Abstract / Summary The vast majority of all secreted and plasma membrane proteins are synthesized on membrane-bound ribosomes at the endoplasmic reticulum (ER). Ribosomes synthesizing this subset of proteins are targeted to the ER membrane by the signal recognition particle (SRP). SRP binds to signal sequences as they emerge from the ribosome as part of the growing polypeptide chain. SRP then hands over the nascent polypeptide to the Sec61 translocon for movement into or across the ER membrane bilayer. Protein folding in the ER lumen proceeds co-translationally. In the event of protein misfolding, a signaling network, the “unfolded protein response (UPR)”, is activated that rebalances the ER’s protein folding capacity with the load of proteins entering the ER. IRE1 is one of the UPR’s main protein misfolding sensors. It is a bifunctional kinase/RNase with a lumenal domain that is activated by unfolded protein binding. It signals through a non-conventional mRNA splicing reaction, utilizing its cytoplasmic RNase domain to excise an intron from the mRNA encoding the transcription factor XBP1. XBP1 drives expression of a host of chaperones, protein folding factors, and component of the protein transport and degradation machinery to reestablish homeostasis. IRE1 also degrades select ER-bound mRNAs, thereby reducing the ER protein-folding load in a reaction termed “regulated IRE1 dependent decay (RIDD)”. IRE1 can directly interact with the co-translational translocation machinery, including the ribosome, SRP, and Sec61 translocon. These data (and supporting evidence) suggest that IRE1, at least in part, performs its functions in physical contact with ER-bound ribosomes. When activated, however, IRE1 forms large oligomeric clusters. We recently visualized these clusters in intact cells using correlative light and electron microscopy. Our high-resolution cryo-tomograms demonstrate that IRE1 clusters are convoluted networks of narrow, anatomosing ER tubes. Surprisingly, these clusters are entirely devoid of membrane-bound ribosomes. This poses a paradox, suggesting that active IRE1 must exist in at least two sub-populations, one composed of small oligomers that can ribosome- and/or SRP-associate, the other composed of large ribosome-free IRE1 clusters. We will map IRE1’s target mRNAs and mechanistic roles to these states. Specifically, we will: (1) Dissect the mechanistic principles of substrate selection by IRE1. (2) Determine the atomic structure of IRE1 complexed to ribosomes, translocon, and/or SRP using single-particle cryo-EM. (3) Identify the sub-cellular birthplaces of IRE1-spliced and RIDD-cleaved mRNAs. (4) Assemble IRE1 cluster of defined stoichiometry in vitro and in cells and assess functional outputs with biochemical and cell-based assays. Defects in proteostasis lie at the heart of an ever-expanding list of diseases. The proposed experiments aim to define how mammalian IRE1 selects its mRNA targets and what key molecular players it associat...