Sensitivity to erlotinib (Tarceva) depends on cyclin-dependent kinase 2 activity

1670 The epidermal growth factor receptor (EGFR, HER1) is highly expressed in a variety of solid tumors. Activation of EGFR signaling pathways in tumor cells has been linked with increased proliferation, angiogenesis and metastasis and decreased apoptosis, and hence EGFR has been proposed as a target for anticancer therapy. Erlotinib, an orally available, quinazoline-based agent, inhibits EGFR tyrosine kinase (TK) by competing with ATP, inhibiting phosphorylation of this kinase. Some reports indicate that EGFR expression level does not predict sensitivity to EGFR-TK inhibitors such as erlotinib (R. Bailey et al, Lung Cancer 2003; Sirotnak FM et al, Clin Cancer Res 2000; Baselga J, J Clin Oncol 2004), while other reports suggest that somatic mutations in the TK domain of EGFR reflect the sensitivity of lung cancer cells to EGFR-TK inhibitors (Lynch TJ, et al, N Engl J Med 2004; Paez JG, et al, Science 2004). Thus far, such mutations have been found only in non-small cell lung cancer, and other factors for predicting response to EGFR-TK inhibitors are yet to be defined. We assessed whether cyclin-dependent kinase 2 (CDK2), a final downstream target in EGFR signaling pathways, is involved in sensitivity to erlotinib in 1 epidermoid carcinoma cell line (A431) and 10 breast cancer cell lines (MDA-231, -361, -435, -453, -468; SKBr-3, BT-20, BT-474, T47D, and MCF-7). MTT assays showed that A431 and SKBr3 cells were quite sensitive to erlotinib (A431, IC50 = 1.531 uM, IC25 = 4.748 uM; SKBr-3, IC50 = 3.980 uM, IC25 = 10.628 uM) and BT-474 and T47D cells were moderately so (BT-474, IC50 = 5.005uM, IC25 > 20uM; T47D, IC50 = 9.803uM, IC25 > 20uM). Flow cytometry confirmed that erlotinib-sensitive cell lines showed G1 arrest and increases in sub-G1 cells. EGFR overexpression (detected in A431, MDA-468, and BT-20 cells by western blotting) was not associated with sensitivity to erlotinib, even though EGFR phosphorylation was inhibited in both sensitive (A431) and resistant cell lines (MDA-468, BT-20). However, loss of CDK2 activity after erlotinib treatment strongly correlated with sensitivity to erlotinib. Restoration of CDK2 activity by transfecting cells with an adenovirus containing wild-type CDK2 partially restored proliferation in A431 cells and SKBr3 cells treated with erlotinib (A431, IC50 = 4.630uM, IC25 > 20uM 4.748uM; SKBr3, IC50 > 20uM). These results indicate that sensitivity to erlotinib depends in part on CDK2 activity; that suppression of CDK2 is critical to the activity of erlotinib regardless of erlotinib’s upstream molecular effects; and that measurement of CDK2 activity after a brief exposure to erlotinib may prove to be a predictive marker of sensitivity to erlotinib.