Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Mar 2.
Published before final editing as: Int J Parasitol. 2025 Dec 9:S0020-7519(25)00226-7. doi: 10.1016/j.ijpara.2025.12.001

Genetic Signatures of Plasmodium vivax Circumsporozoite Surface Protein During Malaria Resurgence in Thailand

Parsakorn Tapaopong a, Aurel Holzschuh b,c, Gustavo da Silva d, Palakorn Chintanawiwat a, Sirasate Bantuchai a, Wasinee Rungsarityotin e, Liwang Cui f, Cristian Koepfli d, Jetsumon Sattabongkot a, Wang Nguitragool a,g
PMCID: PMC12949852  NIHMSID: NIHMS2129360  PMID: 41380756

Abstract

The Plasmodium circumsporozoite surface protein (CSP) is the best characterized pre-erythrocytic vaccine target for malaria. It is a multifunctional protein important for sporozoite mobility, mosquito salivary gland invasion, and hepatocyte invasion. We analyzed diversity of Plasmodium vivax CSP gene (pvcsp) during the 2022–2023 malaria resurgence in northwestern Thailand and assessed how pvcsp haplotypes may affect parasite development in the mosquitoes. Amplicon sequencing of 69 P. vivax isolates revealed both canonical pvcsp variants: VK210 (n = 66) and VK247 (n = 3). The VK210 type exhibited high polymorphism within the central repeat region, with 21 haplotypes (H1–H21) composed of 13 to 20 repeat motifs. Haplotype H2 was the most common, accounting for half of all VK210 sequences, and in membrane feeding assays with Anopheles dirus, appeared to produce more salivary-gland sporozoites per oocyst than other haplotypes, suggesting that repeat-region variation may modulate vector competence. Together, these findings report contemporary pvcsp diversity in Thailand’s highest transmission area, provide functional evidence that repeat-region polymorphisms shape vector-parasite interactions, and highlight three globally prevalent motifs (GDRADGQPA, GDRAAGQPA, ANGAGNQPG) as prime targets for future PvCSP vaccines.

1. Introduction

Plasmodium vivax threatens over 2.5 billion people globally and remains the most prevalent malaria parasite species outside of Africa (WHO, 2024b). Its persistence is largely due to the parasite’s ability to remain dormant in the liver for months or years before reactivating. In Thailand, P. vivax is the primary cause of malaria, with the highest burden in Tak Province on the Myanmar border (Department of Disease Control, 2024). Cross-border migration in this region complicates malaria control efforts. Since the endorsement of malaria elimination program in 2016, Thailand has made steady progress in reducing malaria incidence, but resurgence occurred in 2022–2023, with the case number in 2023 nearly quintuple that in 2021 (Department of Disease Control, 2024). This trend suggests a setback in malaria elimination efforts, highlighting the need to reassess and strengthen intervention strategies.

Vaccines are regarded as a key strategy for advancing malaria elimination, and several P. vivax vaccine candidates are in development (Vargas-Parada, 2023). The Plasmodium circumsporozoite protein (CSP) is a leading target, with vaccines being explored under various platforms (Salman et al., 2017; Stoute et al., 1997; Venkatraman et al., 2025). However, the genetic diversity of csp genes poses a challenge to vaccine efficacy, particularly that of P. vivax csp (pvcsp), due to the presence of at least two immunogenically distinct subtypes, VK210 and VK247, which can be found across different regions (Burkot et al., 1992; Need et al., 1993; Wirtz et al., 1992). Understanding its genetic structure in endemic populations is crucial for the design of vaccines targeting circulating variants. The csp gene is expressed during the sporogonic stage and plays a critical role in the invasion of liver cells in the vertebrate host (Cerami et al., 1992). Targeting CSP could block the pre-erythrocytic stage of the parasite’s life cycle, thereby providing sterile protection. CSP also plays a role in immune evasion within the mosquito vector (Zhu et al., 2022) and in the invasion of mosquito salivary glands (Myung et al., 2004), highlighting its critical function in both invertebrate and vertebrate hosts during malaria transmission.

Using amplicon sequencing of the nearly full-length pvcsp gene, we investigated its genetic variation in northwestern Thailand, a region with consistently highest malaria burden in the country, and compared it against patterns from other regions. The potential association between pvcsp variants and parasite sporogonic development within the mosquitoes was also examined.

2. Materials and Methods

2.1. Study site, sample collection, and ethical consideration

Blood samples were collected in Tha Song Yang District, Tak Province, the most malarious hotspot in Thailand, located along the border with Myanmar. In 2024, the malaria incidence rate in this district was 7.522 per 1,000 population, compared to only 0.173 per 1,000 population nationwide (Department of Disease Control, 2024). The study population consisted of residents of the villages near the border who engaged in agricultural activities that increase their exposure to Anopheles mosquitoes.

Twenty milliliters of whole blood were collected from patients attending the malaria clinics, after obtaining informed consent. Parasite species were first confirmed microscopically. Subsequently, P. vivax genomic DNA was isolated from leukocyte-depleted blood samples, and species confirmation was performed using nested PCR targeting the 18S rRNA gene (Ngernna et al., 2019). The use of these samples and associated data for the current study was approved by the Ethics Committee of the Faculty of Tropical Medicine, Mahidol University, Bangkok (approval number MUTM 2021–067).

2.2. PCR amplification of pvcsp

A total of 69 P. vivax genomic DNA samples were collected from Tha Song Yang District, Tak Province, in 2022 (n = 46) and 2023 (n = 23) (Fig. 1A). These samples were collected during the recent P. vivax resurgence (Fig. 1B). The pvcsp gene was amplified using forward (3rd_Fwd_CSP: 5’-GGCCATAAATTTAAATGGAG-3’) and reverse (3rd_Rev_CSP: 5’-ATGCTAGGACTAACAATATG -3’) primers modified from (Bibi et al., 2021) (Fig. 1C). The PCR mixture was prepared using 10 μM of forward and reverse primer stocks, 2X Phusion® High-Fidelity Master Mix with HF Buffer (Thermo Fisher Scientific, USA), and 5 μl of genomic DNA in a final reaction volume of 50 μl. Thermocycling conditions were as follows: pre-denaturation at 98°C for 2 minutes, followed by 35 cycles of denaturation at 98°C for 10 seconds, annealing at 52°C for 1 minute, and extension at 72°C for 1 minute. A post-extension step was performed at 72°C for 10 minutes. The PCR products were then purified using the GeneapHlow™ Gel/PCR Kit (Geneaid, Taiwan) according to the manufacturer’s protocol.

Fig 1.

Fig 1.

Open in a new tab

(A) Samples were collected from Tak Province, the region with the highest malaria endemicity in Thailand. (B) The accompanying line chart shows the number of P. vivax cases in Tak Province and nationwide over the past decade, highlighting a decline from 2015 to 2021, followed by a resurgence starting in 2022. (C) The schematic diagram of pvcsp illustrates the N-terminal, central repeat, and C-terminal regions, along with the locations of forward and reverse primers (red arrows) used for amplicon deep sequencing. A forward primer specific for Sanger sequencing of the repetitive region is indicated by a green arrow.

2.3. DNA library construction, sequencing, and post-NGS data processing

Purified PCR amplicons were used to prepare DNA libraries using the QIAseq® FX DNA Library UDI-A Kit (Qiagen, Germany) according to the manufacturer’s protocol. Briefly, 10 μl of the purified pvcsp amplicons were fragmented to approximately 350 bp by incubating with 2 μl of 10X FX Buffer and 2 μl of FX Enzyme Mix at 32°C for 7 minutes. This step also included end-repair and A-addition. The fragmented amplicons were then subjected to adapter ligation, where barcode sequences were added by mixing the amplicons with 3 μl of DNA adapter, 12 μl of 5X Ligation Buffer, and 6 μl of DNA Ligase. The ligation reaction was incubated at 20°C for 15 minutes, followed by an immediate cleanup using 0.9X QIAseq® Beads (Qiagen, Germany).

For the amplification of library DNA, a mixture was prepared using 25 μl of 2X HiFi PCR Master Mix and 1.5 μl of 10 μM Primer Mix. The cycling conditions for library amplification were as follows: initial denaturation at 98°C for 2 minutes, followed by 2 cycles of denaturation at 98°C for 20 seconds, annealing at 60°C for 30 seconds, and extension at 72°C for 30 seconds, with a final extension at 72°C for 1 minute. The amplified library was then purified using 1X QIAseq Beads before quantification for quality control using the QIAseq® Library Quant Assay (Qiagen, Germany). AmpSeq was performed on an Illumina MiSeq in paired-end mode using the MiSeq® v3 Reagent Kit (600 cycles, PE) (Illumina, CA), with a 5% spike-in of PhiX Control v3 (Illumina, CA).

Raw FASTQ files containing forward (R1) and reverse (R2) reads were generated for 69 parasite isolates using Illumina MiSeq. Sequence quality was assessed using FastQC (Zakeri et al., 2006), and only quality-passed reads were aligned to the P01 reference genome using BWA (Li & Durbin, 2009). The resulting alignments were stored in BAM format, which was then sorted by chromosomal order using samtools sort (Li et al., 2009), and converted to variant call format (VCF) files using bcftools mpileup. Single nucleotide polymorphisms (SNPs) were identified against the P01 reference sequence using bcftools call (Li et al., 2009). Variants that did not meet the read quality (QUAL) and sequencing depth (DP) criteria (QUAL < 30, DP < 20) were filtered out. The quality-passed variants in the VCF files were then used to generate a consensus sequence with bcftools consensus (Li et al., 2009), and the sequences were stored in FASTA format for further analysis.

The repetitive nature of the central repeat region of the pvcsp presents significant challenges for accurate alignment and assembly, primarily due to the ambiguity in read placement relative to the reference sequence. To address these challenges, in addition to sequencing on the Illumina, the central repeat region was also sequenced using standard dye-terminator Sanger sequencing (Macrogen, Korea). The primer 2nd_Fwd_CSP (5’-ACGTGAAAATAAGCTGAAACAACCA-3’) was used for sequencing, providing longer reads that can span the entire repetitive sequences, thereby improving the accuracy of the sequence data for this region.

2.4. pvcsp sequence analysis

The P. vivax P01 and PNG strains were used as the references for the VK210 and VK247 variants, respectively. The VK210 variant (PlasmoDB accession: PVP01_0835600) spans 1,092 bp and is organized into three distinct regions: the N-terminal region (positions 1–288 bp), the central repeat region (positions 289–747 bp), and the C-terminal region (positions 748–1,092 bp). In contrast, the VK247 variant (NCBI accession: M69059) spans 1,188 bp and is divided into the N-terminal region (positions 1–312 bp), the central repeat region (positions 313–852 bp), and the C-terminal region (positions 853–1,188 bp).

All analyses were based on the major clone of each isolate. Sequence polymorphism in the N- and C-terminal regions was analyzed by calculating several genetic diversity metrics: the number of segregating sites (S), number of mutations (η), nucleotide diversity (π), average number of nucleotide differences (k), number of haplotypes (H), haplotype diversity (Hd), and their corresponding standard deviations (SD), all computed using DnaSP v5.1 software. Tests for deviations from neutral evolution were conducted using Tajima’s D (Taj D), Fu and Li’s D* (FLD), and F* (FLF) statistics, along with nucleotide substitution rates, analyzed in MEGA v10.2.6 software (Stecher et al., 2020).

The repetitive central repeat region was processed and visualized separately using SnapGene Viewer v6.2.1, with low-quality sequences outside the repeat regions manually trimmed for accuracy. Distinct haplotypes in the central repeat region were grouped using DnaSP v5.1 software. Each haplotype was then translated into its corresponding protein sequence and classified based on its peptide repeat motifs (PRMs) using ApE v3.0.9 software (Davis & Jorgensen, 2022).

2.5. Quantification of parasitemia, oocyst density, and sporozoite load

Parasitemia was quantified by examining 1 μl of Giemsa-stained thick films from P. vivax malaria patients under a microscope. All blood-stage forms including rings, trophozoites, schizonts, and male and female gametocytes were counted, and parasitemia was calculated per 5 × 106 red blood cells. To quantify sporogonic development, membrane feeding assays were conducted following a published method (Sattabongkot et al., 2015) and the oocyst density and sporozoite load counted. Briefly, Anopheles dirus mosquitoes were reared at the Mahidol Vivax Research Unit (MVRU), Faculty of Tropical Medicine, Mahidol University, Thailand. The infected patient blood was centrifuged at 3,000 rpm for 5 minutes and the plasma supernatant removed. The red blood cell pellet was washed with Roswell Park Memorial Institute (RPMI 1640) medium, centrifuged again at 3,000 rpm for 5 minutes, and resuspended in AB serum. The final suspension was aliquoted for mosquito feeding at a volume of 500 μl per 100 mosquitoes. Mosquito feeding was conducted using a membrane feeder covered with a wet sheep cecum membrane and maintained at 37 °C using a thermocirculator. To assess infection, oocyst density per mosquito was determined on day 7 post-feeding by randomly selecting 10 to 25 infected mosquitoes and calculating the mean oocyst count for each isolate. Midguts were stained with mercurochrome and examined under a microscope to quantify oocysts. From day 14 post-feeding onward, salivary glands from infected mosquitoes were dissected under a microscope, pooled into a 1.5-mL Axygen tube, washed twice with dissecting medium (Ham’s F-12/MEM supplemented with 10% penicillin-streptomycin and 25 μg/mL fungizone), and washed once more with complete medium. The glands were then grinded using a sterile plastic pestle to release sporozoites, which were counted using a hemocytometer to determine sporozoite load per mosquito.

2.6. Sequences of pvcsp global isolates

The pvcsp sequences from diverse parasite populations were retrieved from the NCBI GenBank to compare genetic diversity between northwestern Thailand and global isolates. The dataset included sequences from Brazil (n = 42, DQ978648–DQ978689) (Santos-Ciminera et al., 2007), Colombia (n = 28, GU339059–GU339086) (Hernandez-Martinez et al., 2011), India (n = 79, FJ491063–FJ491141), Iran (n = 50, KT588159–KT588208) (Zakeri et al., 2006), South Korea (n = 39, DQ859734–DQ859772), Pakistan (n = 34, MT222297–MT222330), Papua New Guinea (n = 18, EU031819–EU031836) (Henry-Halldin et al., 2011), and Vanuatu (n = 24, AB539022–AB539045) (Kaneko et al., 2014).

3. Results

3.1. Genetic polymorphism and test of neutrality of the N-terminal and C-terminal regions in northwestern Thailand pvcsp

A total of 69 P. vivax isolates were collected from northwestern Thailand in 2022 (n = 46) and 2023 (n = 23). The pvcsp gene was successfully amplified and sequenced by Illumina for all isolates. Based on the amino acid sequences of the central repeat region, 66 isolates were classified as VK210 variants and 3 as VK247 variants. The sequences of the N- and C-terminal regions can be found in Supplementary Fig. S2.

Sequence analysis revealed that the N-terminal region of pvcsp exhibited greater variation than the C-terminal region, as evidenced by the number of haplotypes (h = 6 vs. 3), nucleotide diversity (π = 0.005 vs. 0.001), and haplotype diversity (Hd = 0.500 vs. 0.058) (Table 1). No significant values were detected for Tajima’s D or Fu and Li’s tests in the overall dataset.

Table 1.

Genetic diversity of pvcsp N- and C-terminal region

Region Year n S η k H Hd ± SD π ± SD Taj D FLD FLF
N-terminus 2022 46 4 4 0.994 4 0.467 ± 0.070 0.005 ± 0.001 0.2135 −0.0758 0.0134
2023 23 3 3 1.306 5 0.581 ± 0.093 0.005 ± 0.001 0.6987 −0.1742 0.0797
Overall 69 4 4 0.995 6 0.500 ± 0.033 0.005 ± 0.001 0.4137 −0.1788 0.0797
C-terminus 2022 46 3 3 0.130 2 0.043 ± 0.041 0.001 ± 0.001 −1.7042 −2.9710* −3.0158*
2023 23 1 1 0.087 2 0.087 ± 0.078 0.001 ± 0.001 −1.1610 −1.5907 −1.6916
Overall 69 3 3 0.115 3 0.058 ± 0.039 0.001 ± 0.001 −1.5719 −1.8543 −2.0647
Open in a new tab

N, number of sequences; S, number of segregating sites; η, number of mutations relative to P01; π, nucleotide diversity; k, average number of nucleotide differences; H, number of haplotypes; Hd, haplotype diversity; SD, standard deviation); Taj D, Tajima’s D value; FLD, Fu and Li’s D* value; FLF, Fu and Li’s F* value

Comparing the two years, the haplotype diversity of the parasite population was higher in 2023 than in 2022 for both the N-terminal region (Hd = 0.581 vs. 0.467) and the C-terminal region (Hd = 0.087 vs. 0.043) (Table 1). Interestingly, Fu and Li’s test for the C-terminal region of pvcsp in the 2022 isolates showed significant negative values, suggesting natural selection or population expansion (Table 1).

3.2. Sequence polymorphism of the central repeat region in northwestern Thailand pvcsp

The central repeat region of all 69 P. vivax isolates was successfully sequenced by the conventional Sanger approach. Sixty-six isolates were classified as VK210 variants (GDRA-like motifs) and 3 as VK247 variants (ANGA-like motifs). The central repeat region of these isolates displayed significant length polymorphism, with 21 distinct haplotypes identified among the 66 VK210 isolates (Fig. 2A) and 3 haplotypes among the VK247 isolates (Fig. 2B).

Fig 2.

Fig 2.

Open in a new tab

Variation within central repeat region. (A) VK210, showing 21 distinct haplotypes (H1–H21), and (B) VK247, revealing 3 haplotypes (A1–A3). Each circle represents a specific nonapeptide, with its color corresponding to the unique amino acid sequence. (C) Among the 21 VK210 haplotypes derived from 66 variants, half belong to haplotype H2, while the remaining variants are classified as non-H2 haplotypes. (D) A haplotype network constructed using the Median-Joining algorithm illustrates the relationships among the 21 VK210 haplotypes (H1–H21) from isolates in northwestern Thailand. Each circle represents a haplotype, with the size of the circle reflecting its frequency. The distances between circles indicate genetic relatedness, with hatch marks denoting the number of nucleotide-base changes.

In the VK210 isolates, 15 different types of peptide repeat motifs (PRMs) were identified: GDRADGQPA, GDRAAGQPA, GNGAGGQAA, GDRANGQPA, GDRADGQAA, GNGAGGQPA, GDGAAGQPA, GDRAAGQAA, GDGAGGQAA, GDRATGQPA, GARADGQPA, GDKAAGQPA, GNKAAGQPA, GDREAGQPA, and GDKADGQPA. Among them, GDRADGQPA and GDRAAGQPA were the most common. The 15 PRMs were arranged in various patterns and copy numbers within the central repeat region, resulting in 21 unique haplotypes (H1–H21). The copy number of PRMs in the central repeat region of the VK210 variants varied from 13 to 20. Haplotype H2 was the most prevalent, with a frequency of 50% and 33 mutual isolates (Fig. 2C and 2D). Haplotype H2 was then assigned as the dominant haplotype, while the remaining 20 haplotypes were noted as minor haplotypes. With exception of a few, most haplotypes started with either GDRADGQPA or GDRAAGQPA, and ended with GNGAGGQAA.

Among the three VK247 isolates, five distinct PRMs were identified: ANGAGNQPG, ANGADDQPG, ANGAGDQPG, ANEAGNQPG, and ANGAGGQAA. All of these motifs have all been reported previously (Henry-Halldin et al., 2011; Hernandez-Martinez et al., 2011; Kaneko et al., 2014; Zakeri et al., 2006). These PRMs combined to form 3 haplotypes (A1–A3), each represented by a single isolate. Haplotype A1 has been reported previously from Iran (Zakeri et al., 2006), whereas haplotypes A2 and A3 appear to represent newly observed variants. The PRM ANGAGNQPG was the predominant motif in the central repeat region of the VK247 variants from northwestern Thailand. Notably, all VK247 haplotypes started with ANGAGNQPG and ended with ANGAGGQAA.

3.3. Phenotypic comparison of northwestern Thailand pvcsp H2 and non-H2 haplotypes

Of the 66 VK210 variants, 54 were evaluated for mosquito infectivity through direct membrane feeding assays, including 25 isolates carrying the H2 haplotype and 29 isolates with non-H2 haplotypes. Blood parasitemia was microscopically counted using Giemsa-stained thick blood films and showed no difference (p = 0.76, Mann-Whitney U test) between patients infected with H2 haplotype isolates (n = 25) and non-H2 haplotype isolates (n = 29) (Fig. 3A, Supplementary material: Table S2). Mosquito infectivity was assessed by measuring oocyst density and sporozoite loads in An. dirus. The number of oocysts per mosquito, assessed by dissecting the midgut on day 7 post-feeding, showed no difference (p = 0.94, Mann-Whitney U test) between mosquitoes fed with H2 haplotype isolates (n = 25) and non-H2 haplotype isolates (n = 29) (Fig. 3B, Supplementary material: Table S2). The sporozoite rate per mosquito, determined by dissecting the salivary glands on day 14 post-feeding, also showed no difference (p = 0.57, Mann-Whitey U test) between H2 isolates (n = 19) and non-H2 isolates (n = 15) (Fig. 3C, Supplementary material: Table S2). Interestingly, the sporozoite conversion rates, estimated by dividing the mean salivary gland sporozoite count by the mean oocyst count from the same parasite isolate, were higher in H2 isolates (n = 16) than in non-H2 isolates (n = 14) (p = 0.05, Mann-Whitney U test) (Fig. 3D, Supplementary material: Table S2).

Fig 3.

Fig 3.

Open in a new tab

Phenotypic differences between parasites carrying haplotype H2 and non-H2 haplotypes. (A) Blood parasitemia was quantified from thick blood smears and standardized to 5 × 106 red blood cells. (B) Oocyst density per mosquito was determined by dissecting midguts on day 7 post-blood feeding. (C) Sporozoite rate per mosquito was assessed by dissecting salivary glands on day 14 post-feeding. (D) The sporozoite conversion rate was calculated by dividing the mean salivary gland sporozoite count by the mean oocyst count for each parasite isolate. Data are presented on a logarithmic scale on the y-axis.

3.4. Global variation and distribution of PRMs within pvcsp

The proportion of VK210 and VK247 variants differs across endemic countries (Fig. 4A). Among the 69 isolates from Thailand, 96% belonged to VK210, while 4% were VK247. A similar trend was observed in Pakistan (n = 34), Iran (n = 50), and Papua New Guinea (n = 18), where VK210 was the dominant variant, with frequencies of 94%, 80%, and 67%, respectively, while VK247 correspondingly accounted for 6%, 20%, and 33% of the isolates (Henry-Halldin et al., 2011; Zakeri et al., 2006). In contrast, all isolates from Brazil (n = 41), India (n = 79), and South Korea (n = 39) were exclusively VK210 (Santos-Ciminera et al., 2007). Notably, VK247 was the predominant variant in Colombia, accounting for 89% (25/28) of the isolates, while the remaining 11% (3/28) were VK210 (Hernandez-Martinez et al., 2011).

Fig 4.

Fig 4.

Open in a new tab

(A) VK210 and VK247 diversity of pvcsp in northwestern Thailand and global P. vivax isolates. Bars represent the proportional distribution of pvcsp allelic types (VK210 and VK247) across endemic countries. The right panel shows length polymorphism in the central repeat region of pvcsp in (B) VK210 and (C) VK247 variants from northwestern Thailand and global P. vivax isolates. Each circle represents an individual isolate. Dash lines represent median values.

The length of PRMs varies among global isolates. In VK210 (Fig. 4B), isolates from Thailand exhibited PRM lengths ranging from 13 to 20, with a median of 17. Brazil had PRM lengths between 18 and 20, with the median at 20. All three VK210 isolates from Colombia had a PRM length of 20. Isolates from India (13 to 19), Iran (16 to 20), South Korea (18 to 20), Papua New Guinea (18 to 19), and Vanuatu (15 to 19) displayed varying PRM lengths, while all Pakistan isolates had a uniform PRM length of 18. In VK247 (Fig. 4C), Thai isolates had PRM lengths ranging from 19 to 20 (median = 20). Colombia exhibited the widest range, with PRM lengths from 5 to 19 (median = 19). All VK247 isolates from Iran and Pakistan had a PRM length of 19. Papua New Guinea isolates ranged from 18 to 21 (median = 20), while those from Vanuatu ranged from 19 to 20, with a median of 19.

Four of the 21 VK210 haplotypes identified in northwestern Thailand (H2, H5, H11, and H13) were also found in global P. vivax populations, including those in Oceania and South Asia (Fig. 5). Distribution across the four endemic regions revealed that each shared at least one haplotype with northwestern Thailand. Specifically, H2 was detected in Papua New Guinea, while Vanuatu exhibited a high frequency of H11 along with H13. In Iran, H13 was the predominant haplotype, while in India, H13 was the predominant haplotype with a minor presence of H5. Notably, none of the VK247 haplotypes from northwestern Thailand were shared with other endemic regions.

Fig 5.

Fig 5.

Open in a new tab

Distribution of shared pvcsp VK210 haplotypes between northwestern Thailand and global isolates. Donut charts show the frequencies of shared haplotypes, labeled with their geographic origins: THA for Thailand, PNG for Papua New Guinea, VUT for Vanuatu, IRN for Iran, and IND for India. The letter “n” represents the number of isolates analyzed from each region.

4. Discussion

Malaria incidence in northwestern Thailand had been in steady decline until 2021, but there was a surge in 2022, peaking in 2023, and the elevated case number persists until now. Our genetic analysis reveals higher diversity in both the N- and C-terminal regions of pvcsp in 2023 compared to 2022. A similar link between genetic diversity and transmission intensity has been observed in other genes such as pvmsp1 and pvdbp (Tapaopong et al., 2025). The N-terminal region of pvcsp in northwestern Thailand in 2022–2023 exhibited somewhat higher haplotype diversity (0.500) than neighboring endemic regions, including Vietnam (0.102 in 2018–2019), Myanmar (0.096 in 2013–2015), and Cambodia (0.500 in 2006–2007). Conversely, haplotype diversity in the C-terminal region was much lower in northwestern Thailand (0.058) compared to Vietnam (0.383 in 2018 – 2019), Myanmar (0.186 in 2013–2015), and Cambodia (0.159 in 2006–2007) (Parobek et al., 2014; Vo et al., 2020; Vo et al., 2022). The disparity in haplotype diversity between the N- and C-terminal regions, relatively higher in the N-terminus but strikingly low in the C-terminus, may reflect a unique situation of resurgence which occurred in Thailand.

The significantly negative Fu and Li’s statistics in the C-terminal region indicate more singleton mutations than expected under neutrality, even with very few segregating sites. This is supported by negative Tajima’s D values, which also reflect an excess of low-frequency variants. Although such patterns can result from either population expansion or purifying selection, the negative but nonsignificant dN - dS estimate and the extremely low overall diversity of this region are more consistent with purifying selection as the primary force shaping variation in the C-terminal domain of pvcsp. Notably, the low diversity of the C-terminal domain in PvCSP contrasts with PfCSP whose T-cell epitopes (Good et al., 1988; Hoffman et al., 1989) are polymorphic hotspots (Lockyer et al., 1989; Zeeshan et al., 2012), suggesting that sequence variation in this region likely poses a smaller challenge for vaccine development in P. vivax than in P. falciparum.

Regarding the biological function of CSP, the N-terminal region was evident to mediate mosquito salivary gland recognition (Sidjanski et al., 1997) and shields the thrombospondin repeat (TSR) of the C-terminal domain to preserve sporozoite motility (Coppi et al., 2011), whereas the C-terminal domain is essential for oocyst egress (Wang et al., 2005), sporozoite gliding, and infectivity in vertebrate host (Tewari et al., 2002).

Analysis of the central repeat region showed that nearly all P. vivax isolates collected in northwestern Thailand during 2022–2023 were of the VK210 variant. Among the 66 VK210 isolates, 21 unique PRM haplotypes were detected, and one of them (haplotype H2) accounted for half of all samples. Importantly, the dominance of H2 does not by itself imply a selective fitness advantage; it may reflect other process such as genetic drift or cross-border introduction. In direct membrane feeding experiments, however, H2 isolates exhibited a higher sporozoite-conversion rate, producing more salivary-gland sporozoites per oocyst in An. dirus than other haplotypes. Although this advantage may help explain the local dominance of H2, its causal role in parasite fitness remains to be established. The elevated conversion efficiency nevertheless suggests that pvcsp polymorphism may influence vector–parasite interactions, possibly through sporozoite development, egress, migration, or invasion of the mosquito salivary glands. However, because the p-value for sporozoite conversion lies exactly at the threshold of statistical significance (p = 0.05), this finding should be interpreted cautiously and regarded as a preliminary signal that warrants confirmation in a larger sample to strengthen the conclusion.

A recent study has reported much lower pvcsp diversity in southern Thailand, with only three haplotypes of the central repeat region identified among 88 isolates (Khulmanee et al., 2024). Haplotype H11 was the most prevalent, found in 81 out of 88 isolates, whereas H2 was entirely absent. These distinct haplotype distributions are consistent with the large geographical separation between the two regions and may reflect differences in epidemiological histories and local vector species compositions (Sriwichai et al., 2016; Yanmanee et al., 2023). Because our feeding assays assessed An. dirus competence only for northwestern isolates, we cannot directly evaluate whether vector-specific compatibility contributes to the contrasting patterns. Future feeding experiments that test multiple mosquito species against a broader panel of pvcsp haplotypes would help clarify whether vector-parasite interactions underlie these regional differences.

The only malaria vaccines currently endorsed by the World Health Organization, RTI,S/AS01 and R21/Matrix-M, target the P. falciparum CSP (PfCSP) (WHO, 2024a). The P. vivax homolog, PvCSP, is the leading target for P. vivax vaccine with multiple platforms being under investigation, including recombinant protein subunit vaccine (Gimenez et al., 2021), virus-like particle vaccine (Salman et al., 2017), and mRNA-lipid-nanoparticle vaccine (our unpublished work). Here we identified haplotype H2 as the dominant VK210 variant in northwestern Thailand, a pattern that mirrors earlier observation from Papua New Guinea (Henry-Halldin et al., 2011).

Other regions are characterized by different major haplotypes, such as H11 in southern Thailand and Vanuatu (Kaneko et al., 2014; Khulmanee et al., 2024) and H13 in Iran and India (Zakeri et al., 2006). Despite the geographical heterogeneity, two PRM motifs (GDRADGQPA and GDRAAGQPA) make up the vast majority of repeat units worldwide, whereas ANGAGNQPG is the principal motif of the VK247 variant. To date, no study has evaluated whether variation in PRM sequence, repeat number, or arrangement within within VK210 or VK247 affects the inhibitory efficacy of antibodies raised against PvCSP. It is therefore prudent for vaccine development efforts to incorporate these globally prevalent motifs into PvCPS vaccine designs to maximize the likelihood for strain-transcendencing immunity.

In summary, the post-2021 rebound of P. vivax in northwestern Thailand is accompanied by a distinctive evolutionary signature at the pvcsp locus: moderate variability persists in the N-terminal domain while the C-terminal domain is strongly conserved. In addition, a single VK210 repeat haplotype (H2) now dominates, possibly supported by its superier oocyst-to-sporozoite transition within the mosquitoes. The sharp geographic contrast with southern Thailand, where H11 prevails, points to the potential role of local vector ecology and epidemiological history in shaping haplotype structure. Yet, on the global scale, only three repeat motifs (GDRADGQPA, GDRAAGQPA, ANGAGNQPG) account for most of the repeat region of pvcsp, providing a concise antigenic focus for next-generation PvCSP vaccines. Ongoing genomic surveillance of pvcsp, coupled with functional assays, could therefore be useful for anticipating transmission shifts and informing rational vaccine design.

Data Deposition

The sequencing data of this project has been deposited at National Center for Biotechnology Information (NCBI) under the accession PV774584 – PV774652.

Supplementary Material

1

Highlights.

  1. Dominant H2 haplotype showed a trend toward increased infectivity in Anopheles dirus, producing more Plasmodium vivax sporozoites per oocyst.

  2. The first study to demonstrate that pvcsp repeat-region polymorphisms was linked to mosquito-stage parasite development.

  3. Three conserved and globally prevalent pvcsp repeat motifs identified as promising targets for future P. vivax vaccine development.

  4. High-resolution pvcsp diversity mapping during malaria resurgence in Thailand, identifying 21 VK210 haplotype locally.

Funding Sources

This research was supported by the National Institute of Allergy and Infectious Diseases, National Institutes of Health (U19 AI181583), and Mahidol University through the Fundamental Fund for fiscal year 2023, provided by the National Science Research and Innovation Fund (NSRF). Additional support was provided by the ICTM grant from the Faculty of Tropical Medicine, Mahidol University.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Bibi Z, Fatima A, Rani R, Maqbool A, Khan S, Naz S, & Waseem S (2021). Genetic characterization of Plasmodium vivax isolates from Pakistan using circumsporozoite protein (pvcsp) and merozoite surface protein-1 (pvmsp-1) genes as genetic markers. Malar J, 20(1), 112. 10.1186/s12936-021-03654-w [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Burkot TR, Wirtz RA, Paru R, Garner P, & Alpers MP (1992). The population dynamics in mosquitoes and humans of two Plasmodium vivax polymorphs distinguished by different circumsporozoite protein repeat regions. Am J Trop Med Hyg, 47(6), 778–786. 10.4269/ajtmh.1992.47.778 [DOI] [PubMed] [Google Scholar]
  3. Cerami C, Frevert U, Sinnis P, Takacs B, Clavijo P, Santos MJ, & Nussenzweig V (1992). The basolateral domain of the hepatocyte plasma membrane bears receptors for the circumsporozoite protein of Plasmodium falciparum sporozoites. Cell, 70(6), 1021–1033. 10.1016/0092-8674(92)90251-7 [DOI] [PubMed] [Google Scholar]
  4. Coppi A, Natarajan R, Pradel G, Bennett BL, James ER, Roggero MA, Corradin G, Persson C, Tewari R, & Sinnis P (2011). The malaria circumsporozoite protein has two functional domains, each with distinct roles as sporozoites journey from mosquito to mammalian host. J Exp Med, 208(2), 341–356. 10.1084/jem.20101488 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Davis MW, & Jorgensen EM (2022). ApE, A Plasmid Editor: A Freely Available DNA Manipulation and Visualization Program. Front Bioinform, 2, 818619. 10.3389/fbinf.2022.818619 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Department of Disease Control, Thailand M. o. P. H.. (2024). Thailand Malaria Elimination Program. https://malaria.ddc.moph.go.th/malariar10 [Google Scholar]
  7. Gimenez AM, Salman AM, Marques RF, Lopez-Camacho C, Harrison K, Kim YC, Janse CJ, Soares IS, & Reyes-Sandoval A (2021). A universal vaccine candidate against Plasmodium vivax malaria confers protective immunity against the three PvCSP alleles. Sci Rep, 11(1), 17928. 10.1038/s41598-021-96986-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Good MF, Pombo D, Quakyi IA, Riley EM, Houghten RA, Menon A, Alling DW, Berzofsky JA, & Miller LH (1988). Human T-cell recognition of the circumsporozoite protein of Plasmodium falciparum: immunodominant T-cell domains map to the polymorphic regions of the molecule. Proc Natl Acad Sci U S A, 85(4), 1199–1203. 10.1073/pnas.85.4.1199 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Henry-Halldin CN, Sepe D, Susapu M, McNamara DT, Bockarie M, King CL, & Zimmerman PA (2011). High-throughput molecular diagnosis of circumsporozoite variants VK210 and VK247 detects complex Plasmodium vivax infections in malaria endemic populations in Papua New Guinea. Infect Genet Evol, 11(2), 391–398. 10.1016/j.meegid.2010.11.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Hernandez-Martinez MA, Escalante AA, Arevalo-Herrera M, & Herrera S (2011). Antigenic diversity of the Plasmodium vivax circumsporozoite protein in parasite isolates of Western Colombia. Am J Trop Med Hyg, 84(2 Suppl), 51–57. 10.4269/ajtmh.2011.09-0785 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Hoffman SL, Oster CN, Mason C, Beier JC, Sherwood JA, Ballou WR, Mugambi M, & Chulay JD (1989). Human lymphocyte proliferative response to a sporozoite T cell epitope correlates with resistance to falciparum malaria. J Immunol, 142(4), 1299–1303. https://www.ncbi.nlm.nih.gov/pubmed/2464643 [PubMed] [Google Scholar]
  12. Kaneko A, Chaves LF, Taleo G, Kalkoa M, Isozumi R, Wickremasinghe R, Perlmann H, Takeo S, Tsuboi T, Tachibana S, Kimura M, Bjorkman A, Troye-Blomberg M, Tanabe K, & Drakeley C (2014). Characteristic age distribution of Plasmodium vivax infections after malaria elimination on Aneityum Island, Vanuatu. Infect Immun, 82(1), 243–252. 10.1128/IAI.00931-13 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Khulmanee T, Thita T, Kritsiriwutinan K, Boonyuen U, Saai A, Inkabjan K, Chakrabarti R, Rathod PK, Krudsood S, Mungthin M, & Patrapuvich R (2024). Low Genetic Diversity of Plasmodium vivax Circumsporozoite Surface Protein in Clinical Isolates from Southern Thailand. Trop Med Infect Dis, 9(5). 10.3390/tropicalmed9050094 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Li H, & Durbin R (2009). Fast and accurate short read alignment with Burrows–Wheeler transform. bioinformatics, 25(14), 1754–1760. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Li H, Handsaker B, Wysoker A, Fennell T, Ruan J, Homer N, Marth G, Abecasis G, Durbin R, & Subgroup GPDP (2009). The sequence alignment/map format and SAMtools. bioinformatics, 25(16), 2078–2079. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Lockyer MJ, Marsh K, & Newbold CI (1989). Wild isolates of Plasmodium falciparum show extensive polymorphism in T cell epitopes of the circumsporozoite protein. Mol Biochem Parasitol, 37(2), 275–280. 10.1016/0166-6851(89)90159-x [DOI] [PubMed] [Google Scholar]
  17. Myung JM, Marshall P, & Sinnis P (2004). The Plasmodium circumsporozoite protein is involved in mosquito salivary gland invasion by sporozoites. Mol Biochem Parasitol, 133(1), 53–59. 10.1016/j.molbiopara.2003.09.002 [DOI] [PubMed] [Google Scholar]
  18. Need JT, Wirtz RA, Franke ED, Fernandez R, Carbajal F, Falcon R, & San Roman E (1993). Plasmodium vivax VK247 and VK210 circumsporozoite proteins in Anopheles mosquitoes from Andoas, Peru. J Med Entomol, 30(3), 597–600. 10.1093/jmedent/30.3.597 [DOI] [PubMed] [Google Scholar]
  19. Ngernna S, Rachaphaew N, Thammapalo S, Prikchoo P, Kaewnah O, Manopwisedjaroen K, Phumchuea K, Suansomjit C, Roobsoong W, Sattabongkot J, Cui L, & Nguitragool W (2019). Case Report: Case Series of Human Plasmodium knowlesi Infection on the Southern Border of Thailand. Am J Trop Med Hyg, 101(6), 1397–1401. 10.4269/ajtmh.19-0063 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Parobek CM, Bailey JA, Hathaway NJ, Socheat D, Rogers WO, & Juliano JJ (2014). Differing patterns of selection and geospatial genetic diversity within two leading Plasmodium vivax candidate vaccine antigens. PLoS Negl Trop Dis, 8(4), e2796. 10.1371/journal.pntd.0002796 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Salman AM, Montoya-Diaz E, West H, Lall A, Atcheson E, Lopez-Camacho C, Ramesar J, Bauza K, Collins KA, Brod F, Reis F, Pappas L, Gonzalez-Ceron L, Janse CJ, Hill AVS, Khan SM, & Reyes-Sandoval A (2017). Rational development of a protective P. vivax vaccine evaluated with transgenic rodent parasite challenge models. Sci Rep, 7, 46482. 10.1038/srep46482 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Santos-Ciminera PD, Alecrim M, Roberts DR, & Quinnan GV Jr. (2007). Molecular epidemiology of Plasmodium vivax in the State of Amazonas, Brazil. Acta Trop, 102(1), 38–46. 10.1016/j.actatropica.2007.02.013 [DOI] [PubMed] [Google Scholar]
  23. Sattabongkot J, Kumpitak C, & Kiattibutr K (2015). Membrane Feeding Assay to Determine the Infectiousness of Plasmodium vivax Gametocytes. Methods Mol Biol, 1325, 93–99. 10.1007/978-1-4939-2815-6_8 [DOI] [PubMed] [Google Scholar]
  24. Sidjanski SP, Vanderberg JP, & Sinnis P (1997). Anopheles stephensi salivary glands bear receptors for region I of the circumsporozoite protein of Plasmodium falciparum. Mol Biochem Parasitol, 90(1), 33–41. 10.1016/s0166-6851(97)00124-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Sriwichai P, Samung Y, Sumruayphol S, Kiattibutr K, Kumpitak C, Payakkapol A, Kaewkungwal J, Yan G, Cui L, & Sattabongkot J (2016). Natural human Plasmodium infections in major Anopheles mosquitoes in western Thailand. Parasit Vectors, 9, 17. 10.1186/s13071-016-1295-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Stecher G, Tamura K, & Kumar S (2020). Molecular evolutionary genetics analysis (MEGA) for macOS. Molecular biology and evolution, 37(4), 1237–1239. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Stoute JA, Slaoui M, Heppner DG, Momin P, Kester KE, Desmons P, Wellde BT, Garcon N, Krzych U, & Marchand M (1997). A preliminary evaluation of a recombinant circumsporozoite protein vaccine against Plasmodium falciparum malaria. RTS,S Malaria Vaccine Evaluation Group. N Engl J Med, 336(2), 86–91. 10.1056/NEJM199701093360202 [DOI] [PubMed] [Google Scholar]
  28. Tapaopong P, da Silva G, Holzschuh A, Rungsarityotin W, Suansomjit C, Pumchuea K, Manopwisedjaroen K, Khamsiriwatchara A, Khuntong P, Cui L, Koepfli C, Sattabongkot J, & Nguitragool W (2025). Molecular epidemiology and genetic diversity of disappearing Plasmodium vivax in southern Thailand. Sci Rep, 15(1), 2620. 10.1038/s41598-025-86578-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Tewari R, Spaccapelo R, Bistoni F, Holder AA, & Crisanti A (2002). Function of region I and II adhesive motifs of Plasmodium falciparum circumsporozoite protein in sporozoite motility and infectivity. J Biol Chem, 277(49), 47613–47618. 10.1074/jbc.M208453200 [DOI] [PubMed] [Google Scholar]
  30. Vargas-Parada L (2023). In search of a vaccine for Plasmodium vivax malaria. Nature, 618(7967), S32–S33. 10.1038/d41586-023-02053-2 [DOI] [PubMed] [Google Scholar]
  31. Venkatraman N, Tiono AB, Bowyer G, Bellamy DG, Stockdale LK, Powlson J, Collins KA, Coulibaly S, Datoo MS, Silman D, Ouedraogo A, Nebie I, Imoukhuede EB, Brod F, Folegatti P, Dickinson-Craig E, Jamieson S, Bougouma EC, Wright D, . . . Hill AVS (2025). Evaluation of a novel malaria anti-sporozoite vaccine candidate, R21 in Matrix-M adjuvant, in the UK and Burkina Faso: two phase 1, first-in-human trials. Lancet Microbe, 6(3), 100868. 10.1016/S2666-5247(24)00084-3 [DOI] [PubMed] [Google Scholar]
  32. Vo TC, Le HG, Kang JM, Moe M, Naw H, Myint MK, Lee J, Sohn WM, Kim TS, & Na BK (2020). Genetic polymorphism and natural selection of circumsporozoite protein in Myanmar Plasmodium vivax. Malar J, 19(1), 303. 10.1186/s12936-020-03366-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Vo TC, Trinh NTM, Le HG, Kang JM, Yoo WG, Quang HH, & Na BK (2022). enetic Diversity of Circumsporozoite Surface Protein of Plasmodium vivax from the Central Highlands, Vietnam. Pathogens, 11(10). 10.3390/pathogens11101158 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Wang Q, Fujioka H, & Nussenzweig V (2005). Exit of Plasmodium sporozoites from oocysts is an active process that involves the circumsporozoite protein. PLoS Pathog, 1(1), e9. 10.1371/journal.ppat.0010009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. WHO. (2024a). Malaria vaccines: WHO position paper. Weekly Epidemiological Record, 99, 225–248. [Google Scholar]
  36. WHO. (2024b). World malaria report 2024. World Health Organization. [Google Scholar]
  37. Wirtz RA, Sattabongkot J, Hall T, Burkot TR, & Rosenberg R (1992). Development and evaluation of an enzyme-linked immunosorbent assay for Plasmodium vivax-VK247 sporozoites. J Med Entomol, 29(5), 854–857. 10.1093/jmedent/29.5.854 [DOI] [PubMed] [Google Scholar]
  38. Yanmanee S, Seethamchai S, Kuamsab N, Karaphan S, Suwonkerd W, Jongwutiwes S, & Putaporntip C (2023). Natural vectors of Plasmodium knowlesi and other primate, avian and ungulate malaria parasites in Narathiwat Province, Southern Thailand. Sci Rep, 13(1), 8875. 10.1038/s41598-023-36017-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Zakeri S, Abouie Mehrizi A, Djadid ND, & Snounou G (2006). Circumsporozoite protein gene diversity among temperate and tropical Plasmodium vivax isolates from Iran. Trop Med Int Health, 11(5), 729–737. 10.1111/j.1365-3156.2006.01613.x [DOI] [PubMed] [Google Scholar]
  40. Zeeshan M, Alam MT, Vinayak S, Bora H, Tyagi RK, Alam MS, Choudhary V, Mittra P, Lumb V, Bharti PK, Udhayakumar V, Singh N, Jain V, Singh PP, & Sharma YD (2012). Genetic variation in the Plasmodium falciparum circumsporozoite protein in India and its relevance to RTS,S malaria vaccine. PLoS One, 7(8), e43430. 10.1371/journal.pone.0043430 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Zhu F, Zheng H, Chen S, Zhang K, Qin X, Zhang J, Liu T, Fan Y, Wang L, Li X, Zhang J, & Xu W (2022). Malaria oocysts require circumsporozoite protein to evade mosquito immunity. Nat Commun, 13(1), 3208. 10.1038/s41467-022-30988-z [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS2129360-supplement-1.docx (53.4KB, docx)

RESOURCES