Path to Discovery

The path leading to successful drug discovery in the twenty-first century comprises many routes, some of which lead to dead ends or are unpromising. Careful planning and navigation are the keys to that success. The Center for Drug Discovery uses state-of-the-art equipment and techniques and sophisticated methodologies that will yield better candidate compounds for drugs and experimental therapeutics. How can we be so certain? We utilize techniques that sample more fully the theoretical total “chemical space” that comprises all possible chemical synthesis reaction products. The use of such techniques and methods to synthesize chemical entities results in compounds that more efficiently fill three-dimensional molecular space and better modulate the functions of specific target proteins. Chief among the technologies we employ are DNA-encoded chemistry technology (DEC-Tec) and fragment-based library discovery (FBLD).

Pharmaceutical companies assemble extremely expensive million-component drug libraries for high-throughput screening (HTS) directed toward possible drug targets. In addition to the expense, these compound collections are often unsuccessful in generating active compounds some drug targets because they are not composed of diverse molecular types.

We utilize a more economical approach enables the generation of greater small-molecule diversity. The approach involves the preparation of DNA-encoded libraries to discover useful lead small-molecule candidates. DEC-Tec, conceived as a small-molecule analogy to high-affinity biopolymer ligand discovery approaches such as phage display and aptamer selection, allows for the efficient interrogation of much larger and more diverse compound sets (collections up to 100 billion have been described) as a single mixture. The DEC-Tec library of drug-like molecules, each of which is attached to a unique DNA sequence that acts as a “bar code” defining its synthesis, will be incubated with our target proteins and processed as shown. No assay or structural knowledge of the target protein is required as compounds within the library are selected by affinity.

This DEC-Tec strategy permits the identification of the small-molecule “needle” in a billion-compound “haystack.” Compared to typical high-throughput (HTS), DEC-Tec requires a relatively small investment to establish, and requires only micrograms of target protein: an amount within the capability of even the smallest academic laboratory. Additionally, the billion-compound collections involving DEC-Tec equate to 1000 times more compounds than the most state-of-the-art HTS facility can screen. This expanded number of small-molecules provides enhanced opportunities for hit discovery against a range of disease targets. There are several published examples of the utility of DEC-Tec in discovering effective small-molecule ligands. We believe that with DEC-Tec, the CDD has the capacity to drastically improve success rate of discovery and subsequent development of small-molecule probes and lead compounds within the academic community.

FBLD is used to efficiently generate high-quality therapeutic leads and involves screens of collections of chemicals called fragments (typically, <300 Da and <25 heavy atoms). Because fragments are much smaller than drugs, they have lower affinity for a target protein.

There are multiple advantages, however, of fragment-based drug discovery:

• Optimizing the binding affinity of drugs is much faster because it is easier to obtain or synthesize analogs;
• Fragments are selected based on their higher solubility, and analogs of the fragments can be designed to be more “drug-like;”
• FBLD is often utilized to overcome “undruggable” protein targets or those with shallow ligand-binding pockets.

Unfortunately, most commercial fragment libraries are limited to highly sp² (2-dimensional) compounds, severely limiting their chemical potential. We utilize diversity-oriented synthesis in conjunction with FBLD to create structurally-diverse libraries of sp³ (3-dimensional) fragments that fill a larger “chemical” space than existing libraries. This is an important distinction because sp³ fragments mimic the interactions observed between proteins, thus leading to better binding molecules. We are working toward the synthesis of a 2000-member fragment library containing unprecedented shape diversity as compared with that of traditional libraries. This fragment collection will allow for the discovery of fragment binders to some of the most interesting disease-protein targets.