In vitro selection is a relatively recent discipline deployed particularly in the field of biotechnologies. One of the principal interests of the technique is that it enables the exploration of DNA and RNA sequences beyond those found in nature to, for example, create new therapies or molecular probes for imaging studies. Entire oligonucleotide libraries can be constructed using combinatorial approaches[1] and contain up to 1015 molecules, all differing from one another by only a part of their sequences. Once constructed, these libraries are subjected to a selection procedure, with the goal of isolating the oligonucleotides capable of interacting with a protein of interest or catalyzing a chemical reaction. The selected few are called "aptamers". The procedure is iterative: at each selection cycle, the population is enriched with sequences adapted to the selection under way. It so happens that this procedure randomly and progressively introduces mutations in the oligonucleotides[2]. In principle, according to Darwinian evolution, the sequences with "beneficial" mutations as concerns the selection under way have a better chance of being enriched. But is this evolution among the oligonucleotides observable? Is it possible to know if the aptamer isolated at the end of the procedure (the most abundant) was present from the start or a result of evolution within the selection process? And, finally, could there be another aptamer even better than the isolated one?
To respond to these questions, a team from MIRCen employed high throughput sequencing and bioinformatics to identify the sequences present at the different selection cycles. Using a program they developed, PATTERNITY-Seq, they were able to reconstitute the evolutionary pathway of in vitro selection. The unequalled resolution of their method sheds new light on the effects of selection pressure, and especially, it allows for the observation of the emergence of new variants in the later cycles of selection and thus the identification of new aptamers with clearly improved properties. Their procedure is currently being used to discover aptamers specific to biomarkers in neurodegenerative diseases.
[1] For example, to create a 10-nucleotide sequence, the number of possible combinations is 410 (each position will hold one of the four nucleotides, G, A, C, T (or U)) and the number of possible sequences approximately 1012.
[2] For the procedure, at each selection round, the oligonucleotide sequences are subjected to an amplification step involving enzymes that, although carefully selected for their fidelity, can introduce errors in the sequences.