| With the goal of understanding how to increase the likelihood of isolating desired functional molecules by in vitro selection, we examined factors that affect the procedure's outcome, including the roles of pool design and selection protocol.; Using a selection for RNA kinase activity, we investigated the effects of pool bias, in particular, the inclusion of an ATP aptamer intended to bind substrate. We performed selections on pools that were either random, or biased by the inclusion of a mutagenized or unmutagenized ATP aptamer. Results of selections performed in parallel, or in a competitive fashion on combined pools, suggest that ATP aptamers confer little advantage.; A second aspect of pool design we considered was the length of the random sequence. Parallel selections were performed on pools containing 30, 70, or 238 random bases. Similar numbers and activities of isolated kinases were observed in each case, suggesting that pool length in this range is not an important factor in determining the outcome of RNA kinase selections.; We also considered factors in the selection protocol that affect the outcome of selections, using a previously completed RNA ligase selection as a model. It was previously observed that collective enrichments per selection round of ligases in a given range of activities are not directly proportional to activity. By examining the enrichments of individual sequences as a function of their activities over several rounds of selection, we observed a highly variable relationship between activity and enrichment of individual clones. We attribute this poor correlation to incidental biases from processes in the selection procedure other than the intended selective step. Differences in reverse transcription or intrinsic PCR efficiency were too small to account for the observed lack of correlation between activity and enrichment. However, under certain conditions, we observed a negative correlation between sequence abundance and PCR amplification efficiency. This may account for enrichments that are lower than would be expected on the basis of activity alone.; In summary, pool design has played a lesser role than expected and incidental biases a greater role in determining the outcome of in vitro selections. |
