Identification of transcriptional signals in Encephalitozoon cuniculi widespread among Microsporidia phylum: support for accurate structural genome annotation

Abstract

BackgroundMicrosporidia are obligate intracellular eukaryotic parasites with genomes ranging in size from 2.3 Mbp to more than 20 Mbp. The extremely small (2.9 Mbp) and highly compact (~1 gene/kb) genome of the human parasite Encephalitozoon cuniculi has been fully sequenced. The aim of this study was to characterize noncoding motifs that could be involved in regulation of gene expression in E. cuniculi and to show whether these motifs are conserved among the phylum Microsporidia.ResultsTo identify such signals, 5' and 3'RACE-PCR experiments were performed on different E. cuniculi mRNAs. This analysis confirmed that transcription overrun occurs in E. cuniculi and may result from stochastic recognition of the AAUAAA polyadenylation signal. Such experiments also showed highly reduced 5'UTR's (<7 nts). Most of the E. cuniculi genes presented a CCC-like motif immediately upstream from the coding start. To characterize other signals involved in differential transcriptional regulation, we then focused our attention on the gene family coding for ribosomal proteins. An AAATTT-like signal was identified upstream from the CCC-like motif. In rare cases the cytosine triplet was shown to be substituted by a GGG-like motif. Comparative genomic studies confirmed that these different signals are also located upstream from genes encoding ribosomal proteins in other microsporidian species including Antonospora locustae, Enterocytozoon bieneusi, Anncaliia algerae (syn. Brachiola algerae) and Nosema ceranae. Based on these results a systematic analysis of the ~2000 E. cuniculi coding DNA sequences was then performed and brings to highlight that 364 translation initiation codons (18.29% of total CDSs) had been badly predicted.ConclusionWe identified various signals involved in the maturation of E. cuniculi mRNAs. Presence of such signals, in phylogenetically distant microsporidian species, suggests that a common regulatory mechanism exists among the microsporidia. Furthermore, 5'UTRs being strongly reduced, these signals can be used to ensure the accurate prediction of translation initiation codons for microsporidian genes and to improve microsporidian genome annotation.

Annotation of microsporidian genomes using transcriptional signals

“…After the re-annotation, we observed 42.56% of A and 42.56% of G in the + 4 position from the TIS position and 50% of A in the + 5 position, compared with 29.05% of A and 24.32% of G in the + 4 position and 28.37% of A in the + 5 position for most likely falsely predicted TISs. These values were in accordance with the Kozak sequence bias of E. cuniculi 24 and reinforced the pertinence of the E. intestinalis TIS revision. Moreover, new spliceosomal introns were characterized in nine CDS (Supplementary Data 1).…”

“…Finally, the N. ceranae sequence similarity search and clustering approaches identified 475 genes that encode proteins corresponding to transposable elements (TEs, Supplementary Data 4). 24 . †nine identified by TIs re-annotation, five in new predicted genes and seven defined by comparison with those identified in E. cuniculi 9,26 .…”

“…The genomic comparative analysis and TIS re-annotation of the CDSs from E. intestinalis, N. ceranae and E. bieneusi allowed the revision of 97 AUG initiation codons for the E. cuniculi genome, compared with the previous re-annotation 24 . The reannotations also confirmed the presence of three additional spliceosomal introns characterized and validated by a rapid amplification of cDNA ends-PCR (RACE-PCR) approach 26 (Supplementary Data 1).…”

“…Because the E. intestinalis proteome (Supplementary Data 9 and 10) was very similar to that of E. cuniculi, with only 28 additional proteins detected 10 , it was not included in the following analysis for simplicity and to improve the readability of our presentation. 32 showed that 41 proteins were encoded by genes that harbour an AAATTT-like motif or AT-rich sequence upstream of their TISs, both sequences believed to be regulatory elements involved in gene expression enhancement 24,33 . Among these 41 genes, we found proteins contributing to specific major structural and/or physiological characteristics of the parasitic cell, such as the spore wall, polar tube formation and the parasite-host surface interactions.…”

“…Our recent studies show that CCC-like or GGG-like motifs are present immediately upstream from the start codon 24,25 for all E. cuniculi genes as well as for genes encoding ribosomal proteins in the microsporidian species Antonospora locustae, E. bieneusi, A. algerae and N. ceranae 24,25 . The presence of these motifs in close proximity to the TISs was considered to significantly improve microsporidian genome annotation and re-annotation.…”