Transcriptomic analysis reveals novel downstream regulatory motifs and highly transcribed virulence factor genes of Entamoeba histolytica.

BACKGROUND: Promoter motifs in Entamoeba histolytica were earlier analysed using microarray data with lower dynamic range of gene expression. Additionally, previous transcriptomic studies did not provide information on the nature of highly transcribed genes, and downstream promoter motifs important for gene expression. To address these issues we generated RNA-Seq data and identified the high and low expressing genes, especially with respect to virulence potential. We analysed sequences both upstream and downstream of start site for important motifs. RESULTS: We used RNA-Seq data to classify genes according to expression levels, which ranged six orders of magnitude. Data were validated by reporter gene expression. Virulence-related genes (except AIG1) were amongst the highly expressed, while some kinases and BspA family genes were poorly expressed. We looked for conserved motifs in sequences upstream and downstream of the initiation codon. Following enrichment by AME we found seven motifs significantly enriched in high expression- and three in low expression-classes. Two of these motifs (M4 and M6) were located downstream of AUG, were exclusively enriched in high expression class, and were mostly found in ribosomal protein, and translation-related genes. Motif deletion resulted in drastic down regulation of reporter gene expression, showing functional relevance. Distribution of core promoter motifs (TATA, GAAC, and Inr) in all genes revealed that genes with downstream motifs were not preferentially associated with TATA-less promoters. We looked at gene expression changes in cells subjected to growth stress by serum starvation, and experimentally validated the data. Genes showing maximum up regulation belonged to the low or medium expression class, and included genes in signalling pathways, lipid metabolism, DNA repair, Myb transcription factors, BspA, and heat shock. Genes showing maximum down regulation belonged to the high or medium expression class. They included genes for signalling factors, actin, Ariel family, and ribosome biogenesis factors. CONCLUSION: Our analysis has added important new information about the E. histolytica transcriptome. We report for the first time two downstream motifs required for gene expression, which could be used for over expression of E. histolytica genes. Most of the virulence-related genes in this parasite are highly expressed in culture.

Post-transcriptional regulation of ribosomal protein genes during serum starvation in Entamoeba histolytica.

Ribosome synthesis involves all three RNA polymerases which are co-ordinately regulated to produce equimolar amounts of rRNAs and ribosomal proteins (RPs). Unlike model organisms where transcription of rRNA and RP genes slows down during stress, in E. histolytica rDNA transcription continues but pre-rRNA processing slows down and unprocessed pre-rRNA accumulates during serum starvation. To investigate the regulation of RP genes under stress we measured transcription of six selected RP genes from the small- and large-ribosomal subunits (RPS6, RPS3, RPS19, RPL5, RPL26, RPL30) representing the early-, mid-, and late-stages of ribosomal assembly. Transcripts of these genes persisted in growth-stressed cells. Expression of luciferase reporter under the control of two RP genes (RPS19 and RPL30) was studied during serum starvation and upon serum replenishment. Although luciferase transcript levels remained unchanged during starvation, luciferase activity steadily declined to 7.8% and 15% of control cells, respectively. After serum replenishment the activity increased to normal levels, suggesting post-transcriptional regulation of these genes. Mutations in the sequence -2 to -9 upstream of AUG in the RPL30 gene resulted in the phenotype expected of post-transcriptional regulation. Transcription of luciferase reporter was unaffected in this mutant, and luciferase activity did not decline during serum starvation, showing that this sequence is required to repress translation of RPL30 mRNA, and mutations in this region relieve repression. Our data show that during serum starvation E. histolytica blocks ribosome biogenesis post-transcriptionally by inhibiting pre-rRNA processing on the one hand, and the translation of RP mRNAs on the other.

Ribosomal RNA and protein transcripts persist in the cysts of Entamoeba invadens.

In most organisms rDNA transcription ceases under conditions of growth stress. However, we have earlier shown that pre-rRNA accumulates during encystation in Entamoeba invadens. We labeled newly-synthesized rRNA during encystation, with [methyl-(3)H] methionine in the presence of chitinase to enable uptake of isotope. Incorporation rate reduced after 24h, and then increased to reach levels comparable with normal cells. The label was rapidly chased to the ribosomal pellet in dividing cells, while at late stages of encystation the ratio of counts going to the pellet dropped 3-fold. The transcript levels of selected ribosomal protein genes also went down initially but went up again at later stages of encystation. This suggested that rRNA and ribosomal protein transcription may be coordinately regulated. Our data shows that encysting E. invadens cells accumulate transcripts of both the RNA and protein components of the ribosome, which may ensure rapid synthesis of new ribosomes when growth resumes.

Analysis of U3 snoRNA and small subunit processome components in the parasitic protist Entamoeba histolytica.

In the early branching parasitic protist Entamoeba histolytica, pre-rRNA synthesis continues when cells are subjected to growth stress, but processing slows down and unprocessed pre-rRNA accumulates. To gain insight into the regulatory mechanisms leading to accumulation, it is necessary to define the pre-rRNA processing machinery in E. histolytica. We searched the E. histolytica genome sequence for homologs of the SSU processome, which contains the U3snoRNA, and 72 proteins in yeast. We could identify 57 of the proteins with high confidence. Of the rest, 6 were absent in human, and 4 were non-essential in yeast. The remaining 5 were absent in other parasite genomes as well. Analysis of U3snoRNA showed that the E. histolytica U3snoRNA adopted the same conserved secondary structure as seen in yeast and human. The predicted structure was verified by chemical modification followed by primer extension (SHAPE). Further we showed that the predicted interactions of Eh_U3snoRNA boxes A and A' with pre-18S rRNA were highly conserved both in position and sequence. The predicted interactions of 5'-hinge and 3'-hinge sequences of Eh_U3 snoRNA with the 5'-ETS sequences were conserved in position but not in sequence. Transcription of selected genes of SSU processome was tested by northern analysis, and transcripts of predicted sizes were obtained. During serum starvation, when unprocessed pre-RNA accumulated, the transcript levels of some of these genes declined. This is the first report on pre-rRNA processing machinery in E. histolytica, and shows that the components are well conserved with respect to yeast and human.