Obatoclax (GX15-070) Is a Potent Antagonist of Constitutive Mcl-1/Bak Interactions in Intact Mitochondrial Membrane and Synergizes with Bortezomib in Mantle Cell Lymphoma.

Obatoclax (GX15-070), a synthetic small molecule pan-Bcl-2 family inhibitor, is currently under investigation in multiple Phase II trials directed at hematologic and solid tumor malignancies. Inhibition of pro-survival Bcl-2 members antagonizes their ability to neutralize the death effectors Bax and Bak, which are required for obatoclax-induced cellular toxicity. The hydrophobic character of obatoclax is consistent with its predicted occupation of a hydrophobic pocket in the lower region of the Bcl-2 BH3 binding groove. Based on comparisons with the published results of 10 known small molecule inhibitors, obatoclax is computationally predicted to have a significant binding affinity for Bcl-2 (Ki = 220 nanoM). A recent report (Zhai et al., Cell Death Differ. 2006 13: 1419-21), describing the interactions of synthetic BH3 peptides with Bcl-2 family proteins by fluorescence polarization in aqueous medium (PBS), indicated that obatoclax selectively antagonizes such interactions for all pro-survival members tested (Bcl-2, Bcl-Xl, Mcl-1, Bcl-w, A1, and Bcl-b), but did not affect control protein interactions. IC50s for obatoclax ranged from 1 to 7 microM. Under the aqueous conditions of this assay, however, obatoclax is largely insoluble and therefore this report likely underestimates the potency of the compound. In situ, Bcl-2 proteins reside integrated in the hydrophobic membranes of mitochondria and ER. We took advantage of the fact that Mcl-1 constitutively interacts with Bak in the mitochondrial outer membrane, thus presenting the opportunity to investigate the potency of obatoclax in situ on a native Bcl-2 protein-protein interaction. Employing chemical crosslinking to detect Mcl-1/Bak dimers in intact mitochondria, obatoclax was found to disrupt these interactions with an IC50 less than 10 nanoM, similar to that observed for inhibition by synthetic BH3 peptides corresponding to Noxa and Bim. In contrast to Bim peptide, which induced release of cytochrome c from isolated mitochondria, obatoclax did not, indicating that it is a sensitizer rather than an activator of the intrinsic mitochondrial apoptosis program. Bax or Bak are toxic to yeast cells, but are restrained by either Mcl-1, Bcl-2, Bcl-Xl, or Bcl-w. In each case, obatoclax treatment re-instated Bax or Bak toxicity, confirming the pan-inhibitor properties of the compound. Among Bcl-2 proteins, Mcl-1 rapidly turns over at steady state and is subject to ubiquitin-mediated degradation by the 26S proteasome. Not surprisingly, the proteasome inhibitor bortezomib significantly enhances Mcl-1 protein levels in various settings, including mantle cell lymphoma (MCL) cells. Bortezomib also induced Noxa upregulation in MCL, and obatoclax and Noxa appeared to cooperate to displace Bak from Mcl-1. As predicted, obatoclax and bortezomib exhibited synergistic toxicity in MCL lines in vitro and in cells derived from MCL patients ex vivo. Collectively, therefore, a biochemical rationale has been established to interrogate obatoclax-bortezomib combinations in clinical trials directed at mantle cell lymphoma.