Profiling antimalarial drug-resistant haplotypes in Pfcrt, Pfmdr1, Pfdhps and Pfdhfr genes in Plasmodium falciparum causing malaria in the Central Region of Ghana: a multicentre cross-sectional study

Abstract

Background:

The proliferation of Plasmodium parasites resistant to antimalarial drugs poses a serious threat to human life and remains an obstacle to managing and eradicating Plasmodium falciparum. The surveillance of molecular markers has become necessary to monitor the spread of resistant haplotypes and discover emerging mutations.

Objective:

This molecular epidemiological study aimed to evaluate the prevalence of known mutations in the drug resistance genes Pfcrt, Pfmdr1, Pfdhfr and Pfdhps in the Central Region of Ghana.

Design:

A multi-centre cross-sectional study.

Methods:

This prospective study utilised dried blood spots from individuals with P. falciparum-infection from five districts in the Central Region of Ghana. Selective Whole Genome Amplification (sWGA) and Single Nucleotide Polymorphisms (SNPs) in P. falciparum chloroquine transporter genes (Pfcrt), P. falciparum multidrug resistance 1 (Pfmdr1), P. falciparum dihydropteroate synthase (Pfdhps) and P. falciparum dihydrofolate reductase (Pfdhfr) were analysed.

Results:

Whole genome sequencing was carried out on 522 samples. Of these, 409 (78%) samples were successfully sequenced. Six (6) of the sequenced samples were of co-infection of other parasite species with P. falciparum and excluded from the analysis. Analysis of the Pfcrt gene revealed 0.5% were CVIET (C72, V73, M74I, N75E, K76T) while the Pfcrt CVMNK (C72, V73, M74, N75, K76) wild-type haplotypes were 97% with (2.5%) (CV[M/I][N/E][K/T]) being mixed haplotypes. In the Pfmdr1 gene, monoclonal haplotypes; NFD (N86, Y184F, D1246) and YFN (N86Y, Y184F, D1246N) occurred at 44% and 9.8%, respectively, whereas mixed- haplotypes (N[Y/F]D and [N/Y][Y/F]D) were 23.5% and 0.3%, respectively. Combined Pfdhfr/Pfdhps genes yielded about 88% Pfdhfr IRNI (N51I, C59R, S108N, I164) + Pfdhps A437G haplotypes (conferring partial resistance to Sulphadoxine-Pyrimethamine (SP)) while 9% of the parasites had Pfhdfr IRNI + Pfdhps A437G + K540E haplotypes (conferring full resistance to SP). The wild-type haplotype, Pfdhfr (N51, C59, S108, I164) and Pfdhps (S436, A437, K540, A581, A613) was not observed.

Conclusion:

The findings show a low prevalence of CVIET and relatively higher rates for Pfmdr1 NFD and parasites with Pfdhfr IRNI (N51I, C59R, S108N, I164) + Pfdhps A437G haplotypes. These observations advocate for enhanced surveillance which is inimical to malaria management in an endemic area.

Introduction

Resistance of Plasmodium parasites to chloroquine and fansidar (Sulphadoxine-Pyrimethamine (SP)) antimalarial medications has hampered malaria case treatment, leading to a global switch to artemisinin-based combination therapy (ACT) for the treatment of uncomplicated falciparum malaria. ¹ In Sub-Saharan Africa, chloroquine and fansidar (SP) resistance were reported to have been imported from Southeast Asia. ²

Currently, Artemether–lumefantrine (AL) and artesunate–amodiaquine (AS-AQ) are the most widely utilized ACTs in all African countries including Ghana.^3,4 Interestingly, in vitro and ex vivo susceptibility assays have documented elsewhere that these two drugs appear to exert conflicting selection pressures on single nucleotide polymorphisms (SNPs) in the Plasmodium falciparum multidrug resistance transporter 1 (Pfmdr1) gene.^5,6

Genomic surveillance provides a rapid and efficient monitoring tool for studying drug resistance patterns. Through genetic surveillance, the K76T mutation in the chloroquine (CQ) transporter of P. falciparum (Pfcrt) was found to be associated with CQ resistance, while single nucleotide polymorphisms in the Pfmdr1 genes select for drug resistance of ACT partner drug resistance.^7–9

In Africa, SNPs in codons N86, Y184 and D1246 of Pfmdr1 are the most prevalent, and various haplotype combinations result in decreased sensitivity to multiple drugs. For instance, AQ alone and AS-AQ have been shown to select for Pfmdr1 86Y, Y184 and 1246Y (the YYY haplotype) is associated with parasite recrudescence and reinfection following reduced sensitivity to amodiaquine while the AL combination selects for N86, 184F and D1246 (the NFD haplotype) associated with an increased tolerance of lumefantrine.^10,11 There are reports of other Pfmdr1 mutations, such as S1034C and N1042D, elsewhere but infrequent in Africa.^9,12

Single nucleotide polymorphisms in the P. falciparum dihydropteroate synthase (Pfdhps) and P. falciparum dihydrofolate reductase (Pfdhfr) genes have been shown to cause resistance to the combination of SP drugs. ¹³ Key factors in the development of resistance to pyrimethamine in vitro are specific amino acid alterations at codon 108 (S108N), 51(N51I) and/or 59(C59R) and 164(I164L) on the Pfdhfr gene.^13–16 Regarding the Pfdhps gene, point mutations S436A/F, A437G, K540E, A581G and A613T/S provide resistance to Sulphadoxine in vitro. ¹⁷ Novel mutations such as K540T and K540N have been reported in Pfdhps in Indonesia, Cameroon, and India. ¹⁴ Even though K540T has not been explicitly implicated in resistance to Sulphadoxine, ¹⁴ K540N has previously been associated with lower levels of sulpha drug resistance. ¹⁸ Combinations of several mutations in both Pfdhps and Pfdhfr are used to distinguish between different levels of SP resistance.^19,20 Here, we use the nomenclature partial, complete/full and super resistance to describe the three levels of resistance, which are represented by Pfdhfr codons 51, 59, and 108, plus Pfdhps codons 437), (Pfdhfr codons 51, 59, and 108, plus Pfdhps codons 437 and 540) and (Pfdhfr codons 51, 59, and 108 and Pfdhps A437G K540E A581G) respectively.^20,21 The observed resistance in A437G is enhanced by the mutations Pfdhps K540E, Pfdhps A581G and Pfdhps A613S. ²⁰

In Ghana, malaria still persists in the southern (coastal) and middle (forest) belts and remains the leading cause of morbidity and mortality.^22,23 Studies on the molecular markers of resistance showed an increasing trend in the prevalence of haplotype Pfmdr1 N86-F184-D1246 from 2003 to 2010. ²⁴ In another study in the Western region of Ghana, frequencies of the point mutations implicated in the emergence of antifolate resistance were reported to be very high at codons 108, 59 and 51 on the Pfdhfr gene and moderate prevalence at codons 540 and 437 on the Pfdhps gene. However, a synergistic analysis of mutations in both genes yielded about 13.0%, quadruple mutations in the sample set at codons I51, R59, N108 and G437 (IRNG) and 3.7% at codons I51, R59 N108 and E540 (IRNE). ¹³ Similarly, the percentages of the Pfdhfr 51I, 59R, 108N and Pfdhps 437G mutant alleles reported in selected regions in Ghana were reported to be high in several other studies.^25,26 In the central region, the percentage prevalence of mutations in Pfcrt and pfmdr1 stood at 29% for chloroquine resistance mutations in 2017, Cape Coast ²⁷ and 66.36% prevalence of Pfcrt K76 in 2021. ⁸ Reports on the high prevalence of Y184F of the Pfmdr1 gene was also reported in 2017. ²⁸

Considering the high prevalence of malaria in the Central Region of Ghana^23,29 amidst all interventions against the disease as well as the paucity of genetic data on the spread of resistance to current chemotherapy in Ghana, it is imperative to determine the prevalence of known drug-resistance-related point mutations in Pfcrt, pfmdr1, Pfdhfr and Pfdhps in clinical isolates from the central region.

This study is meant to unearth the rate of parasites resistant circulating in the study sites and add to the P. falciparum genetic data reported within the region and throughout the country aiding in policy direction.

Materials and methods

Selection of study areas and sampling period

This multicentre cross-sectional study was done in Agona West (AW), Assin Central (AC), Gomoa East (GE), Abura-Asebu-Kwamankese (AAK), and Cape Coast (CC). The optimum allocation method was used to choose three districts at random from the wooded districts (AW, AC, and GE) and two districts from the coastal districts (AAK and CC). From these districts, study participants with malaria (RDT positive) were concurrently recruited from the respective district hospitals (Figure 1). The study participants were randomly selected by systematically selecting patients with malaria on each clinic day. Study participants were recruited from September 2020 to February 2021.

Figure 1.

Flow chart for participant recruitment and sample selection for genotyping.

RDT- Rapid Diagnostic Tests.

Study design

This was a cross-sectional study conducted in the Central Region of Ghana. Malaria samples were obtained from study participants residing in randomly chosen districts within the region.

Criteria for study population selection

In all districts, participants who started oral or parenteral antimalarial medication before receiving microscopy results were excluded. Furthermore, research participants who had ever left Ghana, to countries in Central and Northern Africa or South-East Asia, were equally excluded (Figure 1). Lastly, the inclusion of laboratory-confirmed Plasmodium infection along with the participant and parental consent was crucial.

Sample size determination and blood sample sampling procedure

According to our previous research, 38.2% of cases of malaria were confirmed by microscopy out of 2495 (62.5%) suspected malaria cases. ²³ Thus, using Cochrane’s formula; n = z²p(1 − p)/d², where n = sample size, z = confidence level at 95% (standard value of 1.96), d = error margin at 5% (standard value of 0.05), a minimum sample size of 363 was reached calculated, setting the power of the study at 80%. At the end, 1525 participants were recruited. The large study participants used in this study were to cater for false negative microscopy results, non-falciparum Plasmodium, missing samples, samples with missing data and samples with very low parasitaemia (<500 parasites/µL) which do not yield enough genomic DNA for sequencing. At the end, 522 samples were eligible for sequencing.

Criteria for selecting samples for genotyping

The criteria for selecting a sample for genotyping were parasitaemia > 500 parasites/µL. In addition, samples must not be haemolysed. Further, samples should be haemoglobinated enough to produce clear and thick blood spots (Figure 1).

Laboratory procedures

Blood sample collection, malaria screening and preparation of dried blood spots

Blood samples were taken from study participants suspected of malaria. Four millilitres of whole blood were drawn into an Ethylenediaminetetraacetic acid (EDTA) tube and mixed uniformly. Initially, CareStart mRDT (Access Bio, Somerset, USA) was used to test the samples to identify malaria-positive cases of malaria parasites. Parasitaemia was determined as earlier published.^30,31 In summary, 6 μL of whole blood was used to prepare thick blood films, air dried, stained with 10% Giemsa for 10 min, and examined using the light microscope. Four dried blood spots were made from the detectable microscopy samples according to the Malaria Genome Laboratory protocols of the Wellcome Sanger Institute (WSI).

Plasmodium DNA extraction and species identification

Selective whole genome amplification (sWGA) was employed in this study to enrich to preferentially amplify P. falciparum DNA over human DNA background as described in Oyola et al. ³² A BDS 600PLUS robotic puncher (Microelectronic System, Brendale, Australia) punched 6–8 dried blood spots (DBS) into Eppendorf™ Deepwell™ plate of 96 wells (Fisher Scientific, UK). P. falciparum genomic DNA was extracted from DBS using the QIAamp DNA Investigator Kit (Qiagen, CA, USA) following the kit manufacturer’s instructions. At least 5 ng of DNA were obtained per sample, for whole genome sequencing.

Library preparation and sequencing of the genome of P. falciparum

The reaction mix consisted of the following: template DNA, 1 × bovine serum albumin, 1 mM dNTP, 2.5 μM of each amplification primer, (primers used for this study are available at https://www.malariagen.net/wp-content/uploads/2023/10/GbS01_Tag_plate_preparation.pdf) 1 × Phi29 reaction buffer, and 30 units of Phi29 polymerase enzyme (New England Biolabs). Isothermal amplification conditions (35°C for 5 min, 34°C for 10 min, 33°C for 15 min, 32°C for 20 min, 31°C for 30 min, 30°C for 16 h before denaturing Phi29 polymerase enzyme at 65°C. After being cleansed with 200 μL of 80% ethanol, the purified XP amplicons were eluted using 50 μL of elution buffer. DNA libraries were created using New England Biolabs’ NEBNext^® Ultra™ DNA library preparation kit (New England Biolabs) before being sequenced on an Illumina HiSeq 2500 DNA sequencer. This protocol has been published.^32,33

Genetic analysis of gene alleles

Standard Illumina QC was applied to the sequence data derived from every sample and each dataset was independently analysed by mapping sequence reads to the 3D7 reference genome using Burrows-Wheeler Aligner. ³² Prior to that, Torrent Suite pipeline software was used to de-multiplex and filter the raw reads using conventional quality filtering parameters. Read quality was assessed using the Torrent Suite FastQC plugin v0.10.1, and high - quality reads were aligned to the reference genome. ³⁴ In the analysis of the genomic data, only successfully sequenced data obtained from each of the sample was analysed for Pfcrt, Pfdhfr, Pfdhps, Pfmdr1 genes.

Data processing and statistical analysis

The data was organised in Microsoft Excel 2016 and imported into SPSS Version 24 software (Chicago, IL, USA) for analysis. Descriptive statistics (numbers and corresponding percentages) were carried out.

Ethics approval and consent to participate

This study received ethical approval from Ghana Health Service Ethics Review Committee (GHS-ERC Number: GHSERC017/03/20). In addition, the parent or legal guardian of each child provided a written informed consent and a child assent form. Written self-consent was obtained from study participants 18 years and over.

Results

Sequencing success rates

Out of 522 samples, 409 (78%) were successfully sequenced. Successful sequencing was where the entire amino acids denoting Pfdhps, Pfcrt, Pfdhfr, and Pfmdr1 haplotypes were present. Of the 409 samples that successfully sequenced parasites, six (6) were excluded from the analysis due to co-infection of other parasite species with P. falciparum. Therefore, analysis described in this publication is based on the successful sequencing of SNPs within 403 Pfcrt, Pfmdr1, Pfdhps and Pfdhfr genes. In this section, putative markers of antimalarial drug resistance focused on amino acid changes in Pfcrt M74, N75, K76; Pfdhfr N51, C59, S108 and I164; Pfdhps S436, A437, K540, A581 and A613; Pfmdr1 N86, Y184 and D1246. Different amino acid polymorphisms in the same genetic loci made it possible to find that 47 different parasite clones were circulating in the study sites. Table 1 Shows the distribution of successfully sequenced genes by districts.

Table 1.

Distribution of successfully sequenced genes by districts.

District	Pfdhps	Pfcrt	Pfdhfr	Pfmdr1
Abura-Asebu-Kwamankese	39 (12.4)	57 (14.1)	48 (13.4)	44 (13.1)
Agona Swedru Municipality	21 (6.7)	29 (7.2)	24 (6.7)	27 (8.0)
Cape Coast Metropolis	61 (19.4)	73 (18.1)	67 (18.7)	61 (18.2)
Assin Central Municipality	128 (40.8)	154 (38.2)	139 (38.7)	132 (39.3)
Gomoa East District	65 (20.7)	90 (22.3)	81 (22.6)	72 (21.4)
Total	314 (77.9)	403 (100)	359 (89.1)	336 (83.4)

Source: Authors analysis.

Data are presented as numbers and proportions.

n, number of genes.

Distribution of Pfcrt gene mutations

The Pfcrt SNPs were successfully sequenced in 403 samples. Of this number, 391/403 (97%) had the wild-type CVMNK (C72, V73, M74, N75, K76) haplotype, 2/403 (0.5%) were of CVIET (C72, V73, M74I, N75E, K76T) haplotype, while the rest 10/403 (2.5%) had mixed haplotypes /infections (CV[M/I][N/E][K/T]). No haplotype of SVMNT was found.

Prevalence of mutant haplotypes in Pfmdr1 genes

Regarding the Pfmdr1 gene, 336 parasites were successfully sequenced at all three loci, namely, amino acid positions N86, Y184 and D1246. Of the number of parasites successfully sequenced, 75 (22.3%) were of the NYD (N86, Y184, D1246) wild type, while the rest had mutations. The majority of the parasites (44.05%) had a single mutation NFD (N86, Y184F, D1246) haplotype while the YFN (N86Y, Y184F, D1246N) triple mutation was 9.8% (33/336). The rest (23.8%) were mixed haplotypes (N[Y/F]D and [N/Y][Y/F]D) (Figure 2).

Figure 2.

Distribution of the Pfmdr1 gene haplotypes at amino acid positions N86, Y184, D1246.

Prevalence of Pfdhfr mutations

Regarding the Pfdhfr gene, 359 parasites were successfully sequenced for all four amino acids that make up the Pfdhfr haplotype. The amino acids are located at positions 51, 59, 108 and 164. Some parasites (1.4%, 5/359) had the wild-type haplotype (N51, C59, S108 and I164). The rest (98.6%) had various mutations. The triple mutant, IRNI, was in the majority (76.6%, 275/359). The dominant double mutant haplotype was NRNI (N51, C59R, S108N, I164) at a rate of 5.8% (21/359). Figure 3 contains the other mutant haplotypes in mixed haplotype infections.

Figure 3.

Distribution of point mutations in the P. falciparum dihydrofolate reductase (Pfdhfr) gene. The amino acid positions for the haplotypes are N51, C59, S108, I164.

Prevalence of Pfdhps mutations

The haplotype for the P. falciparum dihydropteroate synthase gene (Pfdhps) has five amino acids at positions 436, 437, 540, 581 and 613. Of the total parasites sequenced, 314 were fully sequenced for all five amino acids. The wild-type Pfdhps haplotype is represented as S436, A437, K540, A581, and A613, denoted as SAKAA haplotype. The most prevalent haplotype was SGKAA (A437G) at a rate of 32.8% (104/314). The dominant double and triple mutant haplotypes AGKAA and AGKAS are prevalent in 19.4% (61/314) and 6.7% (21/314). The only quadruple mutant haplotype was AGKGS with a prevalence of 0.6% (2/314) (Figure 4). Figure 5 presents the distributions of mutant haplotypes in mixed haplotype infections.

Figure 4.

Distribution of monoclonal haplotypes in the P. falciparum dihydropteroate synthase gene (Pfdhps) at amino acid positions S436, A437, K540, A581 and A613.

Figure 5.

Mixed haplotype distribution in the P. falciparum dihydropteroate synthase gene (Pfdhps) at amino acid positions S436, A437, K540, A581 and A613.

Distribution of the Pfdhfr/Pfdhps haplotypes

Analysing SP resistance, 294 parasites had both Pfdhfr/Pfdhps genes fully sequenced. In these combined genes, the wild-type haplotype (51N-59C-108S 164I-436S-437A-540K-581A-613A) was not identified. In monoclonal infections, the dominant mutations of the Pfdhfr/Pfdhps gene were the quadruple mutation – N51I-C59R-S108N- 164I-436S-A437G- 540K-581A-613A (27.9%) associated with partial resistance to fansidar. Only one double mutation in the Pfdhfr gene was found in 0.3% together with the wild-type Pfdhps gene (Pfdhfr/Pfdhps NRNI-SAKAA). The dominant triple, quintuple, and sextuple mutations were Pfdhfr/Pfdhps NRNI-SGKAA (2.0%), Pfdhfr/Pfdhps IRNI-AGKAA (15.7%) and Pfdhfr/Pfdhps IRNI-AGKAS (5.4%) respectively. Only two septuplet mutations (Pfdhfr/Pfdhps IRNI-AGKGS) were found at (0.7%) (Table 2). A total of 41 different mutations in the Pfdhfr/Pfdhps genes were found in mixed haplotype infections. Forty of the parasites harboured Pfdhfr/Pfdhps IRNI-[S/A]GKAA haplotypes, while 65% (26/40) of them were found at a single frequency (Supplemental File 1).

Table 2.

Distribution of Pfdhfr/Pfdhps haplotypes associated with Sulphadoxine-pyrimethamine resistance in monoclonal infections.

Pfdhfr/Pfdhps gene mutations	Total number of Pfdhfr/Pfdhps genes successfully sequenced N = 294 (%)
Wild types (Pfdhfr NCSI – Pfdhps SAKAA)	0
Single Mutation
NCSI-AAKAA	2 (0.68)
NCSI-SGKAA	1 (0.34)
Double Mutation
NRNI-SAKAA	1 (0.34)
Triple Mutation
NRNI-SGKAA	6 (2.04)
IRNI-SAKAA	1 (0.34)
Quadruple Mutation
IRNI-SGKAA	82 (27.89)
NRNI-AGKAA	5 (1.70)
IRNI-AAKAA	3 (1.02)
ICNI-AGKAA	2 (0.68)
Quintuple Mutation
IRNI-AGKAA	46 (15.65)
IRNI-SGEAA	1 (0.34)
NRNI-FGKAS	1 (0.34)
Sextuple Mutation
IRNI-AGKAS	16 (5.44)
IRNI-FGKAS	3 (1.02)
IRNI-AGKSA	1 (0.34)
Septuplet Mutation
IRNI-AGKGS	2 (0.68)

Distribution of Pfdhfr/Pfdhps haplotypes at amino acid positions 51N-59C-108S 164I-436S-437A-540K-581A-613A).

Elucidating the mixed haplotypes

In the mixed haplotype infections, it was difficult to clarify the clones (wild-type or mutants) observed. It was rather easy to determine the clones present when the amino acid polymorphism occurred at only one gene locus. In Pfmdr1 N[Y/F]D, the likely clones were NYD (a wild-type clone) and NFD (a mutant clone). The same applied to Pfdhps [S/A]KAA, where SAKAA (wild type) and AAKAA (mutant), and Pfdhfr NC[S/N]I (NCSI – wild-type and NCNI – mutant), could be the case. However, in the case of Pfcrt CV[M/I][N/E][K/T], Pfmdr1 [N/Y][Y/F]D and Pfdhps [S/A][G/A]KAA, it was difficult to tell whether a wild-type haplotype existed.

Discussion

The prevalence of malaria in the Central Region of Ghana has been found to be high. ²³ For this reason, parasite mutations that have been associated with antimalarial drug resistance, known as putative antimalarial drug-resistant genes, were studied in the Central Region. The study focused on the Pfcrt, Pfmdr1, Pfdhfr and Pfdhps genes. Our findings revealed a low prevalence (3%) of Pfcrt mutations involving the chloroquine-resistant marker, K76T, as occurred in the CVIET (C72, V73, M74I, N75E K76T) haplotype. This finding is quite interesting, as an extensive reverse mutation in chloroquine-resistant parasites has increased in prevalence in Ghana^8,27 and elsewhere. ³⁵ In 2020, the prevalence of chloroquine resistance in the Central and Eastern regions stood at 21% and 5%, respectively. ²⁷ In the same year, the prevalence was 11.6% in the Greater Accra Region of Ghana, ⁷ an adjoining region to this study region. Also, Samples analysed from three other regions in Ghana (Greater Accra Region < Upper East Region-Navrongo < Bono East Region- Kintampo) produced a prevalence of 8% for Pfcrt K76T in 2018. ²⁵ The decline in chloroquine-resistant parasites observed in this study could be attributed to total withdrawal of chloroquine and continuous adherence to ACT in Ghana. Despite the high prevalence of chloroquine-sensitive parasites (97%) identified in this study, the reintroduction of chloroquine for the treatment of malaria should be considered carefully. Elsewhere, the genetic makeup of the parasite and site-specific epidemiology may explain the differences between studies in Nigeria ³⁶ and Angola ³⁷ where the prevalence of chloroquine-resistant haplotype was recorded at 61.1% and 73%, respectively.

After chloroquine was withdrawn, and ACT was introduced in 2009, ³⁸ certain mutations in the Pfmdr1 gene were found to confer resistance to artemisinin partner drugs. In our study area, the alleles of the Y184F mutant were the most prevalent and resulted in a relatively high prevalence of Pfmdr1 NFD (N86, Y184F D1246) with a relatively low prevalence of YFN (N86Y, Y184F, D1246N). Elsewhere, the Y184F mutation has been found to reduce the sensitivity of parasites to quinine, amodiaquine, chloroquine, mefloquine, and lumefantrine. ³⁹ In Ghana, artemether-lumefantrine is the drug of choice for the treatment of malaria. However, more than 75% of the parasites analysed in this study bore the Y184F mutation in the YFN or NFD haplotypes. This observation is worrying for a country where antimalarial drug-resistant parasites have not yet been confirmed. The prevalence of Pfmdr1 Y184F in P. falciparum seems to be high, especially, in Africa. The prevalence was found to be 71.4% in Niger, ⁴⁰ 66.4% in Cote d’Ivoire, ⁴¹ 68.7% in Burkina Faso ⁴² and elsewhere, 70% in India. ⁴³

The analysis for the Pfhdfr/Pfdhps genes was also carried out. About 77% of the samples had triple Pfhdfr IRNI haplotype. This rate was higher than the rate (> 60%) reported by Amenga-Etego et al. ⁴⁴ This study observed single to quadruple mutations, but Amenga-Etego et al. observed up to quintuple mutations. Mutations in the Pfhdfr/Pfdhps genes have been shown to render SP less effective against P. falciparum parasites. SP treatment was introduced in Ghana in 2003.^45,46 Since then, it has been used to prevent malaria in pregnant women ⁴⁷ and children under 5 years. ⁴⁸ To pregnant women, SP is administered at five different times during the course of the pregnancy. ⁴⁹ Despite the use of SP to prevent malaria, the parasites are less sensitive to the drug to due to N51I, C59R, and S108N mutations in the Pfdhfr gene together with Pfdhps A437G (partial resistance), Pfdhps A437G + K540E (full resistance) and Pfdhps A437G + K540E + 581G/164L (super resistance), as published elsewhere. ²¹ The haplotype contributed by the Pfdhfr gene is IRNI, and that of Pfdhps A437G (partial), Pfdhps A437G + K540E (full), and Pfdhps A437G + K540E + A581G/A581L (super). This study observed IRNI in approximately 77% of the parasites, while in the case of the Pfdhps gene, A437G was observed in SGKAA, AGKAA, SGEAA, AGKAS, FGKAS, AGKSA, and AGKGS in 99% of the monoclonal infections. In monoclonal infections, one (0.5%) Pfdhps A437G + K540E was observed, while in mixed haplotype infections, 12 (11%) parasites were found to harbour the mutations Pfdhps A437G + K540E. Analysis of mutations in Pfdhfr and Pfdhps genes revealed that about 88% and 9% of the parasites had putative haplotypes that have been found elsewhere to confer partial or full resistance to SP, respectively. ²¹ No super-resistant haplotype was identified. In Africa, the full resistant haplotype (IRNI-SGEGAA) was first identified in Kenya, ⁵⁰ therefore, this study will provide a recent report of the identification of this haplotype. SP failure in Kenya is attributable to this haplotype. Although these haplotypes have been found to confer resistance to SP, ²¹ this has not been proven in parasites in Ghana. Therefore, the clinical significance of these haplotypes must be determined, in a prospective chemotherapeutic efficacy study.

Limitations

The major limitation of this study was our inability to expand the complex mixed haplotype into individual parasite clones. This is because only amino acids detected at each gene were available to the team at the time of this publication. Malaria Genome Laboratory of the WSI did not provide the sequence reads but the amino acids detected at each gene loci of the gene target of interest. Secondly, the drug-resistant haplotypes reported herein are a result of a survey of molecular markers but not the result of therapeutic efficacy studies. Further, this publication does not have the Kelch 13 polymorphisms associated with antimalarial drug resistance. Finally, sequencing failure was observed among 113 (21.6%) of the 522 parasites. The reasons for the sequencing failure were not investigated.

Conclusion

The study offers valuable information on the genetic alterations caused by the therapeutic and preventive interventions used in the Central Region of Ghana. The resurgence of parasites with CQ (Chloroquine) sensitivity confirms the reverse mutation of CQ-resistant genes that were seen in Ghana. However, reintroduction of CQ for malaria treatment should be done cautiously since the parasite could easily become resistant in the face of poor therapeutic practices. Significantly prevalent were haplotypes carrying the Pfmdr1 Y184F mutated allele commonly associated with decreased sensitivity to Artemether partner drug lumefantrine. Pfmdr1 D1246N were observed, which requires further investigations and will provide additional information for understanding their effect on susceptibility to artemisinin partner drugs. Furthermore, the very low prevalence of mutations Pfdhfr I164L, Pfdhps K540E and Pfdhps A581G must be monitored and contained, as these mutants can lead to the generation of full and super-resistant haplotypes, compromising the efficacy of SP use for intermittent preventive treatment during pregnancy (IPTp) and seasonal malaria chemotherapy in the region.

Appendix

Abbreviations

ACT artemisinin-based combination therapy

DBS dried blood spot

Pfcrt P. falciparum chloroquine resistance transporter

Pfdhfr P. falciparum dihydrofolate-reductase

Pfdhps P. falciparum dihydropteroate synthetase

Pfmdr P. falciparum multi-drug resistance

mRDT Malaria rapid diagnostic test

SNPs single nucleotide polymorphisms

sWGA Selective Whole Genome Amplification

Amino acid designations

A alanine

C cysteine

D aspartic acid

E glutamic acid

F phenylalanine

G glycine

I isoleucine

K lysine

L leucine

M methionine

N asparagine

R arginine

S serine

T threonine

V valine

Y tyrosine