Genome wide comparison of Ethiopian Leishmania donovani strains reveals differences potentially related to parasite survival

Abstract

Leishmania donovani is the main cause of visceral leishmaniasis (VL) in East Africa. Differences between northern Ethiopia/Sudan (NE) and southern Ethiopia (SE) in ecology, vectors, and patient sensitivity to drug treatment have been described, however the relationship between differences in parasite genotype between these two foci and phenotype is unknown. Whole genomic sequencing (WGS) was carried out for 41 L. donovani strains and clones from VL and VL/HIV co-infected patients in NE (n = 28) and SE (n = 13). Chromosome aneuploidy was observed in all parasites examined with each isolate exhibiting a unique karyotype. Differences in chromosome ploidy or karyotype were not correlated with the geographic origin of the parasites. However, correlation between single nucleotide polymorphism (SNP) and geographic origin was seen for 38/41 isolates, separating the NE and SE parasites into two large groups. SNP restricted to NE and SE groups were associated with genes involved in viability and parasite resistance to drugs. Unique copy number variation (CNV) were also associated with NE and SE parasites, respectively. One striking example is the folate transporter (FT) family genes (LdBPK_100390, LdBPK_100400 and LdBPK_100410) on chromosome 10 that are single copy in all 13 SE isolates, but either double copy or higher in 39/41 NE isolates (copy number 2-4). High copy number (= 4) was also found for one Sudanese strain examined. This was confirmed by quantitative polymerase chain reaction for LdBPK_100400, the L. donovani FT1 transporter homolog. Good correlation (p = 0.005) between FT copy number and resistance to methotrexate (0.5 mg/ml MTX) was also observed with the haploid SE strains examined showing higher viability than the NE strains at this concentration. Our results emphasize the advantages of whole genome analysis to shed light on vital parasite processes in Leishmania.

Fig 1. Comparison of aneuploidy profiles of Leishmania donovani from northern and southern Ethiopia.

Chromosome numbers are listed down the right side of the heat map. Across the bottom the L. donovani isolates, clones and parental patient strains, analyzed are indicated. The dendrogram on the top shows clusters of the strains and clones based on similarity of aneuploidy profiles and was generated by comparison of the Euclidean distances between aneuploidy profiles among strains (R packages; stats, gplots). Bar under dendrogram indicates the parasite source: Red–southern Ethiopia and Blue–northern Ethiopia. Insert: color key indicating ploidy of chromosomes in heat map. The minimum number of chromosomes, 2, is diploid. Green indicates chromosomal copy number > 4 and pink ≤ 4.

Fig 2. Phylogentic tree and similarity analysis based on SNP.

Panel A. Cluster analysis based on Euclidian distance between SNP similarity profiles among all 41 NE (black bar under name) and SE (red bar under name) lines. The resulting phylogenetic tree describes the genetic diversity among Ethiopia isolates within the two main Leishmania populations. B. Comparison between SNP profiles of Leishmania donovani strains and clones from northern and southern Ethiopia. Heat map allows visualization of similarities between all SNPs from SE and NE strains and clones. Each colored square in the matrix indicates the percent SNP similarity for a strain/clone listed on the right compared to strain/clone listed along the bottom of the matrix. The heat map was generated with R packages; stats and gplots. The green range indicates on identity level ≥ 70% and the pink < 70%. Darker green/pink point to higher or lower identity, respectively.

Fig 3. Principal component analysis (PCA) based on SNP of Leishmania donovani NE and SE lines.

PCA analysis was performed after pruning SNPs showing high linkage disequilibrium (LD). SNPs with LD>0.9 are not included. Panel A. PCA of all 41 NE and SE lines. The two atypical SE parasites, AM422 and AM553, fall in eigenvectors 1 and 2 relatively far from both populations (EV1 = 0.229 EV2 = -0.954 and EV1 = 0.945 EV2 = 0.265, respectively). Two NE parasites, GR364sk/cl.II and the atypical LDS373bm, fall with the SE population. Despite the short distance between typical SE (red circles) and NE (black circles) population, these two populations are physically well defined. Panel B. PCA of 39 NE and SE lines without atypical SE strains. The typical SE population (red circles) is very homogenous and falls in a tight cluster unlike the NE population, which is very heterogeneous but still distinct from the SE population. LDS373bm falls equidistance between the SE population and LDS373sp both isolated from the same patient. Line GR364sk/cl.II still clusters together with the SE population (black circle on the middle of the SE cluster). Interestingly lines from spleen and skin taken from the same patients (GR363sp and sk; and GR364 sp and sk) are clustered together and create a well-defined population.

Fig 4. Scatterplots after reduction of semantic redundancy in enriched Gene Ontology (GO) terms for unique SNP containing genes of northern and southern Leishmania donovani populations.

Removal of redundant GO terms was carried out using the web program REVIGO with C set at 0.7 [27]. Enriched GO terms are graphed in two-dimensional semantic space with terms that semantically similar closer together. The semantic space units have no intrinsic meaning. Enrichment p-values are shown by circle color as indicated in the key to the right of each panel, while circle diameter indicates the frequency of the GO term i.e. general terms are larger. Numbers in each circle refer to the GO terms listed in S7 Table. A. Northern Ethiopia (NE). Panel B. Southern Ethiopia (SE).

Fig 5. Comparison of folate transporter 1 (LdBPK_100400.1) genomic DNA copy number in northern and southern Ethiopian Leishmania donovani strains by quantitative polymerase chain reaction (qPCR) and cn.mops (GCNV).

L. donovani strains and clones are shown on the x-axis, while estimated FT1 copy number is given on the y-axis. SE strains (n = 9, line under x-axis), and NE (n = 9) and Sudanese (S3570) strains (dashed line under x-axis).

Fig 6. Folate transporter gene copy number is correlated with greater sensitivity to methotrexate.

Effect of Methotrexate on parasite growth (%Dead) was plotted as a function of the total gene copy number of folate (LdBPK_100380; LdBPK_100390; LdBPK_100400 and LdBPK_100410) and biopterin transporters (LdBPK_355160) showing variation on chromosomes 10 and 35, respectively (FT+BT). CN was predicted by cn.mops. %Dead—percentage parasites killed following incubation with 0.5mg/ml methotrexate as described in Material and Methods. Average of n = 3 independent experiments ± s.e. Isolates from southern Ethiopia (SE) have codes beginning with AM. Isolates from northern Ethiopia (NE) have codes beginning with GR.