A dynamical model for ribozyme function based on the sequential folding of pre-mRNA transcripts.

Making use of parallel Monte Carlo simulations of competing folding pathways, we determine the specific stages in the in vivo sequential folding of group I pre-mRNA transcripts where participation of a trans-acting factor or an already-assembled portion of the transcript itself is required to generate a catalytically competent structure and subsequently shape the 3' splicing site. Thus, the model for ribozyme function revealed by the simulations should be regarded as dynamical since it is based on sequential folding. Our main aim is to prove that sequential folding warrants the chronology of splicing events required for ribozyme function, a feature which cannot be reproduced in existing static models of folding based on free energy minimization. The effect of trans-acting factors on the catalytically relevant folding pathway is assessed by comparing the in vitro folding pathway with the pathway that leads to the structure that splices the 5' extremity. The latter has been inferred previously by other authors using phylogenetic analysis. Since our model is rooted in multiprocessed folding algorithms, we concentrate on mitochondrial pre-mRNA transcripts belonging to group I which undergo no detectable self-splicing in vitro. As an illustrative example, the results have been specialized to the fourth intron of the yeast apocytochrome b gene (YCOB4). A crucial feature of our approach, irreproducible in previous models, is that it accounts for a meaningful scenario of competition between hydrolysis at the 3' extremity of the intron and exon-exon ligation. We prove that this scenario explains how the premature formation of conserved helix P10 is prevented until 5' cleavage has taken place.(ABSTRACT TRUNCATED AT 250 WORDS)