Surveying functional nucleic acid sequence space

Insight into natural evolution can be gained from studying systems in vitro that are subject to adaptation and selection. Nucleic acids are a convenient system in which to study evolution since they are unique among extant biopolymers in that genotype and phenotype are joined in a single molecule. Furthermore, the abundance and distribution of functional molecules in nucleic acid sequence space (the ensemble of all possible sequences) may even have relevance to the origin of life.; Chapter 1 discusses an iterative, function-based amplification process known as in vitro selection that was previously developed to examine the functional repertoire of ribonucleic acids (RNAs). Functionality is rare in nucleic acid sequence space. For example numerous searches of ribonucleic acid sequence space with random sequence unbiased libraries that contain molecules scattered throughout the space have previously identified only a single RNA motif (Sassanfar aptamer) that can bind the ubiquitous metabolic cofactor ATP. Many questions remain about the abundance and distribution of other ATP aptamers in RNA sequence space.; Chapter 2 examines if there are other RNA ATP aptamers by sampling subsets of sequence space with random sequence biased libraries comprised of different genetic alphabets (AUGC, AUG_, _UGC, and AU_). Some of these libraries excluded certain nucleotides and consequently the Sassanfar motif. More than 11 putative motifs with highly biased AUG_ and _UGC compositions were isolated. Analysis of the functional sequence abundance in the different subsets of sequence space indicated that functional sequence are unevenly distributed in the space, and the relative abundance is AUGC > AUG _ ≅ _UGC >> AU_. The high information content of the new functional motifs relative to the Sassanfar motif suggests that 63 nucleotide RNA sequence space contains at least hundreds of high information content motifs and only a few low information content ones.; Chapter 3 describes the distribution of ATP binding motifs in sequence space by employing yet another sampling strategy. Libraries were constructed that contained molecules only from a focused subset of sequence space called a neighborhood. An algorithm was designed to determine the average distribution of the Sassanfar motif in these neighborhoods and consequently find neighborhoods that are unlikely to contain the Sassanfar motif. Experimental searches in these low probability Sassanfar neighborhoods determined that in 63 nucleotide RNA sequence space, ATP binding motifs are distributed such that 10 mutations on average are sufficient to mutate most sequences into a functional motif. Chapter 4 synthesizes these results into a description of the abundance and distribution of ATP binding motifs in RNA sequence space.