Evolutionary analysis of in vitro selected ribozymes

We wanted to gain an understanding of how functional molecules are distributed in random sequence space and how they are obtained by in vitro selection processes. In particular, we investigated a portion of RNA sequence space for a preliminary idea of the number and variety of RNA ligase ribozymes, using a previously completed ribozyme ligase selection. Our study would allow us assess the capacity of RNA as a catalytic biomolecule, and also inform us about factors affecting the outcome of in vitro selections, to help us design more efficient selections. To understand how active molecules respond to selective pressures, we followed a particular ligase through various selection rounds.;In Chapter 2, we attempted to determine the information content for ribozyme RNA ligases. We assayed ligation activity for seventy-nine ribozymes obtained from an early selected pool of the ligase selection. We observed a large range of activities. We estimated at least six thousand different ligases (many with low activity) in the starting pool, which is representative of random sequence space. This is a revision of the previous estimate of about sixty-five ligases. Our study represents the first attempt to analyze the distribution of activities in a pool of molecules at the resolution of single sequences. In a collaborative effort with J. M. Urbach, we observed that, barring the first two rounds of selection, enrichment of individual ribozymes in successive rounds did not correlate in any simple way with their relative activities. We tested various factors that might have contributed to the above effect. Our study emphasizes the importance of a large number of factors besides catalytic activity that influence the outcome of an in vitro selection.;In Chapter 3, we attempted to classify the sixty novel ligases that we obtained from early selection rounds. We have identified putative new families for the three previously described ligase classes. We suggest a novel class based on sequence considerations. Ligases with relatively simple secondary structure occur numerous times scattered throughout sequence space whereas the more structurally complex Class I ligase probably occurs only two times. We have also identified promising new ligases that could be starting points for evolving an RNA dependent RNA polymerase ribozyme, the replicase, an important activity in the putative pre-protein RNA world.;In Chapter 4, we studied the evolution of the Class I ligase through successive selection rounds. We obtained the ancestral sequence for this ribozyme from round three. We observed a modest improvement in its activity, specifically, ribozyme folding, between round three and ten. There are six mutational differences between the round three and round ten sequence. Two of these seem to be important in increasing ribozyme activity.;We discovered that there are many ligase ribozymes that lie in a continuum of a large range of activity. Most of these ribozymes are of low activity. Many distinct RNA sequences catalyze ligation. Additionally, we can better appreciate the influence of factors other than the intended activity on the progression and outcome of in vitro selections.