Prolonged treatment of HER2 positive breast cancer cells with tyrosine kinase inhibitors (TKIs) leads to the emergence of acquired resistance. However, the effects of continuous TKI exposure on cell fate, and the steps leading to the acquisition of a resistant phenotype are poorly understood. To explore this, we exposed five HER2 positive cells lines to HER2 targeted therapies for periods of up to 4 weeks and examined senescence associated β-galactosidase (SA-β-gal) activity together with additional markers of senescence. We found that lapatinib treatment resulted in phenotypic alterations consistent with a senescent phenotype and strong SA-β-gal activity in HER2-positive cell lines. Lapatinib-induced senescence was associated with elevated levels of p15 and p27 but was not dependent on the expression of p16 or p21. Restoring wild type p53 activity either by transfection or by treatment with APR-246, a molecule which reactivates mutant p53, blocked lapatinib-induced senescence and caused increased cell death. In contrast to lapatinib, SA-β-gal activity was not induced by exposing the cells to trastuzumab as a single agent but co-administration of lapatinib and trastuzumab induced senescence, as did treatment of the cells with the irreversible HER2 TKIs neratinib and afatinib. Neratinib- and afatinib-induced senescence was not reversed by removing the drug whereas lapatinib-induced senescence was reversible. In summary, therapy-induced senescence represents a novel mechanism of action of HER2 targeting agents and may be a potential pathway for the emergence of resistance.

CDK7, a transcriptional cyclin-dependent kinase, is emerging as a novel cancer target. Triple-negative breast cancers (TNBC) but not estrogen receptor-positive (ER+) breast cancers have been reported to be uniquely sensitive to the CDK7 inhibitor THZ1 due to the inhibition of a cluster of TNBC-specific genes. However, bioinformatic analysis indicates that CDK7 RNA expression is associated with negative prognosis in all the major subtypes of breast cancer. To further elucidate the effects of CDK7 inhibition in breast cancer, we profiled a panel of cell lines representing different breast cancer subtypes. THZ1 inhibited cell growth in all subtypes (TNBC, HER2+, ER+, and HER2+/ER+) with no apparent subtype selectivity. THZ1 inhibited CDK7 activity and induced G1 arrest and apoptosis in all the tested cell lines, but THZ1 sensitivity did not correlate with CDK7 inhibition or CDK7 expression levels. THZ1 sensitivity across the cell line panel did not correlate with TNBC-specific gene expression but it was found to correlate with the differential inhibition of three genes: CDKN1B, MYC and transcriptional coregulator CITED2. Response to THZ1 also correlated with basal CITED2 protein expression, a potential marker of CDK7 inhibitor sensitivity. Furthermore, all of the THZ1-inhibited genes examined were inducible by EGF but THZ1 prevented this induction. THZ1 had synergistic or additive effects when combined with the EGFR inhibitor erlotinib, with no outward selectivity for a particular subtype of breast cancer. These results suggest a potential broad utility for CDK7 inhibitors in breast cancer therapy and the potential for combining CDK7 and EGFR inhibitors.