In silico and in vitro approaches to aptamer enrichment, selection and evolution

Abstract

Aptamers are short oligonucleotide sequences capable of specific, high-affinity binding. Given this property, aptamers are attractive candidates as active components in therapeutics and diagnostics. Aptamers are isolated from a library of random ssDNA or RNA using systematic evolution of ligands by exponential enrichment (SELEX), which involves successive rounds of selection for analyte binding and amplification. The aptamer selection process is complex and may not always result in aptamers with the desired properties. In order to enhance existing aptamers and improve upon the aptamer selection process, both computational and laboratory methodologies have been developed and tested in this thesis.

To model aptamer binding, theoretical fitness landscapes were evolved by a new process of phenome selected growth. The similarity of phenome selected binding landscapes to aptamer-target binding was demonstrated with several metrics, one involving empirical aptamer binding data. Phenome selected binding landscapes more accurately represent aptamer-target binding than previously described models.

AptSim is a computer program, with method and code developed in this project, to simulate the aptamer selection process using the phenome selected theoretical binding model. Stochastic effects were shown to result in the loss of the highest affinity binders in initial SELEX rounds. Mutation and recombination were each shown to benefit the selection process with a synergy of the two outperforming all other simulations. Furthermore, the properties of recombination size and rate were optimised using an evolutionary algorithm and shown to outperform unoptimised properties in terms of enrichment by 2.75 times. These in silico optimisations of SELEX were obtained at a fraction of the time and cost of empirical optimisation.

Aptamer affinity maturation is a novel method described herein, that enhances the binding affinity of previously selected aptamers. Low sequence space coverage for initial libraries means that the highest affinity aptamer for a target is often excluded from the selection. The bioinformatics tool ‘Resample’ was developed, which takes the sequencing result from a previous SELEX experiment and generates a new library representing every member of the target binding aptamer family which includes the highest affinity aptamer with high probability. DNA microarray can then select and characterise this new library. Aptamer affinity maturation enhanced previously selected aptamers against a malarial antigen with an order of magnitude increase in affinity.

The novel technique DNA display was demonstrated. DNA display involves the capture and display of an RNA strand to the DNA which encodes it during transcription. This linking of RNA phenotype to DNA genotype may be used with particle display to perform FACS selection of RNA aptamers. DNA aptamers were enriched against the spinach fluorophore using FACS selection.

Both in silico and in vitro methods have been developed in this thesis. Much of the benefit in these methods comes from their joint application, in silico plus in vitro, leading to acceleration of aptamer selection, and enhancement of existing aptamers.