High Throughput Screening of Aptamers

Aptamers are functional oligonucleotides that are discovered through directed evolution. With affinities for their targets similar to antibodies, aptamers are useful in biosensors, diagnostics and therapeutics. Three elements of natural selection--variation, selection, and replication afford aptamer discovery methods the ability to enrich and distinguish aptamers from randomized pools. Here I describe an aptamer discovery method we are calling High Throughput Screening of Aptamers (HTSA), which incorporates two additional elements that eliminate the cyclic evolution bottleneck in traditional aptamer discovery methods, and expedite the discovery process with no complex automation required. We designed a method that (1) affords multiple representation of every possible sequence (oversampling) or alternatively the multiple representation of each present sequence (reduced complexity) in a library pool and (2) employs massively parallel sequencing after the selection process. The first element facilitates the rapid expulsion of low affinity binding sequences while retaining the high affinity binders at frequencies far above the background probable frequencies. The second element then enhances the observation of the enrichment by providing a comprehensive picture of the enriched pool that cannot be achieved by conventional sequencing of a few clones. We essentially used high throughput sequencing as a sequence reader of enriched library members and a directional sensor of the selection process. We could predict the relative affinity of a sequence for its desired target by simply counting its frequency in the enriched pool. The goal of the work reported in this volume was to develop a high throughput screening method for aptamers that is viable across a variety of potential targets in single-target and multiplexedtarget environments using whole organisms/cells (live Cryptosporidium Parvum oocysts), protein targets (a thrombin, as well as Cryptosporidium Parvum\u27s surface proteins, Cp23 and Cp15) and small molecule targets (a thrombin\u27s glycan moiety and saccharides). We demonstrated the efficacy of our method to rapidly screen and distinguish minimal aptamers by co-isolating, in a single selection step, a previously known DNA aptamer with nanomolar affinity for human á thrombin that was originally discovered after 5 selection cycles and an unprecedented aptamer with apparent specificity for hexose sugars with affinity in the top third of known aptamers for small molecules, from an oversampled library. We also demonstrated HTSA\u27s superior ability to enrich undersampled RNA libraries without any cycling, by successfully isolating high affinity binding sequences against Cryptosporidium Parvum oocysts and two of its surface wall proteins Cp23 and Cp15. Undersampled libraries theoretically have a single copy of each present sequence and traditionally require cyclic enrichment for the copy number to increase enough to result in an observable enrichment; which is the goal of every aptamer discovery method. We circumvented the cycling by translating the undersampled library into a reduced complexity library, which is essentially an undersampled library depleted in lowaffinity sequences by partitioning against the target, then amplified to create multiple copies of each present sequence; this basically functions like an oversampled library that effects an observable enrichment after a single selection step. Additionally, as a result of the dramatic increase in throughput from Sanger to second generation sequencing, we also demonstrate HTSA\u27s ability to exhaustively search the space of sequences within an enriched library which simplifies the characterization of the core binding domain. Ultimately, HTSA simplifies and shortens the discovery process from weeks to months of standard SELEX to less than a week, exhaustively searches the space of sequences within a library, may eliminate the need to truncate long aptamer sequences to find the core binding domain, reduces the quantity of the target required and cycling artifacts since it eliminates cycling altogether, and allows multiplexing of targets and experiments