Virtual approach hits inhibitor jackpot

Model three-dimensional structure of Bcl-2 in complex with the Bak BH3 peptide.

Recent studies have suggested that Bcl-2 blocks programmed cell death by binding the pro-apoptotic peptides Bak, Bad, or Bax to its BH3 binding pocket. For years, researchers have tried to discover nonpeptide small-molecule inhibitors that competitively bind to this BH3 domain. But despite the availability of powerful computerized search tools, only four distinct classes of Bcl-2 nonpeptide small-molecule inhibitors have been discovered so far.

Remarkably, in a single study, Shaomeng Wang and colleagues at the Georgetown University Medical Center (Washington, DC) and National Cancer Institute (Frederick, MD) have identified seven new classes of Bcl-2 inhibitors using a structure-based computerized search of the National Cancer Institute’s three-dimensional (NCI 3D) database (J. Med. Chem. 2001, 44 (25), 4313–4324). Lacking an experimental structure of Bcl-2, the researchers modeled the Bcl-2 BH3 binding pocket using a molecular-modeling program and a high-resolution NMR solution structure of Bcl-X_L, a protein with similar structure and function to Bcl-2.

They then used a database search program to screen the modeled BH3 binding site against 206,876 compounds in the NCI 3D database, taking into account the conformational flexibility of each compound. Of the 35 candidates identified from the search, 7 demonstrated inhibitory activity toward the pro-apoptotic peptide Bak during in vitro binding assays, and one showed potent inhibition of both cell viability and cell growth in several different cancer cell lines that express Bcl-2.

Wang credits the success of his group’s strategy to two things. First, he says, “Our recent analysis of several large chemical databases, including the Available Chemical Directory, showed that the NCI database has by far the highest number of compounds that are unique to it.” Second, he adds, “By considering the conformational flexibility of small molecules in structure-based database searching, I believe that we have significantly improved the chance of identifying small-molecule inhibitors of Bcl-2, as compared to the rigid search approach which only uses one conformation for each small molecule.”

Since publishing this study, the researchers have completed extensive structure-based design and optimization experiments, with promising results. Soon, small-molecule Bcl-2 inhibitors may form an indispensable new class of weapons against cancer.