Widespread PERK-dependent repression of ER targets in response to ER stress

Nir Gonen,Niv Sabath,Christopher B Burge,Reut Shalgi

doi:10.1038/s41598-019-38705-5

Nir Gonen, Niv Sabath + Show 2 more

Open Access

https://doi.org/10.1038/s41598-019-38705-5

Copy DOI

Abstract

The UPR (Unfolded Protein Response) is a well-orchestrated response to ER protein folding and processing overload, integrating both transcriptional and translational outputs. Its three arms in mammalian cells, the PERK translational response arm, together with the ATF6 and IRE1-XBP1-mediated transcriptional arms, have been thoroughly investigated. Using ribosome footprint profiling, we performed a deep characterization of gene expression programs involved in the early and late ER stress responses, within WT or PERK −/− Mouse Embryonic Fibroblasts (MEFs). We found that both repression and activation gene expression programs, affecting hundreds of genes, are significantly hampered in the absence of PERK. Specifically, PERK −/− cells do not show global translational inhibition, nor do they specifically activate early gene expression programs upon short exposure to ER stress. Furthermore, while PERK −/− cells do activate/repress late ER-stress response genes, the response is substantially weaker. Importantly, we highlight a widespread PERK-dependent repression program, consisting of ER targeted proteins, including transmembrane proteins, glycoproteins, and proteins with disulfide bonds. This phenomenon occurs in various different cell types, and has a major translational regulatory component. Moreover, we revealed a novel interplay between PERK and the XBP1-ATF6 arms of the UPR, whereby PERK attenuates the expression of a specific subset of XBP1-ATF6 targets, further illuminating the complexity of the integrated ER stress response.

Highlights

Protein homeostasis is one of the hallmarks of cellular viability and a well-known factor in health and disease
protein kinase RNA-like endoplasmic reticulum kinase (PERK) −/− Mouse Embryonic Fibroblasts (MEFs) showed no change in their global overall translation levels, and only at the 8 h timepoint a slight repression of translation was observed (Fig. 1B)
Even at 8 h, the degree of translational repression observed in PERK −/− MEFs did not resemble nearly any of the treatments of WT MEFs

Summary

Introduction

Protein homeostasis is one of the hallmarks of cellular viability and a well-known factor in health and disease. ER-stress has long been known to elicit a complex cellular program, termed the Unfolded Protein Response (UPR), which has evolved to allow cells to cope with dynamic changes in the protein folding and processing demands in the ER2,4,5. Previous genome-wide studies have used mRNA expression profiling to define a transcriptional response following a 6 h ER-stress in PERK −/− and ATF4 −/− cells[11,12]. These experiments have shown PERK-dependent metabolic changes enabling the maintenance of redox potential under ER-stress[12]. While PERK is known to elicit an eIF2α phosphorylation-mediated global translational repression in response to ER stress, its role in controlling the translation of specific gene expression programs still remains elusive. We used Thapsigargin (Tg), a SERCA inhibitor, for short (1 h or 2 h) or long (5 h or 8 h) treatments, to examine the gene expression programs that govern either early or late responses to ER stress

Methods

Results

Conclusion