areas such as Bangladesh, China, Colombia, Korea, India, Pakistan, and Thailand (Bruce et al. 1999; Imwong et al. 2005; Kang et al. 2014; Kibria et al. 2015; Kim et al. 2006; Li et al. 2015; Moon et al. 2009; Raza et al. 2013; Rungsihirunrat et al. 2011; Yang et al. 2006). Both genes serve as reliable genetic tools to distinguish diverse P. vivax isolates and study population dynamics. The present study aimed to investigate the genetic diversity of PvMSP1 and PvMSP3 in P. vivax isolates in malaria-endemic areas of Thailand using PCR-RFLP.

Material and methods

Sample collection

One hundred blood samples were collected during 2014–2015 from patients with P. vivax infection (finger prick) who attended malaria clinics in the western (Kanchanaburi and Tak Provinces) and southern (Ranong Province) regions along the Thai-Myanmar border. Blood samples were spotted onto glass slides for blood smear and filter papers for DNA extraction. Thin and thick blood smears were prepared and examined microscopically for the presence of P. vivax. Approval of the study protocol was obtained from the Human Ethical Review Board of Thammasat University (no. 110/2556).

DNA extraction

The genomic DNA of P. vivax was extracted from dried blood-spotted filter paper using DNA Geneius™ Micro gDNA Extraction Kit (Geneaid Biotech Ltd., New Taipei City, Taiwan) and stored at −20 °C until use.

Amplification of PvMSP1 polymorphisms

The three polymorphic regions of PvMSP1 (PvMSP1-F1, PvMSP1-F2, and PvMSP1-F3) were analyzed according to the previously described method with modification (Imwong et al. 2005). The PvMSP1-F1 polymorphism was performed using nested polymerase chain reaction (nested-PCR). Briefly, the reaction was performed in a total volume of 10 μl consisting of 1× Taq buffered with KCl, 2 mM of MgCl2, 0.125 mM of dNTP, 0.5 μM of each oligonucleotide primer, and 0.125 unit of Taq DNA polymerase (Thermo Scientific, MA, USA). The gene was amplified using 0.5 μl of the extracted genomic DNA template. First amplified product was used as a template for the nested reaction with the same components as first amplification. The PvMSP1-F2 gene was amplified using the nested PCR-RFLP. The reaction (10 μl) consisted of 1× Taq buffered with KCl, 1 mM of MgCl2, 0.125 mM of dNTP, 0.25 μM of each oligonucleotide primer, and 0.125 unit of Taq DNA polymerase (Thermo Scientific, MA, USA). The amplified DNA product was digested (at 37 °C for 16 h) with the restriction enzymes Alu I and Mnl I (New England Biolabs Inc., Hertfordshire, UK). For PvMSP1-F3 amplification, conventional PCR was performed using similar reaction condition as that was described for PvMSP1-F2 except the primers used. The PCR and RFLP products were separated on 1.2 and 1.8% agarose gel, respectively, and visualized under UV illumination after staining with ethidium bromide.

Amplification of PvMSP3 polymorphisms

The PvMSP3 polymorphisms were analyzed using the nested polymerase chain reaction-restriction fragment length polymorphism (nested PCR-RFLP). The primers and reaction conditions used for amplification of the two polymorphic PvMSP3 regions (PvMSP3-α and PvMSP3-β) were according to the previously described method with modification (Rungsihirunrat et al. 2011). Briefly, the total volume of 10 μl reaction consisted of 1× Taq buffer with KCl, 2.5 mM of MgCl2, 0.2 mM of dNTP, 0.2 mM of each oligonucleotide primer, and 0.05 unit of Taq DNA polymerase (Thermo Scientific, MA, USA). The extracted genomic DNA (1 μl) was used as the template for amplification, and the first PCR product was used as the template for the nested amplification. The PCR products were separated on 1.5% agarose gel and visualized under UV illumination after staining with ethidium bromide. After individual digestion with restriction enzymes, Hha I (for PvMSP3-α) and Pst I (for PvMSP3-β) (New England Biolabs Inc., Hertfordshire, UK) in total volume of 20 μl (37 °C for 16 h), the digested products were separated on 1.8% agarose gel electrophoresis and visualized under UV illumination after staining with ethidium bromide.

Purification of PCR products and DNA sequencing

The selected PCR amplification products were purified using QIAquick PCR extraction kits (QIAGEN, Germany) and sequenced by ABI 3730XL DNA Analyzer (Thermo scientific, MA, USA). The DNA sequences were translated to protein sequences by using the translated online program, the Expert Protein Analysis System (ExPASy) biology server (https://web.expasy.org) (SIB, Geneva, Switzerland), and sequence alignment were performed by Clustal X (EBI, UK).

Analysis of allele frequencies and multiple infections

The allele frequency was analyzed according to the PCR patterns. Size of the PCR products and restriction fragments were estimated based on their mobility compared to the standard DNA ladder marker (Fermentas, Thermo Scientific, MA, USA).

The multiple infections were initially considered when more than one PCR products with different sizes were detected in a sample, or when the summed size of the restriction fragments exceeded the size of the PCR products. The multiplicity of infection (MOI) was applied for the determination of the frequency of multiple infections. The MOI is implying the average of the number of different parasite genotype coinfecting in the sample that was calculated by dividing the total number of clones by the number of PCR positive samples (Koepfli et al. 2009).

Determination of parasitemia

Parasite parasitemia was determined according to the WHO guideline (WHO 2014) of which hyperparasitemia is indicated if parasite density is > 100,000 parasites/μl. In present study, parasitemia ranged from 355 to 74,667 parasites/μl and, therefore, was classified as low (< 4286 parasites/μl) and moderate (4286–100,000 parasites/μl) parasitemia.

Statistical analysis

Data analysis was performed using SPSS software version 21.0 (IBM Corporation, NY, USA). The difference in the frequencies of the polymorphic gene alleles and association between parasite density and multiple genotype infections (multiple clone infection) were determined using chi-square and Fisher’s exact test. The statistical significance level was set at α = 0.05 for all tests.

Results

Out of 100 samples, 74 samples were obtained from the western and 26 samples from the southern regions of Thailand. Mono-P. vivax infection was found in all patients.

Genetic diversity of PvMSP1

The three segments (F1, F2, and F3) polymorphisms of PvMSP1 gene were analyzed in 100 P. vivax isolates (Fig. 1a). Eighty-seven and 72 isolates were successfully amplified using specific primers for F1 and F3 segments, respectively (Table 1). Five allelic variants identified based on PCR product sizes were type A (450 bp), B (425 bp), C (400 bp), D (375 bp), and E (350 bp) from the F1 segment (Fig. 1b), of which type C was the predominant variant (n = 31, 35.63%). Eight isolates were identified as multiple infections with more than a single band of PCR fragments (MOI = 1.05). Three allelic variants were classified as type A (325 bp), B (275 bp), and C (250 bp) from the F3 segment (Fig. 1d), of which the predominant variant was B (n = 41, 56.94%).

Fig. 1
figure 1

a Plasmodium vivax merozoite surface protein 1 gene (PvMSP1) consisting of seven conserved region (black) and six variable regions (white) (Imwong et al. 2005). b allelic variants of PvMSP1-F1 (350–450 bp): lanes 1–5 represent type C (400 bp), E (350 bp), A (450 bp), D (375 bp), and B (425 bp), respectively. c Allelic variants of PvMSP1-F2 (1090-1150 bp): lanes 1 and 4 = type B (1090 bp), lanes 2 and 3 = type A (1150 bp). d Allelic variants of PvMSP1-F3 (250–325 bp): lanes 1 and 2 = type C (250 bp), lanes 3 and 4 = type A (325 bp) and B (275 bp), respectively. M = 100 bp markers

Full size image
Table 1 Analysis of the frequency of PvMSP1-F1, PvMSP1-F2, and PvMSP1-F3 allelic variants in 100 P. vivax isolates from malaria-endemic areas of Thailand. Data are presented as percentage values
Full size table

Twenty-seven of selectively amplified product from PvMSP1-F1 (n = 20) and PvMSP1-F3 (n = 7) representative of all the size variants were sequenced and compared to the previously published sequences, Belem and Salvador-1 (AF435594.1 and AF435593.1) (Fig. 2). Nineteen allelic variants were identified from PvMSP1-F1 segment, while seven allelic variants were identified from PvMSP1-F3 fragments.

Fig. 2
figure 2

Alignment of amino acid sequences of PvMSP1-F1 and PvMSP1-F3. a Amino acid sequences of PvMSP1-F1 obtained from selected isolates represent in MSP1-F1-1, MSP1-F1-2, etc. b Amino acid sequences of PvMSP1-F3 obtained from selected isolates represent in MSP1-F3-1, MSP1-F1-2, etc. Sequences were aligned together with P. vivax Belem and Salvador-1 (Sal-1) strains obtained from GenBank (AF435594.1 and AF435593.1, respectively); dots represent identical residues and dashes represent gaps

Full size image

Ninety-seven isolates were successfully amplified using specific primer for the F2 segment with an attempt of repeats. Two types of F2 variants were identified based on the PCR product sizes as A (1,150 bp) and B (1,090 bp) (Fig. 1c). The main allelic variant was B (n = 38, 59%). Due to the large fragment size of PCR product with more than 1 kb and low diversity of this variant, RFLP was used to improve genotyping resolution at nucleotide level by digesting PCR products of the F2 variants with Alu I and Mnl I restriction enzymes (Imwong et al. 2005). Twenty-nine and 25 RFLP patterns were obtained after digestion with Alu I and Mnl I, respectively (Fig. 3), of which 26 and four RFLP patterns identified from the western sample after digested with Alu I and Mnl I, respectively. Twenty-three and three RFLP patterns of Alu I and Mnl I were identified from the samples collected from the southern region, respectively. Twenty-nine and 17 isolates were identified as multiple infections after digested with Alu I and Mnl I, respectively (Table 1). The distribution of PvMSP1-F1 and PvMSP1-F2 digested with Alu I and Mnl I were significantly different between the two endemic areas (p = 0.02, p < 0.001, and p < 0.001, respectively) (Table 1).

Fig. 3
figure 3

Restriction fragment length polymorphism of PvMSP1-F2 fragments digested individually with Alu I and Mnl I restriction enzyme. M1 and M2 represented 100 bp and HyperLadder™ V 25 bp markers, respectively. Lanes 1–11 represented genotypic patterns of PvMSP1-F2 fragments digested with Alu I, and lanes 12–22 was genotypic patterns PvMSP1-F2 fragments digested with Mnl I

Full size image

When combining all the three variables markers of PvMSP1 (F1, F2, and F3), 49 genotypic patterns were found (Fig. 4), 41 and eight genotypic patterns from western and southern isolates, respectively.

Fig. 4
figure 4

Frequencies of PvMSP1 allelic variants in 58 P. vivax isolates from western (a) and southern (b) regions of Thailand. Variants were classified by combination of PvMSP1 (PvMSP1-F1, RFLP allelic variant of PvMSP1-F2 fragments from both Alu I and Mnl I restriction enzyme and PvMSP1-F3)

Full size image

Genetic diversity of PvMSP3

Analysis of the polymorphisms of the two segments (PvMSP3-α and PvMSP3-β) of the PvMSP3 regions was performed in 100 P. vivax isolates (Fig. 5a). Ninety-seven isolates were successfully amplified using specific primer for PvMSP3-α. Three allelic variants were identified as type A (1.9 kb), B (1.5 kb), and type C (1.2 kb) with corresponding frequencies of 80.41%, 8.25% and 9.28%, respectively (Fig. 5b). Two isolates (2.6%) were identified as multiple infections with more than a single band of PCR fragments. For PvMSP3-β, 75 isolates were successfully amplified using the specific primer. Two allelic variants were identified as type A (1.7–2.2 kb) and type B (1.4–1.5 kb) with corresponding frequencies of 52.70 and 29.73%, respectively (Fig. 5c). Thirteen isolates (17.57%) were identified as multiple infections. Type C (0.65 kb) variant was not found as a single genotype but in combination with other genotypes. RFLP was further performed with Hha I enzyme digestion for PvMSP3-α and Pst I for PvMSP3-β after PCR amplification to improve efficiency in the differentiation of PvMSP3 variants at the nucleotide level (Fig. 6). Twenty-seven different genotypic patterns of the PvMSP3-α RFLP fragment were detected with a clear common fragment size of 1000 bp in all isolates, together with distinct smaller fragments ranging from 100 to 500 bp (20 and 13 genotypic patterns from western and southern isolates, respectively). Twenty-five distinct genotypic patterns of the PvMSP3-β RFLP fragment were detected with smaller fragments ranging from 200 to 1500 bp (23 and 8 genotypic patterns from western and southern isolates, respectively) (Table 2). Multiple P. vivax genotype infections with PvMSP3-α and PvMSP3-β were identified in five and 15 isolates from the west and southern regions, respectively. The frequency of PvMSP3-α and PvMSP3-β was significantly different in two endemic areas (p < 0.001 and p = 0.006) (Table 2).

Fig. 5
figure 5

a Plasmodium vivax merozoite surface protein 3 gene (PvMSP3) (Jiang et al. 2013). Ten of PvMSP3 genes labeled as 3.1, 3.2, etc. are represented with black boxes. Areas of two amplified segments including PvMSP3-α (3.10) and PvMSP3-β (3.5) are present in horizontal bold lines. b Allelic variants of PvMSP3-α fragments (1.2–1.9 k bp): lanes 1–2 = type A (1.9 kb), lane 3 = type B (1.5 k bp), and lane 4 = type C (1.2 k bp), respectively. c Allelic variants of PvMSP3-β (0.65–2.2 k bp): lane 1 = type A (1.7–2.2 k bp), lane 2 = type B (1.4–1.5 kb). Multiple infections of type A (1.7–2.2 k bp) and type C (0.65 k bp) isolates are demonstrated in lanes 3 and 4. M = 100 bp marker

Full size image
Fig. 6
figure 6

Restriction fragment length polymorphism of Plasmodium vivax merozoite surface protein 3 includes PvMSP3-α (a) and PvMSP3-β (b). M represented 100-bp marker. a The different allelic variation of PvMSP3-α after PCR-RFLP using Hha I enzyme. b The different allelic variation of PvMSP3-β after PCR-RFLP using Pst I enzyme

Full size image
Table 2 Analysis of the frequency of PvMSP3-α and PvMSP3-β allelic variants in 100 P. vivax isolates from malaria-endemic areas of Thailand. Data are presented as percentage values
Full size table

When combining both marker genes (PvMSP3-α and PvMSP3-β), 54 genotypic patterns were found (Fig. 7), 43 and 11 genotypic patterns from western and southern isolates, respectively.

Fig. 7
figure 7

Frequency of PvMSP3 allelic variants in 73 P. vivax isolates from western (a) and southern (b) parts of Thailand. Variants were classified by combination of RFLP allelic variants patterns between PvMSP3-α and PvMSP3-β

Full size image

Genetic diversity of PvMSP1-PvMSP3 combination

The polymorphic patterns of PvMSP1 (F1, F2, and F3) and PvMSP3 (PvMSP3-α and PvMSP3-β) were successfully analyzed in 46 P. vivax isolates. The combined genotypic analysis of both genes improved efficiency to distinguish genetic variants of P. vivax isolates in Thailand. Results revealed the high genetic diversity of the combined markers with 44 genotypic patterns (95.65%); of these, 38 and six genotypic patterns were identified from western and southern isolates, respectively (Fig. 8).

Fig. 8
figure 8

Frequency of allelic variants by combination of two markers, PvMSP1 and PvMSP3, genotypic from western (a) and southern (b) parts of Thailand

Full size image

Multiplicity of infection

Overall, the majority of P. vivax in this study had a single-clone infection (43%), followed by two clones (34%), three clones (17%), and four clones (6%), respectively. Results of the multiplicity of infection (MOI) of each marker (PvMSP1-F1, PvMSP1-F2, PvMSP1-F3, PvMSP3-α, and PvMSP3-β) were presented in Table 3. High frequencies of multiple genotypes were observed with PvMSP1-F2 (Mnl I) alone (MOI = 1.38) and PvMSP3-β alone (MOI = 1.46). The combination of the subfamily of PvMSP1 (PvMSP1-F1, F2 with Alu I and Mnl I, and F3) and PvMSP3 (PvMSP3-α and PvMSP3-β) showed higher MOI values of 1.5 and 2.0, respectively.

Table 3 Genetic diversity and multiplicity of infection of variable markers (PvMSP1-F1, PvMSP1-F2, PvMSP1-F3, PvMSP3-α, and PvMSP3-β)
Full size table

Association between parasite density and multiple genotype of P. vivax infection

The analysis of parasite density and the multiple genotype infection showed no significant association in individual markers (PvMSP1-F1, p = 0.597; PvMSP1-F2 with Alu I, p = 0.552; PvMSP1-F2 with Mnl I, p = 0.391; PvMSP3-α, p = 0.182; and PvMSP3-β, p = 0.283). However, significant association was found between parasite density and multiple genotypes of P. vivax when combined PvMSP1 (PvMSP1-F1, PvMSP1-F2, and PvMSP1-F2) polymorphic markers (p = 0.032) (Fig. 9).

Fig. 9
figure 9

The relationship between parasite density and multiple genotype infections (multiplicity of infection) by individual markers (PvMSP1-F1, PvMSP1-F2 with Alu I and Mnl I, PvMSP3-α, and PvMSP3-β) (a) and polymorphic markers (PvMSP1 and PvMSP3) (b). Gray bars represent low parasitemia and black bars represent medium parasitemia. Numbers of multiple genotype infections are given in X axis. The asterisk represents significant differences between parasite density and multiple genotypes, p < 0.032

Full size image

Discussion

Genetic diversity of malaria parasites is influenced by several factors including the complex interplay of parasite-parasite, parasite-host, parasite-vector, and geographic regions. The polymorphic genetic markers have been used to describe the diversity of malaria parasite in different endemic areas. Furthermore, the knowledge of parasite diversity could be basic knowledge for developing targets of the malaria vaccine. In the current study, polymorphisms of the two polymorphic genes PvMSP1 and PvMSP3 were investigated for their potential application as marker genes for P. vivax diversity in endemic areas of Thailand. In addition, PvMSP1-F2 fragments, PvMSP3-α, and PvMSP3-β were further analyzed by PCR-RFLP using different endonuclease restriction enzymes to improve genotyping resolution at the nucleotide level. Both genes have been applied as genetic markers to identify the genetic diversity of P. vivax isolates using simple molecular techniques (PCR or PCR-RFLP) without sequencing (Rungsihirunrat et al. 2011).

Three variable markers of PvMSP1 including PvMSP1-F1, PvMSP1-F2, and PvMSP1-F3 were performed in this study. The F1 fragments of PvMSP1, located on variable block 2, showed five different allelic variants (types A–E) with infrequent multiple genotype infection. The F3 fragments, located on variable block 10, showed three different allelic variants observed with the absence of multiple genotype infections. However, results of the F3 fragments were contradictory to that reported in previous studies. The study in Thailand in 2005 showed multiple genotype infections in the F1 and F3 fragments of P. vivax isolates from Bangkok Province or central region (Imwong et al. 2005). Furthermore, two studies conducted in Thailand in 2011 and 2016 revealed multiple genotype infections of F3 fragments in isolates from Tak Province, the west region of Thailand (Kosaisavee et al. 2011; Maneerattanasak et al. 2016). These results suggest the parasite genetic diversity in different time periods and endemic areas even in the same region. Furthermore, the genetic polymorphism of PvMSP1-F1 and PvMSP1-F3 investigated in many regions around the world, particularly in hyperendemic areas such as India (Kim et al. 2006), southern Pakistan (Raza et al. 2013), and Bangladesh (Kibria et al. 2015) demonstrated higher genetic diversity than that observed in this study. According to moderate allelic size variation of PvMSP1-F1 and PvMSP1-F3, sequencing of the amplified fragments was performed, and a higher number of P. vivax with different allelic variants was demonstrated compared with the results obtained from gel electrophoresis. This suggests that sequencing method and alignment of amino acid sequences provide high efficiency to distinguish various allelic variants, although the specific equipment is required with high-cost reagents. On the other hand, the conventional PCR with gel electrophoresis could be a convenient method for general laboratory and research unit. The method is relatively cost-effective and straightforward without the requirement of specific equipment.

The PvMSP1-F2 is located on variable blocks 6–8. The nested PCR analysis of the PvMSP1-F2 revealed two major allelic variants (types A and B) with the large fragment and low genetic diversity. Further analysis using Alu I and Mnl I restriction enzymes revealed 29 and 25 distinct allelic variants of the parasite isolates, with a high proportion of multiple genotype infections (MOI = 1.37 and 1.38). Moreover, a combination of the three subunit variable markers of PvMSP1 gene (PvMSP1-F1, the combination of F2 with Alu I and Mnl I, and F3) improved genotyping resolution to distinguish 49 distinct allelic variants and enhance efficiency to identify multiple infection (MOI of PvMSP1 (F1, F2 and F3) = 1.50). Several genotypic allelic variant patterns of PvMSP1 in the current study showed a marked change in patterns compared with that previously reported study in Thailand in 2005, of which 36 were distinguishable PvMSP1 allelic variants (Imwong et al. 2005). The difference in study areas might explain the change in the pattern of genetic variants, and selective drug pressure of the parasite isolates for their survival leading to genetic change over the different time periods.

The polymorphic markers encoding parasite surface proteins play an essential role in vaccine development. Although high genetic diversity of PvMSP1 was observed in several areas of the world (Kim et al. 2006; Raza et al. 2013; Soares et al. 1997), some studies demonstrated high antigenicity of PvMSP1 (Soares et al. 1997; Yeom et al. 2008) and effective protection against P. vivax (Perera et al. 1998). The study in the animal model by immunization with the antibody against PvMSP1 showed partial protection against P. vivax infection (Valderrama-Aguirre et al. 2005). This suggests that even the PvMSP1 is a potential polymorphic marker and a promising target vaccine. Nevertheless, high genetic diversity may also affect host’s immunity against P. vivax through immune escape. Effective strategies to generate the active antibody with broad activity is required.

PvMSP3 corresponding to tetrapeptide repeat region domain of the merozoite surface protein gene (MSP) consists of three variable markers, PvMSP3-α, PvMSP3-β, and PvMSP3-γ (Galinski et al. 2001). Among these, PvMSP3-α and PvMSP3-β were mostly investigated for genetic markers in epidemiological studies and vaccine candidates. In the present study, three different allelic variants of PvMSP3-α (types A, B, and C) were observed with infrequent multiple genotype infections. Digestion of the PCR product of PvMSP3-α with Hha I restriction enzyme revealed 27 distinct allelic variants. In the previous studies conducted in Thailand in 2003 and 2011, 13 and 14 allelic variants were reported, respectively (Cui et al. 2003; Rungsihirunrat et al. 2011). The observation of high genetic diversity of PvMSP3-α observed in this study agreed with that reported from other endemic regions such as Myanmar (Moon et al. 2009), Pakistan (Khan et al. 2014), and China (Li et al. 2015). For PvMSP3-β, two major allelic variants (types A and B) were found with 13 isolates of multiple infections and 25 different allelic variants. The frequencies of these allelic variants were higher than that previously reported in Thailand in 2006 (Yang et al. 2006) and 2011 (Rungsihirunrat et al. 2011), of which 12 and 14 different variants, respectively, were found. The high genetic diversity of PvMSP3-β in this study was similar to that reported in other hyperendemic areas such as India (Gupta et al. 2013) and Pakistan (Khan et al. 2014) and hypoendemic area such as China (Li et al. 2015). Nevertheless, a study in other hypoendemic areas such as Korea revealed low genetic diversity which may be accounted for by suboptimal seasonal conditions for the transmission of the mosquito vector as well as the reemergence of malaria in Korea in that years (Kang et al. 2014). The high proportion of multiple genotype infections was found in PvMSP3-β after digestion with Pst I (MOI = 1.46). The combination of PvMSP3-α and PvMSP3-β also improved the efficiency to distinguish 54 allelic variants and to identify multiple genotype infections (MOI of PvMSP3 (α and β) = 2.0).

PfMSP3 produces a protective effect from malaria infection through the antibody-dependent cell-mediated mechanism (Audran et al. 2005; Hisaeda et al. 2002). Antibodies against PfMSP3 provide a high level of protection in an Aotus monkey model (Hisaeda et al. 2002; Jones et al. 2001). Various studies on the protective effect of PvMSP3 and its homologs were conducted. Some showed high immunogenicity of PvMSP3-α (Lima-Junior et al. 2012; Stanisic et al. 2013), while most studies showed the high genetic diversity and high prevalence of antibody against PvMSP-3α (Lima-Junior et al. 2011; Lima-Junior et al. 2012).

Multiple genotype infections of P. vivax were observed in this study as indicated by the MOI values. MOI is an average number of distinct parasite genotype coinfection in a patient (Pacheco et al. 2016) associated with the transmission intensity, duration of infection, and individual P. vivax number of blood stages. Although Thailand is considered the hypoendemic area of malaria, the high level of MOI was found in this study. This may suggest the continued persistence of hypnozoites that could be reactivated any time, or the new infection of P. vivax at the same time with the preexistence of P. vivax infection (Hisaeda et al. 2002). The resistance of P. falciparum to antimalarials is of concern.

Apart from the association between MOI and P. vivax transmission, multiple genotype infection was supposed to be associated with severity of infection (de Roode et al. 2005a; de Roode et al. 2005b; Pacheco et al. 2016). Malaria parasite density is one of the important criteria for severe malaria, particularly that caused by P. falciparum. Although, the characterization of severe P. vivax patients has not been well defined compared with P. falciparum due to accumulation of P. vivax-infected RBC in bone marrow (Imirzalioglu et al. 2006; Ru et al. 2009) and spleen (Machado Siqueira et al. 2012), correlation between peripheral vivax parasitemia and P. vivax endothelial activation factors (angiopoietin-2, and von-Willebrand-Factor) were demonstrated (Barber et al. 2015). Enhancing of these endothelial activation factors was associated with the severity of malaria (Barber et al. 2015; Conroy et al. 2010; de Jong et al. 2016). In the present study, no isolates were identified as hyperparasitemia. In order to investigate for the association between the multiple genotypes and severity of malaria infection, the severity of P. vivax parasitemia by parasitemia was classified as low (< 4286 parasites/μl) and moderate (4286–100,000 parasites/μl) levels. A significant association was found when using PvMSP1 as a polymorphic marker. The results support previous studies in the animal model in mice that malaria severity and virulence in multiple genotype infections were higher than single genotype infection (de Roode et al. 2005a, b). In addition, the association between multiple genotype infections of P. vivax and disease severity was also reported in complicated malaria patients in Colombia (Pacheco et al. 2016).

Conclusion

The high genetic diversity and multi-clonogenicity of the P. vivax populations in Thailand were demonstrated using PCR-RFLP. Both PvMSP1 and PvMSP3 can be used as efficient genetic markers to determine genetic polymorphisms of the P. vivax population, as well as to differentiate between multiple clone infection. Results showed high efficiency of the molecular technique to differentiate the genetic diversity of P. vivax that could be useful for identification of relapse, reinfection, and multiple infections in P. vivax isolates in Thailand. Moreover, regardless of the limited number of high parasitemia isolates, the significant association between parasitemia and multiple genotype infections was which could contribute to disease severity in P. vivax infection was found.