Study Sites and Sample collection
Samples were collected in September 2018 following four deaths that included a case of atypical malaria in the village of Mbounguiel in the department of Kanel, in the region of Matam. The district of Kanel is bordered in the north by the district of Matam which is a malaria elimination site (identified by green) and the high malaria endemic region in the South (identified by red) (Figure1). Malaria transmission in Senegal is highly seasonal with transmission occurring from July to December, and the main causative agent being P. falciparum [22]. Recent studies based on serologic data have shown the presence of P. vivax in the region of Matam [17], and Aedes mosquitoes competent for different dengue serotype transmission have also been isolated in the north [23].
The Ethics Committee of the Ministry of Health in Senegal approved this study. All samples were collected with informed consent per ethical requirements of the National Ethics Committee of Senegal. Venous blood (approximately 5 mL in EDTA) obtained as part of the clinical work up and case investigation from nine suspected malaria cases were used to perform molecular and serological investigation. Clinical and demographic data was also collected from all suspected cases and included: age, sex, household, treatment received, and outcome. Samples collected by the regional medical team and PATH under the direction and coordination of the NMCP were sent to the Laboratory of Parasitology and Mycology at Aristide Le Dantec Hospital, Dakar.
Nucleic Acid Extraction
Blood was separate by centrifugation, RNA was extracted from the pellet using the QiAmp® Viral RNA Mini kit (Qiagen). And the whole blood were directly used for DNA extraction using the QIAamp® DNA Blood Mini kit (Qiagen ®) according to manufacturer’s instructions.
PET-PCR
A multiplex photo-induced electron transfer polymerase chain reaction (PET-PCR) assay was used for P. falciparum species typing, as previously described [24]. A cycle threshold (CT) value of 40 was used as a cut-off, with samples having a CT value less than 40 considered positive and samples with a CT of 40 or higher were considered negative.
Molecular Barcoding
Samples were pre-amplified using previously described assay [25]. Molecular barcode assay was performed on the LightCycler 96 Roche system. All 24 single-nucleotide polymorphisms (SNPs) were amplified as follows; 2.0 μL of Lightscanner Master Mix (BioFire Defense), 2.5 μL of a 1:100 dilution DNA template, and 0.5 μL of primers and probes. Genomic DNA from cultured P. falciparum strains (3D7, Dd2, 7G8, Tm90) was used for assay validation and as genotyping controls for all reaction plates. Molecular barcode assays 10, 11, 13, 21, and 24 were performed optimally under asymmetric forward to reverse primer ratios of 5:1; all other assays required a 1:5 primer asymmetry. Amplification conditions were 95°C denaturation for 2 minutes, 50 cycles of 94°C for 5 seconds and 66°C for 30 seconds, and a pre-melt cycle of 5 seconds each at 95°C and 37°C. Two or more N’s among the 24 SNPs assayed was taken to indicate that >1 P. falciparum genome was present [26].
Serological assays
All antigens were coupled to magnetic beads (Luminex Corp, TX, USA) for a multiplex bead-based assay (MBA) using MAGPIX technology. For each sample, total immunoglobulin G (IgG) response was measured to four P. falciparum antigens, circumsporozoite protein (CSP), merozoite surface protein 1-fragment 19 (MSP1-19), liver stage antigen type 1 (LSA-1), and glutamate-rich protein (GLURP), as well as viruses including Chikungunya virus envelope glycoprotein -1 (CHIKV E1), yellow fever (YFV), dengue virus serotype 2 and 3 ( DENV2, DENV3), Zika virus- like particle (ZIKV-VLP), and West Nile virus (WNV) as described previously [27],[28]. First, a 6mm circular punch (corresponding to 10μL whole blood) was taken from the center of each blood spot, eluted in 200μL protein elution buffer B containing: PBS (pH 7.2), 0.05 % Tween-20, 0.05 % sodium azide and stored overnight at 4 °C until analysis. Next, 50μL of the bead mixture was added to each well for each plate to assay. The bead mixture contained 6μL of each bead in 5mL total Buffer A (0.5% Polyvinyl alcohol (Sigma), 0.8% polyvinylpyrrolidine (Sigma), 0.1% casein (ThermoFisher), 0.5% BSA (Millipore), 0.3% Tween-20, 0.1% sodium azide, and 0.01% Escherichia coli extract to prevent non-specific binding). The plate was put on the magnet to allow the fixation of the beads and washed 2x with 100μL wash buffer (PBS-T). Then, 50μL of Reagent Mix (Mix of 10μL anti-hIgG-BIOT, 8μL anti-hIgG4-BIOT, and 25μL streptavidin-PE to 5mL Buffer A) was added to each well along with 20μL of the eluted sample from the Axygen storage plate and 30μL of Buffer B. In control wells, 50μL of positive control was added instead. The plate was covered with loose aluminum foil to protect from light, and shaken overnight. The next morning the plate was washed 3x with PBS and Tween. Finally, 100μL PBS was added to each well, the plate was shaken for 5min to resuspend the beads and the assay read on a MAGPIX machine. Seropositive were determined as previously described [28] based upon the distribution of the data (and the antigen), mixture model, US non-exposed population, or a serum standard used to determine mean fluorescence intensity (MFI) by subtracting MFI values from blank background beads (MFI-bg). Conventionally, a sample was considered seropositive using a cut-off of rate of antibody expression for each antigen from the US non-exposed population [28].
Metagenomic sequencing
Extracted RNA was DNAse-treated, cDNA was synthesized, and sequencing libraries were prepared using the Nextera XT kit (Illumina) as previously described [29]. Sequencing libraries were directly constructed from clinical samples without culture or other intervention. Samples were sequenced using Illumina MiSeq with 101 nucleotide paired-end reads at LPM Dakar.
Sequencing data analysis
Sequencing data were analyzed using a publicly available software viral-ngs v1.25.0 [30], implemented on the DNAnexus cloud-based platform. Individual samples were demultiplexed and reads mapping to the human genome and to other known technical contaminants were removed. We used KrakenUniq [31], implemented in viral-ngs, to identify taxa present in the samples using a database that encompassed the known diversity of all viruses that infect humans, as previously described [30]. A taxon was considered to be present if a greater number of reads from that species were detected in a sample compared to the negative control, after normalizing read counts for the total depth of sequencing, and these reads showed a high k-mer diversity.A de novo genome assembly from raw sequencing reads were further performed, following depletion of human reads, using SPAdes [32], and classified the resulting contigs using Kaiju. Contigs with a length > 300nt and coverage depth > 10 reads were considered.