* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Signal Transduction From the Endoplasmic Reticulum to the Cell
Survey
Document related concepts
Cellular differentiation wikipedia , lookup
Histone acetylation and deacetylation wikipedia , lookup
Cell nucleus wikipedia , lookup
Magnesium transporter wikipedia , lookup
Protein phosphorylation wikipedia , lookup
Hedgehog signaling pathway wikipedia , lookup
Endomembrane system wikipedia , lookup
Protein moonlighting wikipedia , lookup
G protein–coupled receptor wikipedia , lookup
List of types of proteins wikipedia , lookup
VLDL receptor wikipedia , lookup
Proteolysis wikipedia , lookup
Transcript
PHYSIOLOGICAL REVIEWS Vol. 79, No. 3, July 1999 Printed in U.S.A. Signal Transduction From the Endoplasmic Reticulum to the Cell Nucleus HEIKE L. PAHL Department of Experimental Anesthesiology, University Hospital Freiburg, Freiburg, Germany 683 684 684 689 690 693 697 Pahl, Heike L. Signal Transduction From the Endoplasmic Reticulum to the Cell Nucleus. Physiol. Rev. 79: 683–701, 1999.—The endoplasmic reticulum (ER) serves several important functions. Cholesterol, an essential component of cellular membranes, is synthesized on the ER surface. Inside the organelle, proteins destined for secretion or transport to the cell surface are folded and become glycosylated. Because these processes are essential for cell viability, a disturbance in ER function presents significant stress to the cell. In response to ER stress, three distinct signal transduction pathways can be activated. Two of these, the unfolded protein response and the ER-overload response, respond to disturbances in protein processing. The third, the sterol regulatory cascade, is activated by depletion of cholesterol. This review summarizes the recent advances in our understanding of these ER-nuclear signal transduction pathways. In addition, it points to novel regulatory mechanisms discovered in these pathways, which may be widely used in other systems. I. INTRODUCTION The endoplasmic reticulum (ER) is one of the largest cell organelles, its membranes constituting over one-half of the total membranes in a cell. The ER lumen, the internal space, comprises over 10% of the cell volume. This vast structure has two essential functions. 1) Proteins destined for transport to other organelles, secretion, or expression on the cell surface are synthesized on the ER surface. During translation, they are translocated into the ER lumen through a pore in the ER membrane. Inside the organelle, they are folded, sometimes with the aid of chaperone proteins, and become glycosylated. A quality control mechanism ensures that only correctly folded proteins exit the ER. Incorrectly folded proteins are retained and ultimately degraded. 2) Synthesis of lipids and cholesterol takes place on the cytoplasmic side of the ER membrane (70). Therefore, the ER is the production site for all components of cellular membranes, proteins, lipids, and sterols. Given the importance of this organelle, its proper function is essential to the cell. Therefore, it seems only 0031-9333/99 $15.00 Copyright © 1999 the American Physiological Society natural that mechanisms have evolved to signal a disturbance in ER function. Various conditions can interfere with ER function, and for the purpose of this review, these are collectively called ER stress. Endoplasmic reticulum stress can arise from a disturbance in protein folding, leading to an accumulation of un- or misfolded proteins in the organelle. Starvation of glucose also leads to protein accumulation in the ER, since glycosylation cannot proceed at the required rate. Likewise, the ER can become saturated when too many proteins are produced that need to be processed by the organelle. Finally, starvation of cholesterol, an essential component of cellular membranes, also elicits ER stress. Cells respond to most stress situations by changing protein production; synthesis of those proteins that may help to alleviate the stress is increased (9). This upregulation occurs mainly at the level of transcription so that more mRNA molecules encoding the required proteins are produced. Increased transcription is in turn achieved by activating transcription factors, proteins which bind specific DNA sequences in the promoter or enhancer 683 Downloaded from http://physrev.physiology.org/ by 10.220.33.6 on May 2, 2017 I. Introduction II. Signaling From the Endoplasmic Reticulum to the Nucleus A. The unfolded protein response B. A novel role for transcription factor NFkB C. The ER-overload response D. ER-nuclear signaling by SREBP III. Summary 684 HEIKE L. PAHL 1. Conditions that induce the unfolded protein response TABLE Drugs perturbing ER function Tunicamycin 2-Deoxyglucose Brefeldin A Castanospermine Glucosamine Thapsigargin AlF42 Glucose starvation Reducing agents 2-Mercaptoethanol Dithiothreitol Heavy metals Cobalt Nickel Overexpression of mutant proteins that cannot fold correctly Calcium ionophores A-23187 Ionomycin ER, endoplasmic reticulum. It was subsequently shown that an accumulation of correctly folded proteins, which cannot be processed, such as the immunoglobulin heavy chain in the absence of light chain, also induces the UPR. Yeast cells contain proteins homologous to the mammalian GRP and can also mount an UPR (82). Because of the elegant genetic experiments that can be performed in yeast, more is known about the UPR in this system than about its mammalian counterpart (90). Therefore, the Saccharomyces cerevisiae UPR is discussed first. Because it is not yet clear whether individual elements of the pathways are evolutionarily conserved, the mammalian UPR is described separately. 1. The UPR in yeast II. SIGNALING FROM THE ENDOPLASMIC RETICULUM TO THE NUCLEUS A. The Unfolded Protein Response Kozutsumi et al. (38) were the first to propose a signal transduction pathway that is activated by an ER stress signal. This group showed that the expression of a mutant influenza hemagglutinin, which does not fold correctly, but not of the wild-type protein, induced the expression of genes for several ER-resident proteins. These proteins, first named glucose-regulated proteins (GRP) because of their induction by glucose starvation (40), include the heavy-chain binding protein BiP/GRP78, GRP94, protein-disulfate isomerase (PDI/ERp59), and ERp72. The GRP facilitate protein folding in the ER. In this way, they reduce the number of misfolded proteins and thus alleviate ER stress. Because a common stimulus for the induction of GRP is the presence of misfolded proteins in the ER (38), this pathway was named the UPR. Yeast cells can be treated with a variety of agents that perturb ER function, for example, glycosylation inhibitors such as 2-deoxyglucose or reducing agents such as 2-mercaptoethanol, which prevents disulfide bond formation (see Table 1). All these treatments induce the transcription of a group of ER-resident proteins (82, 90). These include a member of the 70-kDa heat shock protein family called KAR2, a peptidyl-prolyl cis-trans isomerase (PPIase) encoded by the FKB2 gene, a protein disulfide isomerase-like protein named EUG1, and the protein disulfide isomerase PDI1. These proteins function to facilitate protein folding. Therefore, their increased expression, induced by the presence of unfolded proteins in the ER, should reduce stress to the organelle. How does the presence of misfolded proteins in the ER activate transcription in the cell nucleus? Deletion analysis of the KAR2 promoter identified a 22-bp element, the UPR element (UPRE), which is conserved in the promoters of all UPR target genes and is required for their induction by ER stress (60). Moreover, the UPRE confers Downloaded from http://physrev.physiology.org/ by 10.220.33.6 on May 2, 2017 regions of genes. Their binding increases the rate of transcription of this gene. Until recently, however, very little was known about how cells respond to ER stress, which transcription factors are activated, and by what mechanism they are induced. This review focuses on three different types of ER stress: 1) the presence of mis- or unfolded proteins in the organelle, 2) the overloading of the ER with correctly folded proteins, and 3) the starvation of cholesterol. Each stress situation triggers a unique cellular response, using various signal transduction pathways to induce specific transcription factors. The first of these pathways to be described, the unfolded protein response (UPR), has been known for 10 years. It is activated, as the name suggests, by the presence of mis- or unfolded proteins in the ER. Of the three pathways, it is the only one conserved, at least in part, between humans and yeast. The more recently discovered sterol regulatory element binding protein (SREBP) pathway regulates cholesterol synthesis in the ER membrane in response to sterol plasma levels. In both of these pathways, novel mechanisms of transcription factor activation have been described, which may prove to be paradigms for processes widely used in other pathways. The most recently described ER-nuclear signal transduction pathway, ER-overload response (EOR), a stress triggered by congestion of the organelle with too many proteins, activates a well-characterized transcription factor, nuclear factor kB (NFkB). However, it uses a novel pathway in the process. This review compares and contrasts these signal transduction pathways, pointing out questions that need to be addressed in the future. Volume 79 July 1999 ENDOPLASMIC RETICULUM SIGNAL TRANSDUCTION 685 ER stress inducibility on a heterologous promoter. Mori et al. (57) recently conducted a detailed mutational mapping of the UPRE. They identified an E-box-like palindrome, separated by one cytosine residue (59-CAG C GTG-39), which is essential for UPR function. All five yeast UPR target genes (see Fig. 1) contain a functional copy of this motif in their promoters (59). Recently, a novel transcription factor, alternatively named HAC1 (HAC 5 homologous to ATF and CREB) or ERN4 (ERN 5 ER nuclear signaling pathway), was identified that binds the UPRE E-box motif during ER stress (19, 57). The HAC1 gene is required for activation of the UPR, since yeast strains deleted in this gene no longer induce the response. Activity of HAC1 is regulated in an unusual and complex manner, about which there is some debate in the literature. Two groups have shown that the HAC1 mRNA is spliced at a nonconsensus splice site upon induction of the UPR (19, 33). Splicing occurs within the coding region so that the spliced mRNA gives rise to a different protein. Unspliced HAC1 mRNA encodes a 230-amino acid protein. Splicing removes the sequence encoding the last 10 amino acids and adds a novel COOH terminus of 18 amino acids, giving rise to a 238-amino acid Hac1p (19, 33). In transient transfection assays, the 238amino acid Hac1p has an ;12-fold higher transcriptional transactivation capacity than the 230-amino acid form of the protein (33). Expression of the spliced form of Hac1p, the 238-amino acid protein, constitutively activates the UPR. Only this larger 238-amino acid protein can be detected in Western blots; the shorter 230-amino acid form is not found. Likewise, the Hac1 protein is only detectable after induction of the unfolded protein response; it is not present in unstressed yeast cells. There was considerable disagreement in the literature over the regulation of HAC1 splicing. Cox et al. (19) proposed that although the 230-amino acid Hac1p is constitutively produced, it is a highly unstable protein that is rapidly degraded by the ubiquitin-proteasome pathway and therefore has no effect in the cell. They suggested that the spliced 238-amino acid Hac1p, in contrast, is a Downloaded from http://physrev.physiology.org/ by 10.220.33.6 on May 2, 2017 FIG. 1. Unfolded protein response (UPR) in yeast. Presence of mis- or unfolded proteins in endoplasmic reticulum (ER) is sensed by a yet unknown mechanism. This leads to dimerization and activation of Ire1p kinase, which autophosphorylates. COOH terminus of Ire1p has endonuclease activity that becomes activated and splices HAC1 mRNA. Spliced HAC1 is religated by tRNA ligase RLG1. Upon exit from nucleus, HAC1 mRNA is translated into Hac1p transcription factor. Hac1p translocates to nucleus and binds UPR element (UPRE), transactivating expression of UPR target genes, ER resident enzymes which aid in protein folding. 686 HEIKE L. PAHL in tRNA splicing, is used. A mutation in RLG1 [rlg-100, a C442T point mutation, which changes a conserved histidine (His-148) to tyrosine] completely prevents induction of the UPR. In these strains, the HAC1 mRNA is not spliced but rather becomes rapidly degraded. However, this mutation has no effect on tRNA splicing, proving that the tRNA ligase encodes two distinct enzymatic activities. Sidrauski et al. (91) have proposed that tRNA ligase is required to join the two segments of a cleaved HAC1 mRNA. In the absence of this enzymatic activity, HAC1 becomes cleaved but is then rapidly degraded, since the second step of splicing, the religation, cannot take place. Blocking spliceosome-mediated mRNA splicing, in contrast, has no effect on UPR induction (91). Therefore, HAC1 mRNA is spliced by a novel mechanism requiring tRNA ligase. In addition to the transcription factor HAC1, a second component of the UPR signal transduction pathway has been described. Genetic analyses have identified a Ser/Thr kinase (Ire1p/Ern1p) whose activity is required for induction of the UPR in yeast (18, 58). Ire1p acts upstream of Hac1p, since the HAC1 mRNA is not spliced in yeast strains lacking the kinase (19). However, the spliced, 238-amino acid Hac1 is constitutively active, even in IRE1 mutant strains. Because its structure is similar to that of mammalian transmembrane receptor kinases, the following model for Ire1p activation has been proposed (58, 82): the kinase resides in the membrane of the ER with its NH2 terminus facing the lumen. This domain acts as a sensor of ER stress. Mammalian receptor kinases are activated by dimerization, and Shamu and Walter (83) have recently shown that Ire1p activity is regulated similarly. Under nonstressed conditions, Ire1p is present as an inactive monomer. During periods of ER stress, Ire1p oligomerizes, activating the kinase which trans-autophosphorylates and induces the UPR signaling cascade. In support of this model, COOH-terminally truncated Ire1p molecules, which lack the kinase domain, exert a dominant-negative effect (58). These mutants most likely form kinase-inactive heterodimers with wild-type Ire1p. Sidrauski and Walter (92) were recently able to show that Ire1p is a bifunctional enzyme; its COOH terminus contains, in addition to its kinase activity, an endoribonuclease activity that is responsible for splicing the HAC1 mRNA at both nonconsensus splice sites. The authors (92) were able to reproduce accurate HAC1 splicing in vitro, by supplying only Ire1p and tRNA ligase. With this observation, the entire signal transduction pathway can now be followed from the ER lumen into the nucleus (Fig. 1). The accumulation of un- or misfolded proteins initiates Ire1p dimerization, which induces autophosphorylation, presumably also activating the endoribonuclease activity. Activated Ire1p, whose COOH terminus can protrude either into the cytoplasm or into the nucleus, cleaves the HAC1 mRNA. The RLG1 tRNA ligase religates the cleaved Downloaded from http://physrev.physiology.org/ by 10.220.33.6 on May 2, 2017 stable protein and is therefore capable of inducing the unfolded protein response. In this model, yeast cells constantly translate the unspliced HAC1 mRNA, producing a very low level of 230-amino acid Hac1p. Upon ER stress, cells use this small amount of preformed 230-amino acid Hac1p to initiate the UPR and sustain this response by producing the more stable 238-amino acid Hac1p through inducible mRNA splicing (19). Kawahara et al. (33) were unable to reproduce some of the results reported by Cox et al. (19). Most importantly, they used point mutants to show that only those HAC1 mRNA that have undergone splicing are translated into a protein. Mutation of the 59-splice site, which prevents splicing, completely abrogated protein production in response to ER stress. Kawahara et al. (33) proposed an alternative model based on these results. They suggest that Hac1 expression is controlled posttranscriptionally through ER stress-induced mRNA splicing, because only the spliced mRNA is translated into protein. Recently, a consensus has been reached. It is now agreed that Hac1p production is regulated at the posttranscriptional level. After reinvestigating their experimental system, Chapman and Walter (15) now report that both forms of Hac1, the spliced 238-amino acid form and the unspliced 230-amino acid form, are equally stable. Furthermore, unspliced Hac1 is not stabilized in yeast strains deficient in protein degradation, suggesting that absence of detectable 230-amino acid Hac1 is not due to instability of the protein. Rather, in their newest study, Chapman and Walter (15) show that the 252-nucleotide intron in the unspliced HAC1 mRNA attenuates translation of the protein. Surprisingly, transferring the Hac1 intron to an unrelated green fluorescent protein mRNA also attenuates translation of this message. Therefore, presence of the Hac1 intron seems sufficient to prevent translation of a polyribosome-associated mRNA. Two possible mechanisms have been suggested (15): the intron may prevent progress of the ribosome either at a distance, by looping back, or directly, in both cases causing the ribosome to stall and ultimately release from the mRNA. The newest model compiled from these data argues that unspliced HAC1 mRNA is exported from the nucleus, bound to polyribosomes, but prevented from translation into the 230-amino acid protein by the presence of the intron. Upon induction of the UPR, newly produced HAC1 mRNA is spliced and becomes translated into the 238-amino acid protein, which functions as a transcription factor to induce UPR target genes (15). This model accounts for all observations made by the two labs investigating Hac1. Sidrauski et al. (91) have shown that HAC1 splicing at the nonconsensus splice site occurs via a novel mechanism, distinct from the splicing of other mRNA. Unlike the previously described mRNA splicing, HAC1 splicing does not require the spliceosome (91); rather, a tRNA ligase (RLG1), which has previously only been implicated Volume 79 July 1999 ENDOPLASMIC RETICULUM SIGNAL TRANSDUCTION ligands can be hypothesized, and all models must now await experimental verification. In yeast, inositol-containing phospholipids and lipids constitute a major proportion of membrane components, and the level of free inositol regulates phospholipid biosynthesis (111). A fall in the level of free intracellular inositol induces the transcription of genes encoding enzymes required for phospholipid synthesis. These include INO1, encoding inositol 1-phosphate synthase, CHO1, encoding phosphatidylserine synthase, and OPI3, encoding phospholipid methyltransferase. Interestingly, yeast strains carrying mutations in IRE1, HAC1, RGL1, and ADA5 are inositol auxotrophs; they require inositol in the culture medium for growth (18, 91, 110). It is therefore possible that the UPR coordinately regulates both the synthesis of ER membranes and the transcription of ERresident proteins. Cox et al. (17) have recently provided evidence to this effect. They have shown that inositol starvation induces the transcription of UPR target genes. The authors demonstrate that this is not because of the accumulation of underglycosylated proteins under inositol starvation; rather, the UPR is specifically induced by a decrease in free cellular inositol. The following observations suggest that the transcription of UPR and inositol target genes may be linked. In response to inositol starvation, IRE1 mutant cells induce much lower levels of INO1 mRNA than wild-type cells. Although some INO1 mRNA is produced, in IRE1 mutants, cells cannot sustain mRNA production and therefore never produce the large amounts required for inositol prototrophy (17). The authors propose a two-step mechanism of INO1 transcriptional activation: an initial Ire1p-independent induction followed by a second Ire1p-dependent maintenance of transcription. Interestingly, induction of the UPR by tunicamycin also induces INO1 transcription in an Ire1p- and Hac1p-dependent manner. Strains lacking either Ire1p or Hac1p were unable to induce INO1 in response to tunicamycin treatment. Therefore, Ire1p and Hac1p participate in two signal transduction pathways: the UPR and the inositol response. This redundancy allows cells to expand both the size and the protein content of their ER during stress. 2. The UPR in mammalian cells Induction of the mammalian GRP genes is caused by a variety of agents including glycosylation inhibitors, reducing agents, heavy metals, amino acid analogs, glucose starvation, and perturbance of intracellular Ca21 homeostasis (see Table 1) (40, 41, 47). It has been proposed that the common condition caused by all these treatments is the accumulation of misfolded proteins in the ER and that this represents the UPR-inducing stimulus (38). However, recent data by McCormick et al. (52) suggest that two separate signal transduction pathways may lead to Downloaded from http://physrev.physiology.org/ by 10.220.33.6 on May 2, 2017 ends, producing an open reading frame of 238 amino acids. This mRNA is translated into 238-amino acid Hac1, which binds the UPRE, initiating transcription of the UPR target genes. Recently, Welihinda et al. (110) identified the first negative regulator of the UPR. Using a modified yeast two-hybrid screen, the authors (110) showed that the serine/threonine phosphatase Ptc2p, a type 2C phosphatase, directly interacts with Ire1p. The interaction is dependent on Ire1p phosphorylation and is mediated by the kinase interaction domain within Ptc2p. Ptc2p dephosphorylates Ire1p, thereby downregulating the UPR. Consequently, strains carrying null mutations in PTC2 show a three- to fourfold increase in UPR activity and display augmented HAC1 splicing. Likewise, overexpression of wild-type Ptc2p attenuates HAC splicing and downregulates the UPR. Ire1p represents the first identified substrate for type 2C serine/threonine phosphatases. Welihinda et al. (110) have recently suggested that Hac1p uses the cofactors Ada2, Ada3, Ada5, and Gcn5, which form a multisubunit complex, to activate transcription (110). Gcn5 was found to interact with Ire1p in a modified two-hybrid screen. Subsequently, these data were corroborated by in vitro coimmunoprecipitation. Yeast strains deleted in GCN5, ADA2, or ADA3 showed reduced induction of the KAR2 and PDI1 genes, targets of the UPR. However, these effects are very weak. Stronger evidence comes from the observation that strains deficient in ADA5 completely lack KAR2 and PDI1 induction. In contrast, induction of the heat shock response is normal in these yeast. The Gen5p/Ada2p/Ada3p/Ada5p complex may form a bridge between Hac1p, bound to the UPRE, around 100 bp upstream of the promoter, and the TATA-binding protein, bound to the TATA box in the promoter. However, the significance of these data needs to be verified. Although many components of the yeast UPR have now been identified, one question remains: What is the sensor of stress in the ER lumen? Somehow the presence of un- or misfolded proteins must lead to Ire1p dimerization. This step is not well understood, but several models have been proposed. The GRP78/BiP protein itself has been suggested as a sensor of ER stress. Because GRP78/ BiP binds to unfolded proteins, thereby preventing their exit from the ER, the level of free BiP may inversely correlate with the extent of protein misfolding. In the simplest model, GRP78/BiP serves directly as a ligand for monomeric Ire1p, thus preventing oligomerization of the kinase. As unfolded proteins accumulate, most GRP78/ BiP will form novel complexes with misfolded proteins, thus allowing Ire1p to oligomerize. In support of this hypothesis, Morris et al. (61) have shown in mammalian cells that the overexpression of GRP78/BiP prevents the induction of GRP transcription. Of course, many other 687 688 HEIKE L. PAHL ever, the CCAAT element is not sufficient for promoter activity on its own. In the grp78 promoter, an adjacent CCAC motif also appears necessary for ER stress induction (73). Therefore, in mammalian cells, no single element seems responsible for ER stress-mediated induction of transcription; rather, this response is mediated by multiple, functionally redundant elements (112). The identification of the Ire1p kinase in yeast raises the question of whether the mammalian UPR is also mediated by a kinase, perhaps even by an Ire1p homolog. Cao et al. (11) investigated this hypothesis by preparing stably transfected Chinese hamster ovary cell lines overexpressing yeast Ire1p. Basal and thapsigargin-induced expression of GRP78 and GRP94 was only increased twofold in these cells. Treatment with okadaic acid, a Ser/Thr phosphatase inhibitor, increased GRP induction in response to thapsigargin by twofold. In contrast, the tyrosine kinase inhibitor genistein completely inhibited induction of GRP transcription by thapsigargin. The authors concluded that the mammalian UPR may require both Ser/Thr and Tyr protein kinases, but these appear not to be functionally homologous to Ire1p. Tirasophon et al. (101), however, recently reported cloning of the mammalian Ire1p homolog. Using a degenerate oligonucleotide from the yeast Ire1p kinase domain, the authors (101) were able to amplify a PCR fragment from a human fetal liver cDNA library. With the use of this fragment as a probe, the entire cDNA was cloned. Like yeast Ire1p, the human protein contains a transmembrane domain, a Ser/Thr protein kinase domain, as well as a domain homologous to the endonuclease RNAse L (101). Consistent with these observations, human Ire1p autophosphorylates and is able to cleave yeast HAC1 mRNA at the 59-splice site. Human Ire1p is required for UPR activation as judged by reporter gene assays using the rat BiP promoter. Overexpression of wild-type Ire1p constitutively activated the reporter gene, whereas expression of a catalytically inactive form of the protein inhibited ER stress-induced gene activation in a dominant negative fashion (101). Cloning of the human Ire1p homolog will surely prompt the search for a human Hac1. Recently, Brewer et al. (7) were able to show that basal transcription of the GRP genes under non-ERstressed conditions is regulated by a novel signal transduction pathway, distinct from the unfolded protein response. This pathway is stimulated by cytokine growth factors, since cytokine starvation drastically decreased the basal level of GRP expression. The observation that the UPR remains intact during starvation and the distinct kinetics by which the two pathways are activated support the conclusion that GRP transcription is indeed stimulated by two independent signaling pathways, one for basic transcription and one, the UPR, for stress-mediated induction. Downloaded from http://physrev.physiology.org/ by 10.220.33.6 on May 2, 2017 GRP induction in mammalian cells. Many cell types respond to both the glycosylation inhibitor tunicamycin and the ER Ca21-ATPase inhibitor thapsigargin by increasing GRP transcription. McCormick et al. (52) reported that the WEHI72 lymphoma cell lines do not alter GRP transcription after thapsigargin treatment. Interestingly, treatment of these cells with tunicamycin led to a strong increase in GRP transcription. The authors (52) therefore propose that two separate signaling pathways, one elicited by ER Ca21 depletion and one by the accumulation of misfolded or underglycosylated proteins in the organelle, may induce GRP transcription. The promoters of mammalian GRP proteins contain a conserved 28-bp DNA element, the GRP core (14). This sequence has been shown to bind complex proteins in electrophoretic mobility shift assays and DNA footprinting analyses. Stress-inducible changes were observed in the 39-half of the core region by in vivo footprinting assays, and this sequence was therefore named the stressinducible change region (SICR) (46). A constitutive factor, p70CORE, was first shown to bind this region (46). p70CORE was recently identified as the transcription factor YY1 (45) and shown to enhance stress induction of the grp78 promoter. Two additional factors, which bind the GRP core SICR, were recently identified by southwestern cloning: YB-1 and dbpA (44). Both are members of the Y-box family of proteins, which share a conserved DNA binding motif as well as a cold shock domain. Li et al. (44) have shown that binding of YB-1 and dbpA to the GRP core prevents YY1 binding to this site in gel shift assays. Consistent with this observation, in reporter gene assays, transfection of the Y-box proteins represses stress induction through the GRP core element. Moreover, YB-1 and YY1 were shown to interact in yeast-interaction trap assays. However, gel shift assays, transient transfection experiments, and yeast interaction trap all create artificial conditions for protein-protein or protein-DNA interactions. Therefore, further experiments such as conditional knockout mice are required to confirm the physiological role of these proteins in regulating GRP expression. Despite the fact that the GRP core shares 50% overall homology to the yeast 22-bp UPRE and contains a stretch of 80% identity, this sequence does not function as a UPRE in S. cerevisiae (60). Deletion analysis and linker scanning of the rat GRP78 promoter showed that the GRP core element is functionally redundant (112). Its deletion or mutation does not alter gene induction by misfolded proteins. Deletion of a CCAAT motif, which binds members of the constitutive CBF/NF-Y family (72), reduces both basal transcription and stress inducibility of the rat GRP78 promoter. A similar motif was identified in the promoter of two additional UPR target genes: GRP94 and ERp72 (51, 69). Using dominant negative mutants, Zhou and Lee (117) have shown that the CAAT-binding factor CBF/NF-Y is required for grp78 promoter activity. How- Volume 79 July 1999 ENDOPLASMIC RETICULUM SIGNAL TRANSDUCTION B. A Novel Role for Transcription Factor NFkB 2. ER stress conditions that activate nuclear factor kB TABLE ER overload by protein overexpression Hepatitis B virus truncated middle HB surface antigen Influenza virus hemagglutinin Immunoglobulin m-heavy chain MHC class I Adenovirus E3/19K EPO receptor Drugs perturbing ER function Tunicamycin 2-Deoxyglucose Monensin Brefeldin A Thapsigargin Cyclopiazonic acid MHC, major histocompatibility. inhibition of NFkB activation by overexpresssion of a dominant negative form of IkB renders cells more sensitive to apoptosis by tumor necrosis factor (TNF) as well as by the chemotherapeutic agent daunorubicin and by ionizing radiation (106, 107). Thus, in addition to its role as a mediator of the immune response, NFkB seems to regulate apoptosis, at least in some cell types. 1. A novel ER-nuclear signaling pathway distinct from the UPR All of the previously described activators of NFkB share two properties: they represent a stress to the cell, and they act from the outside. We wondered whether an internally arising stress situation would also activate the transcription factor. A perturbance in the function of the ER, the largest cell organelle, could represent such an internal stress. We observed that expression of the hepatitis B virus protein MHBSt, which accumulates in the ER, activates NFkB (56). Two models can explain this result: either NFkB activation is a unique property of MHBSt, or it represents a more general phenomenon, perhaps related to MHBSt’s accumulation in the ER. We therefore investigated whether, in general, a perturbance of ER function would activate NFkB. Indeed, treatment of HeLa cells with a variety of agents that either inhibit N-glycosylation (tunicamycin and 2-deoxyglucose) or protein transport by causing the formation of a mixed ER-Golgi compartment (brefeldin A) strongly induce the transcription factor (64). Moreover, the overexpression of various membrane proteins, which accumulate in the ER, can activate NFkB (56, 63, 64, 66) (Table 2). Activation of NFkB by these agents was observed both in gel mobility shift assays and by induction of reporter genes controlled by NFkB DNA binding motifs (56, 63, 64). Supershift assays revealed that the NFkB complex induced by ER stress is composed in equal parts of p50/p65 and p50/c-rel heterodimers (63, 64). Downloaded from http://physrev.physiology.org/ by 10.220.33.6 on May 2, 2017 A number of conditions that perturb ER function were recently shown to activate NFkB, a transcription factor previously characterized as a central mediator of immune and inflammatory responses (64, 66). In a great variety of cell types, NFkB is found in an inactive, cytoplasmic complex bound to IkB, its inhibitory subunit (1). Upon exposure of cells to pathological agents, such as proinflammatory cytokines, antigen, and ultraviolet or g-irradiation, or upon bacterial and viral infections, NFkB activity is rapidly induced (for a review, see Baeuerle and Henkel, Ref. 3). Activation occurs via phosphorylation of IkB-a on two serine residues, Ser-32 and Ser-36, and subsequent degradation of the inhibitor by the 26S proteasome (1, 6, 28, 96, 104, 105). Two kinases responsible for IkB phosphorylation, IKKa and IKKb, were recently cloned (55, 71, 113, 116). Together with the NFkB-inducing kinase (NIK), they form a high-molecular-weight complex of 700 –900 kDa. The IKKa and IKKb heterodimerize, but all three kinases appear necessary for IkB phosphorylation, targeting the inhibitor for proteolysis. Degradation of IkB releases the NFkB dimer, which subsequently translocates to the nucleus, where it activates transcription of its target genes. These include important proinflammatory proteins, such as interferons, cytokines, chemokines, cell-adhesion molecules, major histocompatibility (MHC) class I molecules and hematopoetic growth factors (3). A total of five NFkB DNA-binding subunits, RelA, RelB, c-Rel, p50 and p52, are currently known. These can hetero- and homodimerize and may bind distinct NFkB motifs with slightly differing affinities. This could provide a mechanism for selectively activating target genes. A total of five IkB proteins control the activity of NFkB dimers: IkB-a, -b, -g, -d, and -e (3). Given the nature of its target genes, the major role of the NFkB system appears to be the induction of a rapid and coordinated transcription of proinflammatory genes in response to external, primarily pathogenic stimuli. This is strongly supported by the phenotype of mice in which various members of the NFkB family were subject to targeted disruption (reviewed in Baeuerle and Baltimore, Ref. 2). For example, mice lacking the c-rel gene fail to mount effective immune responses to various pathogens, and their immune cells produce less cytokines in response to stimulation. Likewise, a knockout of the RelA (p65) subunit causes death early in embryogenesis due to apoptosis of the liver. In addition, recent experiments using cell lines derived from mice lacking RelA suggest that NFkB induction may protect against apoptosis (5). Fibroblasts and macrophages from these mice undergo programmed cell death when treated with TNF, whereas cells from control mice, which can induce NFkB in response to the cytokine, remain viable. These data are supported by independent experiments that show that 689 690 HEIKE L. PAHL 2. Defining the NFkB-activating signal in the ER The functionally impaired ER must be able to emit two distinct signals: one that activates NFkB and one that induces the UPR. A number of stimuli appear to cause the production of both signals. Because overexpression of a number of unrelated, wild-type, integral membrane proteins induces NFkB activation (see Table 2), it seems that the mere accumulation of proteins in the ER membrane provides an NFkB-activating stimulus. Overexpression of a bacterial protein in the cytosol (chloramphenicol acetyltransferase) or an insect protein in peroxisomes (firefly luciferase) had no such effect, arguing against a more general artifact of transient protein overexpression (64). We assume that NFkB activation by drugs blocking glycosylation and thus protein exit from the ER, also results from accumulation of proteins in the ER membrane, overloading the organelle. The ER-overload hypothesis was investigated using the adenovirus E3/19K protein as a model (66). Wild-type E3/19K resides in the ER, where it binds to MHC class I molecules, thereby preventing their transport to the cell surface. The viral protein possesses a COOH-terminal retention signal sequence, which causes the protein to be continuously retrieved to the ER from post-ER compartments (32). As expected, expression of even small amounts of E3/19K strongly activates NFkB (66). Because the sequence requirements of E3/19K for MHC class I binding and ER retention are precisely known, the NFkBactivating ER signal could be investigated using point mutants that abolish either property. Two point mutants that no longer bind MHC class I molecules activate NFkB as effectively as the wild-type protein (66). Thus the interaction between E3/19K and another protein, MHC class I, is not necessary for NFkB activation. Titration experiments confirmed that the NFkB-activating effects of overexpressing E3/19K and MHC class I are additive, not synergistic (66). However, there is a stringent requirement of ER retention for NFkB activation. Two mutant proteins, which are equally well expressed as the wild type, but escape ER retention and are expressed on the cell surface, no longer activate the transcription factor even when highly overexpressed (66). Hence, the sole determinant for the NFkB-activating effects of E3/19K is its ER retention. The dilysine ER retrieval motif in the cytoplasmic tail of E3/19K was not required, since proteins that carry mutations in this motif but nonetheless accumulate in the ER were equally potent in NFkB activation (66). These studies suggest that the NFkB-activating signal is generated by the accumulation of proteins within the ER membrane, a process we refer to as ER overload. At present, we do not know of any membrane protein that does not activate NFkB when overexpressed. Therefore, the ER-overload stimulus seems biophysical in nature rather than dependent on a particular protein sequence, except, of course, a signal sequence required for ER import and perhaps a transmembrane domain. We have not yet investigated whether the overexpression of ER luminal proteins also activates NFkB. Likewise, it is not clear whether a transmembrane domain is necessary or may even be sufficient for transcription factor activation. Experiments to answer these questions are currently being pursued. C. The ER-Overload Response 1. Messengers of the EOR Endoplasmic reticulum overload must release a signal from the organelle that reaches NFkB in the cytosol. Because the ER stores large amounts of Ca21, which can be released upon stimulation, we investigated whether this ion serves as a second messenger of the EOR (see Fig. 2). However, Ca21 had not been previously implicated in NFkB activation, except as a cofactor in T cells. Nonetheless, two lines of pharmacological evidence suggest that the efflux of Ca21 from the ER is required for EOR-mediated NFkB activation. First, NFkB induction by Downloaded from http://physrev.physiology.org/ by 10.220.33.6 on May 2, 2017 A number of the stimuli that activate NFkB also induce the UPR (see Table 1). For example, both NFkB and the UPR are activated by overexpression of the m-heavy chain, tunicamycin, 2-deoxyglucose, brefeldin A, or thapsigargin (64, 66, 82) (Tables 1 and 2). It was therefore important to determine whether NFkB participated in the UPR or defined a novel, independent signal transduction pathway between the ER and the nucleus. Several pharmacological and biological inducers distinguish between NFkB induction and UPR activation. For instance, there are inducers of GRP expression such as Ca21 ionophores, blockers of cytoplasmic Ca21 increase and 2-mercaptoethanol, that do not activate NFkB (21, 35, 42, 114). Calcium ionophores and the rise of cytoplasmic Ca21 have only a costimulatory effect on NFkB in T cells but cannot activate NFkB in other cell types, whereas 2-mercaptoethanol even inhibits NFkB activation (81, 102). Likewise, the glucosidase inhibitor castanospermine potently induces GRP expression but not NFkB activity (64). Conversely, a number of classical NFkB-inducers, such as TNF-a, do not induce GRP expression and, therefore, act independently of ER stress (64). Moreover, NFkB is potently induced by the phosphatase inhibitor okadaic acid (OA) (100). Treatment with OA does not activate GRP transcription on its own but can potentiate GRP induction in combination with the glycosylation inhibitor thapsigargin (68). Thus a variety of conditions allow a distinction between NFkB and UPR activation, supporting the hypothesis that NFkB functions in a novel signal transduction pathway between the ER and the nucleus. Volume 79 July 1999 ENDOPLASMIC RETICULUM SIGNAL TRANSDUCTION 691 ER stress is prevented by preincubation of cells with two intracellular Ca21 chelators: 3,4,5-trimethoxybenzoic acid 8-(diethylamino)octyl ester (TMB-8) and BAPTA-AM (65, 66). Second, inhibition of the ER-resident Ca21-ATPase by thapsigargin or cyclopiazonic acid (CPA), which causes a rapid efflux of Ca21 from the ER, potently induces NFkB (66). How the accumulation of proteins in the ER membrane increases its Ca21 permeability is unknown. Perhaps an accumulation of membrane proteins impairs Ca21-ATPase function. Alternatively, an increased protein-to-lipid ratio may increase the Ca21 permeability of the lipid bilayer. Both models would explain how a num- ber of membrane proteins that are completely unrelated in primary structure can activate the same response. Calcium is not the sole signal required for NFkB activation by ER overload. Activation of NFkB by all inducers tested to date is inhibited by a variety of antioxidants (80, 81). This finding, together with the observation that exogenously applied H2O2 can activate NFkB in some cell lines (4, 79), suggests that reactive oxygen intermediates (ROI) can act as second messengers during NFkB activation. This is also true for NFkB activation by ER overload. We observed that pretreatment of cells with a variety of structurally unrelated antioxidants abolishes Downloaded from http://physrev.physiology.org/ by 10.220.33.6 on May 2, 2017 FIG. 2. ER-overload response. Accumulation of wild-type or misfolded proteins in ER leads to a release of Ca21 from organelle. This causes production of reactive oxygen intermediates (ROI), which activate transcription factor nuclear factor kB (NFkB). Activation can be blocked by antioxidants and by Ca21 chelators, or it can be activated by drugs that cause an active release of Ca21 from ER. NFkB induces transcription of genes involved in inflammatory and immune responses. CPA, cyclopiazonic acid. 692 HEIKE L. PAHL TABLE 3. Diseases caused by protein misfolding Disease Carbohydrate-deficient glycoprotein syndrome a1-Antitrypsin deficiency Alzheimer’s disease Cystic fibrosis Charcot-Marie-Tooth type 1A Congenital hyperthyroid goiter Osteogenesis imperfecta Familial hypercholesterolemia class 2 von Willebrand disease Marfan syndrome Tay-Sachs disease Retinitis pigmentosa Leprechaunism Misfolded Protein Multiple a1-Antitrypsin Presenillin CFTR Peripheral myelin protein PMP22 Thyroglobulin Type I procollagen LDL receptor von Willebrand factor Fibrillin b-Hexosaminidase Rhodopsin Insulin receptor CFTR, cystic fibrosis transmembrane conductance regulator; LDL, low-density lipoprotein. the production of either second messenger reduces the toxicity of the alkylating agent (48). 2. Physiological Role of the EOR A) THE ANTIVIRAL RESPONSE. What physiological role does the EOR serve? Upon viral infection, cells become programmed to produce large amounts of viral capsid proteins that are processed through the ER. This may elicit an EOR just as seen upon expression of individual viral membrane proteins using transient transfection methods. Transient expression of three unrelated viral proteins, influenza hemagglutinin (63), hepatitis B virus MHBSt (56), and adenovirus E3/19K (66), to levels barely detected by immunofluorescence staining activates NFkB by an EOR (Table 2). Because the NFkB-activating stimulus appears to simply be the accumulation of membrane proteins, we expect that a large variety of membrane proteins from unrelated viruses can elicit this signal. Several NFkB target genes encode important antiviral defense proteins, for example, the cytokine b-interferon and proteins involved in viral peptide presentation to T cells, such as the proteasome subunit LMP2, the TAP1 peptide transporter, MHC class I molecules, and b2microglobulin (3). By inducing their expression via NFkB, the EOR may elicit a fast, nonspecific and therefore broad antiviral response. Direct evidence for this role of the EOR remains to be obtained. B) DISEASES OF PROTEIN FOLDING. Evidence is quickly growing to suggest that many diseases are caused by mutations that cause misfolding of proteins (Table 3). These mutant proteins have been shown to accumulate in the ER. Interestingly, several of these conditions are accompanied by inflammation, of which the pathomechanism remains unknown. We have hypothesized that the accumulation of wild-type or mutant proteins in the ER may lead to NFkB activation. The activated transcription Downloaded from http://physrev.physiology.org/ by 10.220.33.6 on May 2, 2017 NFkB activation by both pharmacological inducers of ER stress and by viral proteins that accumulate in the organelle (64, 65). At the time we interpreted these data to mean that ROI also serve as second messengers in ER overload-mediated NFkB activation. However, the recent cloning of two IkB kinases has renewed the discussion about the role of ROI as second messengers in NFkB activation (55, 71, 113, 116). All signal-transducing proteins required for TNF-mediated NFkB activation have been cloned (71), and they appear to participate in a kinase cascade, each activating the next by phosphorylation. Therefore, the role of ROI in this process is not apparent. Nonetheless, many laboratories have observed that a variety of structurally unrelated antioxidants inhibit TNF-stimulated NFkB induction. It is currently not clear whether ROI indeed serve as second messengers or whether antioxidants interfere with the activity of one of the signaling molecules, perhaps the IkB kinase. Alternatively, these agents may act downstream of IkB phosphorylation, perhaps inhibiting ubiquitinylation or degradation by the proteasome. Because ER stress-mediated NFkB activation apparently requires two second messengers, Ca21 and ROI, we wished to determine in what order they are produced and whether one causes the formation of the other. Two lines of evidence suggest that the release of Ca21 precedes the formation of ROI during ER overload. First, activation of NFkB by thapsigargin and CPA, which directly cause a Ca21 release from the ER, is inhibited by the antioxidant pyrrolidine dithiocarbamate (PDTC) (65). Second, fluorescence-activated cell sorting analysis measuring intracellular peroxide levels with DCFH showed that treatment of HeLa cells with either thapsigargin or CPA increases intracellular peroxide levels. Pretreatment with the Ca21 chelator BAPTA-AM prevented this response, suggesting that the Ca21 release preceded ROI production in these cells (65). Which enzymes produce ROI in response to Ca21 released from the ER? Various enzymes, such as cyclooxygenases, lipoxygenases, and several oxidases, are capable of inducibly producing ROI as a side product of their intrinsic peroxidase activities. Using specific inhibitors for these enzymes, the peroxidase activities of cyclooxygenases and lipoxygenases were identified as possible ROI-producing enzymes during ER stress. Tepoxalin, a specific inhibitor of peroxidases (98), prevented NFkB activation by thapsigargin but not by TNF or 12-O-tetradecanoylphorbol 13-acetate (TPA) (65). Whether tepoxalin inhibits NFkB induction by other ER stress-eliciting agents remains to be investigated. Interestingly, another stress-induced signal transduction pathway also involves both the redistribution of intracellular Ca21 and the generation of oxidative stress (48). Exposure of renal epithelial cells to the alkylating agent iodoacetamide induces both a rise in intracellular Ca21 concentration and the generation of ROI. Blocking Volume 79 July 1999 ENDOPLASMIC RETICULUM SIGNAL TRANSDUCTION accumulate in the ER of the producing liver cell, rather than being secreted into the bloodstream. Several different AAT mutations have been characterized, and their effects vary in severity. Although some mutations only cause a slight decrease in serum AAT levels, patients homozygous for the PiZ allele exhibit only 10–15% of normal AAT levels in their serum and patients carrying “null” mutations, such as Pi Null Hong Kong, have no detectable serum AAT. The PiZ mutant has been shown to accumulate in the ER of liver cells, causing the organelle to enlarge and distend (84, 93). Fifteen percent of children homozygous for the PiZ mutation develop hepatitis, and one-quarter of these progress to liver cirrhosis and die before the age of eight (97). To study this severe disease, transgenic mice carrying the PiZ allele were generated. These animals suffer neonatal hepatitis and show accumulation of the mutant AAT protein in the ER of liver cells (12, 24). We are currently investigating whether the hepatitis resulting from AAT accumulation is due to ER overload-mediated NFkB activation and subsequent induction of proinflammatory mediators. Several additional rare diseases are caused by incorrect folding or processing of proteins in the ER. These include the carbohydrate-deficient glycoprotein syndrome, which is caused by mutations in carbohydrateprocessing enzymes. As a result, most of the patient’s proteins are incorrectly glycosylated. The main clinical manifestation of this disease is a severe malformation of the nervous system, but these children can suffer from sterile inflammation of the pericardium, which is often lethal. Likewise, a group of peripheral neuropathies, which include the Charcaot-Marie-Tooth disease, are caused by mutations in the peripheral myelin protein PMP22. It has been hypothesized that the pathophysiology of these diseases may also involve ER overload-mediated NFkB activation (23). Medeiros-Neto et al. (54) have described a congenital hypothyroid goiter with deficient thyroglobulin and have shown that this disease is due to mutations in the thyroglobulin gene. In four patients, they were able to show accumulation of thyroglobulin in the ER of thyroid cells, which was accompanied by a microscopically visible distension of the ER. However, this condition has not been shown to trigger inflammation of the thyroid gland. Likewise, several other “ER storage diseases,” in which misfolded proteins accumulate in the ER, for example, osteogenesis imperfecta and familial hypocholesterolemia, may not involve inflammatory processes. Because the discovery of the EOR is rather recent, additional time will be required to elucidate the role of this novel pathway in the pathophysiology of ER storage diseases. D. ER-Nuclear Signaling by SREBP Cholesterol, an essential cell membrane component, is made available to cells in two ways. One is the uptake Downloaded from http://physrev.physiology.org/ by 10.220.33.6 on May 2, 2017 factor would lead to expression of its target genes, many of which encode inflammatory cytokines such as interleukin (IL)-1 and TNF-a and chemokines such as IL-8, monocyte chemattractant protein-1 (MCP-1), monocyte inhibitory protein-1a (MIP-1), and RANTES. Secretion of these proinflammatory mediators would result in neutrophil, macrophage, and ultimately T-cell recruitment. These cells would become activated and in turn themselves release inflammatory cytokines, stimulating and sustaining the inflammatory process. This model could explain how the accumulation of a misfolded or a wild-type protein may lead to inflammation; however, it is entirely hypothetical and needs to be investigated experimentally. Nonetheless, several examples of diseases caused by misfolded proteins and accompanied by sterile inflammations are listed (see also Table 3). For example, cystic fibrosis patients develop progressive, cytokine-mediated inflammatory lung disease. Pulmonary inflammation is the major cause of morbidity and mortality in these patients (36). Cystic fibrosis is caused by mutations in the cystic fibrosis transmembrane conductance regulator (CFTR). The most common mutation, termed DF508, causes the protein to be entirely retained in the ER (16). A proportion of wild-type CFTR is also found in the ER; only ;25–50% of the protein reaches the cell membrane (50). Interestingly, it was recently shown that the airway epithelium, those cells in which the mutant CFTR accumulates, produce large quantities of inflammatory cytokines (53, 62). We are currently investigating whether cytokine production is because ER overload-mediated NFkB activation. Alzheimer’s disease usually strikes patients in their 70s and 80s. Tragically, there is a familiar form of the disease that causes an early onset, before the age of 60. The vast majority of these cases are caused by mutations in the two recently cloned presenillin genes. Both the wild-type and the mutant forms of these proteins are located in the ER (37). Like cystic fibrosis, Alzheimer’s disease involves the expression of several proinflammatory genes known to be upregulated by NFkB, for example, IL-1, IL-6, and IL-8 (43, 67). Interestingly, PC-12 pheochromocytoma cells expressing the L286V mutant form of presenillin-1 (PS-1) display increased levels of ROI and intracellular Ca21 in response to apoptotic stimuli compared with control cells expressing wild-type PS-1 (27). In addition, thapsigargin treatment of PS-1 mutant cell lines caused a larger increase in intracellular Ca21 than treatment of cells expressing the wild-type protein. Cells carrying the PS-1 mutant were hypersensitive to the induction of apoptosis by several stimuli. Treatment with antioxidants or blockers of Ca21 influx relieved mutant PS-1 cells of this hypersensitivity (27). a1-Antitrypsin (AAT) deficiency, which affects ;1 in 1,800 live births, is the most common genetic cause of liver disease in children (99). The deficiency results from mutations that cause the molecule to fold incorrectly and thereby 693 694 HEIKE L. PAHL Volume 79 of cholesterol in food. After resorption in the small intestine, cholesterol is transported in the bloodstream bound in low-density lipoproteins (LDL). The LDL receptors on cell surfaces bind and internalize LDL from the plasma. Alternatively, almost all cell types can synthesize cholesterol de novo (10). Both cholesterol uptake and synthesis are subject to feedback repression. Transcription of the genes encoding the LDL receptor, lipoprotein lipase, as well as cholesterol and fatty acid biosynthetic enzymes are inhibited by high intracellular sterol levels and induced upon sterol depletion (26). Regulation is mediated at the level of transcription. This requires a 10-bp DNA sequence, the sterol regulatory element 1 (SRE-1), in the promoters of sterol-regulated genes (94). A basic-helix-loop-helix-leucine zipper (bHLH-LZ) protein, SREBP-1, was cloned and shown to activate transcription via the SRE-1 (115). The SRE-1 sequence does not fit the CANNTG consensus, the so-called E box, described for other bHLH-LZ proteins. Interestingly, SREBP-1 was cloned independently from an expression library by virtue of its ability to bind the E-box motif (103). Spiegelmann and colleagues (34), who named the protein adipocyte determination- and differentiation-dependent factor 1 (ADD-1), show that the ability of SREBP1/ADD-1 to bind both the SRE and the E box depends on a unique tyrosine residue at position 335 (34). Mutation of this amino acid to an arginine, conserved in other bHLHLZ proteins, abolishes binding to the SRE but retains E-box binding. Conversely, substitution of the consensus arginine by a tyrosine in another bHLH-LZ protein called USF confers the ability to bind both sequences on this heterologous protein (34). Both SREBP-1 and the homologous SREBP-2 (31) are synthesized as large precursor molecules of ;1,150 amino acids, which are inserted into the ER/nuclear membrane (109) (see Fig. 3). The NH2 terminus is cytoplasmic and contains the bHLH-LZ motif. The middle of the protein contains two transmembrane domains, which are inserted into the ER or nuclear membrane, causing the intervening 30 amino acids to form a “luminal loop” that protrudes Downloaded from http://physrev.physiology.org/ by 10.220.33.6 on May 2, 2017 FIG. 3. Sterol response pathway. Synthesis and uptake of cholesterol is subject to feedback repression within cells. How cellular cholesterol level is sensed is not yet known. Depletion of cholesterol leads to proteolysis of membranebound transcription factor sterol regulatory element binding protein (SREBP). Proteases that cleave SREBP at 2 well-characterized sites within “luminal loop” in ER lumen and within first transmembrane domain have not been described. SREBP cleavage requires activity of SREBP cleavage-activating protein (SCAP), which has been shown to physically interact with COOH terminus of SREBP. SCAP possesses no enzymatic activity; however, its NH2terminal transmembrane domains may serve as cellular sterol sensor. Cleaved, active SREBP transcription factor homodimerizes and activates transcription of genes involved in cholesterol and fatty acid synthesis and uptake. SRE, sterol regulatory element. July 1999 ENDOPLASMIC RETICULUM SIGNAL TRANSDUCTION 1. The SREBP cleavage-activating protein Despite the fact that the proteases remain elusive, a critical regulator of the SREBP pathway was recently identified. The SREBP cleavage-activating protein (SCAP) was cloned from a cell line, which fails to downregulate SREBP cleavage after sterol saturation. These cells were shown to have a constitutively activating mutation in SCAP (29). Like SREBP, SCAP is an integral membrane protein, since its NH2 terminus contains eight putative transmembrane domains. Interestingly, these segments show 55% homology to the transmembrane domains found in the NH2 terminus of hydroxymethylglutaryl (HMG)-CoA reductase. In HMG-CoA reductase, the NH2terminal transmembrane region functions as a sterol sensor. The catalytic activity of HMG-CoA reductase is also regulated by the intracellular sterol concentration. In sterol-depleted cells, the half-life of the enzyme is 10 h. This decreases dramatically to 1.5 h when cells are saturated with cholesterol. The NH2-terminal transmembrane domains are responsible for this regulated degradation, whereas the COOH-terminal cytoplasmic portion bears the enzymatic activity (39). Although there is no evidence to suggest that SCAP half-life varies with intracellular cholesterol levels, it is nonetheless intriguing to suggest 4. Proteins whose expression is activated by SREBP TABLE Proteins involved in cholesterol synthesis HMG-CoA synthase HMG-CoA reductase Farnesyl diphosphate synthase Squalene synthase Proteins involved in cholesterol and fatty acid uptake LDL receptor Lipoprotein lipase Proteins involved in fatty acid synthesis Acetyl-CoA carboxylase Fatty acid synthase Stearoyl-CoA desaturase 1 Glycerol-3-phosphate acyltransferase Membrane components Caveolin Regulatory proteins Leptin HMG, hydroxymethylglutaryl; SREBP, sterol regulatory element binding protein. that its transmembrane domains also serve as a sterol sensor. Interestingly, SCAP does not possess any enzymatic activity, in particular, no protease activity; rather, its COOH terminus contains four so-called WD domains, which have been shown to mediate protein-protein interaction in other proteins. Recently, Sakai et al. (75) were able to show that SCAP and SREBP-2 interact in coimmunoprecipitation experiments and that the COOH terminus of both proteins is required for complex formation. Interestingly, a COOH-terminal truncation of SREBP-2 leads to a reduction in proteolytic cleavage, suggesting that a SCAP/SREBP complex is required for proteolysis and may therefore represent the target of the SREBP protease. Cleaved SREBP, which contains the bHLH-LZ motif and apparently some of the transmembrane domain, translocates to the nucleus where it activates transcription of target genes. The SRE-1 sites were first reported in the promoters of the LDL receptor and the HMG-CoA synthase genes. However, it has recently become clear that SREBP participates in the transcription of many more genes. These can be classified into three groups (see Table 4): 1) genes encoding enzymes involved in cholesterol biosynthesis, such as HMG-CoA synthase; 2) genes encoding proteins involved in cholesterol and fatty acid uptake from the plasma, for example, the LDL receptor and lipoprotein lipase; and 3) genes encoding proteins mediating fatty acid synthesis, for example, the fatty acid synthase (FAS) gene. Interestingly, FAS is expressed in a differentiation-dependent manner in adipocytes (103). Together with the observation that SREBP/ADD-1 mRNA expression increases substantially during adipocyte differentiation, this raises the intriguing possibility that the transcription factor exerts a dual function in regulating Downloaded from http://physrev.physiology.org/ by 10.220.33.6 on May 2, 2017 both into the ER and into the nuclear envelope (30, 77). The COOH terminus again faces the cytoplasm (77). Upon sterol depletion, SREBP precursors are cleaved by two sequential proteolytic steps (74), releasing an ;500-amino acid fragment that contains the basic DNA-binding domain as well as the leucine zipper dimerization motif. The first cleavage occurs at a Leu-Ser bond in the middle of the luminal loop (22, 74). Sakai et al. (74) have very elegantly shown that his step is required before a second cleavage can take place, this time in the middle of the first transmembrane spanning region. The first cleavage is sterol dependent: in cells saturated with sterols, proteolysis of the Leu-Ser bond is abolished. Because the second cleavage requires the first proteolysis, it also does not take place in sterol-overloaded cells. Neither of the proteases that cut SREBP has been described biochemically, let alone cloned. Both cleavage sites have been characterized quite carefully, and the precise amino acid requirements are known (22). The sequence and location of these sites, one within a luminal loop and the other within a transmembrane segment, suggest that these enzymes cleave by unusual mechanisms and may be unlike any proteases known to date. The cysteine protease called sterol cleavage activator (SCA) that was shown to cleave SREBP-1/ADD-1 between the leucine zipper and the first transmembrane domain in vitro (108) appears not to function as a SREBP cleaving enzyme in vivo. 695 696 HEIKE L. PAHL 2. Diseases of sterol metabolism The importance of this regulatory loop becomes apparent in patients with Niemann-Pick disease. Some of these patients, suffering from the type A and type B disease, have defects in their acidic sphingomyelinase. In these patients, shingomyelin is not degraded effectively; it accumulates and causes a large increase in the amount of free cholesterol. The patients suffer from cholesterol accumulation, causing as the primary clinical symptom neuronal degradation. The gene whose mutation causes the distinct Niemann-Pick disease type C was cloned last year and was shown to have homology to SCAP (13). It shares the transmembrane domains thought to act as the sterol sensor. The majority of the mutations identified in Niemann-Pick type C patients affect these transmembrane domains. These results suggest that NPC1, as the gene was named, may represent an additional regulator of cholesterol synthesis. Increased understanding of the regulation of sterol synthesis, for example, the elucidation of the role of NPC1 and SCAP and the identification of the proteases that cleave SREBP, should allow the develop- ment of a specific therapy for these devastating group of diseases. 3. SREBP in vivo In addition to the SRE-1, the E box (59-CATGTG-39), and the ACC promoter site (59-ATCACGTGA-39), SREBP was recently shown to bind to a fourth sequence, named SRE-3 (25). This sequence, which is found in the promoter of the gene for farnesyl diphosphate synthase, shows only 60% identity to either the SRE-1 or the E box. These data raise the possibility that SREBP may be involved in the transcription of a plethora of genes mediating cholesterol, fatty acid, and lipid biosynthesis, which do not contain consensus SRE-1 binding sites. For example, it has been pointed out that 13 genes encoding enzymes for lipid biosynthesis contain a conserved sequence named UASINO, which bears a 60% sequence homology to both SRE-1 and the E box (25). All SREBP-regulated promoters studied contain binding sites for additional regulatory factors. Although mutation of the SRE-1 abolishes sterol responsiveness, the SRE-1 cannot function efficiently on its own (8, 76). In the LDL receptor promoter, two adjacent binding sites for the ubiquitous transcription factor Sp1 are required for SRE function. Both Sp1 and SREBP have been shown to synergistically activate the gene for the LDL receptor (76, 95). Like the LDL receptor promoter, the ACC promoter also contains a binding site for Sp1 adjacent to the SREBP binding site (49). SREBP has recently been shown to interact directly with the CCAAT-binding factor/nuclear factor Y (20). The CCAAT boxes are found in both the HMG-CoA synthase and the farnesal diphosphate synthase promoters. All results described above were obtained in cultured cell lines. It therefore remained to be investigated whether these data could be corroborated in animal models. Perhaps not surprisingly, there are several significant differences between SREBP function in vitro and in vivo. For example, in cell lines, proteolysis of SREBP-1 and SREBP-2 is regulated in parallel; both proteins are cleaved during sterol deprivation and remain intact when sterols are saturated. In hamster livers, in contrast, SREBP-1 is cleaved and active when the animals are fed a low-fat diet. However, SREBP-2 is not active under these conditions (85). When these livers are depleted of cholesterol by feeding the animals the cholesterol synthesis inhibitor lovastatin together with the bile acid-binding resin Colestipol, SREPB-2 proteolysis is induced (85). However, contrary to observations in cell lines, SREBP-1 is processed much less effectively under these circumstances. The net result is that livers of hamsters depleted of cholesterol contain much more active SREBP-2 than SREBP-1 (85). Within the two isoforms of SREBP-1, SREBP-1a and Downloaded from http://physrev.physiology.org/ by 10.220.33.6 on May 2, 2017 both sterol responsiveness and adipocyte differentiationspecific gene expression (103). It has been suggested that because SREBP regulates both lipid and cholesterol synthesis, both components of LDL, the use of a single transcription factor may serve to provide a balanced supply of both components. A similar hypothesis was made concerning a role for SREBP in maintaining a balance of fatty acid and cholesterol levels. It was shown recently that SREBP activates transcription of the promoter for acetyl-CoA carboxylase (ACC), the rate-limiting enzyme in fatty acid biosynthesis (49). Because the transcription of HMG-CoA reductase, the ratelimiting enzyme in cholesterol biosynthesis, is also regulated by sterols, the levels of these two cellular lipids may be regulated coordinately. Perhaps SREBP/ADD-1 also regulates membrane biosynthesis, since both cholesterol and fatty acids are essential membrane components. Recent evidence suggests that SREBP play a critical role not only in regulating membrane biosynthesis, but in maintaining its optimal lipid composition. It has been known since 1980 that sphingomyelin, a membrane lipid, affects the rate of cholesterol biosynthesis. Addition of sphingomyelin increases cholesterol synthesis, whereas its depletion, through treatment with sphingomyelinase, decreases cholesterol synthesis. Recently, Scheek et al. (78) were able to show that sphingomyelin depletion in cultured cells blocks proteolysis of SREBP-2 at the first site, inside the luminal loop. Thus, in the absence of sphingomyelin, cholesterol biosynthesis is decreased because SREBP is inactive, preventing transcription of biosynthetic enzymes. Volume 79 July 1999 ENDOPLASMIC RETICULUM SIGNAL TRANSDUCTION remains to be investigated whether the low percentage of surviving animals have compensated for the SREBP-1 deletion by some additional mutation. III. SUMMARY The investigation of ER-nuclear signal transduction has led to the discovery of three distinct signaling pathways: the UPR, the EOR, and the SREBP pathway. Each is activated by a distinct ER stress signal. The UPR is induced by the presence of misfolded proteins in the organelle. In yeast, this stress is sensed by the Ire1p kinase, which becomes activated and autophosphorylates, thereby inducing its second enzymatic activity, an endonuclease. In unique reactions not previously described, the Ire1p endonuclease splices the HAC1 mRNA at nonconsensus splice sites. Unlike other characterized mRNA, HAC1 mRNA is religated by the tRNA ligase Rlg1. Spliced HAC1 mRNA is translated to a 238amino acid protein, which functions as a transcription factor to induce expression of the UPR target genes. The EOR is activated by several stimuli that also induce the UPR, but the common activating signal is the congestion of the ER with proteins. This causes a release of Ca21 from the organelle and the production of ROI, ultimately resulting in the activation of transcription factor NFkB. Because NFkB is a central mediator of the human immune response, the EOR may play a role in the pathophysiology of several ER storage diseases. Cells obtain cholesterol either through uptake or through de novo synthesis. Both mechanisms are subject to feedback repression. This is possible because production of both the rate-limiting enzyme of cholesterol biosynthesis and of the LDL receptor, required for cholesterol uptake, is controlled by the same transcription factor. Activity of this protein, the SREBP, is tightly regulated by cellular sterol concentrations. Inactive SREBP is inserted into the ER-nuclear membrane as a large precursor molecule. Sterol depletion leads to two sequential cleavages of the membrane-bound SREBP, releasing an active transcription factor that activates expression of genes required not only for cholesterol uptake and biosynthesis but also genes involved in fatty acid synthesis and uptake. Thus a single transcription factor coordinately regulates the synthesis of two major membrane components. Investigation of these three ER-nuclear signal transduction pathways has not only taught us much about how cells respond to various forms of ER stress. Perhaps more importantly, they have revealed that these pathways contain regulatory mechanisms unknown in other systems. The unconventional splicing of the Hac1 transcription factor may turn out to be the first example of many mRNA species, not processed by the splicesome. This opens up Downloaded from http://physrev.physiology.org/ by 10.220.33.6 on May 2, 2017 -1c, there are differences between in vitro and in vivo observations as well. SREBP-1a activates transcription of the LDL receptor and the HMG-CoA synthase promoters more potently than does the shorter SREBP-1c, but both are almost equally active in stimulating the transcription of the FAS promoter (87). Cells in tissue culture express much more SREBP-1a than SREBP-1c; some even exclusively produce SREBP-1a (89). However, intact organs, for example the liver or adipose tissue, express much higher levels of SREBP-1c than -1a (89). Because of the different effect of the two isoforms of the transcription factor on the expression of enzymes responsible for cholesterol synthesis versus those responsible for fatty acid synthesis, the following hypothesis has been raised. Cell lines in tissue culture need cholesterol constantly for membrane synthesis. Therefore, they contain both active SREBP-1a and SREBP-2, to allow for active cholesterol synthesis. Cells in vivo, however, only produce a basal level of active SREBP-1c to support fatty acid synthesis. Only upon cholesterol depletion do these cells induce proteolysis of SREBP-2. The differences in SREBP-1a and SREBP-1c activity were further demonstrated in transgenic mice, which overexpress the active form of either protein under the control of a strong promoter. Whereas animals overexpressing SREBP-1a accumulate large amounts lipids in their livers, leading to a fourfold enlargement of this organ (86), animals overexpressing SREBP-1c show only a 33% increase in liver mass (87). The mRNA for many SREBP target genes were strongly upregulated in SREBP-1a transgenic animals, despite the fact that the accumulation of cholesterol and triglycerides should have led to a feedback repression (86). In contrast, SREBP-1c transgenic mice showed no increase in mRNA for cholesterol biosynthetic enzymes (87). The only abnormality in these animals was a slight, two- to fourfold increase in mRNA for enzymes responsible for fatty acid synthesis (87). These enzymes, however, were elevated 20-fold in mice overexpressing SREBP-1a (86). Therefore, in vivo, SREBP-1c seems to be responsible for the low level synthesis of mRNA required for fatty acid synthesis, whereas SREBP-1a is strongly induced following sterol deprivation to upregulate cholesterol biosynthesis. SREBP-2 transgenic animals are not yet available for analysis. Recently, mice carrying targeted deletions in the gene for SREBP-1 were generated (88). Interestingly, animals homozygous for the disruption show two distinct phenotypes, although they are genetically identical. Although between 50 and 85% of the mice die in utero at day 11 of gestation, the surviving animals are phenotypically normal. Although they express no SREBP-1, these animals survive and reach adulthood. Their only phenotype is a slight (1.5-fold) increase in SREBP-2 expression and a resulting two- to threefold increase in the level of cleaved SREBP-2 protein found in the nucleus. Consequently, it 697 698 HEIKE L. PAHL countless possibilities of variation and regulation not previously imagined. Likewise, the unconventional proteolysis of the SREPB transcription may find parallels in other regulatory pathways. Characterization and cloning of the unusual proteases involved in SREBP cleavage will indicate whether these proteins uniquely function to regulate SREBP activity or whether they belong to a family of proteases that regulates other cellular processes as well. We look forward to the discovery of additional cellular regulatory systems, based on the paradigms first discovered while investigating ER-nuclear signaling. REFERENCES 1. BAEUERLE, P. A., AND D. BALTIMORE. I kappa B: a specific inhibitor of the NF-kappa B transcription factor. Science 242: 540 – 546, 1988. 2. BAEUERLE, P. A., AND D. BALTIMORE. NF-kB: ten years after. Cell 87: 13–20, 1996. 3. BAEUERLE, P. A., AND T. HENKEL. Function and activation of NF-kB in the immune system. Annu. Rev. Immunol. 12: 141–179, 1994. 4. BAEUERLE, P. A., R. A. RUPEC, AND H. L. PAHL. Reactive oxygen intermediates as second messengers of a general pathogen response. Pathol. Biol. 44: 29 –35, 1996. 5. BEG, A. A., AND D. BALTIMORE. An essential role for NF-kB in preventing TNF-a-induced cell death. Science 274: 782–784, 1996. 6. BEG, A. A., T. S. FINCO, P. V. NANTERMET, AND A. S. BALDWIN, JR. Tumor necrosis factor and interleukin-1 lead to phosphorylation and loss of IkB-a: a mechanism for NF-kB activation. Mol. Cell. Biol. 13: 3301–3310, 1993. 7. BREWER, J. W., J. L. CLEVELAND, AND L. M. HENDERSHOT. A pathway distinct from the mammalian unfolded protein response regulates expression of endoplasmic reticulum chaperones in nonstressed cells. EMBO J. 16: 7207–7216, 1997. 8. BRIGGS, M. R., C. YOKOYAMA, X. WANG, M. S. BROWN, AND J. L. GOLDSTEIN. Nuclear protein that binds sterol regulatory element of LDL receptor promoter. I. Identification of the protein and delineation of its target nucleotide sequence. J. Biol. Chem. 268: 14490 –14496, 1993. 9. BROSTROM, M. A., C. CADE, C. R. PROSTKO, D. GMITTERYELLEN, AND C. O. BROSTROM. Accommodation of protein synthesis to chronic deprivation of intracellular sequestered calcium. A putative role for GRP78. J. Biol. Chem. 265: 20539 –20546, 1990. 10. BROWN, M. S., AND J. L. GOLDSTEIN. A receptor-mediated pathway for cholesterol homeostasis. Science 232: 34 – 47, 1986. 11. CAO, X., Y. ZHOU, AND A. S. LEE. Requirement of tyrosine- and serine/threonine kinases in the transcriptional activation of the mammalian grp78/BiP promoter by thapsigargin. J. Biol. Chem. 270: 494 –502, 1995. 12. CARLSON, J. A., B. B. ROGERS, R. N. SIFERS, M. J. FINEGOLD, S. M. CLIFT, F. J. DEMAYO, D. W. BULLOCK, AND S. L. C. WOO. Accumulation of PiZ a1-antitrypsin causes liver damage in transgenic mice. J. Clin. Invest. 83: 1183–1190, 1989. 13. CARSTEA, E. D., J. A. MORRIS, K. G. COLEMAN, S. K. LOFTUS, D. ZHANG, C. CUMMINGS, J. GU, M. A. ROSENFELD, W. J. PAVAN, D. B. KRIZMAN, J. NAGLE, M. H. POLYMEROPOULOS, S. L. STURLEY, Y. A. IOANNOU, M. E. HIGGINS, M. COMLY, A. COONEY, A. BROWN, C. R. KANESKI, E. J. BLANCHETTE-MACKIE, N. K. DWYER, E. B. NEUFELD, T. Y. CHANG, L. LISCUM, AND D. A. TAGLE. Niemann-Pick C1 disease gene: homology to mediators of cholesterol homeostasis. Science 277: 228 –231, 1997. 14. CHANG, S. C., A. E. ERWIN, AND A. S. LEE. Glucose-regulated protein (GRP94 and GRP78) genes share common regulatory domains and are coordinately regulated by common trans-acting factors. Mol. Cell. Biol. 9: 2153–2162, 1989. 15. CHAPMAN, R. E., AND P. WALTER Translational attenuation mediated by an mRNA intron. Curr. Biol. 7: 850 – 859, 1997. 16. CHENG, S. H., R. J. GREGORY, J. MARSHALL, S. PAUL, D. W. SOUZA, G. A. WHITE, C. R. O’RIORDAN, AND A. E. SMITH. Defective intracellular transport and processing of CFTR is the molecular basis of most cystic fibrosis. Cell 63: 827– 834, 1990. 17. COX, J. S., R. E. CHAPMAN, AND P. WALTER. The unfolded protein response coordinates the production of endoplasmic reticulum protein and endoplasmic reticulum membrane. Mol. Biol. Cell 8: 1805–1814, 1997. 18. COX, J. S., C. E. SHAMU, AND P. WALTER. Transcriptional induction of genes encoding endoplasmic reticulum resident proteins requires a transmembrane protein kinase. Cell 73: 1197–1206, 1993. 19. COX, J. S., AND P. WALTER. A novel mechanism for regulating activity of a transcription factor that controls the unfolded protein response. Cell 87: 391– 404, 1996. 20. DOOLEY, K. A., S. MILLINDER, AND T. F. OSBORNE. Sterol regulation of 3-hydroxy-3-methylglutaryl-coenzyme A synthase gene through a direct interaction between sterol regulatory element binding protein and the trimeric CCAAT-binding factor/nuclear factor Y. J. Biol. Chem. 273: 1349 –1356, 1998. 21. DRUMMOND, I. A., A. S. LEE, E. RESENDEZ, AND R. A. STEINHARDT. Depletion of intracellular calcium stores by calcium ionophore A23187 induces the genes for glucose-regulated proteins in hamster fibroblasts. J. Biol. Chem. 262: 12801–12805, 1987. 22. DUNCAN, E. A., M. S. BROWN, J. L. GOLDSTEIN, AND J. SAKAI. Cleavage site for sterol-regulated protease localized to a Leu-Ser bond in the luminal loop of sterol regulatory element-bindingprotein 2. J. Biol. Chem. 272: 12778 –12785, 1997. 23. D’URSO, D., R. PRIOR, R. GREINER-PETTER, A. A. W. M. GABREELS-FESTEN, AND H. W. MÜLLER. Overloaded endoplasmic reticulum-Golgi compartments, a possible pathomechanism of peripheral neuropathies caused by mutation of the peripheral myelin protein PMP22. J. Neurosci. 18: 731–740, 1998. 24. DYCYICO, M. J., S. G. GRANT, K. FELTS, W. S. NICHOLS, S. A. GELLER, J. H. HAGER, A. J. POLLARD, S. W. KOHLER, H. P. SHORT, F. R. JIRIK, D. HANAHAN, AND J. A. SORGE. Neonatal hepatitis induced by a1-antitrypsin: a transgenic mouse model. Science 242: 1409 –1412, 1988. 25. ERICSSON, J., S. M. JACKSON, B. C. LEE, AND P. A. EDWARDS. Sterol regulatory element binding protein binds to a cos element in the promoter of the farnesyl diphosphate synthase gene. Proc. Natl. Acad. Sci. USA 93: 945–950, 1996. 26. GOLDSTEIN, J. L., AND M. S. BROWN. Regulation of the mevalonate pathway. Nature 343: 425– 430, 1990. 27. GUO, Q., B. L. SOPHER, K. FURUKAWWA, D. G. PHAM, N. ROBINSON, G. M. MARTIN, AND M. P. MATTSON. Alzheimer’s presenilin mutation sensitizes neural cells to apoptosis induced by trophic factor withdrawal and amyloid beta-peptide: involvement of calcium and oxyradicals. J. Neurosci. 17: 4212– 4222, 1997. 28. HENKEL, T., T. MACHLEIDT, I. ALKALAY I., Y. BEN-NERIAH, M. KRÖNKE, AND P. A. BAEUERLE. Rapid proteolysis of IkB-a is necessary for activation of transcription factor NF-kB. Nature 365: 182–184, 1993. 29. HUA, X., A. NOHTURFFT, J. L. GOLDSTEIN, AND M. S. BROWN. Sterol resistance in CHO cells traced to a point mutation in SREBP cleavage activating protein (SCAP). Cell 87: 415– 426, 1996. 30. HUA, X., J. SAKAI, Y. K. HO, J. L. GOLDSTEIN, AND M. S. BROWN. Hairpin orientation of sterol regulatory element-binding protein-2 in cell membranes as determined by protease protection. J. Biol. Chem. 270: 29422–29427, 1995. 31. HUA, X., C. YOKOYAMA, J. WU, M. R. BRIGGS, M. S. BROWN, AND J. L. GOLDSTEIN. SREBP-2, a second basic-helix-loop-helixleucine zipper protein that stimulates transcription by binding to a sterol regulatory element. Proc. Natl. Acad. Sci. USA 90: 11603– 11607, 1993. 32. JACKSON, M. R., T. NILSSON, AND P. A. PETERSON. Identification of a consensus motif for retention of transmembrane proteins in the endoplasmic reticulum. EMBO J. 9: 3153–3162, 1990. Downloaded from http://physrev.physiology.org/ by 10.220.33.6 on May 2, 2017 I express my deep gratitude to Patrick Baeuerle for the continuous support, encouragement, and help I have received from him. The EOR was discovered and elucidated in his lab. Thanks also to Alexander Knorre for critical review of the manuscript. Volume 79 July 1999 ENDOPLASMIC RETICULUM SIGNAL TRANSDUCTION 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. murine ERp72 gene to protein traffic. DNA Cell Biol. 16: 1123–1131, 1997. MC CORMICK, T. S., K. S. MCCOLL, AND C. W. DISTELHORST. Mouse lymphoma cells destined to undergo apoptosis in response to thapsigargin treatment fail to generate a calcium-mediated grp78/grp94 stress response. J. Biol. Chem. 272: 6087– 6092, 1997. MCELVANEY, N. G., H. NAKAMURA, P. BIRRER, C. A. HERBERT, W. L. WONG, M. ALPHONSO, J. B. BAKER, M. A. CATALANO, AND R. G. CRYSTAL. Modulation of airway inflammation in cystic fibrosis: in vivo suppression of interleukin-8 levels on the respiratory epithelial surface by aerosolization of recombinant secretory leukoprotease inhibitor. J. Clin. Invest. 90: 1296 –1301, 1992. MEDEIROS-NETO, G., P. S. KIM, S. E. YOO, J. VONO, H. M. TARGOVNIK, R. CAMARGO, S. A. HOSSAIN, AND P. ARVAN. Congenital hypothyroid goiter with deficient thyroglobulin. J. Clin. Invest. 98: 2838 –2844, 1996. MERCURIO, F., H. ZHU, B. W. MURRAY, A. SHEVCHENKO, B. L. BENNETT, J. W. LI, D. B. YOUNG, M. BARBOSA, M. MANN, A. MANNING, AND A. RAO. IKK-1 and IKK-2: cytokine-activated IkB kinases essential for NF-kB activation. Science 278: 860 – 866, 1997. MEYER, M., W. H. CASELMANN, V. SCHLUTER, R. SCHRECK, P. H. HOFSCHNEIDER, AND P. A. BAEUERLE. Hepatitis B virus transactivator MHBst: activation of NF-kappa B, selective inhibition by antioxidants and integral membrane localization. EMBO J. 11: 2991–3001, 1992. MORI, K., T. KAWAHARA, H. YOSHIDA, H. YANAGI, AND T. YURA. Signalling from endoplasmic reticulum to nucleus: transcription factor with a basic-leucine zipper motif is required for the unfolded protein-response pathway. Genes Cells 1: 803– 817, 1996. MORI, K., W. MA, M.-J. GETHING, AND J. SAMBROOK. A transmembrane protein with a cdc21/CDC28-related kinase activity is required for signalling from the ER to the nucleus. Cell 74: 743–756, 1993. MORI, K., N. OGAWA, T. KAWAHARA, H. YANAGI, AND T. YURA. Palindrome with spacer of one nucleotide is characteristic of the cis-acting unfolded protein response element is Saccharoomyces cerevisiae. J. Biol. Chem. 272: 9912–9920, 1996. MORI, K., A. SANT, K. KOHNO, K. NORMINGTON, M.-J. GETHING, AND J. F. SAMBROOK. A 22 bp cis-acting element is necessary and sufficient for the induction of the yeast KAR2 (BiP) gene by unfolded proteins. EMBO J. 11: 2583–2593, 1992. MORRIS, J. A., A. J. DORNER, C. A. EDWARDS, L. M. HENDERSHOT, AND R. J. KAUFMAN. Immunoglobulin binding protein (BiP) function is required to protect cells from endoplasmic reticulum stress but is not required for the secretion of selective proteins. J. Biol. Chem. 272: 4327– 4334, 1997. NAKAMURA, H., K. YOSHIMURA, N. G. MCELVANEY, AND R. G. CRYSTAL. Neutrophil elastase in respiratory epithelial lining fluid of individuals with cycstic fibrosis induces interleukin-8 gene expression in a human bronchial epithelial cell line. J. Clin. Invest. 89: 1478 –1484, 1992. PAHL, H. L., AND P. A. BAEUERLE. Expression of influenza hemagglutinin activates transcription factor NF-kB. J. Virol. 69: 1480 – 1484, 1995. PAHL, H. L., AND P. A. BAEUERLE. A novel signal transduction pathway from the endoplasmic reticulum to the nucleus is mediated by transcription factor NF-kB. EMBO J. 14: 2580 –2588, 1995. PAHL, H. L., AND P. A. BAEUERLE. Activation of NF-kB by ER stress requires both Ca21 and reactive oxygen intermediates as messengers. FEBS Lett. 392: 129 –136, 1996. PAHL, H. L., M. SESTER, H.-G. BURGERT, AND P. A. BAEUERLE. Activation of transcription factor NF-kB by adenovirus E3/19K requires its ER-retention. J. Cell Biol. 132: 511–522, 1996. PATTERSON, P. H. Cytokines in Alzheimer’s disease and multiple sclerosis. Curr. Opin. Neurobiol. 5: 642– 646, 1995. PRICE, B. D., L. A. MANNHEIM-RODMAN, AND S. K. CALDERWOOD. Brefeldin A, thapsigargin, and AlF42 stimulate the accumulation of GRP78 mRNA in a cycloheximide dependent manner, whilst induction by hypoxia is independent of protein synthesis. J. Cell. Physiol. 152: 545–552, 1992. RAMAKRISHNAN, M., A. H. SCHÖNTHAL, AND A. S. LEE. Endoplasmic reticulum stress-inducible protein grp94 is associated with Downloaded from http://physrev.physiology.org/ by 10.220.33.6 on May 2, 2017 33. KAWAHARA, T., H. YANAGI, T. YURA, AND K. MORI. Endoplasmic reticulum stress-induced mRNA splicing permits synthesis of transcription factor Hac1p/Ern4p that activates the unfolded protein response. Mol. Biol. Cell 1997, 8: 1845–1862, 1997. 34. KIM, J. B., G. D. SPOTTS, Y.-D. HALVORSEN, H.-M. SHIH, T. ELLENBERGER, H. TOWLE, AND B. M. SPIEGELMAN. Dual DNA binding specificity of ADD1/SREBP1 controlled by a single amino acid in the basic helix-loop-helix domain. Mol. Cell. Biol. 15: 2582– 2588, 1995. 35. KIM, K. Y., K. S. KIM, AND A. S. LEE. Regulation of the glucoseregulates protein genes by beta-mercaptoethanol requires de novo protein synthesis and correlates with inhibition of protein glycosylation. J. Cell. Physiol. 133: 553–559, 1987. 36. KONSTAN, M. W., AND M. BERGER. Infection and Inflammation of the Lung in Cystic Fibrosis, edited by P. B. Davis. New York: Dekker, 1993, p. 219 –276. 37. KOVACS, D. M., H. J. FAUSETT, K. J. PAGE, T.-W. KIM, R. D. MOIR, D. E. MERRIAM, R. D. HOLLISTER, O. G. HALLMARK, R. MANCINI, K. M. FELSENSTEIN, B. T. HYMAN, R. E. TANZI, AND W. WASCO. Alzheimer-associated presenilins 1 and 2: neuronal expression in brain and localization to intracellular membranes in mammalian cells. Nature Med. 2: 224 –229, 1996. 38. KOZUTSUMI, Y., M. SEGAL, K. NORMINGTON, M.-J. GETHING, AND J. SAMBROOK. The presence of malfolded proteins in the endoplasmic reticulum signals the induction of glucose-regulated proteins. Nature 332: 462– 464, 1988. 39. KUMEGAI, H., K. T. CHUN, AND R. D. SIMONI. Molecular dissection of the role of the transmembrane domain in the regulated degradtion of 3-hydroxy-3-methylglutaryl coenzyme A reductase. J. Biol. Chem. 270: 19107–19113, 1995. 40. LEE, A. S. Coordinated regulation of a set of genes by glucose and calcium ionophores in mammalian cells. Trends Biochem. Sci. 12: 20 –23, 1987. 41. LEE, A. S. Mammalian stress response: induction of the glucoseregulated protein family. Curr. Opin. Cell Biol. 4: 267–273, 1992. 42. LEE, A. S., J. BELL, AND J. TING. Biochemical characterization of the 94- and 78-kilodalton glucose-regulated proteins in hamster fibroblasts. J. Biol. Chem. 259: 4616 – 4621, 1984. 43. LEVINE, S. J. Bronchial epithelial cell-cytokine interactions in airway inflammation. J. Invest. Med. 43: 241–249, 1995. 44. LI, W. W., Y. HSIUNG, V. WONG, K. GALVIN, Y. ZHOU, Y. SHI, AND A. S. LEE. Suppression of grp78 core promoter element-mediated stress induction by the dbpA and dbbpB (YB-1) cold shock domain proteins. Mol. Cell. Biol. 17: 61– 68, 1997. 45. LI, W. W., Y. HSIUNG, Y. ZHOU, B. ROY, AND A. S. LEE. Induction of the mammalian GRP78/BiP gene by Ca21 depletion and formation of aberrant proteins: activation of the conserved stress-inducible grp core promoter element by the human nuclear factor YY1. Mol. Cell. Biol. 17: 54 – 60, 1997. 46. LI, W. W., L. SISTONEN, R. I. MORIMOTO, AND A. S. LEE. Stress induction of the mammalian GRP78/BiP protein gene: in vivo genomic footprinting and identification of p70CORE from human nuclear extracts as a DNA-binding component specific to the stress regulatory element. Mol. Cell. Biol. 14: 5533–5546, 1994. 47. LITTLE, E., M. RAMAKRISHAN, B. ROY, G. GAZIT, AND A. S. LEE. The glucose-regulated proteins (GRP78 and GRP94): functions, gene regulation and applications. Crit. Rev. Eukaryotic Gene Exp. 4: 1–18, 1994. 48. LIU, H., R. C. BOWES, B. VAN DER WATER, C. SILLENCE, J. F. NAGELKERKE, AND J. L. STEVENS. Endoplasmmic reticulum chaperones grp78 and calreticulin prevent oxidative stress, Ca21 disturbances, and cell death in renal epithelial cells. J. Biol. Chem. 272: 21751–21759, 1997. 49. LOPEZ, J. M., M. K. BENNETT, H. B. SANCCHEZ, J. M. ROSENFIELD, AND T. F. OSBORNE. Sterol regulation of acetyl coenzyme A carboxylase: a mechanism for coordinate control of cellular lipid. Proc. Natl. Acad. Sci. USA 93: 1049 –1053, 1996. 50. LUKACS, G. L., A. MOHAMED, N. KARTNER, X.-B. CHANG, J. R. RIORDAN, AND S. GRINSTEIN. Conformational maturation of CFTR but not its mutant counterpart (DF508) occurs in the endoplasmic reticulum and requires ATP. EMBO J. 13: 6076 – 6086, 1994. 51. MARCUS, N., AND M. GREEN. NF-Y, a CCAAT box-binding protein, is one of the trans-acting factors necessary for the response of the 699 700 70. 71. 72. 73. 74. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. an Mg21-dependent serine kinase activity modulated by Ca21 and GRP78/BiP. J. Cell. Physiol. 170: 115–129, 1997. REDDY, P. S., AND R. B. CORLEY. Assembly, sorting and exit of oligomeric proteins from the endoplasmic reticulum. Bioessays 20: 546 –554, 1998. RÉGNIER, C. H., H. YEONG SONG, X. GAO, D. V. GOEDDEL, Z. CAO, AND M. ROTHE. Identification and characterization of an IkB kinase. Cell 90: 373–383, 1997. ROY, B., AND A. S. LEE. Transduction of calcium stress through interaction of the human transcription factor CBF with the proximal CCAAT regulatory element of the grp78/BiP promoter. Mol. Cell. Biol. 15: 2263–2274, 1995. ROY, B., W. W. LI, AND A. S. LEE. Calcium-sensitive transcriptional activation of the proximal CCAAT regulatory element of the grp78/ BiP promoter by the human nuclear factor CBF/NF-Y. J. Biol. Chem. 271: 28995–29002, 1996. SAKAI, J., E. A. DUNCAN, R. B. RAWSON, X. HUA, M. S. BROWN, AND J. L. GOLDSTEIIN. Sterol-regulated release of SREBP-2 from cell membranes requires two sequential cleavages, one within a transmembrane segment. Cell 85: 1037–1046, 1996. SAKAI, J., A. NOHTURFFT, D. CHENG, Y. K. HO, M. S. BROWN, AND J. L. GOLDSTEIN. Identification of complexes between the COOH-terminal domains of sterol regulatory element-binding proteins (SREBPs) and SREBP cleavage-activating protein. J. Biol. Chem. 272: 20213–20221, 1997. SANCHEZ, H. B., Y. LYNN, AND T. F. OSBORNE. Cooperation by sterol regulatory element-binding protein and Sp1 in sterol regulation of low density lipoprotein receptor gene. J. Biol. Chem. 270: 1161–1169, 1995. SATO, R., J. YANG, X. WANG, M. J. EVANS, Y. K. HO, J. L. GOLDSTEIN, AND M. S. BROWN. Assignment of the membrane attachment, DNA binding, and transcriptional activation domains of sterol regulatory element-binding protein-1 (SREBP-1). J. Biol. Chem. 269: 17267–17273, 1994. SCHEEK, S., M. S. BROWN, AND J. L. GOLDSTEIN. Sphingomyelin depletion in cultured cells blocks proteolysis of sterol regulatory element binding proteins at site 1. Proc. Natl. Acad. Sci. USA 94: 11179 –11183, 1997. SCHRECK, R., AND P. A. BAEUERLE. A role of oxygen radicals as second messengers. Trends Cell. Biol. 1: 39 – 42, 1991. SCHRECK, R., R. GRASSMANN, B. FLECKENSTEIN, AND P. A. BAEUERLE. Antioxidants selectively suppress activation of NFkappa B by human T-cell leukemia virus type I Tax protein. J. Virol. 66: 6288 – 6293, 1992. SCHRECK, R., B. MEIER, D. N. MANNEL, W. DROGE, AND P. A. BAEUERLE. Dithiocarbamates as potent inhibitors of nuclear factor kappa B activation in intact cells. J. Exp. Med. 175: 1181–1194, 1992. SHAMU, C. E., J. S. COX, AND P. WALTER. The unfolded-proteinresponse pathway in yeast. Trends Cell Biol. 4: 57– 60, 1994. SHAMU, C. E., AND P. WALTER. Oligomerization and phosphorylation of the Ire1p kinase during intracellular signaling from the endoplasmic reticulum to the nucleus. EMBO J. 15: 3028 –3039, 1996. SHARP, H. L. Alpha1-antitrypsin deficiency. Hosp. Pract. 6: 83–96, 1971. SHENG, Z., N. OTANI, M. S. BROWN, AND J. L. GOLDSTEIN. Independent regulation of sterol regulatory element binding proteins 1 and 2 in hamster liver. Proc. Natl. Acad. Sci. USA 92: 935–938, 1995. SHIMANO, H., J. D. HORTON, R. E. HAMMER, I. SHIMOMURA, M. S. BROWN, AND J. L. GOLDSTEIN. Overproduction of cholesterol and fatty acids causes massive liver enlargement in transgenic mice expressing truncated SREBP-1a. J. Clin. Invest. 98: 1575– 1584, 1996. SHIMANO, H., J. D. HORTON, I. SHIMOMURA, R. E. HAMMER, M. S. BROWN, AND J. L. GOLDSTEIN. Isoform 1c of sterol regulatory element binding protein is less active than isoform 1a in livers of transgenic mice and in cultured cells. J. Clin. Invest. 99: 846 – 854, 1997. SHIMANO, H., I. SHIMOMURA, R. E. HAMMER, J. HERZ, J. L. GOLDSTEIN, M. S. BROWN, AND J. D. HORTON. Elevated levels of SREBP-2 and cholesterol synthesis in livers of mice homozygous 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. Volume 79 for a targeted disruption of the SREBP-1 gene. J. Clin. Invest. 100: 2115–2124, 1997. SHIMOMURA, I., H. SHIMANO, J. D. HORTON, J. L. GOLDSTEIN, AND M. S. BROWN. Differential expression of exons 1a and 1c in mRNAs for sterol regulatory element binding protein-1 in human and mouse organs and cultured cells. J. Clin. Invest. 99: 838 – 845, 1997. SIDRAUSKI, C., R. CHAPMAN, AND P. WALTER. The unfolded protein response: an intracellular signalling pathway with many surprising features. Trends Cell Biol. 8: 245–249, 1998. SIDRAUSKI, C., J. S. COX, AND P. WALTER. tRNA ligase is required for regulated mRNA splicing in the unfolded protein response. Cell 87: 405– 413, 1996. SIDRAUSKI, C., AND P. WALTER. The transmembrane kinase Ire1p is a site-specific endonuclease that initiates mRNA splicing in the unfolded protein response. Cell 90: 1031–1039, 1997. SIFERS, R. N., M. J. FINEGOLD, AND S. L. C. WOO. Molecular biology and genetics of alpha-1-antitrypsin deficiency. Semin. Liver Dis. 12: 301–310, 1992. SMITH, J. R., T. F. OSBORNE, J. L. GOLDSTEIN, AND M. S. BROWN. Identification of nucleotides responsible for enhancer activity of sterol regulatory element in low density lipoprotein receptor gene. J. Biol. Chem. 265: 2306 –2310, 1990. SUEDHOFF, T. C., D. R. VAN DER WESTHUYZEN, J. L. GOLDSTEIN, M. S. BROWN, AND D. W. RUSSELL. Three direct repeats and a TATA-like sequence are required for regulated expression of the human low density lipoprotein receptor gene. J. Biol. Chem. 262: 10773–10778, 1987. SUN, S.-C., P. A. GANCHI, C. BÉRAUD, D. W. BALLARD, AND W. C. GREENE. Autoregulation of the NF-kappa B transactivator RelA (p65) by multiple cytoplasmic inhibitors containing ankyrin motifs. Proc. Natl. Acad. Sci. USA 87: 1346 –1350, 1994. SVEGER, T. Liver disease in alpha1 antitrypsin deficiency detected by screening 200,000 infants. N. Engl. J. Med. 294: 1316 –1321, 1976. TAM, S. S., D. H. LEE, E. Y. WANG, O. G. MUNROE, AND C. Y. LAU. Tepoxalin, a novel dual inhibitor of the prostaglandin H synthase cyclooxygenase and peroxidase activities. J. Biol. Chem. 270: 13948 –13955, 1995. TECKMAN, J. H., AND D. H. PERLMUTTER. The endoplasmatic reticulum degradation pathway for mutant secretory proteins a1antitrypsin Z and S is distinct from that for an unassembled membrane protein. J. Biol. Chem. 271: 13215–13220, 1996. THEVENIN, C., S.-J. KIM, P. RIECKMANN, H. FUJIKI, M. A. NORCROSS, M. B. SPORN, A. S. FAUCI, AND J. H. KEHRL. Induction of nuclear factor-kB and the human immunodeficiency virus long terminal repeat by okadaic acid, a specific inhibitor of phosphatases 1 and 2A. Nature New Biol. 2: 793– 800, 1991. TIRASOPHON, W., A. A. WELIHINDA, AND R. J. KAUFMAN. A stress response pathway from the endoplasmic reticulum to the nucleus requires a novel bifunctional kinase/endoribonuclease (Ire1p) in mammalian cells. Genes Dev. 12: 1812–1824, 1998. TONG-STARKSEN, S. E., P. A. LUCIW, AND B. M. PETERLIN. Signalling through T lymphocyte surface proteins, TCR/CD3 and CD28, activates the HIV-1 long terminal repeat. J. Immunol. 142: 702–707, 1989. TONTONOZ, P., J. B. KIM, R. A. GRAVES, AND B. M. SPIEGELMAN. ADD1: a novel helix-loop-helix transcription factor associated with adipocyte determination and differentiation. Mol. Cell. Biol. 13: 4753– 4759, 1993. TRAENCKNER, E. B.-M., H. L. PAHL, T. HENKEL, K. N. SCHMIDT, S. WILK, AND P. A. BAEUERLE. Phosphorylation of human IkB-a on serine 32 and 36 controls IkB-a proteolysis and NF-kB activation in response to diverse stimuli. EMBO J. 14: 2876 –2883, 1995. TRAENCKNER, E. B.-M., S. WILK, AND P. A. BAEUERLE. A proteasome inhibitor prevents activation of NF-kB and stabilizes a newly phosphorylated form of IkB-a that is still bound to NF-kB. EMBO J. 13: 5433–5441, 1994. VAN ANTWERP, D. J., S. J. MARTIN, T. KAFRI, D. R. GREEN, AND I. M. VERMA. Suppression of TNF-a-induced apoptosis by NF-kB. Science 274: 787–789, 1996. WANG, C.-Y., M. W. MAYO, AND A. S. BALDWIN. TNF- and cancer therapy-induced apoptosis: potentiation by inhibition of NF-kB. Science 274: 784 –787, 1996. Downloaded from http://physrev.physiology.org/ by 10.220.33.6 on May 2, 2017 75. HEIKE L. PAHL July 1999 ENDOPLASMIC RETICULUM SIGNAL TRANSDUCTION 113. 114. 115. 116. 117. which interacts with CTF/NF-1. Mol. Cell. Biol. 11: 5612–5623, 1991. WORONICZ, J. D., X. GAO, Z. CAO, M. ROTHE, AND D. V. GOEDDEL. IkB kinase-b: NF-kB activation and complex formation with IkB kinase-a and NIK. Science 278: 866 – 869, 1997. WU, F. S., Y.-C. CLARK, D. ROUFA, AND A. MARTONOSI. Selective stimulation of the synthesis of an 80,000-Dalton protein by calcium ionophores. J. Biol. Chem. 256: 5309 –5312, 1981. YOKOYAMA, C., X. WANG, M. BRIGGS, A. ADMON, J. WU, X. HUA, J. L. GOLDSTEIN, AND M. S. BROWN. SREBP-1, a basic-helix-loophelix-leucine zipper protein that controls transcription of the low density lipoprotein receptor gene. Cell 75: 187–197, 1993. ZANDI, E., D. M. ROTHWARF, M. DELHASE, M. HAYAKAWA, AND M. KARIN. The IkB kinase complex (IKK) contains two kinase subunits IKKa and IKKb necessary for IkB phosphorylation and NF-kB activation. Cell 91: 243–252, 1997. ZHOU, Y., AND A. S. LEE. Mechanism for the suppression of the mammalian stress response by genistein, an anticancer phytoestrogen from soy. J. Natl. Cancer Inst. 90: 381–388, 1998. Downloaded from http://physrev.physiology.org/ by 10.220.33.6 on May 2, 2017 108. WANG, X., J. PAI, E. A. WIEDENFELD, J. C. MEDINA, C. A. SLAUGHTER, J. L. GOLDSTEIN, AND M. S. BROWN. Purification of an interleukin-1b converting enzyme-related cysteine protease that cleaves sterol regulatory element-binding proteins between the leucine zipper and transmembrane domains. J. Biol. Chem. 270: 18044 –18050, 1995. 109. WANG, X., R. SATO, M. S. BROWN, X. HUA, AND J. L. GOLDSTEIN. SREBP-1, a membrane-bound transcription factor released by sterol-regulated proteolysis. Cell 77: 53– 62, 1994. 110. WELIHINDA, A. A., W. TIRASOPHON, S. R. GREEN, AND R. J. KAUFMAN. Gene induction in response to unfolded protein in the endoplasmic reticulum is mediated through Ire1p kinase interaction with a transcriptional coactivator complex containing Ada5. Proc. Natl. Acad. Sci. USA 94: 4289 – 4294, 1997. 111. WHITE, M. J., J. M. LOPES, AND S. A. HENRY. Inositol metabolism in yeast. Adv. Microb. Physiol. 32: 1–51, 1991. 112. WOODEN, S. K., L.-J. LI, D. NAVARRO, I. QADRI, L. PEREIRA, AND A. S. LEE. Transactivation of the grp78 promoter by malfolded proteins, glycosylation block, and calcium ionophores is mediated through a proximal region containing a CCAAT motif 701