Epithelial-to-Mesenchymal Transition Activates PERK–eIF2α and Sensitizes Cells to Endoplasmic Reticulum Stress

Yuxiong Feng(Broad Institute), Ethan S. Sokol(Broad Institute), Catherine Del Vecchio Fitz(Broad Institute), Sandhya Sanduja(Broad Institute), Jasper H.L. Claessen(Broad Institute), Theresa A. Proia(Broad Institute), Dexter X. Jin(Broad Institute), Ferenc Reinhardt(Broad Institute), Hidde L. Ploegh(Broad Institute), Qiu Wang(Broad Institute), Piyush B. Gupta(Broad Institute)
Cancer Discovery
April 5, 2014
Cited by 324

Abstract

UNLABELLED: Epithelial-to-mesenchymal transition (EMT) promotes both tumor progression and drug resistance, yet few vulnerabilities of this state have been identified. Using selective small molecules as cellular probes, we show that induction of EMT greatly sensitizes cells to agents that perturb endoplasmic reticulum (ER) function. This sensitivity to ER perturbations is caused by the synthesis and secretion of large quantities of extracellular matrix (ECM) proteins by EMT cells. Consistent with their increased secretory output, EMT cells display a branched ER morphology and constitutively activate the PERK-eIF2α axis of the unfolded protein response (UPR). Protein kinase RNA-like ER kinase (PERK) activation is also required for EMT cells to invade and metastasize. In human tumor tissues, EMT gene expression correlates strongly with both ECM and PERK-eIF2α genes, but not with other branches of the UPR. Taken together, our findings identify a novel vulnerability of EMT cells, and demonstrate that the PERK branch of the UPR is required for their malignancy. SIGNIFICANCE: EMT drives tumor metastasis and drug resistance, highlighting the need for therapies that target this malignant subpopulation. Our findings identify a previously unrecognized vulnerability of cancer cells that have undergone an EMT: sensitivity to ER stress. We also find that PERK-eIF2α signaling, which is required to maintain ER homeostasis, is also indispensable for EMT cells to invade and metastasize.


Related Papers

No related papers found

Powered by citation graph analysis