Modulation of Aneuploidy in Leishmania donovani during Adaptation to Different In Vitro and In Vivo Environments and Its Impact on Gene Expression

Abstract

Aneuploidy is usually deleterious in multicellular organisms but appears to be tolerated and potentially beneficial in unicellular organisms, including pathogens. Leishmania, a major protozoan parasite, is emerging as a new model for aneuploidy, since in vitro-cultivated strains are highly aneuploid, with interstrain diversity and intrastrain mosaicism. The alternation of two life stages in different environments (extracellular promastigotes and intracellular amastigotes) offers a unique opportunity to study the impact of environment on aneuploidy and gene expression. We sequenced the whole genomes and transcriptomes of Leishmania donovani strains throughout their adaptation to in vivo conditions mimicking natural vertebrate and invertebrate host environments. The nucleotide sequences were almost unchanged within a strain, in contrast to highly variable aneuploidy. Although high in promastigotes in vitro, aneuploidy dropped significantly in hamster amastigotes, in a progressive and strain-specific manner, accompanied by the emergence of new polysomies. After a passage through a sand fly, smaller yet consistent karyotype changes were detected. Changes in chromosome copy numbers were correlated with the corresponding transcript levels, but additional aneuploidy-independent regulation of gene expression was observed. This affected stage-specific gene expression, downregulation of the entire chromosome 31, and upregulation of gene arrays on chromosomes 5 and 8. Aneuploidy changes in Leishmania are probably adaptive and exploited to modulate the dosage and expression of specific genes; they are well tolerated, but additional mechanisms may exist to regulate the transcript levels of other genes located on aneuploid chromosomes. Our model should allow studies of the impact of aneuploidy on molecular adaptations and cellular fitness.IMPORTANCE Aneuploidy is usually detrimental in multicellular organisms, but in several microorganisms, it can be tolerated and even beneficial. Leishmania-a protozoan parasite that kills more than 30,000 people each year-is emerging as a new model for aneuploidy studies, as unexpectedly high levels of aneuploidy are found in clinical isolates. Leishmania lacks classical regulation of transcription at initiation through promoters, so aneuploidy could represent a major adaptive strategy of this parasite to modulate gene dosage in response to stressful environments. For the first time, we document the dynamics of aneuploidy throughout the life cycle of the parasite, in vitro and in vivo We show its adaptive impact on transcription and its interaction with regulation. Besides offering a new model for aneuploidy studies, we show that further genomic studies should be done directly in clinical samples without parasite isolation and that adequate methods should be developed for this.

Keywords: Leishmania; aneuploidy; gene dosage; genomics; life cycle.

FIG 1

Flowchart showing the experimental histories of the 3 L. donovani strains, BPK282/0 (A), BPK275/0 (B), and Ld1S (C). Black arrows indicate in vitro passage (R, with the number of subinoculations since isolation from the patient or animal); red arrows indicate passage through hamster (P, with the number of passages in the animal); green arrows indicate a transmission through the natural vector named in green. Samples that were submitted to DNA and RNA sequencing are in solid-line boxes, while samples for which only DNA sequencing was performed are in dashed-line boxes. ProM, promastigotes; aM, amastigotes. Letters within brackets indicate the life stage and environmental conditions from which the parasites came, as follows: I, in vitro promastigotes; A, amastigotes from hamster; P, promastigotes from in vitro culture; P/sf, sand fly-originated promastigotes.

FIG 2

Dynamics of aneuploidy of L. donovani BPK282/0 during adaptation to different in vitro and in vivo environments. (A) Promastigotes reisolated from sand flies. (B) Amastigotes during short-term adaptation. (C) Amastigotes during long-term adaptation. (D and E) Promastigotes reisolated from sand flies infected by hamster-derived amastigotes (D) and during adaptation to in vitro culture of amastigote-derived promastigotes (E). Heat maps show median normalized read depths of chromosomes found within each cell population for each of the 36 chromosomes (y axis) and each sample (x axis). The color key shows the normalized chromosome read depth (S) and the distribution frequency. S ranges are as follows: monosomy, S < 1.5 (dark blue); disomy, 1.5 ≤ S < 2.5 (light blue); trisomy, 2.5 ≤ S < 3.5 (green); tetrasomy, 3.5 ≤ S < 4.5 (orange); pentasomy, 4.5 ≤ S < 5.5 (red). A filled black triangle in an upper right corner indicates a significant change of S (≥0.5, with a shift from one S range to another and a P value of ≤10⁻⁵) in comparison to the S value of the sample framed in a black box at the bottom.

FIG 3

Dynamics of aneuploidy of L. donovani BPK275/0 (A) and Ld1S (B) during adaptation to different in vivo environments. See legend to Fig. 2 for details. See Data Set S1A1 and A2 for all of the pairwise comparisons.

FIG 4

Effects of variable aneuploidy of L. donovani BPK282 on transcriptomes during adaptation to different environments. Three life stages/environments in which aneuploidy differs are considered: in vitro promastigotes (R20) (A), amastigotes adapted to hamster (P4) (B), and sand fly-derived promastigotes (C). The black and blue lines correspond to S values and RNA-S values, respectively. The error bars correspond to median absolute deviations. Details on the correlation between S and RNA-S can be found in Table S4, and the values are in Data Set S1B.

FIG 5

Individual gene expression in promastigotes showing variable aneuploidy. Samples considered here are ProM (I) 20, ProM (P) sand fly 3, ProM (A) R3, and ProM (A) R10. Transcript depth ratios between samples with trisomic and disomic chromosomes 5 and 8 are shown. Each single dot represents a transcript; blue indicates the plus strand, and red the minus strand. The x axis represents each coding unit along the chromosome, the y axis the average transcript depth ratios, and corresponding histograms show their distribution. The snoRNA cluster on chromosome 5 and the amastin family are marked with boxes. The thick gray horizontal lines indicate the median depth, and the thin gray lines represent 1 MAD (median absolute deviation), 2 MAD, and 2.5 MAD, respectively, away from the median depth.

FIG 6

Aneuploidy-independent changes of transcriptomes during parasite differentiation. The samples considered here have the same aneuploidy pattern and correspond to (i) adapted amastigotes of hamsters (aM P3 and P4 taken together) and (ii) promastigotes freshly differentiated from these [ProM(A) R3 and R10 taken together]. The histogram shows the percentage of upregulated coding units that have a ≥2-fold change and Benjamini-Hochberg-corrected P value of ≤0.05 in each chromosome based on the DEseq analysis.