Systematic evolution of ligands by exponential enrichment
[6] The process begins with the synthesis of a very large oligonucleotide library, consisting of randomly generated sequences of fixed length flanked by constant 5' and 3' ends.
The constant ends serve as primers, while a small number of random regions are expected to bind specifically to the chosen target.
[2] A caution to consider in this method is that the selection of extremely high, sub-nanomolar binding affinity entities may not in fact improve specificity for the target molecule.
Several variations of their screening process, called SELEX have been reported which can yield sequences with desired properties needed for their final use.
[2][3][12] While Ellington and Szostak demonstrated that chemical synthesis is capable of generating ~1015 unique sequences for oligonucleotide libraries in their 1990 paper on in vitro selection,[3] they found that amplification of these synthesized oligonucleotides led to significant loss of pool diversity due to PCR bias and defects in synthesized fragments.
Examples of target immobilization methods include affinity chromatography columns,[3] nitrocellulose binding assay filters,[2] and paramagnetic beads.
[7] Recently, SELEX reactions have been developed where the target is whole cells, which are expanded near complete confluence and incubated with the oligonucleotide library on culture plates.
[3] With unbound sequences washed away, the specifically bound sequences are then eluted by creating denaturing conditions that promote oligonucleotide unfolding or loss of binding conformation including flowing in deionized water,[3] using denaturing solutions containing urea and EDTA,[13][14] or by applying high heat and physical force.
A drawback of this method is that the product should be purified from double stranded DNA (dsDNA) and other left-over material from the PCR reaction.
[26] Immobilization is a necessary component of SELEX; however, it has the potential to inhibit key epitopes, and thus weaken the likelihood of successful binding, particularly when working with small molecules.
[30] Despite the publication of various methods aimed at increasing the affinity and specificity of aptamers,[31][32][33] experimental approaches face limitations in the number and variety of sequences that can be examined and selected.
[35] This means that existing aptamers may not fully cover the diversity of target molecules or may not have optimal properties due to limitations of the underlying method.
RNA and DNA secondary structure prediction by dynamic programming algorithms such as RNAfold (ViennaRNA) [36] and by machine learning models such as SPOT-RNA,[37] MXfold2 [38] provides the opportunity to assess the ability of sequences in the primary library to fold into complex structures, allowing for the selection of only the most promising sequences from the entire pool.
For this reason, algorithms like Ufold from the University of California [39] and AliNA from Xelari Inc. [40] have been developed, which demonstrate a significant increase in computational speed due to their faster architecture, and can be applied for preliminary in silico analysis of these libraries.
The technique has been used to evolve aptamers of extremely high binding affinity to a variety of target ligands, including small molecules such as ATP[41] and adenosine[12][42] and proteins such as prions[43] and vascular endothelial growth factor (VEGF).
