Improved tools to evaluate the sensitivity of cancer cells to CDK4 inhibitors

- Introduction:The complexes between D-type cyclins (CCND1-3) and cyclin-dependent kinases CDK4/6 are key regulators of the cell cycle entry, by phosphorylating the onco-suppressor retinoblastoma protein (pRb). Their deregulated activity is a major oncogenic event in most types of tumours. Our group identified the T172-phosphorylation of CDK4 as the determining step for its activation and have described critical regulations of this event in various cell systems (1,2,3). Due to unprecedented improvement of the progression-free survival (4), FDA approved Palbociclib, the first specific inhibitor of CDK4/6, for the treatment of advanced estrogen receptor-positive breast tumours in combination with letrozole. In order to predict tumour’s response to Palbociclib and develop associated biomarkers, cancer cell lines have been used to evaluate the impact of the drug on cell growth (5,6,7). However, divergent IC50s have been described probably due to the different assays used to quantify the drug effects.- Aims:To develop and validate a drug response assay that more directly assesses the cell cycle responsiveness to CDK4/6 inhibition, and compare this to the expression of cell cycle effectors.- Methods:We have studied a representative subset of 20 breast cancer cell lines. We adapted the 5-Bromo-2’-deoxyuridine (BrDU) proliferation method, which labels neo-synthesized DNA, to a 96-wells plate format. Drug impact was also determined in parallel with the sulforhodamine B (SRB) (which measures protein content) and the MTT assays (which reflects a mitochondrial enzymatic activity). The protein levels of total and phosphorylated pRb, CDK4, Cyclin D1, Cyclin D3, Cyclin E and p16 were also evaluated.- Results and discussion:The BrDU proliferation assay quantifies the proportion of cells actively synthesizing DNA, a readout directly associated with cell cycle progression. The proportion of labeled cells after a 1h BrDU-pulse in control asynchronous growth varied from 20-50%. In 13 of the 20 cell lines, Palbociclib reduced this proportion in a dose-dependent way. In all except one, the cell lines responses to the drug were comparable to Finn’s data (5), although with lower IC50 ranges than the ones reported. Furthermore, our IC50 values are closer to the concentration required to block the CDK4 activity in vitro (15 nM). However, sensitivities measured by the three techniques were distinct. The impact of CDK4/6 inhibition was less pronounced when Palbociclib action was assayed by SRB or MTT assays. Using the BrDU assay, all luminal and most ERBB2-amplified cell lines were sensitive to the drug while, interestingly, 33% basal-like cells were also sensitive. Phosphorylated pRb was absent in four insensitive cell lines but detected in three other resistant ones. p16 was mostly absent in sensitive cell lines while it was highly expressed in completely resistant cells. Overall, none of these G1 phase regulators, used alone, could serve as a qualitative biomarker of CDK4 inhibitor sensitivity.- Conclusions: The BrDU labeling is the best method to evaluate CDK4 inhibitors sensitivity. The breast cancer cell line panel is a useful tool to identify molecular determinants of CDK4 inhibition and proteomic/genomic characterization is on going to define biomarkers of CDK4 inhibitor sensitivity.- References: 1. Bockstaele, L. et al. (2006). Mol. Cell Biol. 26, 5070-5085.2. Paternot, S. et al. (2010). Cell Cycle 9, 689-699.3. Bisteau, X. et al. (2013) PLoS Genet. 9(5):e1003546. doi: 10.1371/journal.pgen.1003546.4. Finn, R.S. et al. (2015). Lancet Oncol.16, 25-35.5. Finn, R.S. et al. (2009). Breast Cancer Res. 11, R77.6. Garnett, M.J. et al. (2012). Nature. 483, 570–575.7. Barretina J. et al. (2012) Nature. 483, 603–607.