Computation-Enabled Structure-Based Discovery of Potent Binders for Small-Molecule Aptamers

Aptamers, functional nucleic acids recognized for their high target-binding affinity and specificity, have been extensively employed in biosensors, diagnostics, and therapeutics. Conventional screening methods apply evolutionary pressure to optimize affinity, while counter-selections are used to minimize off-target binding and improve specificity. However, aptamer specificity characterization remains limited to target analogs and experimental controls. A systematic exploration of the chemical space for aptamer-binding chemicals (targets) is crucial for uncovering aptamer versatility and enhancing target specificity in practical applications, a task beyond the scope of experimental approaches. To address this, we employed a high-throughput three-stage structure-based computational framework to identify potent binders for two model aptamers. Our findings revealed that the l-argininamide (L-Arm)-binding aptamer has a 31-fold higher affinity for the retromer chaperone R55 than for L-Arm itself, while guanethidine and ZINC10314005 exhibited comparable affinities to L-Arm. In another case, norfloxacin and difloxacin demonstrated over 10-fold greater affinity for the ochratoxin A (OTA)-binding aptamer OBA3 than OTA, introducing a fresh paradigm in aptamer-target interactions. Furthermore, pocket mutation studies highlighted the potential to tune aptamer specificity, significantly impacting the bindings of L-Arm or norfloxacin. These findings demonstrate the effectiveness of our computational framework in discovering potent aptamer binders, thereby expanding the understanding of aptamer-binding versatility and advancing nucleic acid-targeted drug discovery.