Molecular surveillance of Plasmodium falciparum drug-resistance markers in Vietnam using multiplex amplicon sequencing (2000-2016)

Abstract

Emergence and spread of Plasmodium falciparum resistance to artemisinin-based combination therapies (ACT) is a major challenge for Greater Mekong Subregion countries in their goal to eliminate malaria by 2030. Tools to efficiently monitor drug resistance beyond resource-demanding therapeutic efficacy studies are necessary. A custom multiplex amplicon sequencing assay based on Illumina technology was designed to target the marker of partial resistance to artemisinin (K13), five candidate modulators of artemisinin resistance, the marker of resistance to chloroquine (crt), and four neutral microsatellite loci. The assay was used to genotype 635 P. falciparum-positive blood samples collected across seven provinces of Vietnam and one of Cambodia between 2000 and 2016. Markers of resistance to artemisinin partner-drugs piperaquine (copy number of plasmepsin-2) and mefloquine (copy number of multidrug-resistance 1) were determined by qPCR. Parasite population structure was further assessed using a 101-SNP barcode. Validated mutations of artemisinin partial resistance in K13 were found in 48.1% of samples, first detection was in 2000, and by 2015 prevalence overcame > 50% in Central Highlands and Binh Phuoc province. K13-C580Y variant became predominant country-wide, quickly replacing an outbreak of K13-I543T in Central Highlands. Mutations in candidate artemisinin resistance modulator genes paralleled the trends of K13 mutants, whereas resistance to piperaquine and mefloquine remained low (≈ 10%) by 2015-2016. Genomic tools applied to malaria surveillance generate comprehensive information on dynamics of drug resistance and population structure and reflect drug efficacy profiles from in vivo studies.

Figure 1

Origin of the samples. (A) Map of Vietnam with the location of administrative provinces where samples were collected. Three main geographical regions were defined for analytical purposes based on geographical location and malaria epidemiology criteria (‘Region 1’, stripes; ‘Region 2’, plain orange; ‘Region 3’, dots). Note that Ratanakiri is an administrative province of Cambodia. (B) Detail on sample size per location and year. Circles are proportional to sample size. Time groups used for analysis are indicated by vertical dashed lines. The bottom panel indicates relevant events regarding antimalarial treatment policy in Vietnam and Greater Mekong Subregion. This map was developed for the purpose of this article using QGIS 3.10.

Figure 2

Read depth for amplicons in the pfTSCA assay. The name of each amplicon corresponds to the gene target followed by a letter in genes that require multiple amplicons to cover the target region. Horizontal lines indicate median read count. Values (%) indicate the proportion of samples with reads above the read depth cut-off for each amplicon. MS, microsatellite.

Figure 3

Frequency of K13 mutations as markers of artemisinin partial resistance in Vietnam (2000–2016). Bar charts indicate the percentage of samples with K13 validated mutations (red), non-validated mutations after codon 440 (orange), and wild-type parasites (green). Specific codon changes for validated mutations are indicated with different fill patterns. Data is shown by region and years. Region 1: Quang Tri, Quang Nam, Gia Lai and Ratanakiri (Cambodia) provinces; Region 2: Khanh Hoa, Ninh Thuan, and Binh Thuan provinces: Region 3: Binh Phuoc province.

Figure 4

Frequency of ART-R genetic background alleles in Vietnam (2000–2016). Bar charts show the frequency of haplotypes constructed from the combination of arps10-V127M, fd-D193Y and crt-N326S. Reference haplotype VDN is shown in white and mutant haplotypes at either one, two or all three positions are colored. Data is shown by region and years. Region 1: Quang Tri, Quang Nam, Gia Lai and Ratanakiri (Cambodia) provinces; Region 2: Khanh Hoa, Ninh Thuan, and Binh Thuan provinces: Region 3: Binh Phuoc province.

Figure 5

Frequency of pm2 and mdr1 multiple copy numbers in Vietnam (2000–2016). Bar charts indicate the percentage of samples with multiple copies (red) or single copies (green) of pm2 (marker of piperaquine resistance; A) and mdr1 (marker of mefloquine resistance; B). Samples with multiple copies also carrying K13 mutations are indicated with different fill patterns. Data is shown by region and years. Note that the y axis in both (A,B) is split in two segments (from 0 to 0.15 and from 0.25 to 1) to improve visualization of low frequencies. Region 1: Quang Tri, Quang Nam, Gia Lai and Ratanakiri (Cambodia) provinces; Region 2: Khanh Hoa, Ninh Thuan, and Binh Thuan provinces: Region 3: Binh Phuoc province.

Figure 6

Discriminant analysis of principal component (DAPC) in samples from Vietnam (2000–2016). DAPC was performed using all markers genotyped in the pfTSCA and the SNP barcode. Scatter plot shows discriminant analysis (DA) eigenvalues 1 and 2, in which populations are differentiated by color (geographical region) and shape (time). Alleles contributing most to the DAPC are listed in Supplementary Table S9.