DNA-Encoded Library Technology (DELT) After a Quarter Century

Abstract

Drug discovery scientists have made great efforts over the years to explore a maximum possible number of chemical structures against a biological target or targets of interest. The advent of automation and miniaturization has enabled significant progress in the high-throughput screening, delivery, and analysis of hundreds of thousands to millions of compounds. This journal has published numerous manuscripts that report on the efficiency of such technologies applied to screening and laboratory sciences. With increased efficiency in any process, one must consider its inputs and outputs. This special collection of SLAS Discovery focuses on DNA-encoded library technology (DELT) to shed light on the new opportunity to make available high numbers of small-molecule structures in the hit-finding process, something that has been a limiting factor in drug discovery. DELT specifically addresses the problem of exploring the interface of chemistry and biology with hundreds of millions of small molecules in an efficient manner that has been neither possible nor feasible before. The venue of this journal is particularly befitting for discussing the technical aspects of the DELT process. Many issues, such as DNA-tag design and informatic analysis of hits, have received less treatment in the scientific literature than reviews of the concept, synthetic transformations, and identification of high-affinity hits. It is important to orient a definition of DELT for some readers. There are many excellent reviews published to date.1Neri D. Lerner R.A. DNA-Encoded Chemical Libraries: A Selection System Based on Endowing Organic Compounds with Amplifiable Information.Ann. Rev. Biochem. 2017; 87 (5.1–5.24 and references therein.)Google Scholar However, briefly, DELT involves the use of DNA oligos that are alternately concatenated as a means to record the introduction of small-molecule reagents or building blocks. Once a cycle of chemistry and DNA-tag ligation has been completed, the separate reactions are mixed and redistributed in a classic split-and-pool process before beginning the next cycle. A typical DELT library will have between three and five cycles with hundreds to thousands of chemical inputs and corresponding tags, thereby generating millions to billions of compounds in a short time and on a small scale. This mixture is then incubated with a protein target of interest to identify compounds that bind with relatively higher affinity, followed by isolation of that enriched mixture. The information encoded by the very small quantity of enriched samples is then amplified by PCR methods prior to sequencing of the DNA tags with next-generation sequence (NGS) technologies. The sequence information is then decoded to indicate which building blocks have been enriched (i.e., have bound with a higher relative affinity) by a target protein. DELT originated from a milestone publication by Brenner and Lerner in 1992.2Brenner S. Lerner R.A. Encoded Combinatorial Chemistry.Proc. Natl. Acad. Sci. U.S.A. 1992; 89: 5381-5383Google Scholar It is interesting to note the similarities to phage display and panning methods for the identification of high-affinity peptides. It is also useful to reflect on the time in which this method was first proposed: at the early days of small-molecule combinatorial chemistry. Perhaps little appreciated at the time, the Brenner and Lerner paper described several key aspects that are still central to the practices of this technology: orthogonal chemical and oligonucleotide bond formation, split-and-pool methodology, encoding with DNA sequences, affinity selection methods, amplification with PCR, installation of restriction sites to avoid primer dominance in selective hybridization, and off-DNA resynthesis. Throughout the 1990s and into the 2000s, many advances in small-molecule combinatorial parallel synthesis and design were developed. The advent of combinatorial chemistry promoted the development of high-throughput, small-molecule analytical and purification technologies. The commonly discussed terminologies of drug-likeness, chemical diversity, and compound collection quality were prompted in part by the generation of small molecules in larger numbers that had been created to date (hundreds to thousands of samples). At the same time, there was a growing understanding of the vastness of chemical space, as well as the inexactness of its definition, which limited effective exploration. DELT provides an economical and feasible means to make progress toward the exploration of larger chemical spaces. Instead of focusing on 1–2 million compounds in a typical high-throughput screen (HTS), a DELT library may present some 100 million compounds for selection by a biological target of interest. The concepts of drug-likeness, chemical diversity, and compound collection are still important, but the efficiency of the DELT process shifts the balance of opportunity costs in a unique manner relative to the traditional HTS compound collection and process. Despite the progress and application of DELT, much remains to be learned and solved. DELT is not a single technology method. Different approaches to this general concept are reported. To this end, it is interesting to have a view of the DELT platform that has been built by Eli Lilly and Company, as reported by Román et al.3Castañón J. Román J.P. Jessop T.C. et al.Design and Development of a Technology Platform for DNA-Encoded Library Production and Affinity Selection.SLAS Discov. 2018; 23: 387-396Google Scholar These authors describe a means to automate the creation of multiple libraries in an efficient process. This paper is an excellent counterpoint to recent disclosures by GlaxoSmithKline (GSK) scientists who describe the DELT platform at that company for the analysis of many DELs against multiple targets, thereby providing an early understanding of the feasibility to perform a small-molecule-based drug discovery program for a specific target. DELT does not consist of a single DNA encoding method.4Goodnow Jr., R.A. Dumelin C.E. Keefe A.D. DNA-Encoded Chemistry: Enabling the Deeper Sampling of Chemical Space. Nat. Rev.Drug Discov. 2017; 16: 131-147Google Scholar Rather, the robustness of DNA as an information storage medium has allowed for the evolution of different DNA architectures and constructs; thus, it is instructive to learn of tagFinder, as reported by Amigo et al.5Amigo J. Rama-Garda R. Bello X. et al.tagFinder: A Novel Tag Analysis Methodology That Enables Detection of Molecules from DNA-Encoded Chemical Libraries.SLAS Discov. 2018; 23: 397-404Google Scholar A highly useful and interesting analysis of randomness in the decoding of high-throughput sequencing analysis is also included in this issue. Kuai and other scientists at GSK explain that despite the noisiness of selection informatics, results can be modeled accurately according to a Poisson distribution.6Kuai L. O’Keeffe T. Arico-Muendel C. Randomness in DNA Encoded Library Selection Data Can Be Modeled for More Reliable Enrichment Calculation.SLAS Discov. 2018; 23: 405-416Google Scholar Although the Poisson distribution concept has been noted for DELT selections before, this is a thorough analysis that increases confidence in the reproducibility of the selection outputs. GSK scientists also report on information methods for the decoding of DEL selections. In another paper by scientists at Purdue University, the statistical details of selection are analyzed.7Denton K.E. Wang S. Gignac M.C. et al.Robustness of In Vitro Selection Assays of DNA-Encoded Peptidomimetic Ligands to CBX7 and CBX8.SLAS Discov. 2018; 23: 417-428Google Scholar The understanding of the enrichment of molecules in a selection process has been in need of statistical analysis, such as the calculation of a Z′ factor. In this report, this analysis is done for selections against the chromodomains from CBX8 and CBX7, conducted over three protein concentrations and indicating sufficient robustness for “for both ligand discovery and determination of quantitative structure–activity relationships.” Many DELT publications focus on the identification of high-affinity ligands for targets of biological interest. Such targets must usually be soluble proteins. In an intersection of cutting-edge technologies, authors at X-Chem working with AstraZeneca and Heptares scientists report on selection against a mutant-stabilized PAR2 GPCR.8Brown D.G. Brown G.A. Centrella P. et al.Agonists and Antagonists of Protease-Activated Receptor 2 Discovered within a DNA-Encoded Chemical Library Using Mutational Stabilization of the Target.SLAS Discov. 2018; 23: 429-436Google Scholar Such targets are normally outside the reach of DELT selections, but these authors show the promise of working with such modified protein targets. At this quarter-century milestone, the science and practice of DELT is resulting in a regular flow of unique, high-affinity hits, and it is evolving as a technology. Such vigorous activity is likely to continue. Special journal publications such as these help to highlight and mark such progression to new and broader audiences of scientists. The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.