RNA is a versatile molecule and due to its wide range of biochemical properties it is capable of multifarious functions. The linear sequence of RNA makes it a simple source of genetic information, whereas the property to form secondary and tertiary structures allows its interaction with other macromolecules and provides environments for catalytic activities. Thus, besides the role of RNA molecules as information-carrying intermediaries in gene expression, they act as key catalytic, structural, and regulatory elements in the cell. In bacteria, the discovery of a staggering number of small regulatory RNAs (sRNAs) by systematic searches of sequenced genomes over the last years led to an increasing recognition of the potential impact of sRNAs on bacterial physiology. These sRNAs act as post-transcriptional regulators of bacterial gene expression in response to diverse growth and environmental stress conditions. In contrast to cis-encoded antisense RNAs of mobile elements such as plasmids, the majority of bacterial sRNAs seems to bind by imperfect basepairing to trans-encoded mRNAs and thereby inhibit translation or lead to mRNA degradation. The early studies have often focused on interactions with single target mRNAs, but there is growing evidence that sRNAs can regulate many diverse mRNAs in parallel. However, the understanding of how sRNAs could directly control multiple mRNAs by antisense mechanism has been limited by the low number of validated sRNA-target interactions and, hence, the difficulty to reliably predict new interactions.

In this thesis, two aspects of sRNA-mediated regulation in bacteria are investigated: (1) multiple target regulation and (2) approaches for the identification of novel sRNAs in bacteria. The first part addresses the question how sRNA targets can be identified and how multiple targets can be directly regulated by one sRNA. For this purpose, biocomputational and experimental approaches for the identification and validation of targets of a small RNA, GcvB, from Salmonella typhimurium are presented. Furthermore, it is shown that a conserved region within GcvB RNA directly interacts with multiple mRNAs of genes involved in amino acid transport and biosynthesis. It is shown how the identification of this conserved element can be used to refine experimental and biocomputational target-identification approaches.

The second part deals with the identification of novel sRNAs. Bioinformatics-based approaches often rely on the prediction of orphan transcription signals and primary sequence conservation of sRNA candidates within closely related species or on the conservation of RNA structure. This implies the availability of related genome sequences and well-defined promoter and terminator models. In contrast, approaches based on shotgun-cloning and direct sequencing of RNA (so-called RNomics) allow to identify novel sRNAs without prior knowledge but were, until recently, limited by the cost-intensive Sanger sequencing. In this thesis, the use of high-throughput sequencing for the identification of sRNAs bound to the RNA-binding protein Hfq in Salmonella is demonstrated. Furthermore, deep sequencing reveals the primary transcriptome of the major human pathogen Helicobacter pylori, a bacterium in which no sRNAs have been described. Moreover, an approach based on selective sequencing of cDNA libraries specifically enriched for primary transcripts is developed which allows to define a global map of transcriptional start sites of mRNAs in H. pylori.