Ex vivo RSA and Pfkelch13 targeted-amplicon deep sequencing reveal parasites susceptibility to artemisinin in Senegal, 2017

Introduction. Malaria control is highly dependent on the effectiveness of artemisinin-based combination therapies (ACTs), the current frontline malaria curative treatments. Unfortunately, the emergence and spread of parasites resistant to artemisinin (ART) derivatives in Southeast Asia and South America, and more recently in Rwanda and Uganda (East Africa), compromise their long-term use in Sub-Saharan Africa where most malaria deaths occur. Methods. Here, we evaluated ex vivo susceptibility to dihydroartemisinin (DHA) from 38 P. falciparum isolates collected in 2017 in Thiès (Senegal) expressed with the Ring-stage Survival Assay (RSA). We explored major and minor variants in the full Pfkelch13 gene, the main determinant of ART resistance using a targeted-amplicon deep sequencing (TADS) approach. Results. All samples tested in the ex vivo RSA were found to be susceptible to DHA. Both non-synonymous mutations K189T and K248R were observed each in one isolate, as major (99%) or minor (5%) variants, respectively. Conclusion. Altogether, investigations combining ex vivo RSA and TADS are a useful approach for monitoring ART resistance in Africa.


Results.
All samples tested in the ex vivo RSA were found to be susceptible to DHA. Both non-synonymous mutations K189T and K248R were observed each in one isolate, as major (99%) or minor (5%) variants, respectively.

Conclusion.
Altogether, investigations combining ex vivo RSA and TADS are a useful approach for monitoring ART resistance in Africa.

Background
Prompt management of malaria cases remains a vital component of malaria control and elimination strategies [1]. Over the two last decades, artemisinin (ART)-based combination therapies (ACTs) have contributed signi cantly to the reduction of malaria-related morbidity and mortality in Sub-Saharan Africa [2]. Since 2006, the Senegalese National Malaria Control Program (NMCP) have recommended artemether-lumefantrine (AL) as the rst-line treatment for uncomplicated malaria [3]. However, recent reports of the emergence and expansion of partial resistance to ART (ART-R) in the Greater Mekong Subregion have endangered the long-term e cacy of ACTs. Although ACTs remain clinically and parasitologically effective in most African malaria endemic countries [4], the local emergences of ART-R in Rwanda and in Uganda, respectively reported in 2020 and 2021, are warning signals in the loss of ART e ciency [5][6][7][8]. As ART-R could independently emerge in western Africa or spread from eastern African ATR-R hotspots, the monitoring of P. falciparum parasites susceptibility to ART-based on the current clinical and biological tools is of utmost importance and must be implemented [4]. Clinically, ART-R is de ned as delayed parasite clearance or Day-3 positive parasitemia upon ART-based treatment [9].
Delayed clearance has been associated with decreased in vitro parasite susceptibility (survival rate ≥ 1%) expressed in the Ring-stage Survival Assay (RSA 0-3h ) [10]. These clinical and biological phenotypes were later associated with non-synonymous mutations in the P. falciparum kelch13 gene (Pfkelch13) [11]. Pfkelch13 encodes a 726-amino acid protein containing three structurally conserved domains: a coiledcoil-like (CCC) domain, a Broad-Complex, Tramtrack and Bric a brac (BTB) domain, and a C-terminal kelch-repeat propeller domain where most of non-synonymous mutations associated with ART-R have been described [12]. Since then, the surveillance based on the detection of Pfkelch13 single-nucleotide polymorphisms (SNPs) have been conducted in many countries. In the Greater Mekong Subregion, multiple ART-R Pfkelch13 mutant parasites evolved concomitantly until a multidrug-resistant Pfkelch13 C580Y lineage (named KEL1/PLA1) outcompeted the others and spread across Southeast Asia [13]. In Africa, this variant has been sporadically reported [14][15][16]. Rather, the Pfkelch13 R561H mutant rapidly increased in frequency in Rwanda between 2014-2016 and 2018-2019 (from 8-22%, respectively) [17][18][19] and was associated with in vivo and in vitro ART-R [19]. Similarly, in Uganda, two Pfkelch13 mutants (A675V and C469Y) were recently reported to be associated with in vivo and in vitro ART- R [6, 8, 20]. In Senegal, most investigations have looked for SNPs in the propeller-encoding domain of Pfkelch13 gene using conventional methods such as PCR/Sanger sequencing [21][22][23][24][25][26], except for two studies that used Pfkelch13 targeted-amplicon deep sequencing (TADS) [25,26]. Although the authors did not detect validated or candidate ART-R Pfkelch13 mutations [27], TADS approach is particularly relevant to detect the presence of minor variants in P. falciparum isolates. Here, Pfkelch13 genotype was evaluated by TADS in 38 P. falciparum isolates collected in Thiès in 2017. We also explored the genetic variation of the whole conserved-domains of Pfkelch13 gene owing that mutation in the BTB-encoding domain of Pfkelch13 gene in a western African strain has been shown to be associated with ART-R [28].

Materials And Methods
Ethics. The National Ethics Committee for Health Research and the Ministry of Health of Senegal approved the protocol used for this study under number 00000169/MSAS/DPRS/CNERS (December 2nd, 2016). Written and informed consent was obtained from all participants, before participant recruitment and sample collection.
Study site. The study was conducted during the peak malaria season (September to December) in 2017 in Thiès (Senegal; 14° 47′ 26″ north, 16° 55′ 29″ west), an area belonging to the Sahelian facies de ned by a short malaria seasonal transmission (< 4 months). In this region, the entomological inoculation rate (EIR) is low, estimated to be < 5 [29], and malaria is mainly transmitted by Anopheles arabiensis and Anopheles gambiae mosquito vectors.
Study design. Individuals who visited the Service de Lutte Antipaludique (SLAP) clinic with signs and symptoms suggestive of uncomplicated P. falciparum malaria were screened by microscopic examination on Giemsa-stained blood smears. Malaria infected patients were treated with artemetherlumefantrine (AL, Coartem), according to the treatment guideline of the Senegal National Malaria Control Program (NMCP). For each patient, 5 mL vacutainer tubes of venous blood were collected for ex vivo RSA.
Parasite culture and ring-stage survival assay. Parasitemia was estimated by microscopy examination on Giemsa-stained blood smears. Venous blood samples were then processed by eliminating plasma, leukocytes and anticoagulant from red blood cells (RBCs), washed twice in RPMI 1640 medium (Gibco, Life technologies). Parasitemia were adjusted to 1% if greater by adding uninfected RBCs as previously described [10]. Then, 900 µL RBCs were loaded into wells, exposed to either 100 µL of 700 nM DHA or 0.1% of dimethyl sulfoxide (DMSO) and cultivated at 37°C in incubator for six hours (conditions: 94% N 2 , 5% CO 2 , 1% O 2 ) [30]. Finally, RBCs were washed and cultivated for 66 hours. The proportions of viable parasites were estimated independently by two expert malaria microscopists on Giemsa-stained thin smears. The number of viable parasites that developed into ring/trophozoite stages were determined, pyknotic forms were excluded. The average of the two counts was calculated. If any discrepancy was noted (either difference of parasite density of > 50%), slides were checked by a third independent reader, and parasite densities were calculated by averaging the two most close counts. Survival rates were calculated as the ratio of parasites in exposed and non-exposed cultures. Results were deemed as interpretable if the parasitemia in the sample exposed to DMSO was higher than the initial parasitemia at 0h [31]. following the manufacturer's protocol. Amplicons were puri ed and the library size selection was performed with AMPure Agencourt XPbeads. Libraries quality and quantity control were assessed using Qubit® for concentration and Bio Analyser 2100 Agilent for fragment size. Libraries were pooled at approximately equal concentration and sequenced on an Illumina NextSeq 500 instrument (Illumina Inc, San Diego, CA, USA) to generate 150 bp paired-end reads at the GENOM'IC platform of Cochin Institute (Paris, France). Raw sequences were then demultiplexed and quality trimmed at a Phred score of 30.
Primer sequences were trimmed from the 5′-end of the sequences to avoid primer bias in the sequenced fragments. Base calling was performed by comparing reads with a custom database consisting of the Pfkelch13 sequence retrieved from the 3D7 reference genome. Bioinformatic analyses were performed using the CLC Genomics Workbench 22 software (Qiagen).

Results
Baseline characteristics. A total of 38 patients with uncomplicated P. falciparum malaria meeting the inclusion criteria were enrolled. The sex ratio (M/F) was largely dominated by males (36 males/2 females). The age of the participants ranged from 9 to 70 years (median of 21.5 years). Median weight was 59.5 kg and median body temperature was 38.5°C. The median parasitemia was 1.03%, ranging from 0.68-1.53% (Table 1). Survival rates showed the absence of surviving parasite to the 700nM DHA pulse as for the 3D7 strain (0%) except for three isolates Th54, Th77 and Th95 (0.054%, 0.06% and 0.22% respectively) which were however less than the 1% threshold.
Pfkelch13 genotyping. Among the 38 samples successfully sequenced, two non-synonymous mutations (K189T and K248R) located outside of the propeller-encoding domain were detected (Table 2). Each mutation was observed in a single sample in major (99% for K189T) and minor (5% for K248R) proportions. The Pfkelch13 K248R mutant was detected for the rst time in Senegal (Table 3).

Discussion
ACTs are now the mainstay of treatment for uncomplicated malaria in malaria endemic regions [2]. Unfortunately, the emergence and the clonal expansion of Pfkelch13 mutant parasites have been reported recently in Rwanda (R561H) and Uganda (C469Y and A675V) [5][6][7][8]. Therefore, the World Health Organization (WHO) recommends to closely monitor the susceptibility of P. falciparum to antimalarial drugs and particularly to ART derivatives [4]. As in many Sub-Saharan African countries, ACTs have contributed signi cantly to the decline in malaria incidence and mortality over the past decade. In Senegal, considerable efforts have been done to reduce malaria morbidity and mortality mainly since 2006 when AL was introduced [3]. Consequently, the emergence of ART-resistant parasites is a major threat which can hinder road to malaria elimination.
Partial resistance to ART is associated with non-synonymous mutations in the Pfkelch13 gene. However, the impact of most mutations on ART-R detected in eld samples is largely unknown mainly due to the lack of association between the clinical phenotype (delayed parasite clearance) and the Pfkelch13 genotype. The study presented here aims to ll this gap by combining ex vivo Ring-stage Survival Assay (RSA) with Pfkelch13 genotyping. The ex vivo RSA estimates the susceptibility of P. falciparum to ART using parasite isolates freshly collected from patients with malaria. The ex vivo RSA has been previously used in studies conducted in Central [32] and East Africa [31,33]. Four isolates from Uganda showed high (> 10%) survival rates, levels of which are reported to be closely associated with delayed parasite clearance following artesunate monotherapy [31]. The data presented here showed that none of the tested samples confer in vitro ART-R with a survival rate ≥ 1%, suggesting the absence of decrease susceptibility of Senegalese parasites to ART derivatives.
Partial resistance to ART has also been associated with speci c mutations in the Pfkelch13 gene [1,2]. Since 2015, this molecular marker has been extensively used to seek Pfkelch13 mutants in Sub-Saharan African countries. By using a targeted amplicon deep sequencing approach, we detected here two mutations (K189T and K248R). The K248R mutation was observed in one sample in minor proportion (5%). The mutation is located in the CCC domain of Pfkelch13, but we think the mutation not likely related to ART-R since the survival rate was less than 1%. While K189T had already been reported in Africa [34,35], the non-synonymous mutation K248R was detected for the rst time in Senegal. The isolate carrying the K189T mutation had also a survival rate < 1%, con rming that this allele does not confer in vitro ART-R as previously observed in India [36]. Only three Pfkelch13 wild-type isolates (Th54,Th77 and Th95) were found to have a survival rate above 0%. To date, in vitro susceptibility to ART derivatives by RSA of P. falciparum isolates have been reported in Cameroon and Uganda. Two studies conducted in Cameroon [32,33] showed the absence of Pfkelch13 mutations associated with ART-R while one study in Uganda [31] reported high survival rates of isolates carrying Pfkelch13 A675V and C469Y mutations.
The study presented here has several limitations. First, only 38 samples were tested owing that ex vivo RSA is time consuming and requires skilled personnel. Second, in vitro RSA (from culture-adapted parasites) was not performed to con rm the ex vivo survival rates. And third, the study was conducted only in one site in Senegal (Thiès) and no clinical data on delayed parasite clearance (like Day-3 positivity rate) was collected.

Conclusion
This study shows that combining ex vivo RSA phenotype and Pfkelch13 genotyping can be e ciently carried out to monitor ART-R and providing relevant data to malaria control programs on parasite susceptibility to ART. Particularly, targeted-amplicon deep sequencing used here con rmed to be useful to detect the presence of minor alleles. This study showed that all P. falciparum isolates collected in Thiès were susceptible to DHA. As recommended by the WHO [4], similar studies must be conducted in Senegal and other African countries to strengthen surveillance of antimalarial drug e cacy and resistance and minimize the threat and impact of antimalarial drug resistance in Africa.