Search

Maintaining homeostasis is vital for biological systems. To maintain homeostasis, cells rely on specialized complex mechanisms, known as cell stress responses, that detect specific internal imbalances and react to them. The integrated stress response (ISR) is a fundamental signaling network that reprograms the transcriptome and proteome to leverage the cell’s biosynthetic capacity against different stresses. Signaling plasticity is enabled by distinct ISR sensor kinases that detect specific perturbations. The ISR has two faces, with tailored homeostatic outputs and a terminal one engaged upon overwhelming stress. Through a chemical-genetics approach that uncouples natural stress inputs from ISR actuation, we show that the ISR engages a common, cell-autonomous apoptosis mechanism that requires unconventional signaling by death receptor 5. We also show that the ISR selectively activates the endoplasmic reticulum (ER) stress sensor IRE1 in the absence of ER stress. Together, our results indicate that a common ISR mechanism eliminates terminally injured cells and reveal a new level of intercommunication between the ISR and other stress responses.

ABSTRACTA Physiological Unfolded Protein Response Impacts Progression Through the Cell CycleBy Soham ChowdhuryThe cell cycle is a series of coordinated molecular steps that allow a progenitor cell to produce two daughter cells. During symmetrical cell division, a mother cell’s particular organelle makeup must also be inherited by the daughter cells. Multi-copy organelles such as mitochondria divide and are equally distributed between daughter cells. Single copy organelles such as the endoplasmic reticulum (ER) and the Golgi apparatus however must be first expanded, then fragmented, and finally partitioned to each daughter cell. Such organelle expansion requires the production of new membranes. Being that the ER is the site of endomembrane biosynthesis, we reasoned that ER expansion should precede and be necessary for mammalian cell division. The Unfolded Protein Response (UPR) is an evolutionarily conserved collection of signaling pathways that maintain ER health. Protein folding perturbations in the ER lumen as well disturbances of the ER membrane lipid bilayer trigger ER stress and activate the UPR. The UPR mitigates ER stress by increasing the biosynthetic and protein degradative capacities of the ER, including its physical expansion through endomembrane biogenesis. Here we show that as cells grow through interphase, their ER chaperone and foldase contents increase, as does the ER volume, indicating that ER expansion precedes cell division. Moreover, pharmacological inhibition of the UPR sensors IRE1 and ATF6 at steady-state delayed cell cycle progression. Furthermore, we found that the threshold for UPR activation is lower in the S/G2 stage of the cell cycle suggesting that ER expansion and increased ER protein-processing capacity is subsequent to genome duplication. Finally, our data indicate that IRE1 activity is dampened by the G2/M cell cycle checkpoint kinase PKMYT1 suggesting negative feedback control. Taken together, these findings suggest a physiological role for the UPR in coordinating the mammalian cell cycle

Throughout the cell cycle, genome duplication is coordinated with the multiplication and growth of organelles, which requires membrane biosynthesis at the endoplasmic reticulum (ER). By this reasoning, ER growth and increased ER function would be a pre-requisite for cell division. Because the unfolded protein response (UPR)—a fundamental homeostatic mechanism that maintains ER integrity—increases the size and protein-processing capacity of the ER, I reasoned that it may oversee ER physiology during the cell cycle. To investigate ER growth and activation of the UPR during the cell cycle, I optimized and characterized a well-described fluorescent reporter of cell cycle progression, known as the FAST-FUCCI system. This live-cell reporter enabled me to separate G1 and S/G2 cell populations by fluorescence activated cell sorting (FACS). My data show that mammalian cells increased in size and granularity during interphase. These hallmarks were correlated with an increase in ER-resident protein content, suggesting that the ER enlarges in preparation for cell division. Moreover, I found that inhibition of IRE1 via pharmacological agents delayed progression through the G1/S boundary. While investigating a plausible mechanism that could regulate UPR activity during the cell cycle, I identified PKMYT1, an ER- and Golgi apparatus-associated G2/M cell cycle checkpoint kinase, in a candidate-based approach. I found that IRE1 activity is suppressed by PKMYT1, suggesting that PKMYT1 exerts regulatory control over IRE1 prior to cell division. Preliminary data suggests that PKMYT1 and IRE1 do not physically interact, which suggests regulation via an unidentified intermediate or a transcriptomic regulation by downstream transcription factors, such as XBP1s. Taken together, my results provide evidence to suggest that mammalian cells engage a physiological UPR involving IRE1 signaling during cell cycle progression.

In metazoans, the Unfolded Protein Response (UPR) restores homeostasis in the endoplasmic reticulum (ER) by both increasing the protein-folding capacity of the ER and by reducing the protein folding burden. The UPR in S. cerevisiae exclusively relies on this first solution to survive ER stress. Remarkably, S. pombe utilizes the second, though the UPR in both species is mediated through site-specific RNA cleavage by the evolutionarily conserved kinase/RNase transmembrane protein Ire1. While most of the S. pombe Ire1 mRNA substrates are degraded post cleavage, the mRNA encoding the ER chaperone protein BiP1 evades decay and the 5’ fragment accumulates after truncation. Though the mechanism behind this stabilization remains unresolved, we have discovered that the relevance of bip1 mRNA processing to the cell’s survival may be greater than previously estimated. We have compared the Ire1 kinase/RNase domain from each species in vitro and discovered key structural differences we can modulate to make the S. cerevisiae Ire1 behave like S. pombe Ire1. First, we characterized differences in cofactor sensitivity in the kinase domains of the two enzymes. Additionally, we exploited the toxicity of S. pombe Ire1 in bacterial cells to identify mutations in the more stringent S. cerevisiae Ire1 protein that would cause toxicity and relax substrate stringency. We identified a pair of mutations at the RNase dimer interface that expanded the S. cerevisiae enzyme’s substrate repertoire and coincided with an increase in its propensity to dimerize. This interface could represent an under-utilized target for drug design to modulate the RNase activity of metazoan Ire1.

Although progressive supranuclear palsy (PSP) is a sporadic neurodegenerative disease, a genome-wide association study identified candidate genes as risk factors. One of these genes EIF2AK3 encodes protein kinase RNA (PKR)-like endoplasmic reticulum kinase (PERK), one of three stress sensors, alongside inositol-requiring enzyme 1 (IRE1) and activating transcription factor 6 (ATF6), in the endoplasmic reticulum (ER). Collectively known as the unfolded protein response (UPR), these sensors attempt to maintain protein homeostasis before resorting to programmed cell death. IRE1 splices X-box binding protein 1 (XBP1) mRNAs to form transcription factors that upregulate UPR components. ATF6 is modified to form transcription factors that upregulate ER chaperones like binding immunoglobulin protein (BiP). PERK phosphorylates the alpha subunit of eukaryotic initiation factor 2 (eIF2), attenuating translation with exceptions of activating transcription factor 4 (ATF4) mRNAs. ATF4 then upregulates CCAAT/enhancer-binding protein (C/EBP) homologous protein (CHOP), a proapoptotic protein. Four single nucleotide polymorphisms (SNPs) in PERK are in linkage disequilibrium. The haplotype B variant, with minor alleles for all four, confers deleterious effects. Using patient fibroblasts, this study aimed to characterize the haplotype’s effects on PSP pathology and UPR activation. Haplotype B heterozygous fibroblasts displayed heightened sensitivity to ER stress and excessive cell death. Haplotype B fibroblasts had decreased ER stress sensitivity and insignificant amounts of cell death. Basal levels of eIF2 phosphorylation for the two genotypes were comparable. Haplotype B did not affect IRE1 activity and changes in BiP expression, an indicator of ATF6 activation, were drug-dependent only for haplotype B heterozygous fibroblasts.

Diets rich in saturated fatty acids (SFAs) produce a form of tissue inflammation driven by “metabolically activated” macrophages, which contributes to the development of obesity-associated metabolic diseases. To better understand this form of macrophage activation, we analyzed the transcriptomes of mouse macrophages treated with either SFAs or the classical inflammatory stimulus lipopolysaccharide (LPS). SFA treatment induced a transcriptional signature distinct from that of LPS and strongly enriched by endoplasmic reticulum (ER) stress markers. In particular, SFA treatment increased the expression of the ER stress sensor IRE1α and many of its target genes comprising the adaptive unfolded protein response. SFAs also activate the NLRP3 inflammasome in macrophages, resulting in secretion of the pro-inflammatory cytokine IL-1β. We found that IRE1α mediates SFA-induced IL-1β secretion by macrophages and that its activation by SFAs does not rely on unfolded protein sensing. Instead, the ability of SFAs to stimulate either IRE1α activation or IL-1β secretion can be specifically reduced by preventing their flux into the phosphatidylcholine (PC) class of membrane lipids or by increasing unsaturated PC levels. Thus, IRE1α is an unrecognized intracellular PC sensor critical to the process by which SFAs stimulate macrophages to secrete IL-1β, a driver of diet-induced tissue inflammation and metabolic dysfunction.

Two ER membrane-resident transmembrane kinases, IRE1 and PERK, function as stress sensors in the unfolded protein response. IRE1 also has an endoribonuclease activity, which initiates a non-conventional mRNA splicing reaction, while PERK phosphorylates eIF2alpha. We engineered a potent small molecule, IPA, that binds to IRE1’s ATP-binding pocket and predisposes the kinase domain to oligomerization, activating its RNase. IPA also inhibits PERK but, paradoxically, activates it at low concentrations, resulting in a bell-shaped activation profile. We reconstituted IPA-activation of PERK–mediated eIF2alpha phosphorylation from purified components. We estimate that under conditions of maximal activation, less than 15% of PERK molecules in the reaction are occupied by IPA. We propose that IPA binding biases the PERK kinase towards its active conformation, which trans-activates apo-PERK molecules. The mechanism by which partial occupancy with an inhibitor can activate kinases may be wide-spread and carries major implications for design and therapeutic application of kinase inhibitors.

The unfolded protein response (UPR) maintains protein folding homeostasis in the endoplasmic reticulum (ER). In metazoan cells, the Ire1 branch of the UPR initiates two functional outputs— non-conventional mRNA splicing and selective mRNA decay (RIDD). By contrast, Ire1 orthologs from Saccharomyces cerevisiae and Schizosaccharomyces pombe are specialized for only splicing or RIDD, respectively. The functional specialization lies in Ire1’s RNase activity, which is either stringently splice-site specific (as in S. cerevisiae) or promiscuous (as in S. pombe). Here, we developed an assay that reports on Ire1’s RNase promiscuity. We found that conversion of two amino acids within the RNase domain of S. cerevisiae Ire1 to their S. pombe counterparts rendered it promiscuous. Using biochemical assays and computational modeling, we show that the mutations rewired a pair of salt bridges at Ire1 RNase domain’s dimer interface, changing its protomer alignment. Thus, Ire1 protomer alignment affects its substrates specificity.

Secreted and transmembrane proteins enter the endoplasmic reticulum (ER) as unfolded, nascent polypeptides. Before they continue along the secretory pathway to their final destination, these proteins must fold into their proper structure. The ER is the primary site of quality control for these proteins; terminally misfolded proteins are degraded, while properly folded proteins are packaged into vesicles. This quality control ensures proper folding, and thus proper function, for all of the proteins that mediate cell-environment interactions. However, the mechanisms required for quality control can also cause an accumulation of unfolded proteins when environmental stresses, developmental programs, or drug treatments disrupt protein folding. Cells facing such ER stress activate a signaling pathway, the Unfolded Protein Response (UPR), designed to restore protein folding homeostasis.In S. cerevisiae, the UPR is activated by the transmembrane protein, Ire1, which initiates a transcriptional program to increase the size and capacity of the ER. Ire1 consists of a stress-sensing lumenal domain that oligomerizes in response to ER stress and cytoplasmic domains that transmit signals to the nucleus and cytoplasm. Ire1 and its activation by oligomerization are conserved throughout the Metazoa. Despite the ubiquity of the UPR pathway and its relevance to human disease, the mechanism of Ire1 stress-sensing and subsequent oligomerization were largely unknown. In Chapter Two, we provide biochemical evidence that unfolded proteins are ligands for Ire1 activation by showing that Ire1 forms a complex in vivo with a model misfolded protein, binds short peptide proxies in vitro, and that peptide binding causes Ire1 oligomerization. Chapter Three extends our findings with Ire1 cLD from S. cerevisiae to the Ire1 lumenal domain from Mus musculus. Although the crystal structure of MmIre1 LD revealed a narrower binding groove, we found that MmIre1 LD can also bind peptides in vitro, but with a different binding preference than ScIre1 cLD. In Chapter Four, we crystallized the lumenal domain of the Ire1 homolog from a thermophilic yeast, Chaetomium thermophilum. This structure reveals several intriguing new interactions within the crystal lattice, which may be important for Ire1 activation.

The endoplasmic reticulum (ER) responds to the accumulation of misfolded proteins in its lumen (termed ER stress) by activating collective signal transduction pathways, called unfolded protein response (UPR). The UPR alleviates ER stress through its translational and transcriptional outputs carried out by three distinct ER stress sensors: PERK, ATF6 and IRE1. IRE1 is the best characterized UPR signal transduction molecule, and it is conserved from yeast to mammals. UPR signaling promotes apoptotic cell death if ER stress is not alleviated. However, the exact mechanism of how ER stress induces apoptosis remains unclear. Particularly, each branch of the UPR exerts opposing effects on cell survival under ER stress. To distinguish the effect of each branch of the UPR, we took a chemical-genetic approach, where we artificially prolonged IRE1 signaling by a small molecule, and observed enhanced cell survival under ER stress. Our data suggests a causal link between the duration of UPR branch signaling and life or death cellular decision. To further dissect IRE1 signaling upon prolonged ER stress, we developed a visual assay to assess IRE1 activation in cells. IRE1 is activated by misfolded protein direct binding to its lumenal domain, which then transduces the signal to its cytosolic kinase and endoribonuclease domains. Using the in vivo imaging approach and an in vitro biochemical assay, we demonstrated that IRE1 oligomerizes at the ER membrane upon misfolded protein accumulation, which in turn activates it kinase and RNase activities. IRE1 signaling attenuation correlates perfectly with its oligomers dissociation, suggesting a built-in timer within IRE1 that may be important in switching the UPR from the initially cytoprotective phase to the apoptotic mode. We next took an unbiased proteomic approach to address IRE1 signal attenuation mechanism using mass spectrometry analysis. We identified over 400 proteins and compiled a list of interesting and potential IRE1 interacting partners and modulators.

Diabetes mellitus is a disease caused by a combination of insulin resistance and decline of &beta-cell function. There is a high demand on the pancreatic &beta-cell to synthesize and secrete appropriate amounts of insulin in response to glucose levels. When the protein folding demand exceeds the folding capacity of the endoplasmic reticulum (ER), this results in a condition known as ER stress. The cell responds to ER stress by activating an adaptive signaling pathway called the unfolded protein response (UPR). The UPR first tries to restore homeostasis to the cell by reducing the load of proteins to be folded as well as by increasing the folding capacity of the cell. However, if ER stress is chronic, the UPR switches from a homeostatic pathway to an apoptotic pathway. The molecular mechanisms that guide this switch from adaptation to death remain unclear. Elucidating the underlying mechanism of ER stress induced apoptosis would allow better insight for future development of drugs that could prevent cell death and treat diseases such as diabetes. By utilizing mutations in different domains of the ER stress sensor IRE1&alpha, we were able to induce expression of IRE1&alpha mutants in &beta-cells and look at effects on cell fate. By using flow cytometry and microarray analysis we found that chronic IRE1&alpha overexpression resulted in downregulation of hundreds of mRNAs that localize to the ER or secretory pathway, which we hypothesize destabilizes the ER and mediates apoptosis. We also found that IRE1&alpha overexpression alone results in apoptosis, but expression of either kinase dead or RNase dead mutants of IRE1&alpha did not result in downregulation of mRNAs nor apoptosis. These findings uncovered a novel mechanism whereby IRE1&alpha has not only a cytoprotective function, but a proapoptotic arm as well, which requires both its kinase and RNase activity. Continuing this work, we found that TXNIP is a downstream proapoptotic mediator of ER stress apoptosis induced by IRE1&alpha that promotes oxidative stress and cell death. Through these novel findings, we have gained further insight on the mechanism of ER stress mediated apoptosis and also found a convergence of the oxidative stress and ER stress pathways.

AbstractIn eukaryotic cells, the untoward accumulation of misfolded proteins inside the endoplasmic reticulum (ER) triggers a transcriptional program that restores cellular homeostasis. This program, called the unfolded protein response (UPR), is activated through a series of enzymatic steps that culminate in the production of a transcription factor that upregulates a set of target genes required for cell survival during ER stress. Activation occurs through an ER-resident sensor, Ire1, which is stimulated by misfolded proteins in the ER lumen. Activated Ire1 transmits a signal to its cytosolic portion, containing a kinase and an RNase, which trans-autophosphorylates and activates the RNase to cleave an mRNA encoding the Hac1 transcription factor. The cleavage products are ligated by tRNA ligase, Trl1, to produce a spliced form of HAC1 mRNA that is translated into the transcription factor. Together, the coordinated activation of Ire1 and ligation by Trl1 to produce spliced HAC1 mRNA comprise the central mechanism of UPR activation. This thesis describes my work toward understanding two key regulatory steps in UPR activation. Several lines of evidence suggest that Ire1 kinase plays a role in the UPR that is autonomous from its role in RNase activation. First, RNase L, a close homolog of Ire1, is an RNase with a pseudokinase domain that has lost activity through evolution. Conservation of Ire1's kinase suggests that activity is important for UPR fitness. Moreover, in vitro studies have shown that kinase-inactive Ire1 retains RNase activity and, in vivo, kinase activity can be bypassed by ATP-competitive drugs. In an endeavor to understand the importance of Ire1 kinase we discovered that kinase activity contributes to the attenuation of Ire1 signaling and is imperative for homeostatic adaptation during ER stress. The final chapters of this thesis describe my work to decipher how tRNA ligase achieves pathway specificity for the UPR. Trl1 splits its time between in HAC1 mRNA splicing and in tRNA splicing. Our efforts to discern how Trl1 acts in HAC1 mRNA splicing were unfruitful, however, this work enlightened our knowledge of Trl1 biology and set the stage for future work to elucidate Trl1's role in the UPR.