The emergence of antiestrogen resistance in breast cancer is an important clinical phenomenon affecting long-term survival in this disease. Identifying factors that convey cell survival in this setting may guide improvements in treatment. Estrogen (E2) can induce apoptosis in breast cancer cells that have been selected for survival after E2 deprivation for long periods (MCF-7:5C cells), but the mechanisms underlying E2-induced stress in this setting have not been elucidated. Here, we report that the c-Src kinase functions as a key adapter protein for the estrogen receptor (ER, ESR1) in its activation of stress responses induced by E2 in MCF-7:5C cells. E2 elevated phosphorylation of c-Src which was blocked by 4-hydroxytamoxifen (4-OHT), suggesting that E2 activated c-Src through the ER. We found that E2 activated the sensors of the unfolded protein response (UPR), IRE1α (ERN1) and PERK kinase (EIF2AK3), the latter of which phosphorylates eukaryotic translation initiation factor-2α (eIF2α). E2 also dramatically increased reactive oxygen species (ROS) production and up-regulated expression of heme oxygenase HO-1 (HMOX1), an indicator of oxidative stress, along with the central energy sensor kinase AMPK (PRKAA2). Pharmacological or RNAi-mediated inhibition of c-Src abolished the phosphorylation of eIF2α and AMPK, blocked E2-induced ROS production, and inhibited E2-induced apoptosis. Together, our results establish that c-Src kinase mediates stresses generated by E2 in long-term E2-deprived cells that trigger apoptosis. This work offers a mechanistic rationale for a new approach in the treatment of endocrine-resistant breast cancer.

Stress responses are critical for estrogen (E2)-induced apoptosis in E2-deprived breast cancer cells. Nuclear factor-kappa B (NF-κB) is an important therapeutic target to prevent stress responses in chronic inflammatory diseases including cancer. However, whether E2 activates NF-κB to participate in stress-associated apoptosis in E2-deprived breast cancer cells is unknown. Here, we demonstrated that E2 differentially modulates NF-κB activity according to treatment time. E2 initially has significant potential to suppress NF-κB activation; it completely blocks tumor necrosis factor alpha (TNFα)-induced activation of NF-κB. We found that E2 preferentially and constantly enhances the expression of the adipogenic transcription factor CCAAT/enhancer binding protein beta (C/EBPβ), which is responsible for the suppression of NF-κB activation by E2 in MCF-7:5C cells. Interestingly, NF-κB p65 DNA-binding activity is increased when E2 is administered for 48 h, leading to the induction of TNFα and associated apoptosis. Blocking the nuclear translocation of NF-κB can completely prevent the induction of TNFα and apoptosis induced by E2. Further examination revealed that protein kinase RNA-like endoplasmic reticulum kinase (PERK), a stress sensor of unfolded protein response (UPR), plays an essential role in the late activation of NF-κB by E2. This modulation between PERK and NF-κB is mainly mediated by a stress responsive transcription factor, transducer and activator of transcription 3 (STAT3), independently of the classic canonical IκBα signaling pathway. Thus, inhibition of PERK kinase activity completely blocks the DNA binding of both STAT3 and NF-κB, thereby preventing induction of NF-κB-dependent genes and E2-induced apoptosis. All of these findings suggest that PERK is a key regulator to convey stress signals from the endoplasmic reticulum to the nucleus and illustrate a crucial role for the novel PERK/STAT3/NF-κB/TNFα axis in E2-induced apoptosis in E2-deprived breast cancer cells.

Purpose Our publications demonstrate that physiological concentrations of estrogen (E2) induce endoplasmic reticulum and oxidative stress which finally result in apoptosis in E2-deprived breast cancer cells, MCF-7:5C. c-Src is involved in the process of E2-induced stress. To mimic the clinical administration of c-Src inhibitors, we treated cells with either E2, a c-Src inhibitor PP2, or the combination for 8 weeks to further explore the apoptotic potential of the c-Src inhibitor and E2 on MCF-7:5C cells. Methods Protein levels of receptors and signaling pathways were examined by immunoblotting. Expression of mRNA was detected through real-time PCR. Cell cycles were analyzed by flow cytometry. Results Long-term treatment with PP2 alone or E2 alone decreased cell growth. In contrast, a combination of PP2 and E2 blocked apoptosis and the resulting cell line (MCF-7:PF) was unique, as they grew vigorously in culture with physiological levels of E2, which could be blocked by the pure antiestrogen ICI182,780. One major change was that PP2 collaborated with E2 to increase the level of insulin-like growth factor-1 receptor beta (IGF-1Rβ). Blockade of IGF-1Rβ completely abolished E2-stimulated growth in MCF-7:PF cells. Furthermore, combination treatment up-regulated transcription factors, Twist1 and Snail, and repressed E-cadherin expression which made MCF-7:PF cells display a characteristic phenotype of epithelial-mesenchymal transition (EMT). Conclusions These data illustrate the role of the c-Src inhibitor to block E2-induced apoptosis and enhance E2-stimulated growth. Caution must be exercised when considering c-Src inhibitors in clinical trials following the development of acquired resistance to aromatase inhibitors, especially in the presence of the patient’s own estrogen.

Purpose: Estrogen (E2)-stimulated growth re-emerges after a c-Src inhibitor blocking E2-induced apoptosis. A resulting cell line, MCF-7:PF, is selected with features of functional estrogen receptor (ER) and over-expression of insulin-like growth factor-1 receptor beta (IGF-1Rβ). We addressed the question of whether the selective ER modulator (SERM), 4-hydroxytamoxifen (4-OHT) or other SERMs could target ER to prevent E2-stimulated growth in MCF-7:PF cells. Methods: Protein levels of receptors and signaling pathways were examined by immunoblotting. Expression of mRNA was measured through real-time RT-PCR. Recruitment of ER or nuclear receptor coactivator 3 (SRC3) to the promoter of ER-target gene was detected by chromatin-immunoprecipitation (ChIP). Results: 4-OHT and other SERMs stimulated cell growth in an ER-dependent manner. However, unlike E2, 4-OHT suppressed classical ER-target genes as does the pure antiestrogen ICI 182,780 (ICI). ChIP assay indicated that 4-OHT did not recruit ER or SRC3 to the promoter of ER-target gene, pS2. Paradoxically, 4-OHT reduced total IGF-1Rβ but increased phosphorylation of IGF-1Rβ. Mechanistic studies revealed that 4-OHT functioned as an agonist to enhance the non-genomic activity of ER and activate focal adhesion molecules to further increase phosphorylation of IGF-1Rβ. Disruption of membrane-associated signaling, IGF-1R and focal adhesion kinase (FAK), completely abolished 4-OHT-stimulated cell growth. Conclusions: This study is the first to recapitulate a cellular model in vitro of acquired tamoxifen resistance developed in athymic mice in vivo. Importantly, it provides a rationale that membrane-associated pathways may be valuable therapeutic targets for tamoxifen resistant patients in clinic.

Estrogen (E2) exerts a dual function on E2-deprived breast cancer cells, with both initial proliferation and subsequent induction of stress responses to causes apoptosis. However, the mechanism by which E2 integrally regulates cell growth or apoptosis associated pathways remains to be elucidated. Here, E2 deprivation results in many alterations in stress-responsive pathways. For instance, E2-deprived breast cancer cells had higher basal levels of stress-activated protein kinase, c-Jun N-terminal kinase (JNK), compared with wild-type MCF-7 cells. E2 treatment further constitutively activated JNK after 24 hours. However, inhibition of JNK (SP600125) was unable to abolish E2-induced apoptosis, whereas SP600125 alone arrested cells at the G2-phase of the cell cycle and increased apoptosis. Further examination showed that inhibition of JNK increased gene expression of tumor necrosis alpha (TNFα) and did not effectively attenuate expression of apoptosis-related genes induced by E2. A notable finding was that E2 regulated both JNK and Akt as the downstream signals of insulin-like growth factor-1 receptor (IGF1R)/phosphoinositide 3-kinase (PI3K), but with distinctive modulation patterns: JNK was constitutively activated, whereas Akt and Akt-associated proteins, such as PTEN and mTOR, were selectively degraded. Endoplasmic reticulum-associated degradation (ERAD) was involved in the selective protein degradation. These findings highlight a novel IGF-1R/PI3K/JNK axis that plays a proliferative role during the prelude to E2-induced apoptosis and that the endoplasmic reticulum is a key regulatory site to decide cell fate after E2 treatment. IMPLICATIONS This study provides a new rationale for further exploration of E2-induced apoptosis to improve clinical benefit.