Comparative expression profiling of Leishmania: modulation in gene expression between species and in different host genetic backgrounds

Abstract

Background: Genome sequencing of Leishmania species that give rise to a range of disease phenotypes in the host has revealed highly conserved gene content and synteny across the genus. Only a small number of genes are differentially distributed between the three species sequenced to date, L. major, L. infantum and L. braziliensis. It is not yet known how many of these genes are expressed in the disease-promoting intracellular amastigotes of these species or whether genes conserved between the species are differentially expressed in the host.

Methods/principal findings: We have used customised oligonucleotide microarrays to confirm that all of the differentially distributed genes identified by genome comparisons are expressed in intracellular amastigotes, with only a few of these subject to regulation at the RNA level. In the first large-scale study of gene expression in L. braziliensis, we show that only approximately 9% of the genes analysed are regulated in their RNA expression during the L. braziliensis life cycle, a figure consistent with that observed in other Leishmania species. Comparing amastigote gene expression profiles between species confirms the proposal that Leishmania transcriptomes undergo little regulation but also identifies conserved genes that are regulated differently between species in the host. We have also investigated whether host immune competence influences parasite gene expression, by comparing RNA expression profiles in L. major amastigotes derived from either wild-type (BALB/c) or immunologically compromised (Rag2(-/-) gamma(c) (-/-)) mice. While parasite dissemination from the site of infection is enhanced in the Rag2(-/-) gamma(c) (-/-) genetic background, parasite RNA expression profiles are unperturbed.

Conclusion/significance: These findings support the hypothesis that Leishmania amastigotes are pre-adapted for intracellular survival and undergo little dynamic modulation of gene expression at the RNA level. Species-specific parasite factors contributing to virulence and pathogenicity in the host may be limited to the products of a small number of differentially distributed genes or the differential regulation of conserved genes, either of which are subject to translational and/or post-translational controls.

Figure 1. Expression profiling of gene subsets from three Leishmania genomes.

Scatter plots (generated using Microsoft Excel) show the distribution of log₂ fold changes in RNA expression levels against their statistical significance (−log₂ p-value) for up to 785 genes per genome. Fold change of >1.7 between data points (dashed horizontal lines); significant p-value<0.05 (dashed vertical lines). (A) L. braziliensis amastigotes vs. procyclics; (B) L. braziliensis amastigotes vs. metacyclics; (C) L. braziliensis metacyclics vs. procyclics; amastigotes of (D) L. braziliensis vs. L. major, (E) L. braziliensis vs. L. infantum, (F) L. major vs. L. infantum; (G) L. major amastigotes from BALB/c vs. Rag2^−/−γ_c ^−/− mice.

Figure 2. Distribution of genes preferentially expressed in one or more stages of L. braziliensis.

Venn diagram showing distribution of upregulated genes (>1.7 fold, p<0.05) in stages of the L. braziliensis life cycle. 8.8% (60/678) of the genes probed are regulated in their expression by these criteria, with 58% of these showing increased expression during the metacyclic stage. *Genes that are not differentially expressed between these stages.

Figure 3. Differential expression of amastigote genes between Leishmania species.

Venn diagrams showing pairwise comparisons of amastigote RNA expression from the three target Leishmania species; the number of genes with increased expression (>1.7-fold, p<0.05) in each species is indicated. *Genes that show no significant difference in expression between species.

Figure 4. Calmodulin RNA expression is upregulated in amastigotes of L. braziliensis.

Comparison of amastigote RNA levels by both microarray and RT-qPCR analysis demonstrates increased expression from the chromosome 9 calmodulin gene array in L. braziliensis but decreased expression in L. major and L. infantum. 2 independent biological replicates were used to generate data from each species.

Figure 5. Comparison of L. major infections in BALB/c and Rag2^−/−γ_c ^−/− mice.

Groups of 5 mice of each strain were inoculated with L. major purified metacyclics and infections monitored over a 4 week period as described (Materials and Methods). (A) Lesion development (footpad thickness); (B) parasite burdens in footpad lesions after 4 weeks; (C) parasite burdens in spleen after 4 weeks; (D) parasite burdens in liver after 4 weeks. The data presented are representative of two independent studies; error bars represent the standard error of the mean and the * indicates a significant difference (p<0.05) as determined by Students t-test (unpaired).

Figure 6. HASPB RNA and protein abundance in L. major amastigotes isolated from BALB/c and Rag2^−/−γ_c ^−/− mice.

Amastigotes isolated from the livers of two mice of each strain (BALB/c and Rag2^−/−γ_c ^−/− mice), 4 weeks post-infection with L. major (see Figure 5), were used for RNA and protein extraction, as described. HASPB expression was analysed by (A) RT-qPCR for RNA and (B) immunoblotting with anti-HASPB for protein (L. major HASPB migrates at 38.5 kDa (black arrow); NMT, at 50 kDa (white arrow)). NMT was used as a constitutive control for both RNA and protein expression levels.