Genetic polymorphisms in the circumsporozoite protein of Plasmodium malariae show a geographical bias

A total of 89 P. malariae isolates were collected from 4 different Asian countries, including Thailand, Myanmar, Lao PDR, and Bangladesh (Table 1). This study received ethical clearance from the Faculty of Tropical Medicine, Mahidol University, Thailand (MUTM2011-049-06). Genomic DNA was extracted from the isolates according to the manufacturer’s instruction (Qiagen, Germany) and stored at − 20 °C until further use.

The DNA samples were subjected to nested PCR [13] to confirm the presence of P. malariae and detect the presence of any other Plasmodium species. Sequences of pmcsp corresponding to the accession numbers S69014, U09766, J03992, AJ001525, AJ001523, AJ002582, AJ002578, AJ002580, AJ002576, AJ001526, AJ001524, AJ002583, AJ002581, AJ002577, AJ002579, AJ002575 were retrieved from NCBI (https://​www.​ncbi.​nlm.​nih.​gov/​). Gene-specific primers spanning the complete coding sequence of pmcsp were designed based on the multiple sequence alignment of pmcsp sequences (Table 2). Semi-nested PCR approach was used for the amplification of pmcsp using the conditions described in Table 2. All PCRs were carried out with 10 mM Tris–HCl (pH 8.3), 50 mM KCl, 2 mM MgCl₂, 125 M dNTPs, 250 nM of each primer, and 4 units of Taq Polymerase (Kapa biosystems, USA). The PCR products were examined by gel electrophoresis. All amplified PCR products were purified using the FavorPrep™ Gel/PCR Purification Kit (Favorgen, Taiwan) and sequenced.

DNA sequences of pmcsp were read on both strands and analysed using BioEdit Sequence Alignment Editor Program as described previously [14]. DNA sequence polymorphisms, haplotype diversity, nucleotide diversity, and the rate of synonymous (dS) and nonsynonymous (dN) substitutions were calculated using DnaSP version 5.10.01 [15] and MEGA6 [16]. Phylogenetic analysis was done using the neighbour-joining method [17].

pmcsp of 89 isolates of P. malariae (Thailand = 43, Myanmar = 40, Lao PDR = 5, and Bangladesh = 1; Table 1) was successfully PCR amplified and sequenced. Multiple sequence alignment of these sequences was performed, and sequences corresponding to primer-binding sites, including 12 amino acids at the 5′ end and 6 amino acids at the 3′ end, were removed. Thus, the total length of pmcsp used in this analysis varied from 315 to 403 amino acids. In addition to the pmcsp sequences obtained from the 89 isolates, 58 pmcsp sequences were retrieved from NCBI (https://​www.​ncbi.​nlm.​nih.​gov/​). All 143 pmcsp sequences (Asia = 91 and Africa = 52) were analyzed for DNA sequence polymorphisms. DNA divergence between populations was calculated in area containing more than 30 sequences: Thailand (n = 43), Myanmar (n = 40), and previously published results from Kenya (n = 38). Average nucleotide diversity in Kenya (pi = 0.017) was lower compared with that in Thailand (pi = 0.036) and Myanmar (pi = 0.043). When compared across different continents, the nucleotide diversity in Asia (pi = 0.042) was higher compared with that in Africa (pi = 0.015). A sliding method plot with a window length of 100 bp and a step size of 25 bp using DnaSP v5 revealed a pi value in three different locations (Fig. 1a) and two continents (Fig. 1b). The haplotype diversity of pmcsp was similar across these 3 countries, and ranged from 0.660 to 0.977. However, the ratio of nonsynonymous (dN) to synonymous (dS) substitutions in Kenya (dN/dS = 0.442) was lower compared with that in Thailand (dN/dS = 1.563) and Myanmar (dN/dS = 0.880). The ratio of nonsynonymous (dN) to synonymous (dS) substitutions in Africa (dN/dS = 0.411) was lower compared with that in Asia (dN/dS = 1.017) (Table 3). The neutrality test was performed in different populations though Tajima’s D, Fu and Li’s F* and Fu and Li’s D* tests. Analysis based on the non-repeat regions revealed significant differences in Asia and Africa (P  <  0.05) for Tajima’s D and Fu and Li’s F* tests (Table 3). It was indicated that these population groups have negative value reflect a lower frequency alleles than expected under a neutral model, which can result from population size expansion after a recent bottleneck or purifying selection. A total of 538 positions along pmcsp of all 147 P. malariae isolates were used for inferring their relationship with the neighbor-joining method [17]. The bootstrap consensus tree inferred from 1000 replicates which is taken to represent the evolutionary history of those 147 P. malariae isolates analysed. Branches corresponding to partitions reproduced in less than 50% bootstrap replicates are collapsed. Results showed that most of the P. malariae isolates clustered within their respective continents, with the exception of a couple of isolates from Thailand that clustered together with the four isolates from Venezuela and showed closely related to the African isolates (Fig. 2).

pmcsp sequences obtained from 89 Asian isolates collected in this study were analyzed together with those previously obtained from 58 African isolates. Multiple sequence alignment of 147 pmcsp sequences revealed a pattern of tetrapeptide repeat units that exhibit unique characteristic for each species. The central repeat region of pmcsp contained the NAAG tetrapeptide repeat unit in a majority of the 147 samples, and the number of NAAG repeats varied from 0 to 79 in different samples. P. malariae isolates from Africa (Kenya, Cameroon) carried a higher number of NAAG repeats (42–79), whereas Asian isolates carried two range of repeat number: 0–38 repeats and 40–51 repeats (Additional file 1). The second most prevalent repeat unit was NDAG, which was present in all samples. The number of NDAG repeats varied from 2 to 9 randomly across all regions (Additional file 2). A novel tetrapeptide repeat, NAPG, was identified in this study. The number of NAPG repeats varied from 0 to 51 in samples from Thailand, Myanmar, Lao PDR, and Bangladesh (Additional file 3). Of the 43 isolates collected from Thailand, 17 isolates carried 1–40 repeats of the NAPG unit, whereas 26 isolates did not carry the NAPG repeat. Of the 40 isolates collected from Myanmar, 17 carried 11–20 NAPG repeats, 8 carried 1–10 copies, and 6 carried 21–30 copies, whereas 9 isolates did not contain the NAPG repeat. Of the 5 isolates collected from Lao PDR, 3 carried 41–60 copies of NAPG repeats and 2 did not carry the NAPG repeat. The isolate collected from Bangladesh carry 3 copies of NAPG repeat in the central repeat region.

To compare the average number of the repeat units between the Asian and African samples, the sampling model for each repeat type: NAAG and NDAG, was generated as follow;where r_i refer to the repeat number of the sample i, j indexes country and k indexes continent. The repeat numbers of each repeat type from the sample from any country was assumed to be sampled from a Poisson distribution where its average number was λ. These average numbers λ were thought to be varied between countries and also be distributed normally around the average numbers in the continent level (μ) with the standard deviation σ. The model was fitted to the data using the Markov Chain Monte Carlo (MCMC) technique in WinBUGS [18] (Fig. 3 and Additional file 4).

Sequences of the nonrepeat regions of pmcsp, including conserved regions I and II, Th2R, and Th3R, were analyzed in 143 isolates (Table 4). Limited polymorphisms were observed in all 4 conserved regions. Seven amino acid haplotypes were identified in conserved region I. Among these, the haplotype “AVENKLKQP” was the most predominant and was identified in 82.52% of isolates (118 out of 143 isolates). In conserved region II, the haplotype “ITEEWSPCSVTCG” was identified in 93.01% of isolates (133 out of 143 isolates). The Th2R and Th3R domains showed 6 and 9 haplotypes, respectively. The most prevalent haplotypes in Th2R and Th3R were present in 91.61 and 74.83% of isolates, respectively. All polymorphisms in each Th3R haplotype were nonsynonymous. Contrary to Th2R, synonymous substitutions were found in 2 samples.

In addition, both N- and C-terminal nonrepeat regions of pmcsp were used for estimating the average level of genetic differentiation between each population. The Fst [19] for all pairwise comparisons between population were significant (P < 0.001) (Table 5). The Fst value between Thailand and Myanmar (0.087) was lower than that between Thailand and Kenya (0.518) and Myanmar and Kenya (0.366). High differentiation between Asia and Africa was observed (0.404).