The amazing world of bacterial structured RNAs

See related research article by Weinberg et al.: http://genomebiology.com/2010/11/3/R31 webciteThe compact genomes of bacteria contain 10 to 15% non-coding DNA sequences, which are transcribed into non-coding RNAs. Several classes of non-coding RNAs are small, less than 80 to 150 nucleotides, and act as post-transcriptional regulators by targeting mRNAs. Another large class of non-coding RNAs act in cis by binding structured elements in the 5' untranslated regions of mRNAs. Perhaps the best known are called riboswitches; upon binding a metabolite, the fold of the transcript is modified and this influences either the termination of transcription or the initiation of translation [1].Some longer non-coding RNAs have also been detected in recent years. For example, RNAIII present in several Gram-positive bacteria is 500 nucleotides long and contains structured regions framing an open-reading frame [2]. However, two recent papers from Ron Breaker's group increase the number of large non-coding RNAs astonishingly [3,4]. Several new smaller non-coding RNAs are also identified. Strikingly, most of the new non-coding RNAs are structurally very complex. The complexity of some of the larger ones seems similar to that of the large ribozymes, such as the self-splicing group I and group II introns. These observations show, once again, how little we know about the microbial world: a great proportion of these new non-coding RNAs were identified in metagenomes or in environmental DNA sequences.The search for non-coding RNAs in genomes is far from trivial [5]. Even for homologous and functionally well characterized RNA molecules, such as the ubitquitous RNaseP or the telomerase RNA, the search cannot be reliably automated because of the large and unpredictable variation in the length of the RNA transcript, with new insertions appearing in an otherwise globally similar secondary structure. On the other hand, the de novo search for the presence of non-coding RNAs within intergenic regi