Induction of sequential G1 arrest and synchronous S phase entry by reversible CDK4/CDK6 inhibition sensitizes myeloma cells for cytotoxic killing through loss of IRF-4.

Abstract 299

Dysregulation of the cell cycle is a hallmark of cancer. However, targeting the cell cycle in cancer therapy has only been modestly successful since broad-spectrum cyclindependent kinase (CDK) inhibitors lack specificity and are highly toxic. The critical importance of controlling CDK4/CDK6 in cancer treatment is further exemplified by recent evidence of prominent CDK4/CDK6 dysregulation in human cancers, including breast cancer, metastatic lung adenocarcinoma, glioblastoma, mantle cell lymphoma and multiple myeloma (MM).

To advance mechanism-based targeting of the cell cycle in cancer, we have developed a novel strategy that both inhibits cell cycle progression and enhances cytotoxic killing in tumor cells using PD 0332991(PD), the only known CDK4/CDK6-specific inhibitor that is also reversible, potent and orally bioavailable. We demonstrated by BrdU pulse-labeling that inhibition of CDK4/CDK6 with PD in primary bone marrow (BM) myeloma cells and human myeloma cell lines (HMCL) (IC50 60nM) leads to a complete early G1 arrest in the absence of apoptosis and upon release of the G1 block, synchronous cell cycle progression to S phase. Furthermore, prolonged early G1 arrest enhances cytotoxic killing of MM cells by diverse clinically relevant drugs at low dose, including bortezomib, carfilzomib (PR-171) and dexamethasone, and this is dramatically augmented during synchronous S phase entry. The enhancement of cytotoxic killing in either G1 arrest or synchronous S phase entry is sustained in the presence of BM stromal cells. This killing is caspase-dependent and triggered by the loss of mitochondrial outer membrane potential and activation of the intrinsic apoptosis pathway. Time course studies of cell cycle-specific gene expression by expression profiling, quantitative real time RT-PCR and immunoblotting further revealed that the expression of IRF-4, essential for normal plasma cell differentiation and myeloma cell survival, is strictly cell cycle-dependent: elevated in G1 and markedly declined in S phase. The IRF-4 protein is also markedly reduced (50%) by bortezomib treatment, resulting in a combined 5-fold reduction in S phase. This suggests that differential enhancement of cytotoxic killing in G1 arrest and S phase is mediated by cell cycle-dependent IRF-4 expression. Indeed, shRNA interference confirms that by antagonizing mitochondrial depolarization, IRF-4 is required to protect myeloma cells from cell cycle-dependent enhancement of bortezomib killing.

By timely administration and discontinuation of PD treatment, we have further demonstrated in a human MM 1.S. xenograft myeloma model that it is feasible to induce sequential G1 arrest and synchronous S phase in vivo. This leads to synergistic tumor suppression through amplification of bortezomib killing of myeloma cells, but not normal BM cells. As PD is orally bio-available, specific and low in toxicity, our novel strategy has been implemented in the first phase I/II multi-center clinical trial targeting CDK4/CDK6 with PD in combination with bortezomib and dexamethasone in MM.

Preliminary bone marrow immunohistochemistry demonstrates PD preferentially and completely inhibits CDK4/CDK6-specific phosphorylaton of Rb and DNA replication in tumor cells, but not other bone marrow cells in all patients. One patient achieved VGPR (12.5%) while 1 patient each achieved MR and SD respectively for an ORR 25% (Niesvizky et al, submitted). Collectively, our preclinical and clinical data indicate, for the fist time, that selective targeting of CDK4/CDK6 in combination therapy is a promising mechanism-based therapy for MM and potentially other cancers.