Genetic diversity and population structures of Plasmodium falciparum parasites imported to China from Central and West Africa determined using a SNP barcode

Abstract

Background

Malaria remains a highly threatening infectious disease, with Africa being the main epidemic area. China has achieved the notable feat of eliminating malaria domestically, yet the looming threat of imported malaria cases remains a concern. This study aims to apply the SNP barcode technology to investigate the population structure of P. falciparum imported to China from Central and West Africa.

Methods

We applied a 24-SNP high-resolution melting (HRM) barcode to analyze the diversity of P. falciparum populations imported to China from Central and West Africa.

Results

A total of 181 samples were analyzed using HRM assay to obtain a complete 24-SNP barcode. There was no significant difference in the proportion of multi-clone infections among the four populations. The level of nucleotide diversity observed across all four populations is low. The observed pairwise F_ST values ranged from 0.001 to 0.054, indicating a low to moderate level of genetic differentiation between the four populations. The combined results of principal component analysis and population structure assessment reveal a lack of prominent clustering among the four populations.

Conclusions

The present study validated the potential of the 24-SNP HRM barcode in examining diversity and variations in imported P. falciparum populations. Our studies confirmed that this barcode is insufficient for distinguishing isolates from Central and West Africa, potentially due to the minimal genetic differentiation among P. falciparum populations in these regions.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12879-025-11720-w.

Introduction

According to the World Malaria Report 2023 from the World Health Organization (WHO), there were 249 million malaria cases and 608,000 deaths worldwide in 2022 [1], with Africa accounting for approximately 95% of these cases. Plasmodium falciparum is the predominant malaria parasite in Africa. Malaria was once a key infectious disease in China, and through unremitting efforts, China has maintained zero local malaria cases for several consecutive years. On June 30, 2021, the WHO officially announced that China had eliminated malaria, becoming the 40th country in the world to eliminate malaria. Although China has eliminated local malaria, the significant threat of imported malaria still exists. Under the umbrella of the Belt and Road Initiative, amidst the intensifying globalization, labor migration and commercial collaborations, there is a heightened possibility of local malaria transmission resurfacing in regions that are currently malaria-free. From 2010 to 2019, there were 29,248 imported cases, with 18,032 (61.65%) cases of falciparum malaria. Malaria was mainly imported from Africa and Southeast Asia [2]. As such, it is of paramount importance to continuously monitor these imported malaria cases, ensuring that appropriate measures are in place to prevent and manage any potential outbreaks. Therefore, it is necessary to monitor the dynamics and diversity of imported malaria populations.

Historically, researchers relied on molecular markers like merozoite surface protein 1 (msp1), merozoite surface protein 2 (msp2), and the glutamine-rich protein (glurp) to decipher the population dynamics of P. falciparum [3]. However, a limitation of these methodologies lies in the absence of standardized protocols and/or definitive references for interpreting the outcomes, hindering their widespread applicability [4]. Microsatellites, comprising short tandem repeat (STR) sequences, have also emerged as a tool for characterizing P. falciparum [5]. However, their inherent capacity for rapid evolution presents a formidable obstacle in utilizing them as a dependable genotyping strategy due to the potential for rapid genetic changes [6]. The malaria barcode represents a set of multiple unique single nucleotide polymorphisms (SNPs) embedded within the genome sequence of P. falciparum [7]. When seamlessly integrated with high-resolution melting (HRM) analysis [8], this barcode system significantly enhances the precision in delineating the population diversity of the parasite. The SNP genotyping methodology serves as a powerful tool for probing parasite population dynamics diversity and assessing the effectiveness of intervention strategies [9, 10]. HRM analysis, in particular, leverages the polymerase chain reaction (PCR), employing specialized dyes designed to discern subtle variations in DNA sequences [8]. The cornerstone of HRM lies in the melting of PCR amplicons, where the unique melting curves of SNP alleles serve as the cornerstone for their identification and differentiation [11]. Compared to polymorphic antigenic markers such as msp1, msp2, and glurp, the advantage of the HRM SNP barcode is that they can detect changes in population events, such as decreases or increases in transmission rate. Furthermore, SNP barcode analysis offers a robust means of monitoring alterations in parasite population characteristics by tracking parasite genotypes across time and tracing their introduction routes via human migration patterns [9]. For example, it has been applied to investigate the impact of intervention strategies in Zambia and Zimbabwe [10], outbreaks of P. falciparum clone amplification in Panama [8], and the population structure of parasites in Nigeria [12]. This study aimed to use this genotyping technique to study the population structure of P. falciparum malaria parasites imported to China from Central and Western Africa.

Material and methods

Parasite collection and DNA extraction

Parasite samples were collected from Chinese migrant workers who had returned from various African countries (Cameroon, the Central Africa Republic, the Democratic Republic of Congo, Liberia, and Ghana). Patients with malaria symptoms attended the Guangxi Shanglin Hospital from 2016 to 2019 and were diagnosed with malaria through microscopic examination of thick and thin blood smears stained with Giemsa, and then confirmed through nested PCR in the laboratory [13]. These patients who had been to two or more than two countries are excluded. The blood samples were stored at 4 °C at the time of collection and then transported to the laboratory and stored at −20 °C until analysis.

Parasite DNA was extracted from whole blood using the High Pure PCR Template Preparation Kit (Roche) following instructions from the manufacturer. Then, samples were diluted to a concentration of 1 ng/μL using 1 × Tris EDTA Buffer. Confirmation of P. falciparum infection was done using nested PCR targeting the small subunit ribosomal RNA gene as described previously [14].

HRM SNP barcodes and genotype determination

The P. falciparum 24 SNP molecular barcode assay was performed as described by Daniels et al. [7]. Genomic DNA from cloned strains of P. falciparum, including 3D7, Dd2, HB3, 7G8, and K10, were used as references or alternative control samples. The P. falciparum HRM SNP barcode assay was conducted as outlined by Bankole et al. [12]. Briefly, the HRM SNP barcode assay was executed on the ABI QuantStudio6 real-time PCR system. The PCR reaction mixture consisted of 10 μl, containing 1.0 μL of forward primer, 1.0 μL of reverse primer, 2.0 μL of double-distilled water, 4.0 μL of 2.5 × Light Scanner Master mix (BioFire Diagnostics Inc., Salt Lake City, Utah, USA), and 2 μl of DNA template. The PCR cycling conditions were as follows: an initial denaturation step at 95 °C for 2 min, followed by 40 cycles of 94 °C for 30 s, 64 °C for 60 s, and a final HRM cycle comprising 95 °C, 55 °C, and 95 °C, each for 15 s.

To ascertain the genotype of the samples, we analyzed the derivative melting temperature (Tm) curve for each individual measurement. Since P. falciparum exists in a haploid state within peripheral blood, we interpreted the detection of two alleles at any specific location as indicative of a mixed infection, whereas the presence of a single allele was deemed to represent a monoclonal infection. During the execution of the HRM SNP barcode assay on each test sample, reference and alternative allele control samples were assayed to identify the Tm curve of each reference and alternative SNP. By comparing the Tm peak generated by the control sample with that of the test sample, we identified each allele. The derivative Tm curve of each SNP for each sample was utilized to ascertain the genotype of the sample. Monomorphic genotypes were identified by their singular Tm peak and their alignment with a control Tm curve. Conversely, polymorphic genotypes were distinguished by their skewed or shifted Tm curves.

Population genetic analyses

When classifying a parasite sample as polygenomic, Sisya et al. [15] defined a minimum threshold of at least two heterozygous SNPs (N). We determined the complexity of infection (COI) through the COIL web tool [16], and the dispersion pattern of COI among patients offers insights into malaria transmission dynamics and epidemiology. Samples that possess no or, at most, one heterozygous SNP (N) within the barcode were utilized for subsequent analysis. The Mann–Whitney U test was used to measure differences between two groups. Statistical analysis was conducted using GraphPad Prism software version 6.0, and p-values < 0.05 were considered statistically significant. Samples with no or at most one heterozygous SNP (N) in their barcode, which correspond to monoclonal and biclonal samples, were used for subsequent analysis.

The calculation of Minor Allele Frequency (MAF) and its average follows a methodology outlined by Baniecki et al. [11]. In essence, MAF is derived by assessing the frequency of each allele for each SNP within a given population, based on the allele counts. For polymorphic genotypes, the reference and substitute alleles are assigned values that reflect their contribution, being half of what would be expected in a monotypic genotype. Building upon this, we determine the Average Minor Allele Frequency (AMAF), which is simply the unweighted mean of the MAF values for each SNP across two populations. This approach provides a comprehensive overview of the allelic diversity within and between populations.

We measured nucleotide diversity using the π statistic, which calculates the average pairwise differences among measured SNPs within a population, divided by the total number of SNPs measured. This gives an indication of the genetic variation within the population [11]. Expected heterozygosity (He) was obtained using the POPGENE software Version 1.31.

The genetic divergence among haplotypes was quantified by calculating the pairwise genetic distance (F_ST). This F_ST value was derived utilizing the DnaSP Version 5.0 software [17]. Furthermore, the genetic distance between paired populations was determined by employing Slatkin's linearization method [18]. To reinforce the validity of these F_ST values, we harnessed the COIL network tools for the generation and subsequent analysis of F_ST estimates [16].

To compare the genetic architecture of parasite samples sourced from various locations and time points, we conducted a Principal Component Analysis (PCA) utilizing the online platform ClustVis [19]. Phylogenetic analysis was performed on the distance metric 1-PS using the “Ape” R package and visualized using the online tool iTOL [20]. Parasite sample transmission networks were constructed using the online tool StrainHub [21]. The implementation of the transmission network was based on the metric of degree centrality.

The model-based Bayesian method was used to divide individual parasite samples into genetic clusters by the STRUCTURE V2.3.4 software. The number of clusters (K) is determined by simulating K values from 1 to 5, and the posterior probability value of each K is estimated by the Markov Chain Monte Carlo (MCMC) method. The optimal K value is calculated based on Evanno's ΔK statistical method.

Results

Genotyping the 24-SNP barcode

A total of 181 samples were collected, including 51 from Central Africa [the Central African Republic (n = 8), Cameroon (n = 26), and the Democratic Republic of the Congo (n = 17)] and 130 from West Africa [Liberia (n = 25) and Ghana (n = 105)](Figure S1). All samples were analyzed successfully using the HRM assay to obtain a complete 24-SNP barcode. These samples comprised 31 from Central Africa in 2016–2017 (CA16-17), 20 from Central Africa in 2018–2019 (CA18-19), 75 from West Africa in 2016–2017 (WA16-17), and 55 from West Africa in 2018–2019 (WA18-19). Among the 181 samples, 49 (27.1%) demonstrated monoallelicity at each nucleotide position, indicating monoclonal infections. Additionally, 79 samples (43.6%) showed the presence of two alleles differing by exactly one SNP, suggesting biclonal infections. Furthermore, 53 samples (29.3%) exhibited two or more alleles at least in two SNP locations, indicating multiclonal infections (Table S1). The proportion of multiclonal infections was 35%, 35%, 32%, and 20% for the respective populations from CA16-17, CA18-19, WA16-17, and WA18-19, respectively. The mean COI values for the populations originating from CA16-17, CA18-19, WA16-17, and WA18-19 were 1.29, 1.35, 1.29, and 1.13, respectively. There was no significant difference in the proportion of multi-clone infections among the four populations, but the mean COI values of CA18-19 and WA16-17 were significantly higher than those of WA18-19 (Table 1). The remaining 128 (70.7%) samples, including CA16-17 (i = 20), CA18-19 (n = 13), WA16-17 (n = 51), and WA18-19 (n = 44), displayed either a single allele across all 24 SNPs or two alleles at a single locus, indicating monoclonal infections. These monoclonal samples were subsequently utilized for further genetic analyses.

Table 1.

Genetic characteristics of parasite isolates collected from different years and regions

	CA16-17	CA18-19	WA16-17	WA18-19
No. of isolates	31	20	75	55
No. of SNPs	24	24	24	24
Proportion of multi-clone infections	35%	35%	32%	20%
Mean COI	1.290	1.350a	1.293b	1.127

^aComparison Mean COI between CA18-19 and WA18-19 by the Mann Whitney U test, P = 0.0435

^bComparison Mean COI between WA16-17 and WA18-19 by the Mann Whitney U test, P = 0.0325

Minor allele frequency (MAF)

Upon analysis of the entire population, it was observed that only 1 out of 24 SNPs (4.2%) exhibited an AMAF below 0.1. When examining the four populations, SNP 23 was the only one with MAF values less than 0.1 in both CA18-19 and WA16-17. In the WA18-19 population, two SNPs (3 and 23) showed MAF values below 0.1. Although the populations exhibited a high level of intra-population diversity, a comparison of the MAF values between the four sub-sets revealed a relatively low level of inter-population diversity (Fig. 1).

Fig. 1.

Minor allele frequencies (MAF) of P. falciparum samples collected from different years and regions

Nucleotide diversity (π) and expected heterozygosity (He)

The nucleotide diversity (π) values for CA16-17, CA18-19, WA16-17, and WA18-19 were 0.404, 0.413, 0.418, and 0.344, respectively. The He values for CA16-17, CA18-19, WA16-17, and WA18-19 were 0.404, 0.380, 0.421, and 0.373, respectively (Figure S2). The computed π and He statistic indicates that these populations have a low level of diversity.

Pairwise F_ST values

The observed pairwise F_ST values ranged from 0.001 to 0.054, suggesting a low to moderate level of genetic differentiation between the four populations (Fig. 2), and the populations CA16-17 and WA18-19 exhibited a moderate level of genetic differentiation. This finding supports the similarity in MAF and π values among these four populations.

Fig. 2.

Pairwise F_ST values obtained for the parasite populations for different years and regions

Principal component analysis (PCA)

The PCA revealed an intricate intermingling of samples from the four populations, lacking distinct clustering patterns (Fig. 3A). This finding was further corroborated by the phylogenetic tree (Fig. 3B). This finding suggests minimal population differences among the four populations.

Fig. 3.

Principal component analysis (PCA) and Phylogenetic relationships of P. falciparum samples collected from different geographic regions and years

Transmission networks

The transmission network constructed based on the phylogenetic tree indicates substantial genetic exchanges among the four P. falciparum populations, with the highest degree of exchange observed between WA16-17 and WA18-19 (Figure S3).

Population structure

The Bayesian cluster analysis of SNP genotypes, conducted using STRUCTURE software, revealed that two genetic clusters (K = 2) offered the most optimal fit for the SNP barcode data. The structure analysis, which provides insights into the ancestral proportions within each sample, further disclosed that there was no notable disparity in the distribution of the two clusters across the four sample populations (Fig. 4).

Fig. 4.

The population structure (the membership fraction for each population) of P. falciparum from different geographic regions and years

Discussion

Utilizing 24 SNP molecular barcodes, we presented a successful attempted to evaluate the population diversity of imported P. falciparum samples originating from Central and West Africa. In several regions of the world, the HRM technique has been employed to assess the dynamics of parasite populations, specifically of P. falciparum and P. vivax [7, 9, 11, 15, 22–24].

In this research, the parasite samples were obtained from Chinese immigrant workers who had returned from various African nations of Central and West Africa from 2016 to 2019. We observed that despite that the four sample sets had similar multi-clone infections, the mean COI value in the WA18-19 period was conspicuously lower compared to that in CA18-19 and WA16-17. The result appears to indicate a decrease in malaria transmission intensity in West Africa from 2016 to 2019. However, since these malaria cases involve travelers, their lifestyles and potentially varying exposure patterns could lead to a lower likelihood of contracting malaria compared to the indigenous residents. This trend may not accurately reflect the prevalence of malaria among the local population.

The AMAF parameter, when exceeding 0.1, indicates significant diversity at a specific SNP [11]. In this study, we observed that only one locus exhibited an AMAF value below 0.1. Comparing variations across all SNPs of the barcode, we did not find significant differences among the four populations. The presence of high intra-population diversity with minor inter-population differences aligns well with previous research conducted in Nigeria [12]. The π values also showed a comparatively low level of genetic diversity in the parasites of these four populations.

The genetic differentiation among the four P. falciparum populations, studied using pairwise F_ST values, showed that only the CA16-17 and WA18-19 parasite populations had a moderate level of genetic differentiation with an F_ST value of 0.054, whereas all other pairwise comparisons had F_ST values within the 0–0.05 range. Despite the narrow span of F_ST values, we observed that the F_ST values of populations across different regions at the same time are smaller than those of populations within the same region at different times. This disparity may stem from a certain degree of population exchange between Central and West African groups during the same period.

We employed PCA and population structural analysis to further investigate the genetic differentiation among parasite populations. These analyses did not reveal the clustering of individual populations, indicating relatively minor genetic differentiation of these four populations. A recent study utilizing SNP barcodes demonstrated that P. falciparum populations in West Africa, Central Africa, and East Africa exhibit similar genetic structures, aligning well with the findings of this research [25]. The observed absence of genetic variation in the P. falciparum population in this study could potentially be attributed to the similarity in epidemiological conditions across these countries or to analogous selective pressures resulting from the implementation of control and management measures in these respective nations. In Africa, even on a continental scale, geographic separation does not inherently result in significant genetic divergence among P. falciparum populations [11]. Overall, our results indicate that this barcode is insufficient in distinguishing samples from Central and West Africa, potentially due to the minimal genetic disparities among these P. falciparum populations, and this is consistent with previous findings using genetic barcodes [25].

One main limitation of this research lies in the relatively small sizes of samples gathered from Central Africa. Moreover, the existence of a substantial percentage of multi-clone infections further diminishes the number of viable samples for investigating population structure. Consequently, the implementation of future studies encompassing a broader sample size is pivotal for increasing the robustness of the study. Another limitation of this work lies in the fact that malaria from travelers may inaccurately represent the local parasite populations.

Conclusion

The current study demonstrated that the 24-SNP HRM barcode could be used to investigate the diversity and variations among imported P. falciparum populations. We found that the genetic differences among P. falciparum populations from Central and West Africa were relatively minor.

Abbreviations

SNP

Single nucleotide polymorphism

HRM

High-resolution melting

COI

Complexity of infection

MAF

Minor allele frequency

He

Expected heterozygosity

CA

Central Africa

WA

West Africa

Authors' contributions

Weilin Zeng, Wenya Zhu and Yucheng Qin wrote the main manuscript text. Hongyu Lan, Yulin Cen, Kemin Sun, Yiman Wu, Liang Tao, Ye Mou, Cheng Liu and Xiuya Tang prepared data collection.Xiang Zheng, Yaming Huang and Liwang Cui participated in article editing. Zhaoqing Yang designed the research. All authors reviewed the manuscript.

Funding

This study was supported by the National Science Foundation of China (32370543), a grant (U19AI089672) from the National Institutes of Health, USA, Major science and technology projects of Yunnan Province (2018ZF0081), and International Science and Technology Cooperation-Yunnan International Science and Technology Cooperation Base (202003AE140004). WZ and ZX were sponsored by the Yunnan Applied Basic Research Projects-Union Foundation (202301AY070001-116 and 202401AY070001-060).

Data availability

Data is provided within the manuscript or supplementary information files.

Declarations

Ethics approval and consent to participate

All patients in this study voluntarily signed informed consent forms, and the research project was approved by the Shanglin Hospital Institutional Review Board in accordance with the Declaration of Helsinki (Clinical trial number: not applicable).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.