The Sensor IRE1 Couples Stress Detection to Protein Synthesis

Abstract

Endoplasmic reticulum (ER) stress results in accumulation of misfolded proteins and profoundly regulates cellular protein synthesis. The regulation includes both inhibition of translation of most proteins and upregulation of synthesis of a select group of stress-related polypeptides. This differential regulation has long been attributed to the kinase PERK, but recently the related stress sensor, inositol requiring enzyme 1 (IRE1), was suggested to be linked to translation by virtue of its physical association with both ribosomes1 and the Sec61 translocon2,3. ER stress causes the quiescent IRE1 to dissociate from the inhibitor BiP, self-phosphorylate, oligomerize (an event also called clustering) and activate its RNase activity. Consequently, active IRE1 activates the critical transcription factor XBP1, and also reduces the expression of select other transcripts, to cope with increased ER stress. We previously showed that IRE1 clustering is a response to stress4, but is not related to its RNase activity, raising the question – what is the function of IRE1 clustering? In examining the requirements for IRE clustering we discovered that ongoing protein synthesis and translocation are necessary for both the formation of IRE1 clusters upon ER stress and their maintenance. IRE1 clusters do not develop when cells are treated with either cycloheximide or lactimidomycin, which inhibit nascent chain elongation and ‘freeze’ membrane-bound ribosomes on the ER membrane. If added after clusters were allowed to form under ER stress, both inhibitors caused rapid disappearance of the pre-formed clusters. IRE1 clustering is similarly sensitive to Eeyarestatin I (EEY), an inhibitor of the Sec61 translocon. When translocation is inhibited, IRE1 is dissociated from Sec61 and its interaction with BiP is increased. Concomitantly, more IRE1 is found in high molecular weight species, which perhaps represent ubiquitinylated species. Paradoxically, EEY also induces IRE1 activation (XBP1 splicing). To explain the paradox, we hypothesize the following model: in unstressed conditions, IRE1 binds the translocon and is kept inactive by the interaction with the chaperone BiP. Upon ER stress, different scenarios with distinct IRE1 populations can appear at the same time: 1) IRE1 is released from BiP and clusters on a subset of translocons that are engaged in synthesis of stress-related polypeptides; 2) translocons not engaged in active translocation release IRE1 which is then kept inactive by increased interaction with BiP; 3) some activated IRE1 dimerizes (appearing as non-clustered) and binds the ribosomes that recruit XBP1 mRNA, facilitating its splicing. Acosta-Alvear D et al. eLife 2018 Dec 24;7:e43036. Plumb R et al. eLife 2015 2015 May 20;4:e07426. Sundaram A et al. eLife 2017 May 15;6:e27187. Ricci, D. et al. The FASEB Journal 2019; Sep;33(9):9811-9827.