Interaction between Glycogen Synthase Kinase-3 and Estrogen Receptor-alpha in ligand-dependent activation of the receptor

Abstract

Glycogen synthase kinase-3 (GSK-3), a serine/threonine kinase with docking properties, regulates numerous cellular processes. Two isoforms, GSK-3alpha and GSK-3beta have been described. In vivo, GSK-3beta is the major isoform and plays a key role in the regulation of transcription factors including steroid receptors. The aim of the present work, mainly performed on the estrogen receptor-alpha (ERalpha)-positive MCF-7 human breast cancer cell line, was to unravel the role of GSK-3 regarding ERalpha function. After silencing of GSK-3alpha and/or GSK-3beta isoforms using specific siRNA sequences, increased proteasomal degradation of ERalpha was observed. The use of the proteasome inhibitor MG132 restored ERalpha protein levels in GSK-3 silenced cells, showing that GSK-3 stabilizes ERalpha and protects it from proteasomal degradation. In another approach, specific silencing of the endogenous GSK-3beta of MCF-7 cells using microRNA constructs was accompanied by down-regulation of ERalpha protein content. In these cells, ERalpha protein was rescued after overexpression of wild-type or kinase-inactive xenopus GSK-3beta, which suggests that the docking properties of GSK-3 and not the kinase activity are important for ERalpha stabilization. Then, we found that 17beta-estradiol (E2) -treatment resulted in rapid phosphorylation and consequent inactivation of cytoplasmic GSK-3. This GSK-3 phosphorylation may lead to ERalpha release from the GSK-3/ERalpha complex and ERalpha translocation into the nucleus, where it is phosphorylated at Ser-118 leading to its full activation. Upon E2 stimulation, treatment of the cells with the GSK-3 inhibitor LiCl resulted in a decrease of ERalpha phosphorylation at Ser-118. This decrease was confirmed upon silencing of GSK-3 in the nucleus and show that a nuclear active pool of GSK-3 is required for E2-induced phosphorylation of ERalpha at Ser-118. As a consequence, in GSK-3 silenced cells, E2-induced ERalpha transcriptional activity, studied by ERE-dependent luciferase reporter assays and by measuring transcription of the ERalpha-dependent target genes, pS2 and progesterone receptor, by quantitative real-time PCR, was significantly reduced. In GSK-3 silenced cells, neither Ser-118 phosphorylation nor luciferase activity was restored by use of MG132. Furthermore, overexpression of human GSK-3beta wild-type and mutants inactive towards primed substrate of the kinase in MCF-7 cells stably transfected with an ERE-controlled luciferase reporter confirmed that GSK-3 triggers E2-induced ERalpha activation and suggests that ERalpha is a non-primed substrate of GSK-3 kinase. Taken together, this newly signalling pathway depicted a dual function of GSK-3 regarding ERalpha, GSK-3 stabilises ERalpha in the cytoplasm of unstimulated cells and phosphorylates/activates the receptor in the nucleus upon E2 treatment. This permits the conclusion that GSK-3 represents a link between the rapid cytoplasmic non-genomic and the nuclear genomic actions of E2-liganded ERalpha. Finally, ERalpha signalling pathway plays a crucial role in breast cancer initiation and progression. Therefore, the regulation of ERalpha function and activity by GSK-3 may have an impact on breast cancer progression. Preliminary data from GSK-3beta immunostaining of formalin-fixed human tissue sections suggests a tendency toward an increase of GSK-3beta expression in grade 3 tumors in comparison with grade 1/2 tumors.