SP analogs, such as cotrimoxazole (trimethoprim-sulfamethoxazole), are also widely used in the Sahel region to manage opportunistic infections in people with HIV/AIDS, and this chemoprophylaxis has been associated with multidrug resistance [6]. Trimethoprim-sulfamethoxazole is a combination of a 2,4 diaminopyrimidine (trimethoprim) and a sulfonamide (sulfamethoxazole) and has the same resistance mechanism as SP, namely, mutations in the Plasmodium falciparum dihydrofolate reductase (Pfdhfr) and the P. falciparum dihydropteroate synthase (Pfdhps) genes. Strains associated with treatment failure carry a quintuple mutation: the triple PfdhfrN51I/C59R/S108N mutation and the double PfhpsA437G/K540E mutation [7–9]. Additionally, PfdhfrI164L and PfdhpsK540E mutations confer very high levels of resistance [10]. Thus, the widespread use of SP could lead to genetically diverse strains and the selection of resistant parasites. Against this background, we conducted a comparative study of the polymorphism of the P. falciparum (Pf) merozoite surface protein-1 (Pfmsp1) genetic marker between Mali and Niger and the variability of mutations at codons 59 and 51 of the Pfdhfr gene and codons 436 and 437 of the Pfdhps gene using high-resolution melting analysis (HRM).

The study was approved by the Ethics Committee of the Faculty of Medicine, Pharmacy and Dentistry of Bamako, Mali, and the Ethics Committee of Niger (Deliberation No. 024/2015/CCNE). Informed consent was obtained from the parents or legal guardians of all study participants.

This is a comparative study of the genetic polymorphism of parasitic strains using the msp1 genetic marker and a descriptive study of molecular resistance to SP through analysis of the Pfdhfr and Pfdhps genes. The study was conducted in Mali and Niger, which are neighboring countries in West Africa. All study sites were located in the Sahel region, where malaria transmission is seasonal and lasts for 3–4 months. In Mali, the rural communities of Sélingué (Yanfolila Cercle, Sikasso Region) and Missira (Kolokani Cercle) were selected. In Niger, the Aguié Department (Maradi Region) and Ertcharna (Tahoua Region) were included. The samples were collected during the SMC campaigns of 2016 and 2017. All children aged 3–59 months, having a mono-specific Pf infection and having received SP and AQ doses during the CPS campaign were included in this study. The inclusion criteria were children aged 3–59 months presenting signs of malaria, in whom infection was confirmed by a rapid diagnostic test (RDT) or a thick smear. Children with severe malaria or comorbidity were excluded. The sample size was 400 children, 200 from Niger and Mali, respectively.

The samples consisted of blood spots from filter papers contained in the RDT cassettes used for biological confirmation in Niger and Wattman grade 3 filter papers in Mali. DNA was extracted from the samples using QIAamp DNA Kits (QIAGEN GmbH, Düsseldorf, Germany). The procedure comprised 4 main steps. First, the samples underwent enzymatic digestion with protease. Next, the DNA was adsorbed on the silica gel membrane of a collector. Then, the membrane was washed with AW1 and AW2 buffers. Finally, DNA was eluted using AE buffer. The resultant DNA samples were aliquoted, numbered, and stored at −20°C before analysis by PCR or HRM.

The multiplicity of infection (MOI) refers to the number of distinct parasite genotypes present within an infected individual. This can be assessed using PCR amplification of specific genetic markers to identify different parasite strains. We analyzed the genetic polymorphism of Pf strains using nested PCR with specific primers for conserved blocks 1 and 3 of the Pfmsp1 gene. The first round of PCR was performed in a final reaction volume of 25 μl comprising 10 μl distilled H₂O, 5 μl of buffer, 1.5 μl MgCl₂ (50 mM), 1 μl dNTPs (10 mM), 1 μl forward primer M1-RO33-F (5′-CTAGAAGCTTTAGAAGATGCAGTATTG-3′), 1 μl reverse primer M1-RO33-R (5′CTTAAATAGTATTCTAATTCAAGTGGATCA-3′), 0.5 μl Taq polymerase, and 5 μl of template DNA. The first-round PCR program included 40 cycles of denaturation at 95°C for 1 min, hybridization at 59°C for 2 min, and extension at 72°C for 2 min. The second-round PCR amplified 1 of the 3 allelic families of msp1 (K1, MAD20 or RO33) in a final reaction volume of 25 μl comprising 10 μl distilled H₂O, 5 μl buffer, 1.5 μl MgCl₂ (50 mM), 1 μl dNTPs (10 mM), 0.5 μl Taq polymerase, and 5 μl template DNA, and 1 μl each of the following primer pairs: forward primer M1-K1-F (5′-AAA TGAAGAAGAAATTACTACAAAAGGTGC-3′) and reverse primer M1-K1-R (5′-GCTT GCATCAGCTGGAGGGCTTGCACCAGA-3′), or forward primer M1-MAD20-F (5′-AAATGAAGGAACAAGTGGAACAGCTGTTAC-3′) and reverse primer M1-MAD20-R (5′-ATCTGAAGGATTTGTACGTCTTGAATTACC-3′), or forward primer M1-RO33-F (5′-TAAAGGATGGAGCAAATACTCAAGTTGTTG-3′) and reverse primer M1-RO33-R (5′-CATCTGAAGGATTTGCAGCACCTGGAGATC-3′). The second-round PCR program, used to detect the allelic families, included denaturation at 95°C for 1 min, hybridization at 59°C for 2 min, and extension at 72°C for 2 min. The amplified nucleotide sequences from the nested PCR were electrophoresed in a 1% agarose gel containing ethidium bromide. A UVP program was used to identify polymorphism at different allotypes.

HRM employs a DNA-binding dye that fluoresces when bound to double-stranded DNA and decreases in fluorescence during strand separation to monitor DNA melting behavior upon heating. The resulting fluorescence decrease produces a melting curve, which is a plot of fluorescence intensity versus temperature. The shape of the melting curve provides valuable information about the DNA sequence and genetic variations, such as single nucleotide polymorphisms (SNPs), mutations, and DNA methylation patterns, which are used for genotyping, mutation scanning, and methylation analysis. HRM is a simple, sensitive, and cost-effective technique, making it a valuable tool in molecular biology research and diagnostics [11].

We used HRM to analyze molecular resistance to SP. The PCR program was run in a fi nal reaction volume of 25 μl comprising 10 μl distilled H₂O, 10 μl Light Scanner Master Mix, 2.5 μl primers for the Pfdhfr and Pfdhps genes, and 2.5 μl of template DNA. The 2.5× Light Scanner Master Mix kit contains buffer, dNTPs, Taq polymerase, MgCl₂, and primers. A probe is added for the mutation search specification. To detect mutations in the S436/A437 codons of the Pfdhps gene, we used forward primer 5′-GAATGTTTGAAATGATAAATGAAG-3′, reverse primer 5′-AGGAAACAGCTATGACGAAATAATTGTAATACA GGTACTACTAAATCTCT-3′ and probe sequence ATCCTCTGGTCCTTTTGTTATACC-block. To detect mutations in the N51/C59 codons of the Pfdhfr gene, we used forward primer 5′-ACATTTAGAGGTCTAGGAAATAAAGGAG-3′, reverse primer 5′-ATATTTACATCTCTTATATTTCAATTTTTCATATTTTGATTCATTCAC-3′, and probe sequence AAATGTAATTCCCTAGATATGAAATATTTTTGTGCAG-block. Strain 3D7 (wild-type, melting temperature: 67.5°C) and strain Dd2 (mutant, melting temperature: 62.8°C) were used as controls. The thermocycler program consisted of denaturation at 94°C for 15 sec; annealing at 63°C for 30 sec for Pfdhfr51, Pfdhfr59, Pfdhps436, and Pfdhps437; and extension at 60°C for 1 min. Thereafter, the plate was centrifuged and then subjected to HRM.

Graph Pad Prism v6.00 (GraphPad Software, La Jolla, CA, USA) was used for statistical analysis which included Chi-square tests and determination of odds ratios (ORs).

Of the 400 patient samples, 201 tested positive (50.3%), including 111 (55.2%) from Ni ger and 90 (44.8%) in Mali. In Niger, there were 55 (49.5%) malaria-infected patients from Aguié and 56 (50.5%) from Ertcharna. In Mali, there were 54 malaria-infected patients from Sélingué (60.0%) and 36 from Missira (40.0%).

Sample genotyping revealed a total of 201 allotypes (K1: 68, 33.8%; MAD20: 38, 18.9%; and RO33: 95, 47.3%). Table 1 shows the distribution of allotype frequencies by country. The prevalence of the RO33 allotype in Niger (63.6%) showed a statistically significant difference compared with the prevalence of the RO33 allotype in Mali (39.3%) (OR=0.3696; 95% confidence interval [CI]: 0.2009–0.6791; P=0.0015). However, there were no statistically significant differences between the prevalence of the K1 allotype in Mali (38.5%) versus Niger (24.2%) (OR=1.958; 95% CI: 1.011–3.793; P=0.0566) or the prevalence of the MAD20 allotype in Mali (22.2%) versus Niger 12.1% (OR=2.071; 95% CI: 0.8912–4.815; P=0.1239). However, there was a statistically significant difference in the distribution of allotypes by country (P=0.0049).

Table 2 shows the multiplicity of infections by country. In Mali, the MOI was 2 allotypes per patient, compared with one allotype per patient in Niger. Overall, 57.0% of patients had 1 allotype, 36.3% had 2 allotypes, and 6.7% had 3 allotypes. There was a statistically significant difference in the MOI between the 2 countries (P=0.0049). The proportion of patients with 2 Pf allotypes was statistically higher in Mali (41.1%) compared with Niger (26.7%) (OR=2.907; 95% CI: 1.298–6.512; P=0.0120). The proportion of patients with 1 or 3 allotypes was comparable between the 2 countries (P>0.05). However, the MOI in Mali was higher compared to the MOI in Niger.

Table 3 shows the details of all SNPs identified in the Pfdhr and Pfdhps genes. The prevalence of the SNP Pfdhfr51I was 68.5% for Niger and Mali combined (Niger, 71.8%; Mali, 67%) in Mali. There was no statistically significant difference in the prevalence of mutant Pfdhfr51I alleles between Mali and Niger (OR=0.7993; 95% CI: 0.3494–1.828; P=0.6812). The prevalence of the point mutation Pfdhps59R was 63.8% for Niger and Mali combined (Niger, 71.8%; Mali, 60.2%). There was no statistically significant difference in the prevalence of mutant pfdhfr59R alleles between Mali and Niger (OR=0.5949; 95% CI: 0.2626–1.348; P=0.2356). The prevalence of SNP Pfdhps436F was 27.5% for Niger and Mali combined (Niger, 37.2%; Mali, 23.5%). There was no statistically significant difference in the distribution of mutations at position 436 of the Pfdhps gene between the 2 countries (OR= 0.5207; 95% CI: 0.2227–1.218; P=0.1765). The prevalence of the SNP pfdhps437G was 54.2% for Niger and Mali combined (Niger, 48.6%; Mali, 56.5%). There was no statistically significant difference in the distribution of mutant alleles (437G and 437A/G) of the Pfdhps gene between Mali and Niger (OR=1.374; 95% CI: 0.6236–3.026; P=0.5458).

Table 3 shows the details of all multiple mutations identified in the Pfdhr and Pfdhps genes. The prevalence of the quadruple mutation Pfdhfr51I/Pfdhfr59R/Pfdhps436A/Pfdhps437G was 7.7% for Niger and Mali combined (Niger, 9%; Mali, 7.1%), and there was no statistically significant difference between them. The prevalence of the triple mutations Pfdhfr51/59/Pfdhps436 and Pfdhfr51/59/Pfdhps437 was 18.1% and 30.2%, respectively, in Niger and Mali combined, and there was no statistically significant difference between them (P=0.2369). The prevalence of double mutants Pfdhfr51/59 and Pfdhps436/437 was 58.3% and 14.9%, respectively, in Niger and Mali combined, and there was no statistically significant difference between them (P=0.2776).

The prevalence of allelic families of the msp1 genetic marker and the MOIs were examined by nested PCR, a widely used technique for studying genetic diversity [12,13]. The msp1 marker is well known for its high rate of polymorphism, and it is recommended by the World Health Organization as one of the major markers for determining genetic diversity [14]. msp1 is located on chromosome 9 of Pf and consists of 17 blocks, of which block 2 is the most variable, enabling the distinction of Pf clones. Block 2 includes the K1, MAD20, and RO33 alleles. Our investigation revealed that the prevalence of the msp1 marker alleles was 34%, 19%, and 47% for K1, MAD20, and RO33, respectively. Although all msp1 allelic families (K1, MAD20, and RO33) were found in infected children in both Mali and Niger, the prevalence of the RO33 allelic family was the highest. There was a difference in the distribution of msp1 alleles between Mali and Niger. Epidemiological studies have shown that the genetic diversity of parasite populations can vary from one site to another (e.g., 2 distinct villages in the same country) and according to the state of malaria endemicity (forest zone versus Sahelian zone) [15]. In 2011, 17 different msp1 alleles and 14 different msp2 alleles were reported in Niger. The most prevalent allelic family in the population was 3D7 (63%), followed by K1 (43.2%). The rarest allelic families were MAD20 (28.4%) and RO33 (28.4%) [13]. This illustrates the temporal and spatial variability of allelic families of Pf.

The MOI was 2 (135/90) in Mali compared with 1 (66/45) in Niger. This was identical to that of Sasoun Broum in Niger (MOI=2) in 2011 [13] and similar to that of Ivory Coast (MOI=1.7) in 2016 [12]. The MOI ranged from 1 to 3 in 2016 in Dangassa, Mali [16]. Our study results showed a high rate of malaria transmission in Sélingué and Missira (Mali). In contrast, there was a decrease in the rate of infection transmission in Niger, supported by a study conducted in Niger in 2011, reporting an average MOI of 2.8 [13].

HRM of DNA was described in 2012. It is a sensitive technique for detecting point mutations as well as for differentiating haplotypes [17]. We used strains 3D7 and Dd2 as controls because 3D7 was sensitive and Dd2 was resistant to SP. Studies on SP resistance were performed in the Maradi region of Niger in 2012, well before CPS implementation. These studies showed a very high prevalence of point mutations Pfdhfr51I (68%), Pfdhfr59R (61%), Pfdhfr108N (69%), Pfdhps436A/F/H (65%), and Pfdhps437G (83%). Dinzouna-Boutamba et al. recently reported that the prevalence of antifolate resistance-associated polymorphisms was high among parasite isolates in Gabon [18]. In Dakar (Senegal), Wurtz et al. reported the prevalence of isolates with high-level pyrimethamine resistance as 83.5% and resistance to sulfadoxine as 40.2% [19]. Beshir et al. surveyed 13 countries in the Sahel region in 2016 and 2018. They reported that the dhfr haplotype (51Ile-59Arg-108Asn) was common at both survey timepoints, but the dhps haplotype (431Ile-436Ser-437Gly-540Glu-581Ala-613Ala), crucial for resistance to SP, was rare [20]. These haplotypes are crucial for SP resistance.

Notably, mutations associated with very high levels of resistance, such as PfdhfrI164L and PfdhpsK540E were not reported in south central Niger [8]. Similarly, in Mali, a prevalence of 16.1% for the triple mutation Pfdhfr51I/Pfdhfr59R/Pfdhfr108N and 15% for the double mutation Pfdhps436A/F/H/Pfdhps437G were reported in Dangassa [16]. SP resistance develops easily, and many studies have demonstrated this after SP introduction in 1967, to the extent that some areas, such as the Horn of Africa, have become multiresistant zones. Our findings did not show a significant difference in genetic diversity between the study areas (Mali and Niger), suggesting that the 2 regions have similar patterns of genetic diversity in P. falciparum strains, which may have implications for malaria control and treatment strategies in both areas. The limitation of our study is the absence of mutations at position 108 of Pfdhfr and position 540 of Pfdhps. Furthermore, the determination of Pf resistance to pyrimethamine often requires the study of positions 108 and 164. Additionally, position 540 of the Pfdhps gene is well known for its involvement in sulfadoxine resistance. Our study did not include these mutations, which is a limitation.

In conclusion, our study highlights the genetic diversity of Pf strains in Mali and Niger, revealing the presence of multiple mutations associated with SP resistance. Our findings underscore the importance of careful monitoring of the status of antimalarial drug resistance and the need for continued surveillance to ensure the success of chemoprevention programs in the Sahel region. Our result also demonstrates the utility of HRM as a rapid and cost-effective tool for detecting mutations associated with antimalarial drug resistance. Thus, HRM could be useful for guiding treatment decisions and informing public health strategies aimed at malaria control and elimination malaria.