Deep Learning Approach for the Discovery of Tumor-Targeting Small Organic Ligands from DNA-Encoded Chemical Libraries

DNA-Encoded ChemicalLibraries (DELs) have emerged as efficientand cost-effective ligand discovery tools, which enable the generationof protein-ligand interaction data of unprecedented size. Inthis article, we present an approach that combines DEL screening andinstance-level deep learning modeling to identify tumor-targetingligands against carbonic anhydrase IX (CAIX), a clinically validatedmarker of hypoxia and clear cell renal cell carcinoma. We presenta new ligand identification and hit-to-lead strategy driven by machinelearning models trained on DELs, which expand the scope of DEL-derivedchemical motifs. CAIX-screening datasets obtained from three differentDELs were used to train machine learning models for generating novelhits, dissimilar to elements present in the original DELs. Out ofthe 152 novel potential hits that were identified with our approachand screened in an in vitro enzymatic inhibition assay, 70% displayedsubmicromolar activities (IC50 < 1 & mu;M). To generatelead compounds that are functionalized with anticancer payloads, analoguesof top hits were prioritized for synthesis based on the predictedCAIX affinity and synthetic feasibility. Three lead candidates showedaccumulation on the surface of CAIX-expressing tumor cells in cellularbinding assays. The best compound displayed an in vitro K (D) of 5.7 nM and selectively targeted tumors in mice bearinghuman renal cell carcinoma lesions. Our results demonstrate the synergybetween DEL and machine learning for the identification of novel hitsand for the successful translation of lead candidates for in vivotargeting applications.