Genomes of Leishmania parasites directly sequenced from patients with visceral leishmaniasis in the Indian subcontinent

Abstract

Whole genome sequencing (WGS) is increasingly used for molecular diagnosis and epidemiology of infectious diseases. Current Leishmania genomic studies rely on DNA extracted from cultured parasites, which might introduce sampling and biological biases into the subsequent analyses. Up to now, direct analysis of Leishmania genome in clinical samples is hampered by high levels of human DNA and large variation in parasite load in clinical samples. Here, we present a method, based on target enrichment of Leishmania donovani DNA with Agilent SureSelect technology, that allows the analysis of Leishmania genomes directly in clinical samples. We validated our protocol with a set of artificially mixed samples, followed by the analysis of 63 clinical samples (bone marrow or spleen aspirates) from visceral leishmaniasis patients in Nepal. We were able to identify genotypes using a set of diagnostic SNPs in almost all of these samples (97%) and access comprehensive genome-wide information in most (83%). This allowed us to perform phylogenomic analysis, assess chromosome copy number and identify large copy number variants (CNVs). Pairwise comparisons between the parasite genomes in clinical samples and derived in vitro cultured promastigotes showed a lower aneuploidy in amastigotes as well as genomic differences, suggesting polyclonal infections in patients. Altogether our results underline the need for sequencing parasite genomes directly in the host samples.

Fig 1. Performance of SuSL-sequencing on clinical samples.

A. Enrichment of Leishmania DNA from clinical samples by SureSelect; Y axis shows percentage of reads mapping to the L. donovani reference genome following SureSelect enrichment. X axis is the pre-enrichment percentage of Leishmania DNA, estimated by either whole-genome sequencing or qPCR; dotted lines indicate no enrichment and 10-, 100- and 1000-fold enrichment. B. Read mapping statistics in the 63 samples after SureSelect enrichment. Upper panel shows evenness of genome coverage as the proportion of bases covered with a minimum of number of sequencing reads from 1 to 100 as indicated by the seven different lines. Lower panel shows the percentage of reads mapping to the Leishmania reference genome for each sample, shown on the x-axis. The x-axis is common to both panels. BM, Bone Marrow; SP, Spleen aspirate.

Fig 2. Genomic diversity among clinical samples.

A. SNPs. Phylogenetic tree based on SureSelect enriched clinical samples (black circle) and previously sequenced isolates [8]. Diagram is a neighbor-joining phylogeny based on 197 variable sites. ISC2-ISC10 were sub-populations previously defined [8]; ISC11 is a novel group that includes clinical samples and isolates previously designated as ‘ungrouped’. Full sample identifiers are shown in S5A Fig. B. Karyotypes (High quality samples). Y axis shows normalized somy estimate for each chromosome (x axis). Points show central estimate and bars show one standard deviation around these estimates, calculated by the binned depth method. C. Local CNVs. Average copy number per cell of H- and M-loci per ISC group in SureSelect enriched clinical samples (Bone Marrow or Spleen aspirate; BM/SP) and in cultured isolates (promastigotes, Prom). Error bars show one standard deviation around the mean estimate. Only a single sample was available from bone marrow aspirates in ISC1, hence no standard deviation is shown.

Fig 3. Somy comparison between matched clinical samples and culture parasites.

A. Inferred somy for 3 samples for which SureSelect-enriched bone marrow (BM) samples and cultured isolates (MO) collected at the same time from the same clinical samples had matching ISC genotypes. Y axis shows normalized somy estimate for each chromosome (x axis). Points show central estimate and bars show one standard deviation around these estimates. Somy estimates and standard deviations for the bone marrow samples were based on the binned depth method while corresponding values for isolates were based on depth of each position. B. Evidence for polyclonal infections. Bars represent the proportion of sequencing reads showing the ISC-specific genotype or the reference (REF) genotype at loci with ISC-specific alleles. 1) Genotypes at ISC9 and ISC4 loci for SureSelect-enriched bone marrow sample (BPK471BM). 2) Genotypes at ISC4 and ISC3 loci for SureSelect-enriched bone marrow sample BPK296BM.