Identification of Mutations in Antimalarial Resistance Gene Kelch13 from Plasmodium falciparum Isolates in Kano, Nigeria

Malaria control relies on first-line treatments that use artemisinin-combination therapies (ACT). Unfortunately, mutations in the plasmodium falciparum kelch13 gene result in delayed parasite clearance. Research on what is causing ACT failure is non-existent in northwestern Nigeria. Thus, the presence of mutations in kelch13 in P. falciparum isolates from Kano, Nigeria was investigated in this study. Microscopic examination of 154 blood samples obtained from patients revealed a high prevalence of P. falciparum infection (114 positive individuals, slide positivity rate = 74.03%). The 114 patients were administered Cartef® (ACT) and out of the 50 patients that returned for the 14-day follow up, 11 were positive for P. falciparum (slide positivity rate = 22%). On day 0, 80 samples out of 114 and 11 samples on day 14 (91 out of 125 microscopy-positive samples) were positive with Plasmodium according to the PCR of cytochrome oxidase I, which corresponds to 72.8%. A fragment of the kelch13 gene encompassing the propeller domains was sequenced in 49 samples, alongside samples of the susceptible strain pf_3D7. Low polymorphism was observed, suggesting a lack of selection on this gene, and only six mutations (Glu433Gly, Phe434Ile, Phe434Ser, Ile684Asn, Ile684Thr and Glu688Lys) were found. The epidemiologic impact of these mutations and their potential role in ACT resistance needs to be investigated further.


Introduction
Malaria still takes the lives of approximately 405,000 people every year, 93% of these are in Africa, of which Nigeria alone accounts for 25% [1]. Malaria control relies heavily on vector control and administration of antimalarial medicines [2]. The use of antimalarials and vector control has contributed to a significant reduction in the malaria burden in Africa by about 40% from 2000 to 2015 [3]. Starting from 2005, the World Health Organization (WHO) recommended artemisinin combination therapy (ACT) as a first-line therapy for uncomplicated P. falciparum malaria, a policy that was adopted by many endemic countries [4]. Nigeria first adopted ACT drugs as a first-line treatment of uncomplicated P. falciparum in 2005 [5]. Currently, there are no alternative, fully effective

Drug Treatment and Follow Up
On the day of the first sample collection (day 0), 114 patients were identified as infected with P. falciparum using thick and thin blood film microscopy. Microscopic examination of slides from the rest of the patients (forty individuals) revealed no infection of any of the four major human malaria parasites. The 114 patients infected with Plasmodium were provided with ACT, Cartef ® (GB PHARMA, United Kingdom), an Artemether/Lumefantrin (80 mg/480 mg taken twice daily). The dangers of non-compliance with the drug regimen was explained in detail to all patients. Parasite clearance was evaluated after 14 days of drug therapy in the 50 individuals that returned for follow-up, by blood collection and microscopy (conducted independently in the hospitals and BUK). The positive samples from the microscopy at day 0 (114 individuals) and those from the 14-days after treatment (50 individuals returned and 11 of them were still infected with P. falciparum) were used for downstream molecular analyses.

DNA Extraction and Confirmation of Plasmodium Falciparum Infection Using PCR
DNA was extracted from the whole blood of 125 patients who were positive for P. falciparum. These included all 114 samples collected at day 0, and the 11 positive samples from follow up. DNA isolation was carried out using the QIAamp ® DNA Mini Kit (QIAGEN, Hilden, Germany) according to the manufacturer's instructions. DNA pellet from an antimalarial susceptible parasite strain, 3D7 was provided by Dr Janet Storm from the Parasitology Department, LSTM, UK and used alongside the field samples for DNA extraction. The DNA was eluted in 100 µL of nuclease-free water and its concentration measured using the Qubit 4.0 fluorometer (Invitrogen, Massachusetts, USA). Samples were stored at −20 • C.

Amplification of Propeller Domains of the Kelch13 Gene and Column Purification
The DNA samples were used to amplify a fragment of the kelch13 gene encompassing the propeller domain, based on the nested PCR protocol of Ariey et al. [8] with modifications in the kelch-in primers and thermocycling conditions. In the first round PCR, primers kelch-out-F (5 -gggaatctggtggtaacagc−3 ) and kelch-out-R (5 -cggagtgaccaaatctggga−3 ) were used. PCR was carried out in a 20 µL final volume comprised of 2 µL of the genomic DNA, 10 µL of a GoTaq master mix (Promega, Wisconsin, USA) containing optimized buffer, MgCl 2 and dNTP mixes; 1 µL each of forward and reverse primers and 6 µL ddH 2 0. Thermocycling conditions were initial denaturation at 95 • C for 1 min, followed by 35 cycles each of 20 sec at 95 • C (denaturation), 20 sec at 57 • C (primer annealing), 1.5 min at 60 • C (extension). This was followed with a 3 min final extension at 60 • C. PCR products were separated in 1.5% agarose gel stained with pEqGREEN and visualized for bands. For nested PCR, primers kelch-in-F2 (5 -cataccaaaagatttaagtgaaagtgaagc−3 ) and the kelch-out-R (above) were used. PCR was carried out in a final volume of 20 µL comprised of 2 µL of the genomic DNA, 10 µL of master mix (Promega, Wisconsin, USA), 1 µL each of forward and reverse primers and 6 µL of ddH 2 0. Amplification was carried out using the following conditions: initial denaturation at 95 • C for 1 min, followed by 35 cycles each of 20 sec at 95 • C (denaturation), 20 sec at 57 • C (primer annealing), 1 min at 60 • C (extension). This was followed with a 3 min final extension at 60 • C. PCR products were separated in a 2% agarose gel stained with pEqGREEN and examined for bands. The nested PCR products were purified using the QIAquick ® PCR Purification Kit (QIAGEN, Hilden, Germany) and DNA eluted in 30 µL of nuclease-free water.

Cloning of the Kelch13 Fragment and Sequencing
The purified nested products were ligated into pJET1.2 (CloneJET PCR Kit, ThermoFisher Scientific, UK) vector according to the manufacturer's instruction. Ligated products were transformed into DH5α E. coli cells (Promega, Wisconsin, USA) by mixing 4 µL of the purified nested products to 40 µL of the cells pre-chilled on ice. This was incubated on ice for 30 min followed by heat shocking for 45 sec at 42 • C. Transformants were returned to ice and chilled for 2 min before 950 µL of S.O.C medium was added. Transformants were incubated for 1 h at 37 • C and 200 rpm. Then, 100 µL of the transformants were streaked onto LB plates containing 100 mg/mL ampicillin and incubated at 37 • C overnight. Colonies that had grown in the plates were individually picked and diluted in 20 µL ddH 2 0 and used for colony PCR. Primers, pJET1.2-F (5'-cgactcactatagggagagcggc−3') and pJET1.2-R (5-aagaacatcgattttccatggcag−3') were used in PCR to identify positive colonies. 1.5 µL of 10× Taq A buffer, 2 µL of dNTP mixes, 0.75 µL of 2 mM MgCl 2 , 0.4 µL each of above primers, 0.1 µL of KAPATaq polymerase and 13.9 µL of ddH 2 0 were constituted into a final volume of 19 µL. Then, 1 µL of colony suspended in ddH 2 0 was added to this and PCR was carried out with the following conditions: initial denaturation at 95 • C for 3 min, followed by 35 cycles each of 30 sec at 94 • C (denaturation), 30 sec at 60 • C (primer annealing), 1.5 min at 72 • C (extension). This was followed with 5 min final extension at 72 • C. PCR products were separated in a 1.5 % agarose gel as described above. Positive colonies were mini-prepped overnight. Positive colony (4 µL) in ddH 2 O and 4 µL of 100 mg/mL ampicillin were added into a 15 mL tube containing 6 ml of LB medium. Tubes were incubated at 37 • C and 200 rpm for 14 h and 4.5 mL of overnight culture pelleted at 13000 rpm for 10 min. Plasmid preparation was carried out using the QIAprep ® Spin Miniprep Kit (QIAGEN, Hilden, Germany) and the plasmid concentration was measured using a Nanodrop spectrophotometer (Thermo Fisher Scientific, Massachusetts, USA). Plasmids were sequenced using the pJET1.2-F and pJET1.2-R primers mentioned above.

Analysis of Genetic Variability of Kelch13
Polymorphism analysis of sequences was carried out through manual examination of the sequence traces using Bioedit version 7.2.3.0 [24] and/or nucleotides/amino acid differences from multiple sequence alignments with the CLC sequence viewer v7.6 (http://www.clcbio.com/). Genetic parameters such as the number of haplotypes (h) and its diversity (Hd), the number of polymorphic sites (S) and nucleotide diversity (π) were computed using DnaSP 5.10 [25]. Different haplotypes were compared by constructing a maximum likelihood phylogenetic tree, using MEGA 6.06 [26].

Malaria Slide Positivity Rate
From 154 suspected malaria patients recruited on day 0, 114 were positive for P. falciparum (74.03%) ( Table 1). Analysis of samples from 50 patients who returned for the 14-day follow up (out of the 114 patients that were given Cartef ® ) found that 11 of these patients were still infected with P. falciparum (day−14 slide positivity = 22.0%). Infection in these samples was confirmed with PCR and they were then used along with the initial 114 patients from day 0, making a total of 125 samples for downstream molecular analyses. No. = Number. Day 0 represents the day of malaria diagnosis prior to drug treatment, while day 14 is the 2-week follow up day after initiating drug treatment, to evaluate parasitemia after drug administration.

Cytochrome Oxidase III PCR Confirmation of Plasmodium Infection
Out of the 125 samples used for PCR, only 91 samples (72.8%) were confirmed as positive for P. falciparum/P. vivax (Figure 1). No. = Number. Day 0 represents the day of malaria diagnosis prior to drug treatment, while day 14 is the 2-week follow up day after initiating drug treatment, to evaluate parasitemia after drug administration.

Cytochrome Oxidase III PCR Confirmation of Plasmodium Infection
Out of the 125 samples used for PCR, only 91 samples (72.8%) were confirmed as positive for P. falciparum/P. vivax (Figure 1).

Amplification and Cloning of the Kelch13 Fragment
In first round PCR, DNA from 91 samples was used to successfully amplify fragment of kelch13 gene, with PCR products of 2097 bp (Figure 2a). These include 80 of the 82 day 0 samples that were positive for Plasmodium from the PCR, plus the DNA from 11 follow-up samples.

Amplification and Cloning of the Kelch13 Fragment
In first round PCR, DNA from 91 samples was used to successfully amplify fragment of kelch13 gene, with PCR products of 2097 bp (Figure 2a). These include 80 of the 82 day 0 samples that were positive for Plasmodium from the PCR, plus the DNA from 11 follow-up samples. No. = Number. Day 0 represents the day of malaria diagnosis prior to drug treatment, while day 14 is the 2-week follow up day after initiating drug treatment, to evaluate parasitemia after drug administration.

Cytochrome Oxidase III PCR Confirmation of Plasmodium Infection
Out of the 125 samples used for PCR, only 91 samples (72.8%) were confirmed as positive for P. falciparum/P. vivax (Figure 1).

Amplification and Cloning of the Kelch13 Fragment
In first round PCR, DNA from 91 samples was used to successfully amplify fragment of kelch13 gene, with PCR products of 2097 bp (Figure 2a). These include 80 of the 82 day 0 samples that were positive for Plasmodium from the PCR, plus the DNA from 11 follow-up samples.  From nested PCR, 2849 bp fragments corresponding to nucleotides 1281-2129 (which covers almost the entire sequence of the Broad complex tram track bric a brac (BTB) and six blades of propeller domains, codons 427-709) were amplified successfully in 77 samples (Figure 2b). Because of the non-specific bands that were observed in some samples, e.g., in lanes 12 and 13 in Figure 2b, the 77 nested PCR products were purified and successfully cloned into E. coli DH5α to confirm sizes using PCR. Positive colonies were mini-prepped and 50 samples were sequenced.

Pattern of Genetic Variability of the Kelch13 Fragment
Out of the 77 purified nested PCR products, 50 were successfully sequenced. These comprised 49 field samples and a fragment amplified from the pf_3D7. The 49 field samples comprised seven samples from the follow-up individuals and 42 samples from day 0. Analyses of all sequences revealed eight mutations compared to the pf_3D7 sequences (Table 2). These sequences have been deposited into the Genbank with accession numbers: MT263314-MT263363. Two isolates; MM_12B and MM_83B from the 14-day follow up harbor Glu 433 Gly and Glu 688 Lys, mutations, respectively. These mutations are not seen in the other day 0 isolates or the other five samples from follow up. In addition, the other four non-synonymous mutations include Phe 434 Ile obtained from day 0 isolates MM_31 and MM_21, as well as Phe 434 Ser, Ile 684 Asn, Ile 684 Thr, which were present only in the day 0 isolates MM_114, MM_10 and NSR_59, respectively.  (Figure 3).
From the 49 samples sequenced, the kelch13 was discovered not to be highly polymorphic. It has nine haplotypes (Table 3) with haplotype diversity, Hd of only 0.336. The sequences possess eight polymorphic sites (S); two of which were synonymous and six led to amino acids substitution. A neutrality test of all the sequences revealed Li and Fu's D* as negative and statistically significant, which indicates an excess of singleton mutations. The Tajima D statistic, on the other hand shows a low frequency of polymorphism. Overall, the presence of a dominant haplotype and very low diversity, with most mutations being non-synonymous suggests that this gene is either probably undergoing selection, or these mutations are just rare and endogenous.  From the 49 samples sequenced, the kelch13 was discovered not to be highly polymorphic. It has nine haplotypes ( Table 3) with haplotype diversity, Hd of only 0.336. The sequences possess eight polymorphic sites (S); two of which were synonymous and six led to amino acids substitution. A neutrality test of all the sequences revealed Li and Fu's D* as negative and statistically significant, which indicates an excess of singleton mutations. The Tajima D statistic, on the other hand shows a low frequency of polymorphism. Overall, the presence of a dominant haplotype and very low diversity, with most mutations being non-synonymous suggests that this gene is either probably undergoing selection, or these mutations are just rare and endogenous. A predominant haplotype, which had the largest frequency was observed (Figure 4a). This haplotype 1 is comprised of 40 sequences out of 49. It was followed by haplotype 8 with two sequences, while the rest of the haplotypes have only one sequence each. Figure 4b presents the position of the nucleotide substitution, respectively, with respect to the predominant haplotype 1.  A predominant haplotype, which had the largest frequency was observed (Figure 4a). This haplotype 1 is comprised of 40 sequences out of 49. It was followed by haplotype 8 with two sequences, while the rest of the haplotypes have only one sequence each. Figure 4b presents the position of the nucleotide substitution, respectively, with respect to the predominant haplotype 1.
To establish genetic distances, a phylogenetic tree was constructed with the 49 sequences successfully sequenced and a sequence of pf_3D7 (Figure 4c). Sequences cluster according to the presence of mutation, with the sequences harboring mutations clustering away from sequences of the Hap_1 and that of the pf_3D7. To establish genetic distances, a phylogenetic tree was constructed with the 49 sequences successfully sequenced and a sequence of pf_3D7 (Figure 4c). Sequences cluster according to the presence of mutation, with the sequences harboring mutations clustering away from sequences of the Hap_1 and that of the pf_3D7.

Discussion
This study investigated the presence of polymorphism in the kelch13 gene using P. falciparum isolates from northern Nigeria. Eight polymorphic sites were discovered in the propeller region of this gene, with six of them leading to amino acids substitution. However, none of the four most implicated mutations associated with ACT resistance (Y493H, R539T, I543T, or C580Y) [8] were seen in the field isolates from Kano. Indeed, the absence of these mutations has been reported previously in southeastern Nigeria and other African countries, including Niger, Cameroon and Benin, that share borders with Nigeria [20,27]. However, several other mutations exist across the propeller domain of the kelch13 gene in sub-Saharan African P. falciparum isolates, with 22 major nonsynonymous mutations already described. These include A557S, V566I, A569T, S576L, A578S, L589I, with the A578S being the most common [16,28,29]. None of these mutations were found in the field isolates investigated in this study. In Senegal, N554H, Q613H and V637I were reported [30] and recently in the Niger Republic, five mutations (M472I, Y558C, K563R, P570L and P615S) were also described in the propeller domain of kelch13 gene [27]. However, none of these mutations was detected in any of our isolates, despite the geographic closeness between the Niger Republic and Kano, in northern Nigeria. Several other mutations observed in other African countries include the most frequent mutation in Mali, F446I [31] and a few synonymous mutations reported in Burkina Faso [32]. None of these mutations were found in samples from Kano. The mutations observed in the Kano isolates may be unique to the Sudan/Sahel of northwest Nigeria. The high heterogeneity in the kelch13 gene mutations across sub-Saharan Africa suggests lack of selection in this gene. The results of microscopy strengthen previous observations of P. falciparum being the primary parasite responsible for malaria in northern Nigeria. Previous studies in Nigeria show similar results, for example, in southwestern Nigeria, where malaria parasite positivity was found among 61.1% of the study population [33]. Similarly, a 73.39% prevalence was also documented in another study in southeast of Nigeria [34].

Discussion
This study investigated the presence of polymorphism in the kelch13 gene using P. falciparum isolates from northern Nigeria. Eight polymorphic sites were discovered in the propeller region of this gene, with six of them leading to amino acids substitution. However, none of the four most implicated mutations associated with ACT resistance (Y493H, R539T, I543T, or C580Y) [8] were seen in the field isolates from Kano. Indeed, the absence of these mutations has been reported previously in southeastern Nigeria and other African countries, including Niger, Cameroon and Benin, that share borders with Nigeria [20,27]. However, several other mutations exist across the propeller domain of the kelch13 gene in sub-Saharan African P. falciparum isolates, with 22 major non-synonymous mutations already described. These include A557S, V566I, A569T, S576L, A578S, L589I, with the A578S being the most common [16,28,29]. None of these mutations were found in the field isolates investigated in this study. In Senegal, N554H, Q613H and V637I were reported [30] and recently in the Niger Republic, five mutations (M472I, Y558C, K563R, P570L and P615S) were also described in the propeller domain of kelch13 gene [27]. However, none of these mutations was detected in any of our isolates, despite the geographic closeness between the Niger Republic and Kano, in northern Nigeria. Several other mutations observed in other African countries include the most frequent mutation in Mali, F446I [31] and a few synonymous mutations reported in Burkina Faso [32]. None of these mutations were found in samples from Kano. The mutations observed in the Kano isolates may be unique to the Sudan/Sahel of northwest Nigeria. The high heterogeneity in the kelch13 gene mutations across sub-Saharan Africa suggests lack of selection in this gene. The results of microscopy strengthen previous observations of P. falciparum being the primary parasite responsible for malaria in northern Nigeria. Previous studies in Nigeria show similar results, for example, in southwestern Nigeria, where malaria parasite positivity was found among 61.1% of the study population [33]. Similarly, a 73.39% prevalence was also documented in another study in southeast of Nigeria [34].

Conclusions
This study confirmed a high prevalence of falciparum malaria in Kano, northwestern Nigeria. As established in various previous studies, it was found that microscopy could lead to false positives and its result should be treated with caution. The kelch13 gene in P. falciparum isolate from Kano carries some novel mutations that should be further studied to establish their epidemiologic impact on antimalarial resistance. Funding: This research and associated costs of publishing were privately sponsored by SSI and UFA.