| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 279, Issue 27, 27948-27956, July 2, 2004
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Gene Expression Is Activated by Endoplasmic Reticulum Stress through an Unfolded Protein Response Element Downstream of the Protein Coding Sequence*
From the Department of Biochemistry and Molecular Biology, Center for Mammalian Genetics, and Center for Nutritional Sciences, University of Florida College of Medicine, Gainesville, Florida 32610-0245
Received for publication, December 19, 2003 , and in revised form, April 14, 2004.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
(C/EBP) is a member of the bZIP family of transcription factors that contribute to the regulation of a wide range of important cellular processes. The data in the present study document that transcription from the human C/EBP gene is induced in response to endoplasmic reticulum stress, such as glucose deprivation, or treatment of cells with tunicamycin or thapsigargin. Transient transfection of C/EBP genomic fragments linked to a luciferase reporter gene demonstrated that the C/EBP promoter plays no major regulatory role. Instead, by deletion analysis it was discovered that a 46-bp region, located at a genomic site that corresponds to the 3'-untranslated region of the C/EBP mRNA, harbored an element that was required for the stress response. Mutagenesis demonstrated that a cis-regulatory element located at nt +1614–1621 (5'-TGACGCAA-3') is responsible for activation of the C/EBP gene. Electrophoresis mobility shift analysis revealed that proteins are bound to this element and that the amount of binding is increased following glucose deprivation. This element is homologous to a previously reported mammalian unfolded protein response element that binds XBP-1. Consistent with those data, overexpression of XBP-1 caused an increase in transcription that was mediated by the C/EBP mammalian unfolded protein response element. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
(C/EBP)1 is a member of a family of transcription factors that also includes C/EBP
, -
, -
, and -
and C/EBP homology protein (CHOP) reviewed in Refs. 1–3. The C/EBP members contain a basic leucine zipper (bZIP) domain at their C terminus, which is responsible for selective dimerization with other bZIP family members, as reviewed by Newman and Keating (4). Although C/EBP plays a role in a wide range of important cellular processes, such as adipocyte differentiation, carbohydrate metabolism, inflammation, and cellular proliferation, investigation of the transcriptional control of the C/EBP gene itself is limited. It has been reported that the cAMP-response element-like sequences, located within the rat C/EBP proximal promoter, are required for C/EBP expression and IL-6-mediated induction of the gene during the acute phase response (5).
Disturbance of the protein folding process in the endoplasmic reticulum (ER) activates an ER stress-signaling pathway called the unfolded protein response (UPR), reviewed in Refs. 6–8. In the yeast Saccharomyces cerevisiae, ER stress activates the kinase/endonuclease Ire1p, which in turn mediates an unconventional splicing of HAC1 mRNA. The processed HAC1 mRNA codes for a bZIP protein, Hac1p, which binds to a UPR element (UPRE) in the appropriate target genes. Two mammalian Irelp forms, IRE1 and IRE1 (9, 10), have been identified, and XBPI has been described as the mammalian counterpart to yeast Hac1p (11–13). However, at least two ER stress-responsive pathways exist in mammals, and proteolytic activation of the transcription factor ATF6 also plays an important role (14–16). The precursor form of ATF6 is an ER transmembrane protein that is proteolytically cleaved in response to ER stress, and the N-terminal portion is thereby released as an active transcription factor, which translocates to the nucleus. A bipartite mammalian ER stress response element (ERSE), 5'-CCAATN9CCACG-3', was identified (17, 18), which was shown to bind either XBP-1 or ATF6 at the CCACG half-site (17). However, more recently it has been demonstrated that XBP-1 activates a second set of ER stress-responsive genes by binding to a genomic element composed of the consensus sequence 5'-TGACGTG(G/A)-3' (12, 19), and this element has been referred to as the mammalian UPRE (11).
Asparagine synthetase (ASNS), which catalyzes asparagine biosynthesis, is transcriptionally regulated in response to a variety of cellular stress signals, including either amino acid limitation or ER stress (20–22). A cis-element, termed nutrient-sensing response element-1, located within the ASNS proximal promoter at nt –68 to –60 contributes to this induction under the stress conditions mentioned above, and binding of C/EBP to the NSRE-1 is increased following activation of these stress pathways. Marten et al. (23) was the first to document that the mRNA content of C/EBP is increased by amino acid deprivation, but regulation of the C/EBP gene by ER stress has yet to be examined.
The present study was designed to test the hypothesis that transcription from the human C/EBP gene is induced in response to ER stress and to examine the mechanism of this induction. C/EBP mRNA content was increased following ER stress conditions including glucose deprivation, tunicamycin, or thapsigargin treatment. Given that ER stress did not alter the turnover rate of the C/EBP mRNA, it was proposed that this increase was due to increased transcription. Transient transfection of genomic fragments linked to a luciferase reporter gene demonstrated that a 46-bp region, located at a genomic site that corresponds to the 3'-untranslated region (UTR) of the C/EBP mRNA is required for the induction following ERSR activation, and that the C/EBP promoter plays no major regulatory role. Mutagenesis further indicated that a cis-regulatory element located at nt +1614–1621 (5'-TGACGCAA-3') is responsible for activation of the C/EBP gene by ER stress. This element differs from the consensus mUPRE (see above) by 2 nucleotides (19), but expression of exogenous XPBP1 activated transcription from a C/EBP genomic fragment containing this sequence.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Northern Blot Analysis—HepG2 cells were cultured to 70–80% confluence in 60-mm dishes and then incubated for the indicated time in complete MEM, glucose-free MEM, complete MEM containing either 300 nM thapsigargin (Tg) or 5 µg/ml tunicamycin. Total cellular RNA was isolated using an RNeasy Mini Kit (Qiagen Inc., Valencia, CA). 32P-Radiolabeled cDNA probe synthesis and Northern analysis was performed as described by Aslanian et al. (25). The cDNA probe for C/EBP was nt +1425–1632, which corresponds to a fragment of the 3'-untranslated region of the human C/EBP mRNA. The cDNA probe for human C/EBP was the entire 3'-untranslated region of the human C/EBP mRNA. The ribosomal protein L7a cDNA probe was the entire coding sequence obtained from Dr. Tatsuo Tanaka (University of Ryukyus, Okinawa, Japan), and the cDNA probe for the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was the entire coding sequence obtained from Dr. Anupam Agarwal (University of Alabama, Birmingham, AL).
Real Time Quantitative RT-PCR—25–200 ng of HepG2 total RNA was used in each reaction to measure the mRNA content for C/EBP, Asns, and GAPDH. SYBR green chemistry was used to detect the products of interest (Applied Biosystems Inc., Foster City, CA). The primers for amplification were as follows: C/EBP, 5'-AGAACGAGCGGCTGCAGAAGA-3' and 5'-CAAGTTCCGCAGGGTGCTGA-3'; ASNS, 5'-GCAGCTGAAAGAAGCCCAAGT-3' and 5'-TGTCTTCCATGCCAATTGCA-3'; GAPDH, 5'-TTGGTATCGTGGAAGGACTC-3' and 5'-ACAGTCTTCTGGGTGGCAGT-3'. The RT-PCRs were performed and quantified using a DNA Engine Opticon 2 system (MJ Research, Reno, NV). Each RNA sample was measured in duplicate, and three independent RNA samples were collected for each time point.
Subcloning of the Human C/EBP Gene—A BAC clone (RP11-112L6) containing the human DNA sequence from chromosome 20 was obtained from the Sanger Institute (Cambridge, UK). To obtain a C/EBP-containing genomic fragment, the BAC clone was digested with EcoRI and PvuI, and fragments were separated by preparative field inversion gel electrophoresis and then ligated into the EcoRI site of the pBluescript II SK vector (Stratagene, La Jolla, CA). Using the C/EBP cDNA probe described above, colony hybridization was used to screen the resulting DH5 colonies, and an 11.5-kb C/EBP clone was obtained that contained nt –8451 to +3074.
Deletion Analysis—C/EBP fragments containing nt –8451/+157 and –1595/+157 were obtained by restriction digestion of the –8451/+3074 clone, whereas the C/EBP promoter fragment (nt –325/+157) was prepared by PCR. The C/EBP sequences +1554/+1646, +1423/+2213, and +1423/+3541, which are 3' to the coding sequence, were amplified by PCR using either nt –8451/+3074 or the original BAC clone as template. The promoter fragments were ligated into the SmaI site of the pGL3-basic vector (Promega, Madison, WI), which contains the Firefly luciferase gene as a reporter, whereas the C/EBP gene downstream sequences were ligated into the BamHI site. To test the C/EBP genomic sequences +1554/+1600 and +1601/+1646 with the SV40 promoter, oligonucleotides were synthesized with BamHI linkers (Invitrogen) and ligated into the BamHI site of the pGL3-promoter vector (Promega, Madison, WI).
Mutagenesis—All site-directed mutagenesis was performed using the QuikChange® site-directed mutagenesis kit from Stratagene. For mutagenesis, substitutions were made within the C/EBP +1423 to +2213 or the +1601/+1646 sequence, cloned into BamHI site of the firefly luciferase reporter gene, and expressed under the control of either the C/EBP promoter –1593/+157 or the SV40 promoter. The specific mutations made are given under "Results." The C/EBP-luciferase constructs were transiently transfected into HepG2 cells, and the firefly luciferase activity was measured as described below.
Transient Transfection and Transcription Factor Overexpression— For each transfection, 1 µg of the pGL3 firefly luciferase reporter construct, driven by either the human GRP78 promoter –132/+7 (a generous gift from Dr. Kazutoshi Mori, Kyoto, Japan) or by the C/EBP sequence +1601/+1646 under the control of the SV40 promoter, was co-transfected along with 0.5 ng of a reference Renilla luciferase expression plasmid, phRL-SV40 (Promega). When indicated, 0.5 µg of pcDNA3.1 vector only (Invitrogen) or pcDNA3.1 containing the cDNA sequence for the active and nuclear form of ATF6 (amino acids 1–373) or the spliced and active form of XBP-1 (both gifts from Dr. Kazutoshi Mori, Kyoto, Japan) was included. HepG2 cells were transfected at 
50% confluence in 24-well plates (2 x 105 cells/well) using the SuperFectTM transfection reagent (Qiagen). After transfection and a subsequent 18-h recovery in complete MEM containing 10% fetal bovine serum, cells were then incubated for 12 h in either fresh complete MEM or MEM plus 300 nM Tg, each supplemented with 10% dialyzed fetal bovine serum. The effect of transcription factor overexpression was measured by luciferase activity.
Luciferase Reporter Assay—HepG2 cells were transfected with the C/EBP-firefly luciferase reporter plasmids described above. The relative transfection efficiency was corrected by co-transfection with Renilla luciferase reporter (phRL-SV40) driven by the SV40 promoter. The firefly and Renilla luciferase activities were assayed by the dual luciferase reporter system according to the manufacturer's directions (Promega). The cells were transfected at 50% confluence in 24-well plates (2 x 105 cells/well) using the SuperFect transfection reagent, as described above. The data are expressed as the averages ± S.D. of 3–4 assays, and each experiment was repeated with multiple batches of cells.
Immunoblotting—Total cell extracts (30–60 µg/sample) were separated on an SDS-polyacrylamide gel and electrotransferred to a Protran nitrocellulose membrane (Schleicher & Schuell), as previously described (26). The membrane was stained with Fast Green Stain to check for equal loading between lanes. Primary antibodies (1:100 to 1:1000 dilution) were rabbit polyclonals against either C/EBP (catalog no. sc-150) or XBP-1 (catalog no. sc-7160) from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). The secondary antibody was goat anti-rabbit IgG and was used at a dilution of 1:2000 (XBP-1) or 1:20,000 (C/EBP). The bound secondary antibody was detected using an enhanced chemiluminescence kit (Amersham Biosciences) and exposed to Biomax MR film (Eastman Kodak Co.).
Nuclear Extract Preparation and Electrophoretic Mobility Shift Assay—HepG2 cells were seeded on 150-mm dishes (15 x 106 cells/dish). After 16 h of culture, the cells were washed twice with PBS and incubated for 18 h in either complete MEM or MEM lacking glucose, both supplemented with 5% dialyzed fetal bovine serum. The nuclear extraction was performed as previously described (27). Protein concentration was determined using a modified Lowry assay (28). Singlestranded oligonucleotides were annealed by adding 4.8 nmol of each oligonucleotide, with 20 µl of 10x annealing buffer (100 nM Tris-HCl, pH 7.5, 1 M sodium chloride, 10 mM EDTA) in a total volume of 200 µl. The oligonucleotide solution was heated to 95 °C for 5 min and then allowed to gradually cool to 4 °C over 2 h. The oligonucleotides used either as electrophoretic mobility shift assay probes or competitors are listed in Fig. 10. The double-stranded oligonucleotides were radiolabeled by extension of overlapping ends with Klenow fragment in the presence of [-32P]dATP. For each binding reaction, 10 µg of nuclear extract protein was incubated with 40 mM Tris-base, 200 mM NaCl, 2 mM dithiothreitol, 10% (v/v) glycerol, 0.05% (v/v) Nonidet P-40, 3 µg of poly(dI-dC) (Amersham Biosciences), and 0.05 mM EDTA for 20 min on ice. The radiolabeled probe was added at a concentration of 0.004 pmol/reaction (20,000 cpm), and where indicated, unlabeled competitor oligonucleotides were added at 0.04 pmol/reaction. The reaction mixture (30-µl final volume), was incubated at room temperature for 20 min. The reactions were subjected to electrophoresis as described previously (27).
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
mRNA by Glucose Deprivation—To examine whether or not C/EBP or C/EBP mRNA content was altered in response to glucose deprivation of HepG2 cells, the cells were incubated in glucose-free MEM for 0–12 h, and Northern blot analysis was performed (Fig. 1A). An initial increase in C/EBP mRNA content was observed between 2 and 4 h of glucose deprivation and then reached a maximum of about 7 times the control value (5.5 mM glucose) at 8 h (Fig. 1B). In contrast, C/EBP mRNA content remained relatively steady for 4 h and then declined during the 8–12-h period of glucose deprivation (Fig. 1C).
|
regulation by glucose, the mRNA level for C/EBP was measured for three independent experiments by real time quantitative RT-PCR (data not shown). The induction of C/EBP mRNA measured by Northern blot analysis and quantitative RT-PCR followed a similar pattern, reaching a comparable degree of induction at 8 h.
Induction of C/EBP mRNA and Protein by ER Stress—To further investigate whether or not the ERSR pathway was responsible for the increased C/EBP mRNA content, HepG2 cells were incubated with known ERSR activators such as thapsigargin (Tg), an ER Ca2+-ATPase inhibitor, or tunicamycin, an inhibitor of protein glycosylation (29). Consistent with their common mode of action, tunicamycin treatment and glucose deprivation increased C/EBP mRNA content over a similar time frame, with the initial increase being observed between 2 and 4 h, and reaching a 7- or 8-fold induction at 8 h (Fig. 2). In contrast, thapsigargin treatment also induced C/EBP mRNA content, but the maximal induction of 8-fold was achieved by 4 h.
|
protein content was also increased, immunoblots of control (MEM-incubated cells) or Tg-treated whole cell extracts were subjected to immunoblotting (Fig. 3). ER stress caused an elevation of full-length (liverenriched activating protein) C/EBP protein that was consistent with the rise in mRNA level, although the absolute magnitude of the protein increase was less than that for the mRNA. The peak of C/EBP protein occurred at 4 h and stayed relatively high for the remainder of the 12-h period investigated. To show equal lane loading, the blot was stained with Fast Green, and as a negative control, the blot was probed with antibody specific for the bZIP transcription factor ATF2, which did not show increased expression (Fig. 3, upper panel).
|
mRNA by the ERSR Is Dependent on de Novo Protein Synthesis—To determine whether or not synthesis of an upstream regulatory protein was required for induction of the C/EBP gene, HepG2 cells were incubated with thapsigargin in the presence or absence of 0.1 mM cycloheximide (Fig. 4). Cycloheximide completely prevented the increase in C/EBP mRNA content following ERSR activation, suggesting that de novo protein synthesis was required at some unidentified step leading to activation of the C/EBP gene. Inhibition of protein synthesis may also have a minor effect on the turnover of C/EBP mRNA, as indicated by the small, but consistent elevation in mRNA content in cells treated with cycloheximide in the absence of thapsigargin (Fig. 4).
|
mRNA Content by the ERSR Is Not Due to mRNA Stabilization—To test for increased mRNA stability as a possible mechanism for the ERSR induction, HepG2 cells were incubated in glucose-free MEM for 8 h to elevate the C/EBP mRNA content and were then transferred to either fresh glucose-free MEM or complete MEM, both containing 5 µM actinomycin D (Fig. 5). The results showed that the half-life of C/EBP mRNA with or without glucose was 1.5 h, indicating that the ERSR-dependent elevation in C/EBP mRNA is probably not the result of mRNA stabilization.
|
Genomic 5' Upstream Region Does Not Respond to the ERSR—To investigate the role of genomic 5' upstream sequences in mediating C/EBP transcription in response to ER stress, a fragment corresponding to the human C/EBP proximal promoter nt –325/+157 was tested initially (Fig. 6). Compared with the 7-fold induction of endogenous mRNA by glucose deprivation or Tg treatment (Figs. 1 and 2), this promoter fragment did not respond to the glucose-limited condition and resulted in less than a 70% increase in Tg-treated cells (Fig. 6B). To test the possibility that the cis-acting elements required for full induction may be located farther upstream, much longer promoter fragments (–1595/+157 and –8451/+157) were examined, but similar results were obtained (Fig. 6B).
|
Genomic Sequence 3' to the Protein Coding Sequence Is Essential for the ERSR—As depicted in Fig. 6A, the human C/EBP gene is intronless and relative to the transcription start site, the first of multiple translation start sites is at nt +206, and the single translation stop codon is at +1243 (30). Although not fully characterized, an apparent polyadenylation signal (5'-AATAAA-3') is 1.8 kb downstream from the transcription start (GenBankTM number NM_005194
[GenBank]
). Given that a 5' upstream fragment of nearly 8.5 kbp did not support induction by ER stress, the C/EBP genomic region 3' to the protein coding sequence was investigated (Fig. 7). Sequentially deleted fragments were ligated downstream of the firefly luciferase reporter gene, driven by a C/EBP promoter fragment containing nt –1595 to +157. To approximate the endogenous location, the C/EBP 3' genomic fragments were inserted downstream of the firefly coding sequence within the reporter plasmid. The C/EBP sequence from nt +1423 to +3541, containing the sequence corresponding to the mRNA 3'-untranslated region and some additional 3' genomic sequence, induced transcription by 17-fold when cells were treated with Tg (Fig. 7). The degree of induction was similar when this 2.1-kb genomic sequence was deleted to an 800-bp DNA fragment covering nt +1423–2213. The ER stress-responsive region was narrowed even further by establishing that a 93-bp DNA fragment containing nt +1554–1646 activated transcription (Fig. 7). The 93-bp sequence yielded an induction level of about 8 times the control, less than the longer fragments but nearly identical in magnitude to the maximal increase in endogenous mRNA content following Tg treatment (Fig. 2). The decline in the relative increase was due to an increase in the basal rate rather than a decline in the absolute transcription rate following ER stress. Taken together, the data of Figs. 6 and 7 indicate that the DNA regulatory element necessary to mediate the ERSR activation of the C/EBP gene is located 3' to the protein coding sequence and within nt +1554–1646.
|
3' Genomic Sequence Can Confer ER Stress Responsiveness to an Otherwise Inert Promoter—To test the hypothesis that nt +1554–1646 of the C/EBP gene (Fig. 8A) could confer ER stress-regulated transcription to an unrelated promoter, the C/EBP promoter fragment used in the previous experiments was replaced with the SV40 promoter (Fig. 8B). The SV40 promoter alone was inert to Tg treatment, but when a single copy of the 93-bp C/EBP genomic sequence was present, transcription was induced to 9 times the MEM control.
|
Gene by ER Stress—Computer analysis of the 93-bp region revealed no perfect match for either the ERSE (5'-CCAATN9CCACG-3') (17, 18) or the nutrient-sensing response unit (5'-TGATGAAACN11GTTACA-3') (21, 22, 31), either of which can mediate the ER stress signal. However, a sequence corresponding to the ATF6/XBP-1 half-site of the ERSE (5'-CCACG-3') is present at nt +1623–1627, and a sequence identical to the second element of the nutrient-sensing response unit (NSRE-2; 5'-GTTACA-3') is present between nt +1628 and +1633 (Fig. 8A). A perfect match to the CHOP C/EBP-ATF sequence is also located within the 93-bp C/EBP genomic region at nt +1576–1568 (Fig. 8A). CHOP is also induced by ER stress and contains a functional ERSE (15), but the C/EBP-ATF binding site (nt –310/–302; 5'-ATTGCATCA-3') has been suggested to also influence the regulation of the CHOP gene by the ERSR pathway (32). An additional sequence of interest in the C/EBP gene is present at nt +1614–1621 (5'-TGACGCAA-3') and is highly similar to a consensus sequence referred to as the mammalian UPRE (19, 11). The mUPRE represents the recognition site for an ER stress pathway in mammals and selectively binds XBPI rather than ATF6 and, in this way, mediates the activation of UPR-responsive genes that do not contain the ERSE sequence (11).
To investigate the contribution of the C/EBP-ATF composite site in the C/EBP gene, the core element and surrounding sequence (nt +1554–1582) were deleted from the +1554–1646 genomic fragment (Fig. 8B, Del C/EBP-ATF). Although deletion of the C/EBP-ATF sequence reduced the -fold induction by Tg, it did so because the basal transcription rate was elevated. The absolute luciferase activity in cells incubated in the presence of Tg was also increased. To confirm this result, rather than deletion a block mutation of the C/EBP-ATF core sequence was tested, and similar results were observed (Fig. 8C, Mut-1). Furthermore, this site was proposed to modulate the ERSR by binding ATF4 (32), but induction of C/EBP expression by glucose deprivation still occurs in ATF4–/– mouse embryonic fibroblasts,2 supporting the conclusion that at least a component of the regulation of C/EBP gene expression by ER stress is independent of the C/EBP-ATF site and the ATF4 pathway.
To examine the potential role of the CCACG and NSRE-2 sequences, each was mutated in the context of a C/EBP fragment corresponding to nt +1423/+2213. When the CCACG box and its 5'-flanking sequence GCAAC was mutated (Fig. 8A), a significant loss in Tg-induced luciferase activity was observed (Fig. 8C, Mut-2). However, when the CCACG sequence and its 3'-flanking sequence TGTAACT (underlined nucleotides are identical to the ASNS NSRE-2 site) was substituted, the absolute transcription rate in Tg-treated cells was not significantly reduced compared with the wild type sequence, although there was a decline in the -fold induction due to a increase in basal transcription (Fig. 8C, Mut-3). Collectively, the results indicate that the sequence immediately 5' to the CCACG box is required to mediate the ERSR for the C/EBP gene, but neither the CCACG box itself nor the flanking 3' NSRE-2-like sequence are critical.
To continue the characterization of this region, the sequence covered by nt +1554–1646 was first dissected in half, and the ability of each fragment to respond to ER stress was tested (Fig. 9A). When the 5' half of the 93-bp sequence (nt +1554–1600) was tested, only a slight effect of Tg treatment was detectable. Conversely, when the 3' half was tested (nt +1601–1646), Tg treatment resulted in a 9-fold induction over MEM, a result comparable to the 7–8-fold increase in endogenous C/EBP mRNA (Fig. 2). Collectively, the data of Figs. 8 and 9 suggest that the ER stress-responsive element for the C/EBP gene resides between nt +1601 and +1623, and consequently, additional mutagenesis targeted the region containing the mUPRE-like sequence at nt +1614–1621 (Fig. 9B). When nt +1601–1609 was substituted (Fig. 9C, Mut-4), the relative induction by Tg treatment was only modestly reduced, but mutation of the sequence 5'-GGACTGAC-3' from nt +1610 to +1617 (Fig. 9C, Mut-5) or 5'-GCAACC-3' from nt +1618 to +1623 (Fig. 9C, Mut-6) completely prevented the Tg-induced increase in transcription. These results suggested that nt +1610–1623 (5'-GGACTGACGCAACC-3'), containing the XBP-1-like binding site (5'-TGACGCAA-3'), was the sequence harboring the responsive element. To investigate the boundaries of this potential mUPRE sequence, the 5' (Mut-7) and 3' (Mut-8) flanking nucleotides were mutated (Fig. 9C). Whereas modifying the two 3' nucleotides caused a significant reduction in transcription, mutation of the 5' GGA had only a modest effect (Fig. 9C).
|
mUPRE sequence, with nuclear extracts from HepG2 cells resulted in the formation of two major DNA-protein complexes (Fig. 10, lanes 1 and 6). Formation of these complexes could be blocked by an excess of unlabeled oligonucleotide (Fig. 10, lanes 2 and 7) but not by an oligonucleotide with an unrelated sequence (Fig. 10, lanes 4 and 9). When nuclear extracts from cells incubated in glucose-free medium were tested, the amount of both complexes was increased in abundance (Fig. 10). To determine whether both complexes were associated with the C/EBP mUPRE core sequence, a competitor oligonucleotide (Fig. 10, lanes 3 and 8) was prepared that contained the same mutations used for the functional analysis shown in Fig. 9 (mutants 5 and 6). The competitor oligonucleotide containing the mutated C/EBP mUPRE core did not block formation of either complex. Furthermore, when this mutated oligonucleotide was radioactively labeled and used as the probe, no complexes were observed (Fig. 10, lanes 5 and 10).
Induction of XBP-1 Protein Content by ER Stress—To determine whether XBP-1 protein content in HepG2 cells was increased by ER stress, whole cell extracts were subjected to immunoblotting after a 6-h incubation in either MEM or MEM + Tg. It has been reported that XBP-1 is rapidly degraded, and inhibition of proteosome action aids in detection of the protein (24). This was also the case for HepG2 cells. The level of XBP-1 protein was below detection in control (MEM) or Tg-treated cells, but including the proteosome inhibitor MG132 during the Tg incubation resulted in a readily identifiable, stress-inducible XBP-1 band (data not shown). MG132 itself did not cause an increase in XBP-1 protein content. The increase in XBP-1 protein content following Tg treatment is consistent with the cycloheximide sensitivity of the Tg-dependent induction of C/EBP mRNA shown in Fig. 4.
XBP-1 Modulates C/EBP Transcription—To investigate whether or not transcription from the C/EBP gene was enhanced by increased expression of ATF6 or XBP-1, HepG2 cells were transfected with the active form of each, along with a luciferase reporter construct containing the C/EBP mUPRE sequence (Fig. 11A). Exogenously expressed XBP-1 further enhanced both basal (MEM) and Tg-induced transcription by 5- and 2-fold, respectively. ATF6 overexpression modestly increased the transcriptional activity in the control (MEM) condition but actually blocked the induction by Tg treatment when compared with control (Fig. 11A). As a positive control, the promoter activity of GRP78 (nt –132/+7) containing three ERSE sequences (17) was examined (Fig. 11A). In general agreement with previously published results by others (11), GRP78 promoter activity was induced by Tg treatment or by expressing either ATF6 or XBP-1 in nonstressed cells (MEM) (Fig. 11A). Neither factor produced a further enhancement when coupled with Tg treatment, but it is interesting to note that ATF6 expression did not block the Tg-induced transcription of the GRP78 promoter as it did for the C/EBP element.
|
expression after activation of ER stress, wild type and XBP-1-deficient (XBP–/–) mouse embryonic fibroblasts were subjected to Tg treatment for 4 h and then analyzed for C/EBP mRNA expression by real time quantitative PCR (Fig. 11B). Consistent with the overexpression data (Fig. 11A), the basal C/EBP mRNA content was reduced in the XBP–/– cells relative to the wild type. Although the Tg-dependent induction in the wild type mouse embryonic fibroblasts was not as great as that observed in HepG2 cells, relative to the wild type mouse embryonic fibroblasts, the induction of C/EBP mRNA content by Tg was completely prevented in the XBP–/– cells. |
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
mRNA and protein content is increased by glucose deprivation and other causes of ER stress, whereas C/EBP is decreased. 2) This increase in C/EBP mRNA is dependent on de novo protein synthesis, suggesting the need for synthesis of an upstream regulator. 3) Stabilization of mRNA does not contribute to the mRNA accumulation, and genomic analysis illustrated transcriptional control of the human C/EBP gene. 4) The genomic 5' upstream sequence, up to 8.45 kbp, is not sufficient to induce transcription following ERSR activation, whereas a mUPRE, located 3' to the protein coding sequence, is responsible for transcriptional induction of the human C/EBP gene. 5) The C/EBP mUPRE exhibits enhancer-like properties in that it can convey responsiveness to a heterologous promoter. 6) There are specific DNA-protein complexes formed on the C/EBP mUPRE core sequence, and the abundance of these complexes is increased following activation of the UPR pathway. 7) C/EBP mRNA is not induced in XBP–/– mouse embryonic fibroblasts, and in functional transient transfection assays, the C/EBP mUPRE responds to an elevated level of XBP-1.
The C/EBP gene has not been reported previously to be responsive to, or a mediator of, mammalian ER stress pathways. To respond to ER stress, cells activate genes involved in ER functions such as protein folding, export, and degradation (6–8), and to eliminate cells that have accumulated an irreparable level of damage, ER stress eventually induces genes implicated in growth arrest and apoptosis, such as CHOP (33, 34). CHOP is a member of the C/EBP family of transcription factors, and Fawcett et al. (35) showed that in response to oxidative stress, the expression of both C/EBP and CHOP is induced and that they interact in vivo. Those authors further proposed that C/EBP trans-activates CHOP expression through a C/EBP-ATF composite site in the CHOP promoter (nt –310/–302, 5'-ATTGCATCA-3'). The increased CHOP to C/EBP protein ratio may lead to the heterodimerization of CHOP with C/EBP, and consequently, given the dominant negative effect of CHOP (36), this heterodimerization could block the trans-activation effect of C/EBP on numerous genes, including the CHOP promoter itself, thereby completing an autoregulatory loop for CHOP expression. This proposed mechanism for oxidative stress may parallel the increased C/EBP expression following ER stress, documented in this report, because Ma et al. (32) provided evidence that C/EBP transcriptionally up-regulates CHOP expression following ER stress via the C/EBP-ATF composite site in the CHOP promoter. This sequence, in concert with a functional ERSE site also present in the CHOP promoter region (15), was proposed to activate CHOP expression to an optimal level after ER stress. Although the human C/EBP gene contains a C/EBP-ATF composite site sequence (nt +1568–1576), the data in this report demonstrate that the C/EBP-ATF sequence is not essential for induction by the ER stress. Instead, the results demonstrate that the human C/EBP gene contains a mUPRE that mediates this response.
C/EBP–/– mice (B phenotype) die shortly after birth. These newborn pups have a severe hypoglycemia, probably the result of a lack of the hepatic gluconeogenic enzyme phosphoenolpyruvate carboxykinase (37). Given that the ERSR pathway steps downstream of C/EBP are probably defective in these animals, the possible lack of critical ERSR-dependent mechanisms must also be considered as contributing factors to their mortality. One possible reason that the ERSR pathway might be activated shortly after birth is the adaptive response to oxidative stress that would result from increased oxidative metabolism (38). Although the physiologic importance of increased asparagine biosynthesis during ER stress remains unknown, we have previously demonstrated that ASNS expression is induced by ER stress (21, 22), and that C/EBP modulates ASNS transcriptional rates via binding to the NSRE-1 site within its proximal promoter (39). The results in this report extend those previous observations by illustrating that the expression of C/EBP itself is induced after ER stress. As an upstream regulator of ASNS expression, C/EBP may be a metabolic switch that links carbohydrate metabolism and/or ER stress to amino acid metabolism.
The C/EBP mUPRE is highly conserved across the human, rat, and mouse genomes (data not shown). Why this element is located at the 3' end of the gene is unclear, but it may have to do with the fact that the gene is intronless. The general importance of the 3' downstream region of the C/EBP gene is emphasized by the extensive conservation among species of large stretches of sequence in this region. In contrast, the corresponding region of the human C/EBP gene, which is not induced by ER stress, as documented here, does not contain a UPRE-like sequence, based on computer analysis. To our knowledge, all previous examples of ER stress-activated genes that have been tested have the corresponding ERSE or UPRE sequences located within their promoter region. Therefore, the C/EBP gene may represent a novel model for investigating how such distal elements interact with the general transcription machinery.
|
FOOTNOTES |
|---|
To whom correspondence and reprint requests should be addressed: Dept. of Biochemistry and Molecular Biology, University of Florida College of Medicine, Box 100245, Gainesville, FL 32610-0245. Tel.: 352-392-2711; Fax: 352-392-6511; E-mail: mkilberg{at}ufl.edu.
1 The abbreviations used are: C/EBP, CCAAT/enhancer-binding protein; CHOP, C/EBP homology protein; UPR, unfolded protein response; UPRE, UPR element; mUPRE, mammalian UPRE; ERSR, endoplasmic reticulum stress response; ERSE, ERSR element; GRP, glucose-regulated protein; nt, nucleotide(s); ER, endoplasmic reticulum; ASNS, asparagine synthetase; UTR, untranslated region; MEM, minimal essential medium; Tg, thapsigargin; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; RT, reverse transcriptase.
2 Y.-X. Pan, unpublished data.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
