A J-Protein Co-chaperone Recruits BiP to Monomerize IRE1 and Repress the Unfolded Protein Response

摘要

class="head no_bottom_margin" id="sec1title">IntroductionSecretory and transmembrane proteins enter the endoplasmic reticulum (ER) as unfolded polypeptides and emerge as folded and processed proteins. The protein folding capacity of the ER, as measured by its luminal volume and the levels of its protein-folding and processing machinery, is matched to the inward flux of secretory and transmembrane proteins by an unfolded protein response (UPR) (, ). A stress signal arising from an imbalance between unfolded proteins and the ER machinery is recognized by three ER-localized transmembrane proteins—IRE1, PERK, and ATF6—that affect a rectifying transcriptional and translational response to restore protein-folding homeostasis (reviewed in ). Although a good deal is known about the effector functions of the UPR transducers, the physiological significance of ER stress, and the response to it (reviewed in ), the upstream molecular mechanisms that detect the ER stress signal remain poorly understood.IRE1, the most conserved and studied UPR transducer (, ), detects stress with its ER-luminal domain (IRE1^LD) and dimerizes, leading to dimerization-dependent autophosphorylation of its cytosolic domain () and the subsequent activation of its cytosolic endoribonuclease activity (). Activated IRE1 unconventionally splices the mRNA of the transcription factor XBP1/HAC1 (, , ), promoting XBP1 translation and a conserved XBP1-dependent gene-expression program.Dimerization emerges as a key upstream event both in IRE1 activation and in activation of PERK, which shares with IRE1 a structurally related ER-stress-sensing luminal domain (, ). Two hypotheses have been put forth to explain the coupling of ER stress to dimerization. In one hypothesis, unfolded proteins are proposed to serve as activating ligands by directly binding to the luminal domain and stabilizing it in a dimeric conformation. An alternative hypothesis holds that the UPR is organized along principles similar to its cytosolic counterpart, the heat shock response, in which an imbalance between unfolded proteins and heat shock protein (Hsp)70 chaperones is recognized via the former’s ability to compete for the latter, kinetically disrupting repressive complexes between chaperones and stress transducers (, ).The IRE1 luminal domain crystallizes as a dimer, which, in the yeast protein, is traversed by a groove that could engage an extended peptide as a stabilizing ligand (href="#bib12" rid="bib12" class=" bibr popnode">Credle et al., 2005). Yeast IRE1^LD peptide ligands have been identified, but when added to solutions of yeast IRE1^LD, they principally affect a transition from a collection of oligomers to higher-order oligomers (href="#bib19" rid="bib19" class=" bibr popnode">Gardner and Walter, 2011). In the crystal structure of mammalian IRE1^LD and the related PERK^LD, the groove is occluded (href="#bib7" rid="bib7" class=" bibr popnode">Carrara et al., 2015a, href="#bib61" rid="bib61" class=" bibr popnode">Zhou et al., 2006). Furthermore, mammalian IRE1^LD is a dimer even in pure, dilute conditions (href="#bib31" rid="bib31" class=" bibr popnode">Liu et al., 2000, href="#bib61" rid="bib61" class=" bibr popnode">Zhou et al., 2006). Thus, the role of unfolded protein binding in promoting the monomer-to-dimer transition that initiates UPR signaling remains unclear.The ER lumen has a single Hsp70 chaperone, BiP. Reversible chaperone repression as the regulatory principle of UPR activity is supported by an inverse correlation between IRE1 activity and the amount of the ER-localized BiP recovered in complex with it (href="#bib5" rid="bib5" class=" bibr popnode">Bertolotti et al., 2000, href="#bib38" rid="bib38" class=" bibr popnode">Oikawa et al., 2009, href="#bib39" rid="bib39" class=" bibr popnode">Okamura et al., 2000), a feature that extends to the related UPR transducer PERK (href="#bib5" rid="bib5" class=" bibr popnode">Bertolotti et al., 2000, href="#bib32" rid="bib32" class=" bibr popnode">Ma et al., 2002). Furthermore, mutations in yeast BiP (Kar2p) that stabilize the BiP-IRE1^LD interaction repress the UPR (href="#bib25" rid="bib25" class=" bibr popnode">Kimata et al., 2003), and only unfolded proteins that engage BiP can induce the UPR (href="#bib37" rid="bib37" class=" bibr popnode">Ng et al., 1992). However, beyond plausibility suggested by these correlative cell-biological findings and the argument of evolutionary analogy, this model was otherwise unsupported.Hsp70 chaperones undergo a nucleotide-binding and hydrolysis-dependent cycle that dramatically alters their affinity for substrates. The intrinsic ATPase activity of Hsp70s is low, and ATPase-accelerating J-protein co-chaperones are therefore required for efficient substrate recognition and binding by Hsp70 (reviewed in href="#bib22" rid="bib22" class=" bibr popnode">Kampinga and Craig, 2010, href="#bib34" rid="bib34" class=" bibr popnode">Mayer, 2013). This feature arises from the ability of the co-chaperones to stimulate the ATPase activity of the Hsp70 via their conserved J domains while presenting diverse substrates to the Hsp70 via their divergent targeting domains. As a consequence, the Hsp70 initially interacts with substrates in a high-kon, ATP-bound state and then retains the substrate in a low-koff, ADP-bound state. Cycles of substrate release, accelerated by nucleotide exchange factors (NEFs, reviewed in href="#bib4" rid="bib4" class=" bibr popnode">Behnke et al., 2015) and rebinding, specified by the J domain co-chaperone, result in substrate-selective ultra-affinity (href="#bib15" rid="bib15" class=" bibr popnode">De Los Rios and Barducci, 2014, href="#bib35" rid="bib35" class=" bibr popnode">Misselwitz et al., 1998) that is the basis for formation of Hsp70-substrate complexes. Here, we drew on these well-established principles to examine the possibility that past failures to obtain biochemical support for reversible BiP-mediated repression as the basis for UPR regulation arose from the absence of a suitable ER-localized co-chaperone in the experimental systems used.