Towards the identification of antisense RNAs within genes of transcription regulators

Motivation: Almost one hundred of small regulatory RNAs (sRNAs) have been discovered in bacteria. Most of them are encoded in trans to regulated genes, while the set of known antisense transcripts, generated from within coding sequences (aRNAs) include only few species. Currently it is not clear how accurately this difference reflects the real situation. Thus, most methods used to reveal regulatory RNAs in genome-wide scale were based on searching for evolutionary conserved sequences or were purposefully attuned to intergenic regions. That prohibited identification of novel transcripts within coding sequences. To our knowledge a total of 1493 non-overlapping genes for untranslated RNAs have been suggested in "empty" genomic regions. To estimate the significance of antisense transcription we used pattern recognition software PlatProm, capable of predicting promoters independently on their location. More than a thousand of promoter-like signals for antisense transcription have been found. Direct experimental verification of their activity is required to estimate a reliability of these predictions. Results: Here we analyze functional attribution of genes, possessing putative promoters for antisense transcription, characterize the distribution of promoter-like signals in the genetic locus containing gene hns and provide experimental evidence that RNA polymerase in vitro forms transcriptionally competent complexes with its internal promoter for antisense transcription. We also compare free energies of folding for known sRNAs with those of aRNAs, predicted in genes of transcription factors, and conclude that aRNA which may be expressed from hns has stability typical for other aRNAs.