Introduction
As a biological catalyst, RNA would appear to be underpowered, seemingly outgunned by the diverse chemical side chains available to proteins. But in this first issue of Nature Chemical Biology, Das and Piccirilli1 demonstrate that RNA has more catalytic tricks at its disposal than was originally thought when catalytic RNA was discovered. Using a clever chemogenetic suppression analysis, the authors provide compelling evidence that an RNA enzyme uses general acid catalysis. Their approach, which merges chemical substitution with enzyme kinetics, provides an elegant and potentially general method to test whether a particular functional group acts as a general acid or general base in enzyme catalysis.
General acid or general base catalysis is used by enzymes to facilitate the proton transfers required in biochemical transformations. This catalytic mechanism requires that enzymes position active-site acidic or basic side chains with appropriate pKa values near the substrate. This possibility was considered unlikely ten years ago because the pKas of the nucleobases were viewed as being outside the catalytically useful range2. Recent evidence has suggested that RNA can perturb the pKas of its nucleobase functional groups sufficiently close to neutrality that RNA could use general acid or general base catalysis3, 4. The possibility of general acid-base catalysis by RNA has been most extensively explored in the hepatitis delta virus (HDV) ribozyme, a cis-acting enzyme that cleaves its RNA genome into unit-sized linear fragments as part of rolling-circle replication5. At the cleavage site, a 2'-OH attacks the scissile phosphate to liberate a 5'-OH, with the 2'-OH losing a proton and the O5' leaving group accepting one during the course of the reaction (Fig. 1). A cytidine (C76 in the antigenomic form of the enzyme studied by Das and Piccirilli) is essential to the self-cleavage reaction. Crystal structures of the cleaved and uncleaved forms of the genomic HDV ribozymes have shown that the equivalent cytidine is reasonably positioned to act catalytically6, 7. What remained ambiguous was whether the cytidine acts as a general acid or as a general base.
Figure 1: General acid catalysis by the HDV ribozyme.
(a) Proposed mechanism for general acid catalysis of phosphodiester cleavage by the general acid C76 in the HDV ribozyme. (b) Das and Piccirilli probed the role of C76 through functional-group substitutions in the phosphodiester backbone (X) and the cytosine base (Y and Z).
Full size image (10 KB)Das and Piccirilli used a series of chemical substitutions at residues surrounding the HDV ribozyme active center. Most importantly, they replaced the O5' leaving group with a hyperactivated 5'-bridging sulfur group (5'-PS, Fig. 1), a substitution that eliminates the need for a general acid catalysis during bond cleavage. The 5'-sulfur is an excellent leaving group but is not a particularly good hydrogen-bond acceptor. As a result the P-S linkage shows strong lability toward basic conditions in the uncatalyzed reaction, but is relatively unaffected by acidic pH. Thus, this substitution provides a straightforward means to distinguish between the roles attributed to C76. Das and Piccirilli reported that a C76U mutant ribozyme, which cannot cleave the normal P-O bond, was able to cleave the P-S substrate at wild-type levels. Furthermore, the addition of imidazole, which partially rescues the C76U mutation for P-O bond cleavage, had no effect on the rate of P-S bond cleavage. These results strongly implicate C76 action upon the O5' leaving group.
The authors further refined their conclusions by examining the effect of chemical substitutions of C76 on cleavage of the hyperactivated substrate. First, they eliminated the proton transfer capability of C76 by replacing N3 with a carbon (Fig. 1). Consistent with a general acid mechanism involving the N3 imino group, they observed complete loss of activity for P-O bond cleavage, but virtually no effect for P-S bond cleavage. Second, they enhanced the acidity of the cytidine N3 by replacing C6 with a nitrogen atom (Fig. 1) and explored the pH dependence of P-S and P-O bond cleavage. Divergence in the pH-rate profile of the functional-group mutant provided evidence that deprotonation of C76 N3 is inhibitory. The data fully support the mechanistic proposal originally advanced by Bevilacqua and coworkers, who suggested that the active-site cytidine acts as a general acid4.
These compelling biochemical data seem to contradict conclusions derived from crystal structures of the uncleaved genomic HDV ribozyme. In a previous report, Ke et al. trapped the genomic form of the ribozyme in an unreactive state7. They observed that the C75 N3 (corresponding to C76 in the ribozyme studied by Das and Piccirilli) was closest to the O2' nucleophile, suggestive of general base catalysis by C75. It is conceivable that the mechanisms of the genomic and antigenomic ribozymes differ, but this seems unlikely. Das and Piccirilli do not attempt to reconcile their biochemical data with the previous structural results, arguing simply that the different conformations observed for the substrate and product states render the structures ambiguous. The conclusions of Das and Piccirilli necessitate that the chemical details of the transition state must differ from those inferred from the structural studies of Ke et al. Specifically, the transition state must have a conformation closer to the product than the substrate structures. It may be necessary to obtain structures of transition-state mimics before a convergence between the biochemical and structural results can be achieved.
The HDV ribozyme precedent argues that other catalytic RNAs could use general acid catalysis. For example, a conserved adenosine in the VS ribozyme and a conserved guanosine in the hairpin ribozyme are both crucial for the RNA cleavage reaction and each has been proposed to act as a general acid or general base8, 9. Peptide bond formation by the ribosome may also proceed by general acid-base catalysis, although currently the best candidate for catalytic participation is not the nucleobase of the ribosome but rather the 2'-OH at the end of the tRNA substrate10. This report by Das and Piccirilli provides the means needed to characterize such systems; the ability to ask fine-tuned mechanistic questions using chemically manipulated substrates provides a powerful synergy at the interface of chemistry and biology.
