Combinatorial Chemical Screening with DNA-Encoded Libraries

A robotic multichannel pipette and an assembly line of multi-well plates are on a drug discovery quest. Each of the thousands or hundreds of thousands of wells in this process contains one specific molecule. When a promising candidate is identified, its molecular identity is confirmed by knowing which well it came from.

This is high-throughput screening, the backbone of drug discovery in academic and commercial labs worldwide. Every day, new candidate molecules are identified, selected, and characterized in the hopes they will one day be able to treat a human disease. Biologics, including monoclonal antibodies, recombinant proteins, and gene therapies, are screened similarly but have the advantage that their identity and structure can be determined after the fact by their DNA sequence.

What if we had a different way to identify candidate small molecules—one that didn’t conform to the “one well, one molecule” paradigm? One that could incorporate some of the advantages that discovery using biologics has over small molecules.

DNA-encoded libraries

The idea of using unique DNA sequences as “barcode” identifiers is not new, having been used in species identification, cell lineage tracking, and mapping the brain.^1–3 Using DNA barcodes as labels for combinatorial chemistry was first conceived in the 1990s, about the same time as these other applications.⁴ The idea was to build a chemical library with each compound bearing a readable DNA barcode for identification.

At its simplest, a DNA-encoded library (DEL) uses building block chemistry to expand a molecule, conferring additional properties and building more molecules with this combinatorial chemistry design. A DNA sequence corresponding to the new building block is added with each building block, using the DNA sequence to record the chemistry. Complexity can be added by splitting a chemical pool, adding a new building block (and expanding the barcode) in each split reaction, and re-pooling.⁵ This allows massively parallel combinatorial chemistry.

Each synthesized compound or variant doesn’t need its own well in a plate, and modern PCR can detect and identify a barcode in minute amounts, vastly reducing the amount of each individual test compound needed for screening. DELs most often use an immobilized target, but fully soluble and even screens using overexpressed protein on the cell surface have been used.⁵ Furthermore, some groups have used innovative techniques to explore variants of promising candidates, including directed evolution-like techniques to expand their library.^5,6

Since the 90s, DNA synthesis and next-generation sequencing advances have made DELs more tractable and spurred further innovation. Advancements simplified the process, such as using a universal template so that a chemical building block could be added simultaneously with a DNA barcode addition on the same molecule, requiring fewer chemical steps for addition and making it less error-prone to build increasingly complex molecules and keep track of the synthesis.⁷

DELs leading the way in discovery

DNA-encoded libraries have been gaining popularity. The number of publications describing small molecules discovered through DELs shows it, with over 80 publications since 2015 compared to only 22 between 2004 and 2014.⁵ Clinical candidates discovered using DELs include potential treatments for subarachnoid hemorrhage, drug-resistant infections, insulin sensitivity, rheumatic and autoimmune diseases, and cancer.^5,8

Beyond clinical candidates, discoveries from DELs have yielded new types of compounds for modulating protein targets. A collaboration between Duke University and Nuevolution used DEL screening to identify a negative allosteric modulator for unliganded β-adrenergic receptor and its first known positive allosteric modulator.⁹ With these exciting biological discoveries, Amgen acquired Nuevolution in 2019 and positioned DELs as the future of drug discovery.¹⁰

Challenges and the future

DELs have been hailed as the future of drug discovery; the market for DELs is expected to increase by nearly 25% yearly through this decade.¹¹ One of their key strengths is their versatility, speed, and the sheer size of libraries compared to traditional HTS methods.

However, DELs have some limitations that remain to be overcome. DELs have been used extensively in binding assays but cannot screen for effects like inhibition or activation of a target, warranting a complementary approach to drug discovery using multiple screening techniques.^9,12 The compounds in a DEL cannot rely on harsh synthesis conditions that could damage the DNA barcode, which could limit the pool variety, and it has been noted that DELs occasionally fail to identify candidates for certain classes of protein targets.¹² A solution to these could be to use a DEL to identify candidates. Then, a focused screen could characterize a different parameter, such as inhibition or kinetics. Several biopharma companies have taken this approach, yielding candidates for clinical development.¹²

Another feature that could be a limitation is that the DNA barcodes don’t impart functional characteristics to the molecules. This is ideal for unbiased screening, which means you can’t derive a compound from a sequence, limiting the feasibility of a “directed evolution” approach to modify the synthesis as you would with a DNA-encoded biologic. However, some labs are exploring the possibility of directed evolution with barcoded peptides using an expanded genetic code.⁶

It would be ideal to identify candidates in the complex environment of the cell, but DNA-labeled molecules are membrane impermeant, so screening in cells has been limited to surface proteins. Potential methods to introduce DELs to the cytosol, such as transfection or microinjecting oocytes, are possibilities under investigation.^9,12 Given the advances in the past decade alone, we can anticipate creativity leading to new methods and applications for DELs and more clinical development for DEL-identified compounds in the coming years as this shifts the drug discovery paradigm.

References

1. Huang L, Kebschull JM, Fürth D, et al. BRICseq Bridges Brain-wide Interregional Connectivity to Neural Activity and Gene Expression in Single Animals. Cell. 2020;182(1):177-188.e27.

2. Kebschull JM, Zador AM. Cellular barcoding: lineage tracing, screening and beyond. Nat Methods. 2018;15(11):871-879. 

3. Hebert PDN, Cywinska A, Ball SL, deWaard JR. Biological identifications through DNA barcodes. Proc R Soc Lond B Biol Sci. 2003;270(1512):313-321. 

4. Brenner S, Lerner RA. Encoded combinatorial chemistry. Proc Natl Acad Sci. 1992;89(12):5381-5383. 

5. Peterson AA, Liu DR. Small-molecule discovery through DNA-encoded libraries. Nat Rev Drug Discov. 2023;22(9):699-722. 

6. Krusemark CJ, Tilmans NP, Brown PO, Harbury PB. Directed chemical evolution with an outsized genetic code. Isalan M, ed. PLOS ONE. 2016;11(8):e0154765. 

7. Li Y, Zhao P, Zhang M, Zhao X, Li X. Multistep DNA-templated synthesis using a universal template. J Am Chem Soc. 2013;135(47):17727-17730. 

8. Ma P, Zhang S, Huang Q, et al. Evolution of chemistry and selection technology for DNA-encoded library. Acta Pharm Sin B. Published online October 11, 2023. 

9. Zhao G, Huang Y, Zhou Y, Li Y, Li X. Future challenges with DNA-encoded chemical libraries in the drug discovery domain. Expert Opin Drug Discov. 2019;14(8):735-753. 

10. Amgen. DNA-encoded libraries will drive drug design. Amgen. Published November 22, 2019. Accessed November 16, 2023. 

11. Global DNA-Encoded Library Market: Challenges, Triumphs, and the Road Ahead to 2031 | LinkedIn. Accessed November 16, 2023. 

12. Gironda-Martínez A, Donckele EJ, Samain F, Neri D. DNA-encoded chemical libraries: a comprehensive review with succesful stories and future challenges. ACS Pharmacol Transl Sci. 2021;4(4):1265-1279.