Search

Splicing of the ribosomal RNA precursor of Tetrahymena is an autocatalytic reaction, requiring no enzyme or other protein in vitro. The structure of the intervening sequence (IVS) appears to direct the cleavage/ligation reactions involved in pre-rRNA splicing and IVS cyclization. We have probed this structure by treating the linear excised IVS RNA under nondenaturing conditions with various single- and double-strand-specific nucleases and then mapping the cleavage sites by using sequencing gel electrophoresis. A computer program was then used to predict the lowest-free-energy secondary structure consistent with the nuclease cleavage data. The resulting structure is appealing in that the ends of the IVS are in proximity; thus, the IVS can help align the adjacent coding regions (exons) for ligation, and IVS cyclization can occur. The Tetrahymena IVS has several sequences in common with those of fungal mitochondrial mRNA and rRNA IVSs, sequences that by genetic analysis are known to be important cis-acting elements for splicing of the mitochondrial RNAs. In the predicted structure of the Tetrahymena IVS, these sequences interact in a pairwise manner similar to that postulated for the mitochondrial IVSs. These findings suggest a common origin of some nuclear and mitochondrial introns and common elements in the mechanism of their splicing.

An RNA model system consisting of an oligomer binding to a 4-nt overhang at the 5' end of a hairpin stem provides thermodynamic parameters for helix-helix interfaces. In a sequence-dependent manner, oligomers bind up to 1000-fold more tightly adjacent to the hairpin stem than predicted for binding to a free tetramer at 37 degrees C. For the interface (/) in [formula: see text] additional free energy change, delta delta G 37 degrees, for binding is roughly the nearest-neighbor delta G 37 degrees for propagation of an uninterrupted helix of equivalent sequence, CGGC. When X and Z are omitted, the delta delta 37 degrees is even more favorable by approximately 1 kcal/mol (1 cal = 4.184J). On average, predictions of 11 RNA secondary structures improve from 67 to 74% accuracy by inclusion of similar stacking contributions.

The accuracy of computer predictions of RNA secondary structure from sequence data and free energy parameters has been increased to roughly 70%. Performance is judged by comparison with structures known from phylogenetic analysis. The algorithm also generates suboptimal structures. On average, the best structure within 10% of the lowest free energy contains roughly 90% of phylogenetically known helixes. The algorithm does not include tertiary interactions or pseudoknots and employs a crude model for single-stranded regions. The only favorable interactions are base pairing and stacking of terminal unpaired nucleotides at the ends of helixes. The excellent performance is consistent with these interactions being the primary interactions determining RNA secondary structure.