Hfq-aptamers in antisense transcripts

For a cell to grow in a controlled way, the synthesis of its gene products has to be tightly regulated. Until recently, studies on the regulation of gene expression were mainly focusing on regulatory proteins. However, in recent years, hundreds of regulatory RNAs have been discovered that regulate gene expression at all levels. Using genomic SELEX as an alternative approach to discover novel RNAs, we identified a large number of genomic Hfq aptamers that map to the antisense strand of protein coding genes in E. coli. These aptamers are enriched opposite to translation initiation sites and opposite to intervening sequences between genes in operons. We hypothesize that these aptamers might be involved in regulating translation or stability of their cognate mRNAs or in the differential expression of genes in polycistronic mRNAs by processing. Alternatively, Hfq aptamers might be responsible to keep the expression of antisenses transcripts low. We want to test these hypotheses by characterizing the antisense transcriptome and revealing the role of Hfq aptamers in the regulation of the E. coli transcriptome. This work might establish large antisense transcripts as an additional class of RNA molecules that control translation and/or transcriptional silencing.

Bacteria are simple single cell organisms that need to adapt to rapidly changing environmental conditions. To be able to do so, they need to respond to these changing environmental signals by regulating their transcriptional output. The RNA-binding protein Hfq plays a key role in these adaptation processes. Therefore we performed a very detailed analysis of the Hfq-dependent transcriptome of E. coli to understand the dynamics of genetic adaptation. Bacteria have very compact genomes compared to higher eukaryotes; the information contained in these genomes is compactly packed with protein coding genes often overlapping each other. With the onset of new sequencing technologies making it possible to obtain complete transcriptome information of cells at different growth conditions, it was rapidly realized that there are many more transcripts in bacterial cells than originally thought from the perspective of canonical protein coding genes. Many transcripts are being discovered that are antisense to protein coding genes and many of them bind tightly to the RNA chaperone Hfq. We took advantage of novel high through-put technologies to identify E. colis Hfq-binding RNAs, Hfq-dependent RNAs, functional antisense RNAs and Hfq-dependent processed RNAs. We developed protocols for the construction of specific libraries for RNA seq using specific antibodies and enzymes to enrich for double-stranded RNAs, Hfq-binding RNAs and 5 processed RNAs. In addition we defined the 3 ends of RNA transcripts en masse combining massive 3 RACE and deep sequencing. We performed total RNA sequencing in different genetic backgrounds using Hfq- and RNase III-deficient E. coli strains. Total RNA seq was performed under logarithmically growing cells, stationary cells and in the presence of the rho factor inhibitor bicyclomycin. The deep sequencing data was analysed computationally and biochemically by Northern hybridization. We further developed a very sensitive assay to detect novel antisense promoters and could show that the newly discovered antisense RNAs are expressed from functional promoters and are involved in the formation of RNA double strands suggesting their functionality. In summary, this project is revealing a very detailed picture of the role of Hfq in shaping bacterial transcriptomes. Hfq affects approximately 20% of the E. coli RNAs, being that their expression is up- or down regulated in Hfq-deficient strains, that their 5 ends are differentially processed in the absence of Hfq and that the formation of dsRNAs is affected.