Assessment of Molecular Markers of Anti-malarial Drug Resistance in Pfk13-propeller, Pfcrt and Pfmdr1 Genes in Plasmodium Falciparum Isolated From Patients Returning for Malaria Retreatment in Democratic Republic of Congo

Doudou Malekita Yobi (  doudou.yobi@unikin.ac.cd ) Université de Kinshasa, Faculté de Médecine, Département de Sciences de Base, Service de Biologie Moléculaire https://orcid.org/0000-0002-8703-1269 Nadine K. Kayiba School of Public Health, Faculty of Medecine, University of Kinshasa, Democratic Republic of Congo; School of Public Health and Research Institute of Health and Society, Catholic University of Louvain, 1200 Brussels, Belgium Dieudonné M. Mvumbi University of Kinshasa: Universite de Kinshasa Raphael Boreux Laboratory of Clinical Microbiology, University of Liege, 4000 Liege, Belgium Pius Z. Kabututu Department of Basic Sciences, Faculty of Medicine, University of Kinshasa, Kinshasa, Democratic Republic of Congo Hippolyte NT. Situakibanza Department of internal Medicine, Faculty of Medicine, University of Kinshasa, Kinshasa, Democratic Republic of Cong Solange E. Umesumbu National Malaria Control Programme, Kinshasa, Democratic Republic of Congo Patrick De Mol Laboratory of Clinical Microbiology, University of Liege, 4000 Liege, Belgium Niko Speybroeck School of Public Health and Research Institute of Health and Society, Catholic University of Louvain, 1200 Brussels, Belgium Georges L. Mvumbi Department of Basic Sciences, Faculty of Medicine, University of Kinshasa, Kinshasa, Democratic Republic of Congo Marie-Pierre Hayette Laboratory of Clinical Microbiology, University of Liege, 4000 Liege, Belgium


Introduction
Plasmodium falciparum (P. falciparum) is the most widespread Plasmodium species and is responsible for most severe forms and deaths related to malaria in sub-Saharan Africa. P. falciparum has succeeded in developing resistance mechanisms against almost all existing antimalarial drugs, which is a major threat to malaria control worldwide. The high level of P. falciparum resistance to chloroquine (CQ) and then to sulfadoxine-pyrimethamine (SP) has led the Democratic republic of Congo (DRC), like all endemic countries, to change its anti-malarial drug policy for the treatment of uncomplicated malaria. The DRC national policy currently supports two rst-line artemisinin-based combinations therapies (ACT) for the treatment of uncomplicated P. falciparum malaria: artemether-lumefantrine (AL) and artesunateamodiaquine (ASAQ). In case of con rmed treatment failure by microscopy to both rst-line ACTs, the patient should be given dual therapy of quinine plus clindamycin or plus doxycycline [1]. The World Health Organization (WHO) recommends regular surveillance of ACT e cacy to provide an early warning against the emergence and spread of resistance [2]. Thanks to the discovery of several P. falciparum genes involved in anti-malarial drug resistance, molecular markers have become a precious tool in resistance surveillance. Numerous polymorphisms in the P. falciparum genome have been suggested to provide resistance to ACTs, both to the artemisinin and the associated drugs [3,4]. Mutations in the propeller domain of P. falciparum Kelch 13 gene (pfk13) have been identi ed as associated with in vivo delayed parasite clearance and in vitro artemisinin (ART) resistance in ring stage survival assay (RSA) [5,6]. These mutations spread into the Greater Mekong Sub-region (GMS) of Southeast Asia and have been classi ed as validated markers and candidate/associated markers of ART resistance [7]. Some of these mutations are increasingly detected in some sub-Saharan African countries providing evidence for de novo emergence of resistance to artemisinin in Africa [8][9][10]. Mutations in the chloroquine resistance transporter gene (pfcrt) originally identi ed as a marker of CQ resistance, have also been associated with resistance to amodiaquine (AQ) [11]. The pfcrt gene is highly polymorphic in codon position 72-76 determining different haplotypes that include the key mutation K76T associated with CQ resistance while the SVMNT haplotype has been found associated with AQ resistance [12]. Additionally, increased copy number of the gene encoding the multidrug resistance 1 transporter (pfmdr1) has been postulated to confer resistance to lumefantrine (LU) [13].
Since it is not possible to distinguish routinely between recrudescence and reinfection in endemic areas, the absence of resolution of fever and parasitaemia or their recurrence within 28 days of treatment is considered treatment failure with currently recommended ACTs, while all recurrence of fever and parasitaemia after 28 days of an initial treatment should, from an operational standpoint, be considered as re-infections and treated with rst-line ACT [14]. The distinction between recrudescence and re-infection may be con rmed only by genotyping. This concern is addressed by therapeutic e cacy studies according to WHO protocol [15]. Many factors can contribute to treatment failure, including resistance and inadequate exposure to drug due to sub-optimal dosing, poor patient compliance, poor drug quality, vomiting, poor drug absorption and drug interactions. The present study investigated the resistance to each associated drug in the two rst-line ACTs currently in use in the DRC through molecular markers in pfk13, pfcrt and pfmdr1 genes in case of malaria retreatment.

Study area
Six sentinel sites of National Malaria Control Programme (NMCP) were selected among 26 provinces of the DRC. The sites included 2 of the 3 largest cities of country and 4 other sites selected based on their epidemiological facies and their accessibility. Thus, the following sentinel sites were selected: Kingasani in Kinshasa and Kabondo in Kisangani for the largest cities; Bolenge in Equateur province and Vanga in Kwilu province for the equatorial facies; Fungurume in Lualaba province for the tropical facies and Katana in Sud-Kivu province for the mountainous facies. In each study site, 1 to 3 health structures were selected based on the accessibility and the attendance for the realization of the study.

Study participants
From November 2018 to November 2019, patients of all ages who returned to health structures for fever within 28 days from an ACT-based treatment for an initial episode of malaria that had to be documented in the patient's medical record and who had, during investigation, a positive rapid diagnostic test (RDT) for malaria were enrolled after that informed consent was given. The ACT which was used to treat the initial episode depended on the ACT available on each study site during the investigation.

Blood sample collection
Screening tests were performed on blood samples taken by nger prick. Malaria was detected using RDTs available on site: SD Bioline Malaria Ag P.f (Standard Diagnostics) or CareStart Malaria Pf (Access Bio). After enrollment, a blood sample was taken by nger prick and three spots were deposited on Whatman Grade GB003 lter paper (Whatman, GE Healthcare). Dried blood spots (DBS) were placed in an individual ziploc plastic bag containing silica gel desiccant and were then stored at -20° before molecular analysis.

DNA Extraction
DNA was extracted from blood spots using the QIAamp DNA Mini Kit (Qiagen, Germany) following the manufacturer's recommended protocol for DBS. The extracted DNA was stored at − 20 °C before PCR testing.

Plasmodium falciparum real-time PCR
A real-time PCR for the detection of P. falciparum was performed according to a previously described procedure [16]. Assays were run on an ABI 7500 Fast real-time thermocycler (Applied Biosystems) at the Laboratory of Molecular biology of school of Medicine, University of Kinshasa, DRC.

pfk13 PCR
The target sequence was a 506-nucleotide fragment of the pfk13-propeller gene containing codons 427-595, as previously described [17]. The interest segment included recently described mutations associated with ART resistance [7]. PCR was run using a conventional thermal cycler Hybaid HBPXE 0.2 (Thermo Scienti c) at Laboratory of Molecular biology of school of Medicine, University of Kinshasa, DRC.

pfcrt PCR
The fragment of interest (containing codons 72-76) on the pfcrt gene was ampli ed following a previously described procedure [18]. The PCR was run on the conventional Thermal Cycler cited above.
Estimation of pfmdr1 copy number by real-time PCR Pfmdr1 copy number was assessed by a relative quanti cation real-time PCR using ABI 7500 thermocycler (Applied Biosystems, Warrington, UK) as previously described [19]. In each run of real-time PCR, 3D7 and Dd2 clones were used as calibration controls with single and multiple copies of pfmdr1 gene, respectively 1 copy and 4 copies, and negative control containing no template DNA was also included. Test samples were assayed in triplicate, the copy number of pfmdr1 was determined using the comparative ΔΔCt method and calculated as 2 −∆∆Ct . Samples with a spread of ∆∆Ct among the three triplicates of more than 1.5 or with a Ct > 35 were repeated and the second result used. Analyses were performed at Laboratory of Clinical Microbiology of University of Liege, Belgium.
Pfcrt and pfk13 genotyping After puri cation using AMPure XP magnetic beads (Beckman Coulter, CA, US), the PCR products were added to a mix of Big Dye Terminator V3.1 for the sequencing reaction. The resulting nucleotide sequences were analyzed on an ABI 3730 DNA Analyzer automated sequencer (Applied Biosystems) using the Sanger method at GIGA (University of Liège's Interdisciplinary Research Institute in the Biomedical Sciences). Sequences (forward and reverse) were aligned using Vector NTI (Thermo Fisher Scienti c, US) and compared to the reference sequence PF3D7_0709000 (https://www.ncbi.nlm.nih.gov/gene/term=PF3D7_0709000 accessed on March 2020) for pfcrt and PF3D7_1343700 (http://www.plasmodb.org, accessed on March 2020) for pfk13 using the online Basic Local Alignment Search Tool (BLAST) (National Center for Biotechnology Information-NCBI) for identifying mutations.

Statistical analysis
Data were entered in a 2010 Excel sheet by an independent data clerk. Statistical analysis was performed using SPSS V. 20.0 (IBM corp, Armonk, NY). Samples for which genotype pro le could not be determined were excluded from the analysis. Categorical variables were expressed as frequencies with 95% con dence intervals (95% CI) while quantitative variables as median with interquartile range (IQR). The mutant and wild-type alleles identi ed in the sequenced isolates were used to generate the prevalence of the alleles.

Results
In total, 474 patients were enrolled in the study, their age ranged from 0 to 76 years with a median age of 10 years (IQR: 4-21 years). The male to female sex ratio was 0.74.

Pfmdr1 copy number
In total, 322 P. falciparum isolates were successfully analyzed for copy number variation in the pfmdr1 gene. Using a copy number threshold of 1.5 to de ne multiple copies, all isolates harbored single copy of pfmdr1 gene.

Discussion
The present study has not detected molecular markers associated with resistance to rst-lines ACT in use in the DRC in patients returning to health facilities for malaria retreatment.
Although none of the mutations associated with ART resistance in Southeast Asia has been found by the study, 7 coding substitutions that are of unknown phenotype were observed, among which 3 previously reported notably N498I in Kenya [20], N554K in Comoros and A557S in Togo and DRC [21] and 4 others not yet reported (F506L, E507V, D516E, G538S).
Numerous pfk13-propeller mutations of unknown function are commonly reported in the Sub-Saharan African countries like in the DRC [17,21,22]. There are criteria for prioritizing further laboratory studies notably the frequent observation of a new allele with a non-synonymous mutation, the evidence of dissemination and preliminary association with clinical data whenever possible [22]. Several independent single nucleotide polymorphisms (SNPs) could be responsible of the ART-R in Sub-Saharan Africa, the known mutations that confer drug resistance would differ from one location to another, depending on the parasite genetics. There is the possibility that pfk13 mutations do not cause ART-R in isolation but would act in combination with other genetic or non-genetic factors that are different in African and Southeast Asian parasite populations [23,24]. Since African pfk13-propeller mutations were shown to be different from those found in Southeast Asia, further molecular and biochemical studies should investigate whether other factors such as additional mutations could be associated to alter the functions of PFK13 protein, resulting in altered ART sensibility. However, some of the validated and candidate mutations associated with ART resistance in Southeast Asia have been detected in some neighbouring countries such as in Rwanda [8,9] and in Uganda [10], providing evidence of de novo emergence of ART resistance in Sub-Saharan Africa. Thus, surveillance must be strengthened to avoid the worst.
The global prevalence of the K76T mutation known to be associated with CQ resistance was 41.5%, but it remained variable from one site to another, ranging from 10.0% in Fungurume to 76.9% in Katana. In 2017, a study conducted in 10 sites including 6 sites of the present study reported a global prevalence of K76T mutation of 28.5% but with a high between regions variability ranging from 1.5% in Fungurume to 89.5% in Katana [16]. In the present study, the prevalence of K76T in patients treated with ASAQ (48.9%) was higher than in those treated with AL (37.4%) with low statistical difference (p-value = 0.038). The possibility that AQ would continue to contribute to the selection for the K76T mutant even after discontinuance of CQ usage has been previously raised [25]. The simultaneous presence of very low and high prevalence of CQ resistance could be related to between-regions difference of CQ pressure and also to effect of selection for CQ resistance depending on the genetic structure of parasite populations which have been shown to vary signi cantly across the country [26]. Data concerning current CQ use in the country are not available, further studies at the community level and parasite genetic studies should be conducted to explain the persistence of high CQ resistance rate in some provinces despite the withdrawal of this molecule from the national policy of malaria treatment. The SVMNT haplotype associated with AQ resistance was not detected in the present study, which is encouraging for the DRC national policy for the continued use of AQ in ACT. This haplotype has not yet been reported in the DRC [16,18,27,28] whereas it was found in neighbouring countries such as Tanzania and Angola [29,30].
None of the P. falciparum isolates had multiple copies number of pfmdr1 gene in this study, consistent with the general absence [31,32] or low frequencies [28,33,34] of pfmdr1 copy number variation in P. falciparum from sub-Saharan Africa. Multiple copy number of pfmdr1 gene has been postulated to confer resistance to LU [13], which is the associated drug in AL, one of the ACTs used in the DRC.
The ACT used to treat the initial episode was that available on each study site during investigation and best tolerated by the patient. In practice, AL tends to be primarily used in urban areas because patients have more options to obtain it from private pharmacies, while ASAQ is used in rural areas [1]. The study assessed molecular markers of anti-malarial resistance in patients who returned to health structures for fever within 28 days of the ACT treatment, which is considered as treatment failure [14]. Treatment failure is the inability to clear parasites from a patient's blood or to prevent their recrudescence after the administration of an anti-malarial drug. When possible, treatment failure must be con rmed by microscopy, as histidin rich protein-2 (HRP-2)-based RDT may remain positive for weeks after successful treatment due to persistent antigenemia, even without recrudescence [35,36]. In this study, P. falciparum real-time PCR assay was used afterwards to con rm P. falciparum malaria in patients returning to health structures for fever within 28 days of an initial malaria treatment.
The study contributes to the ongoing surveillance of the resistance to anti-malarial drugs in use in the DRC as recommended to track the emergence and spread of P. falciparum resistance to different molecules used in malaria management. However, although the study was carried out in six provinces of the DRC with varied geography and malaria endemicity, the results were not representative of either any one province or the entire country.

Conclusion
No molecular marker known to date as associated with resistance to rst-line ACTs components in use in the DRC was detected in P. falciparum isolated from patients returning for malaria retreatment. The ndings are encouraging for the current DRC malaria treatment policy, however, the appearance of coding substitutions that are of unknown functions calls for further investigations and other factors, apart from the resistance, that contribute to treatment failure should be assessed in order to ensure the effectiveness of the treatment of malaria in the DRC.

Declarations
Ethics approval and consent to participate The protocol and the informed consent form were approved by the Ethics Committee of the Faculty of Medicine, University of Kinshasa (Approval N°: ESP/CE/111/2018). All participants involved in the study signed an informed consent form.
Where participants were young (children), the consent form was approved and signed by their parents or guardians.