Chemical genetics approach to restoring p27Kip1 reveals novel compounds with antiproliferative activity in prostate cancer cells

BACKGROUND: The cyclin-dependent kinase (CDK) inhibitor p27(Kip1 )is downregulated in a majority of human cancers due to ectopic proteolysis by the ubiquitin-proteasome pathway. The expression of p27 is subject to multiple mechanisms of control involving several transcription factors, kinase pathways and at least three different ubiquitin ligases (SCF(SKP2), KPC, Pirh2), which regulate p27 transcription, translation, protein stability and subcellular localization. Using a chemical genetics approach, we have asked whether this control network can be modulated by small molecules such that p27 protein expression is restored in cancer cells. RESULTS: We developed a cell-based assay for measuring the levels of endogenous nuclear p27 in a high throughput screening format employing LNCaP prostate cancer cells engineered to overexpress SKP2. The assay platform was optimized to Z' factors of 0.48 - 0.6 and piloted by screening a total of 7368 chemical compounds. During the course of this work, we discovered two small molecules of previously unknown biological activity, SMIP001 and SMIP004, which increase the nuclear level of p27 at low micromolar concentrations. SMIPs (small molecule inhibitors of p27 depletion) also upregulate p21(Cip1), inhibit cellular CDK2 activity, induce G1 delay, inhibit colony formation in soft agar and exhibit preferential cytotoxicity in LNCaP cells relative to normal human fibroblasts. Unlike SMIP001, SMIP004 was found to downregulate SKP2 and to stabilize p27, although neither SMIP is a proteasome inhibitor. Whereas the screening endpoint - nuclear p27 - was robustly modulated by the compounds, SMIP-mediated cell cycle arrest and apoptosis were not strictly dependent on p27 and p21 - a finding that is explained by parallel inhibitory effects of SMIPs on positive cell cycle regulators, including cyclins E and A, and CDK4. CONCLUSIONS: Our data provide proof-of-principle that the screening platform we developed, using endogenous nuclear p27 as an endpoint, presents an effective means of identifying bioactive molecules with cancer selective antiproliferative activity. This approach, when applied to larger and more diverse sets of compounds with refined drug-like properties, bears the potential of revealing both unknown cellular pathways globally impinging on p27 and novel leads for chemotherapeutics targeting a prominent molecular defect of human cancers.