Phylogenetic conservation among > 100 group I introns and previous in vitro studies have implicated a G.U pair as defining the 5'-splice site for exon ligation. The U residue defines the 3' end of the 5' exon, and the complementary G residue is part of the internal guide sequence (IGS) that base pairs to the 5' exon. We now quantitate the effect of this pair on individual reaction steps using the L-21ScaI ribozyme, which is derived from the group I intron of Tetrahymena thermophila pre-rRNA. The following results indicate that interactions with this G.U pair contribute to the binding of the 5'-exon, the positioning of the 5'-splice site with respect to the catalytic site, and the chemical step. The oligonucleotide, CCCUCU, binds to the ribozyme approximately 20-fold stronger than CCCUCC despite the fact that the U-containing oligonucleotide forms an approximately 5-fold less stable duplex with an oligonucleotide analog of the IGS, GGAGGG. This and two independent experimental observations indicate that the G.U pair contributes approximately 100-fold (3 kcal/mol, 50 degrees C) to tertiary interactions that allow the P1 duplex, which is formed between the 5'-exon and the IGS, to dock into the ribozyme's core. The approximately 50-80-fold increase in miscleavage of 5'-exon analogs upon replacement of the 3'-terminal U of CCCUCU with C or upon removal of the 3'-terminal U suggests that the tertiary interactions with the G.U pair not only contribute to docking but also ensure correct positioning of the 5'-splice site with respect to the catalytic site, thereby minimizing the selection of incorrect splice sites. Comparison of the rates of the chemical cleavage step with G.U vs G.C suggests that the G.U pair contributes approximately 10-fold to the chemical step. It was previously suggested that the 2'-hydroxyl of this U residue helps stabilize the 3'-oxyanion leaving group in the chemical transition state via an intramolecular hydrogen bond. Relative reactivities of oligonucleotide substrates with ribose and deoxyribose U and C are consistent with a model based on a recent X-ray crystallographic structure in which the exocyclic amino group of G helps orient the 2'-hydroxyl of U via a bridging water molecule, thereby strengthening the hydrogen bond donated from the 2'-hydroxyl group to the neighboring incipient 3'-oxyanion. Finally, kinetic and thermodynamic evidence for the formation of a G.C+ wobble pair is presented.(ABSTRACT TRUNCATED AT 400 WORDS)

The L-21 ScaI ribozyme derived from the Tetrahymena thermophila pre-rRNA group I intron catalyzes a site-specific endonucleolytic cleavage of RNA, DNA, and chimeric RNA/DNA oligonucleotides: CCCUCUA5 + G-->CCCUCU + GA5. The pH-rate dependence was determined for the reaction of the E.G complex with the oligonucleotide substrate d(CCCUC)r(U)d(A5) [(kcat/Km)S conditions]. Although it was shown that the pH dependence is not affected by specific buffers, there is inhibition by specific monovalent cations. The intrinsic pH-rate dependence is log-linear with slope 1 below pH 7, displays an apparent pKa of 7.6, remains nearly level until pH 8.5, and then begins to fall. Two models to explain the apparent pKa were ruled out: (1) the pKa represents loss of a proton from the nucleophilic 3' OH of G, and (2) the pKa arises from a change in rate-limiting step from a pH-dependent to a pH-independent step. In addition, these models, or others involving a single titration, cannot account for the decrease in activity at high pH. A third, unconventional, model is consistent with all of the data. It involves inactivation of the ribozyme by any of several independent titrations of groups with pKa values considerably higher than the apparent pKa of 7.6. The data are consistent with loss of catalytic function upon release of a proton from any one of 19 independent sites with pKa = 9.4 (the unperturbed pKa of N1 of G and N3 of U in solution). Independent experiments investigating the effect of pH on different reaction steps supported this model and suggested the identity of some of the required protons. This mechanism of inactivation is expected to generally affect the behavior of RNAs at pH values removed from the pKa of the titrating bases.

Phylogenetic comparisons and site-directed mutagenesis indicate that group I introns are composed of a catalytic core that is universally conserved and peripheral elements that are conserved only within intron subclasses. Despite this low overall conservation, peripheral elements are essential for efficient splicing of their parent introns. We have undertaken an in-depth structure-function analysis to investigate the role of one of these elements, P5abc, using the well-characterized ribozyme derived from the Tetrahymena group I intron. Structural comparisons using solution-based free radical cleavage revealed that a ribozyme lacking P5abc (E ∆P5abc ) and E ∆P5abc with P5abc added in trans (E ∆P5abc ‚P5abc) adopt a similar global tertiary structure at Mg 2+ concentrations greater than 20 mM [Doherty, E. A., et al. (1999) Biochemistry 38, 2982-90]. However, free E ∆P5abc is greatly compromised in overall oligonucleotide cleavage activity, even at Mg 2+ concentrations as high as 100 mM. Further characterization of E ∆P5abc via DMS modification revealed local structural differences at several positions in the conserved core that cluster around the substrate binding sites. Kinetic and thermodynamic dissection of individual reaction steps identified defects in binding of both substrates to E ∆P5abc , with g25-fold weaker binding of a guanosine nucleophile and g350-fold weaker docking of the oligonucleotide substrate into its tertiary interactions with the ribozyme core. These defects in binding of the substrates account for essentially all of the 10 4 -fold decrease in overall activity of the deletion mutant. Together, the structural and functional observations suggest that the P5abc peripheral element not only provides stability but also positions active site residues through indirect interactions, thereby preferentially stabilizing the active ribozyme structure relative to alternative less active states. This is consistent with the view that peripheral elements engage in a network of mutually reinforcing interactions that together ensure cooperative folding of the ribozyme to its active structure.Catalytic RNAs can provide rate enhancements that rival those of protein enzymes (1-3). To accomplish this, an RNA must contain sufficient information in its primary sequence to meet two challenges: it must be able to form stable tertiary structure, and it must be able to stabilize the active structure relative to alternate inactive and less active conformations. Relative to proteins, RNA is expected to encounter structural difficulties arising from the charged nature and greater rotational freedom of the phosphodiester backbone, the limited number of nucleoside bases, and the sequestration of the bases by base-pairing within secondary structure (4, 5). We have focused on group I introns to study the means by which RNA molecules can achieve functional structures.The wealth of functional and structural information available for group I introns makes them a powerful system in which to examine the structural underpin...