(260a) Predicting Accessible Splice Sites for Trans-Splicing Ribozymes: Modeling and Experiments

Trans-splicing ribozymes are catalytic RNA molecules derived from group I introns and engineered to edit a target messenger RNA by replacing a portion of the latter with a new sequence. Such ribozymes may be useful for repairing disease-related mRNAs before they are translated into protein. Because RNA has a strong propensity to form secondary and tertiary structures, a key problem in the design of trans-splicing ribozymes is the identification of accessible binding sites on the target mRNA. Here we tackle this problem computationally by calculating the change in free energy, ΔG, for the binding of a ribozyme to each possible target site on a given substrate mRNA. The resulting ΔG values include contributions from local unfolding of appropriate regions on both the substrate and the ribozyme. Very negative values of ΔG should correlate with high target site accessibility and consequently with high product franctions. To test this approach, we assayed the in vitro efficiency of several trans-splicing ribozymes that target different sites on the chloramphenicol acetyl transferase (CAT) mRNA. Of the sites predicted to be accessible, ~70% were found to yield significant amounts of trans-splicing products. Thus, our computational approach may provide a rapid assessment of the target sites most likely accessible to trans-splicing ribozymes on a given mRNA.