Control of the intrinsic apoptotic pathway is exercised through the interplay of a 20-member family of proteins related to the founding member, Bcl-2. These proteins are confined to the vicinity of two membrane surfaces, the endoplasmic reticulum and the mitochondrial outer membrane. Like most of cellular biochemistry, reactions are confined in nano time and space which do not allow the approach to equilibrium seen at the macro level of the test tube, motivating us to initiate a dynamic systems biology approach to these non-equilibrium reactions. Our laboratory has investigated the mechanism of action of a non-peptide small molecule TW-37, which represents a new chemical class of compounds like ABT-737 (Abbott) useful in the induction of apoptosis in tumor cells. The drug induces apoptotic death in lymphoma cells with an observed EC50 of 290 nM (

Mohammad et al., Clin. Cancer Res. 13, 2226 (2007)
). TW-37 binds with nanomolar affinity to the hydrophobic groove on the surface of the pro-survival Bcl-2 proteins which can accommodate the hydrophobic face of the amphipathic α-helix of BH3-only family members such as Bid. To model how the drug perturbs the Bcl-2 system, we have immobilized a highly-helical 24-mer Bid-derived BH3-domain peptide via its N-terminus to the surface of a CM5 chip (ligand) and analyzed peptide-protein interaction with flowing analyte proteins Bcl-2, Bcl-XL, Mcl-1, Bcl-w, and Bfl-1/A1 using the Biacore 3000. Using an analysis of Langmuir binding, we observe that the analyte proteins bind Bid with ON-rates (ka) varying modestly (3.01–7.72 E+05 1/Ms), but dissociate from Bid with 17-fold variation in OFF-rate (kd). The Bcl-w complex with Bid is the slowest to dissociate with a kd of 0.0015 (1/s); the most rapid dissociation is seen with Bcl-2 and Bid, where the kd is 0.0385. We find that Bid interacts with the multidomain proteins with dissociation constants (KD) between 4 and 82 nM, intermediate between earlier determinations (
J. Med. Chem. 49, 6139 (2006)
);
Molec. Cell 17, 393 (2005)
), but significantly lower than the KD for the drug TW-37 and the multidomain proteins (260–1100 nM reported previously; 79-333 nM measured here). Contrary to the suggestion of Shangary and Johnson (
Biochemistry 41, 9485 (2002)
), our BH3-mimetic TW-37 is unable to disrupt already-formed Bid heterodimers with the mutidomain survival proteins; when heterodimers on the Biacore chip are challenged with doses of TW-37 up to 10000 nM. Consequently, the only opportunity the drug might have to perturb the Bcl-2 system is to locate the hydrophobic groove before Bid. Thus, a new picture emerges of drug action in which the dynamics of pre-apoptotic protein-protein interactions creates a temporal window for drug binding to the hydrophobic groove. In this light, attempts to develop small molecule inhibitors of protein-protein interaction (
Current Topics in Medicinal Chem 7(10), 922–927 (2007)
) should incorporate surface plasmon resonance measurements of ka and kd into preclinical drug discovery programs in addition to KD estimates of drug binding to target.

Figure
View largeDownload slide

Figure

Topics:
bcl2 gene, bcl-xl protein, binding (molecular function), hydrophobicity, molar tooth, small molecule, surface plasmon resonance, mechlorethamine, peptides, abt-737

Author notes

Disclosure:Ownership Interests: Dr. Shomeng Wang (co-author) is one of the founding members of Ascenta Therapeutics, Inc., of San Diego, CA, a company which features a Bcl-2 mimetic compound (AT-101) as a pipeline target, a compound unrelated in chemical structure to TW-37; Dr Wang is listed as Inventor on US patent application 2006/0084647 A1, “Small molecular inhibitors of anti-apoptotic Bcl-2 family members and the uses thereof”, which described TW-37 (p 51 and Figs. 4, 5). Research Funding: Leukemia and Lymphoma Society grant 6028-07 to R. M. Mohammad.

2007, The American Society of Hematology
2007
Sign in via your Institution