Genome-wide bioinformatic prediction and experimental evaluation of potential RNA thermometers.

Only recently, the fundamental role of regulatory RNAs in prokaryotes and eukaryotes has been appreciated. We developed a pipeline from bioinformatic prediction to experimental validation of new RNA thermometers. Known RNA thermometers are located in the 5'-untranslated region of certain heat shock or virulence genes and control translation by temperature-dependent base pairing of the ribosome binding site. We established the searchable database RNA-SURIBA (Structures of Untranslated Regions In BActeria). A structure-based search pattern reliably recognizes known RNA thermometers and predicts related structures upstream of annotated genes in complete genome sequences. The known ROSE(1) (Repression Of heat Shock gene Expression) thermometer and several other functional ROSE-like elements were correctly predicted. For further investigation, we chose a new candidate upstream of the phage shock gene D (pspD) in the pspABCDE operon of E. coli. We established a new reporter gene system that measures translational control at heat shock temperatures and we demonstrated that the upstream region of pspD does not confer temperature control to the phage shock gene. However, translational efficiency was modulated by a point mutation stabilizing the predicted hairpin. Testing other candidates by this structure prediction and validation process will lead to new insights into the requirements for biologically active RNA thermometers. The database is available on http://www.ruhr-uni-bochum.de/mikrobiologie/ .