- Email: [email protected]
SREBP-1c Mediates Insulin-Induced Gene Expression Linked to the MAP Kinase Pathway
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS ARTICLE NO.
249, 375–379 (1998)
RC989161
ADD1/SREBP-1c Mediates Insulin-Induced Gene Expression Linked to the MAP Kinase Pathway Jo¨rg Kotzka, Dirk Mu¨ller-Wieland,1 Anni Koponen, Dieudonne Njamen, Lorena Kremer, Gunther Roth, Martina Munck, Birgit Knebel, and Wilhelm Krone Klinik II und Poliklinik fu¨r Innere Medizin der Universita¨t zu Ko¨ln at the Center of Molecular Medicine, University of Cologne, Cologne, Germany
Received July 6, 1998
The aim of this study was to define the role of sterol regulatory element binding protein (SREBP)-1c, the human homologue to ADD1 (adipocyte determinationand differentiation-dependent factor 1), in insulin-induced gene expression. Transfection studies using SREBP-1-deficient cells and a LDL receptor promoter fragment containing the ADD1/SREBP-1c binding side showed that the effects of insulin and PDGF were abolished compared to control cells and completely reconstituted by overexpressing ADD1/SREBP-1c. Overexpression of upstream activators of MAP kinases, like MEKK1 or MEK1, demonstrated that ADD1/SREBP-1cmediated effects of insulin and PDGF might be linked to the MAP kinase cascade. The recombinant N-terminal domain of ADD1/SREBP-1c was phosphorylated predominantly on serine and slightly on threonine residues by MAP kinases ERK1 and ERK2 in vitro. This was reversible by alkaline phosphatase. We conclude that ADD1/SREBP-1c mediates gene regulatory effects of insulin as well as PDGF and that this signalling is linked to the MAP kinase cascade. q 1998 Academic Press
Insulin resistance and obesity are tightly related and appear to be associated with glucose intolerance or type 2 diabetes and an increased risk for cardiovascular complications. A major breakthrough in the understanding of the relationship between insulin sensitivity and adipogenesis was the discovery of peroxisome proliferator-activated receptor (PPAR)-g, a member of the nuclear hormone receptor superfamily (for review, see 1, 2). PPAR-g was identified and cloned as a component of an adipocyte differentiation-dependent regulatory factor (ARF6) that binds to the adipose-specific enhancer from the aP2 gene. PPAR-g2, an alternatively spliced product of PPAR-g gene, is most abundantly 1 To whom correspondence should be addressed. Fax: /49/221-4784179/3107. E-mail: [email protected].
expressed in adipocytes and appears to play a major role in adipocyte differentiation and lipid metabolism. One essential factor modulating the PPAR-g2 activity is the adipocyte determination and differentiation-dependent factor (ADD)1. ADD1 was cloned by the group of Spiegelman et al. as an adipogenesis related transcription factor and independently by the group of Brown and Goldstein as a transcription factor, called sterol regulatory binding protein (SREBP)-1c, involved in cellular cholesterol metabolism binding to a sterol regulated cis element in the LDL receptor (LDLR) gene (for review, see 3, 4). Overexpression of ADD1/SREBP-1c in fibroblasts (5) or in the liver of transgenic mice (6) induces lipoprotein lipase and fatty acid synthase. Further studies indicate, that ADD1/SREBP-1c can promote adipocyte differentiation and increases transcriptional activity of PPAR-g2. Therefore, ADD1/SREBP-1c appears to be a key link between cholesterol and triglyceride metabolism, adipogenesis as well as insulin sensitivity. In this study we demonstrate that insulin and PDGF can act via ADD1/SREBP-1c possibly by activating MAP kinase cascade. In accordance to that, ADD1/SREBP-1c is a direct substrate of MAP kinases ERK1 and ERK2 in vitro. EXPERIMENTAL PROCEDURES Plasmids and molecular cloning. Construction of the LDLR promoter reporter gene plasmid phLDL4, containing sterol regulatory element (sre)-1 flanked by two SP1 elements, was described previously (7). Expression vectors pFC-MEKK for dominant-active MEKK1 (aa 380-672) and pFC-MEK1 for activated MEK1 (D32-51/ S218E/S222E) under control of CMV promoter were purchased from Stratagene. Total RNA was isolated from HepG2 cells by guanidine isothiocyanate extraction protocol described previously (7). Using the FirstStrand cDNA synthesis kit (Pharmacia Biotech) the RNA was reverse transcribed following the directions of the manufacturer. A cDNA fragment corresponding to the amino acids 1-460 of SREBP1a, corresponding to transcriptional active domain, was amplified by 0006-291X/98 $25.00
375
Copyright q 1998 by Academic Press All rights of reproduction in any form reserved.
AID
BBRC 9161
/
695b$$$481
08-01-98 15:47:26
bbrcg
AP: BBRC
Vol. 249, No. 2, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Role of ADD1/SREBP-1c in hormonal activation of LDLR promoter activity. HepG2/pcDNA3 cells [A] and SREBP-1(-) cells without [B] or with ectopically expressed ADD1/SREBP-1c-NT (0.05 mg/well) [C] were transiently transfected with LDLR promoter construct phLDL4 (1 mg/well). Before harvesting as described under experimental procedures cells were incubated with insulin (1 x 1007 M) or PDGF (3.3 x 1009 M) for 4 h. Transfection efficiency was monitored and normalized by cotransfecting cells with pSV b-galactosidase vector (1 mg/ well). All transfection experiments were carried out in triplicate. The results are given as means ({S.D.) of five independent experiments, each done in triplicate.
PCR from this single stranded cDNA using the sequence-specific 5*-primer (GGGGATCCCCATGGACGAGCCACCCTTCA) and 3*primer (GGAATTCTCAGTC AGGCTCCGAGTCACTGCCA), which contain additional bases (underlined sequence) to generate a BamHI and a EcoRI restriction site, respectively. This fragment was inserted in frame to glutathione-S-transferase into pGEX 3X (Pharmacia Biotech). The transcriptional active domain of ADD1/SREBP-1c (aa 1-436), the splice variant of SREBP-1a, was amplified by PCR with BamHI/HgaI digested SREBP-1a-NT clone as matrix using sequence-specific 5*-primer (CGGGATCCCGATGGATTGCACTTTCGAAGACATGCTTCAGCTTATC) with BamHI restriction site (underlined sequence) and 3*-primer identical to 3*-primer of SREBP-1a. To obtain ADD1/SREBP-1c-NT in pGEX 3X vector the SREBP-1aNT was replaced by BamHI/SmaI digested PCR product. The ADD1/ SREBP-1c-NT expression vector was constructed by ligating the Nterminal domain of ADD1/SREBP-1c as a 1308 bp BamHI/EcoRI fragment into the BamHI and EcoRI site of pcDNA3 (Invitrogen). The sequence of the constructs was confirmed by using a model 373A DNA sequencer (Applied Biosystem Inc.). Transient transfection assay. HepG2 cells were maintained in RPMI 1640 medium (Sigma) supplemented with 10% fetal calf serum (GibcoBRL) and antibiotics. Mock-transfected (pcDNA3) HepG2 cells and HepG2 cells with reduced intracellular level (õ10%) of SREBP1 protein, named SREBP-1(-) cells, generated by antisense technique as described previously (7), were cultured in RPMI 1640 medium supplemented with 10% fetal calf serum (GibcoBRL), antibiotics and 500 mg/ml G418. Before electroporation, cells were released by trypsinization, washed and suspended in Opti-MEM (GibcoBRL) supplemented with 10% FCS. Cell suspensions (2 x 105 cells/well) were mixed with vectors as indicated in figure legends. Samples were transferred to an electroporation cuvette (interelectrode distance 0,4 cm) and pulsed for 18 msec in GenePulser II (Bio-Rad). Before seeding on 6 well plates (Costar) cell suspension was diluted with RPMI 1640 with 10% (v/v) FCS and antibiotics. If external induction experiments were carried out, cells were serum-starved on day two for 16
h. Before harvesting, cells were incubated with insulin (1007 M) or PDGF (3.3 x 1009 M) for 4 h. For endogenous induction using expression vectors for dominant active MEKK1 or activated MEK1, cells were harvested 16 h after electroporation without further treatment. Luciferase activity was measured according to the supplier’s instructions (Promega). Transfection efficiency was monitored by cotransfection of b-galactosidase expression vector pSV-bGal (Promega). bGalactosidase activity was determined by galactolight assay (Tropix). The data represent relative luciferase activity as x-fold induction of either endogenous or exogenous stimulation relative to unstimulated cells as indicated in the legends to the figures. Protein kinase assay. The glutathione-S-transferase (GST)ADD1/SREBP-1c-NT fusion protein was expressed in E. coli strain BL21 and purified according to the manufacturer’s recommendations (Pharmacia Biotech). Protein phosphorylation by MAP kinases ERK1 or ERK2 (Upstate Biotechnology) was performed with 10 mg GSTADD1/SREBP-1c-NT fusion protein or 0.5 mg MBP and activated GST-MAP kinase ERK1 (100 ng/assay) or ERK2 (100 ng/assay) fusion proteins in kinase buffer (5 mM MOPS (pH 7.2), 6.25 mM bglycerophosphate, 1.25 mM EGTA, 0.25 mM sodium-orthovanadate, 0.25 mM DTT). The reaction was initiated by addition of 50 mM [g-32P]ATP (10 Ci/mmol) in a final volume of 40 ml kinase buffer. The reaction was terminated after 15 min at 257C by addition of Laemmli sample buffer. The phosphorylation of ADD1/SREBP-1c-NT was examined after sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) by autoradiography. Protein phosphatase assay. GST-ADD1/SREBP-1c-NT fusion protein (1 mg), phosphorylated by activated ERK2, was suspended in AP buffer (50 mM Tris/HCl, pH 8.0, 0.1 mM EDTA). The reaction was initiated by addition of 0-0.5 U alkaline phosphatase (Boehringer Mannheim) in a final volume of 50 ml AP buffer. The reaction was terminated after 60 min at 257C by addition of Laemmli sample buffer. The dephosphorylation of ADD1/SREBP-1c-NT was examined after SDS-PAGE by autoradiography.
376
AID
BBRC 9161
/
695b$$$481
08-01-98 15:47:26
bbrcg
AP: BBRC
Vol. 249, No. 2, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 2. MEKK1 and MEK1 enhances transcriptional activation of LDLR gene by ADD1/SREBP-1c.HepG2 cells were transiently transfected with LDLR promoter construct phLDL4 (1 mg/well) along with expression vectors (0.05 mg/well) without (black bars) or containing (gray bars) [A] ADD1/SREBP-1c-NT and dominant active MEKK1 or [B] ADD1/SREBP-1c-NT and activated MEK1, respectively. Conditions are described under experimental procedures. Transfection efficiency was monitored and normalized by cotransfecting cells with pSV bgalactosidase vector (1mg/well). All transfection experiments were carried out in triplicate. The results are given as a mean ({S.D.) of five independent experiments, each done in triplicate.
RESULTS To evaluate the role of ADD1/SREBP-1c in signal transduction of insulin stable HepG2 cell lines were generated using anti-sense RNA technique being deficient in protein levels (õ10%) of SREBP-1 (7). Transient transfection experiments using this cell line showed that the effects of insulin and PDGF on LDLR promoter activity using a construct (phLDL4) containing an intact SREBP-1c binding sterol-regulatory cis element sre-1 were abolished in SREBP-1 deficient cells compared to mock-transfected control cell lines, see Fig. 1. Accordingly, overexpression of the transcriptional active N-terminal region of ADD1/SREBP-1c in SREBP-1-deficient cells led to a complete reconstitution of insulin and PDGF action on LDLR promoter activity. One major pathway linking receptor associated tyrosine kinases, like the receptors of insulin and PDGF, to the transcriptional machinery of various genes is the intracellular MAP kinase cascade. In accordance to that, we could show (Fig. 2) that overexpression of constitutive active MEKK1 or MEK1, which are upstream activators of MAP kinases, stimulated LDLR promoter activity in HepG2 cells several-fold. Ectopic expression of the transcriptional active region of ADD1/ SREBP-1c (ADD1/SREBP-1c-NT) had a stimulatory action on LDLR promoter activity per se and a synergistic effect when coexpressed with MEKK1 or MEK1.
To study, whether SREBP-1c might be a direct substrate of MAP kinases we produced and purified recombinant GST-fusion proteins of SREBP-1c. Incubation of ADD1/SREBP-1c-NT fusion protein with activated recombinant ERK1 or ERK2 showed significant phosphorylation, in vitro (Fig. 3). In contrast to that, the C-terminal region was not phosphorylated (data not
FIG. 3. Phosphorylation of ADD1/SREBP-1c by MAP kinases ERK1 and ERK2 in vitro. Bacterial-synthesized glutathione-S-transferase (GST)-ADD1/SREBP-1c-NT fusion protein was treated with activated GST-MAP kinase ERK1 or ERK2 fusion proteins and separated by SDS-PAGE followed by autoradiography. A representative result of five independent experiments are shown.
377
AID
BBRC 9161
/
695b$$$481
08-01-98 15:47:26
bbrcg
AP: BBRC
Vol. 249, No. 2, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 4. Dephosphorylation of ADD1/SREBP-1c by phosphatase in vitro. (GST)-ADD1/SREBP-1c-NT fusion protein phosphorylated by active GST-MAP kinase ERK2 fusion protein was treated with increasing amounts of alkaline phosphatase (AP), followed by SDSPAGE and autoradiography. A representative result of four independent experiments are shown.
shown). It is interesting to note, that phosphorylation by ERK2 appeared to be more prominent when compared to ERK1. Phosphoamino acid analyses revealed phosphorylation predominantly on serine and slightly on threonine residues (data not shown). This phosphorylation was reversible by incubating phosphorylated ADD1/SREBP-1c-NT with increasing concentrations of alkaline phosphatase (Fig. 4). The autoradiography of the SDS-PAGE showed two bands indicating possibly more than one phosphorylation site.
DISCUSSION The promoter of LDLR gene contains a sterol regulatory element (sre-1/ATCACCCCAC), which is regulated by the intracellular content of sterols (8). This DNA sequence is the target of three transcription factors, named SREBP-1a, SREBP-1c, and SREBP-2 (9,10). Some essential features of the structure-function relationship of SREBPs led them belong to the family of basic helix-loop-helix leucine zipper (bHLH-LZ) transcription factors. The basic domain of these proteins regulates DNA binding to a consensus sequence referred to as E-box motif (CANNTG). This latter motif is found in promoters of many genes. Most bHLH-LZ proteins contain an arginine in the basic region which restricts protein-DNA interaction to this E-box motif. However, SREBPs uniquely contain a tyrosine at this position instead of an arginine. Thereby, SREBPs might have a dual specificity, i.e. not only for E-box but also for sre-1-like (consensus sequence: Py-CA-Py; Py: pyrimidine) elements (11, 12). Sterol-responsive sre-1-like cis elements seem to exist in promoters of many more genes coding for enzymes involved not only in cholesterol metabolism, but also
in triglyceride synthesis and possibly others (for review, see 3, 4, 13). Furthermore, there are some data indicating that SREBPs might not only mediate the effects of intracellular cholesterol levels on gene transcription, but are also integrating signalling cascades activated by various stimuli (7, 14). Interestingly, SREBP-1c is the human homologue to ADD1, which has been cloned by Spiegelman and colleagues in an attempt to identify transcription factors linked to adipocyte differentiation (15). Expression of wild-type and dominant negative forms of ADD1 in preadipocytes and nonadipogenic cells have shown that ADD1/SREBP-1c plays a pivotal role in fat cell differentiation and gene expression (5, 15). Further experiments provided evidence, that ADD1/SREBP-1c promotes PPAR-g2-mediated transcriptional activity. Using a stable SREBP-1 deficient HepG2 cell line this study shows that the transcription factor ADD1/ SREBP-1c integrates the signalling of insulin as well as PDGF receptor associated kinases on the LDLR promoter. The LDLR promoter was used, because it contains the classical SREBPs binding cis element sre-1 (8). SREBP-1a and SREBP-1c are splice variants coded by a single gene (16). One major difference between SREBP-1a and SREBP-1c is an alternative exon 1. The antisense construct used to generate the SREBP-1 deficient cells is directed to exon 2 thereby affecting both splice variants. However, the reconstitution experiments (Fig. 1) were performed with a construct specific for the N-terminal domain of ADD1/SREBP-1c showing that the hormonal effects can be restored completely. This might indicate the existence of a regulatory mechanism affecting ADD1/SREBP-1c activity independent of intracellular sterol levels. Recently, the group of Spiegelman also provided evidence that ADD1/SREBP1c may be modulated by a signalling mechanism related to insulin (14). Since not only insulin but also PDGF action was affected by intracellular levels of ADD1/SREBP-1c we focussed on a common intracellular signalling pathway like MAP kinase cascade. Ectopic expression of upstream activators of MAP kinases like MEKK1 and MEK1 stimulates LDLR promoter activity. Coexpression of ADD1/SREBP-1c has a synergistic effect (Fig. 2). Since several transcription factors appear to be direct substrates of MAP kinases (17), we wanted to prove the hypothesis that ADD1/SREBP-1c might be a direct substrate of the MAP kinases ERK1 and ERK2. Our experiments show, that ADD1/SREBP1c is phosphorylated by ERK1 and ERK2, in vitro, the phosphorylation being reversible by alkaline phosphatase treatment. The fact that only the transcriptional active N-terminal domain is phosphorylated but not the C-terminal sequence involved in precursor cleavage supports the hypothesis that insulin regulates ADD1/ SREBP-1c mediated gene transcription by stimulating the activity rather than cleavage rate at the endoplasmatic reticulum of ADD1/SREBP-1c.
378
AID
BBRC 9161
/
695b$$$481
08-01-98 15:47:26
bbrcg
AP: BBRC
Vol. 249, No. 2, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Taken together, this study shows that ADD1/ SREBP-1c activity is not only regulated by intracellular sterol concentrations, but also by hormones acting possibly via phosphorylation. Accordingly, ADD1/ SREBP-1c might be a unique convergence point integrating metabolic and endocrine signals to gene regulatory events. ACKNOWLEDGMENTS We thank the German Ministry of Education and Research (BMBF/01 KS 9502) for current support. G.R. was supported by Graduiertenkolleg ‘‘Molekularbiologische Grundlagen Pathophysiologischer Vorga¨nge’’ of the German Research Foundation (DFG).
REFERENCES 1. Spiegelman, B. M. (1997) Eur. J. Med. Res. 2, 257–264. 2. Spiegelman, B. M. (1998) Diabetes 47, 507–514.
3. Brown, M. S., and Goldstein, J. L: (1997) Cell 89, 331–340. 4. Spiegelman, B. M., and Flier, J. S. (1996) Cell 5, 377–389. 5. Kim, J. B., and Spiegelman, M. B. (1996) Genes Dev. 10, 1096– 1107. 6. Shimano, H., et al. (1997) J. Clin. Invest. 29, 846–854. 7. Streicher R., et al. (1996) J. Biol. Chem. 271, 7128–7133. 8. Smith, J. R., et al. (1990) J. Biol. Chem. 265, 2306–2310. 9. Wang, X., et al. (1993) J. Biol. Chem. 268, 14497–14504. 10. Briggs, R. B., et al. (1993) J. Biol. Chem. 268, 14490–14496. 11. Kim, J. B., et al. (1995) Mol. Cell. Biol. 15, 2582–2588. 12. Magan˜a, M. M., and Osborne, T. F. (1996) J. Biol. Chem. 271, 32689–32694. 13. Mu¨ller-Wieland, D., et al. (1997) Curr. Opin. Lipidol. 8, 348– 353. 14. Kim, B., et al. (1998) J. Clin. Invest. 1, 1–9. 15. Tontonoz, P., et al. (1993) Mol. Cell Biol. 13, 4753–4759. 16. Hua, X., et al. (1995) Genomics 25, 667–673. 17. Whitmarsh, A. J., and Davis, R. J. (1996) J. Mol. Med. 74, 589– 607.
379
AID
BBRC 9161
/
695b$$$481
08-01-98 15:47:26
bbrcg
AP: BBRC
249, 375–379 (1998)
RC989161
ADD1/SREBP-1c Mediates Insulin-Induced Gene Expression Linked to the MAP Kinase Pathway Jo¨rg Kotzka, Dirk Mu¨ller-Wieland,1 Anni Koponen, Dieudonne Njamen, Lorena Kremer, Gunther Roth, Martina Munck, Birgit Knebel, and Wilhelm Krone Klinik II und Poliklinik fu¨r Innere Medizin der Universita¨t zu Ko¨ln at the Center of Molecular Medicine, University of Cologne, Cologne, Germany
Received July 6, 1998
The aim of this study was to define the role of sterol regulatory element binding protein (SREBP)-1c, the human homologue to ADD1 (adipocyte determinationand differentiation-dependent factor 1), in insulin-induced gene expression. Transfection studies using SREBP-1-deficient cells and a LDL receptor promoter fragment containing the ADD1/SREBP-1c binding side showed that the effects of insulin and PDGF were abolished compared to control cells and completely reconstituted by overexpressing ADD1/SREBP-1c. Overexpression of upstream activators of MAP kinases, like MEKK1 or MEK1, demonstrated that ADD1/SREBP-1cmediated effects of insulin and PDGF might be linked to the MAP kinase cascade. The recombinant N-terminal domain of ADD1/SREBP-1c was phosphorylated predominantly on serine and slightly on threonine residues by MAP kinases ERK1 and ERK2 in vitro. This was reversible by alkaline phosphatase. We conclude that ADD1/SREBP-1c mediates gene regulatory effects of insulin as well as PDGF and that this signalling is linked to the MAP kinase cascade. q 1998 Academic Press
Insulin resistance and obesity are tightly related and appear to be associated with glucose intolerance or type 2 diabetes and an increased risk for cardiovascular complications. A major breakthrough in the understanding of the relationship between insulin sensitivity and adipogenesis was the discovery of peroxisome proliferator-activated receptor (PPAR)-g, a member of the nuclear hormone receptor superfamily (for review, see 1, 2). PPAR-g was identified and cloned as a component of an adipocyte differentiation-dependent regulatory factor (ARF6) that binds to the adipose-specific enhancer from the aP2 gene. PPAR-g2, an alternatively spliced product of PPAR-g gene, is most abundantly 1 To whom correspondence should be addressed. Fax: /49/221-4784179/3107. E-mail: [email protected].
expressed in adipocytes and appears to play a major role in adipocyte differentiation and lipid metabolism. One essential factor modulating the PPAR-g2 activity is the adipocyte determination and differentiation-dependent factor (ADD)1. ADD1 was cloned by the group of Spiegelman et al. as an adipogenesis related transcription factor and independently by the group of Brown and Goldstein as a transcription factor, called sterol regulatory binding protein (SREBP)-1c, involved in cellular cholesterol metabolism binding to a sterol regulated cis element in the LDL receptor (LDLR) gene (for review, see 3, 4). Overexpression of ADD1/SREBP-1c in fibroblasts (5) or in the liver of transgenic mice (6) induces lipoprotein lipase and fatty acid synthase. Further studies indicate, that ADD1/SREBP-1c can promote adipocyte differentiation and increases transcriptional activity of PPAR-g2. Therefore, ADD1/SREBP-1c appears to be a key link between cholesterol and triglyceride metabolism, adipogenesis as well as insulin sensitivity. In this study we demonstrate that insulin and PDGF can act via ADD1/SREBP-1c possibly by activating MAP kinase cascade. In accordance to that, ADD1/SREBP-1c is a direct substrate of MAP kinases ERK1 and ERK2 in vitro. EXPERIMENTAL PROCEDURES Plasmids and molecular cloning. Construction of the LDLR promoter reporter gene plasmid phLDL4, containing sterol regulatory element (sre)-1 flanked by two SP1 elements, was described previously (7). Expression vectors pFC-MEKK for dominant-active MEKK1 (aa 380-672) and pFC-MEK1 for activated MEK1 (D32-51/ S218E/S222E) under control of CMV promoter were purchased from Stratagene. Total RNA was isolated from HepG2 cells by guanidine isothiocyanate extraction protocol described previously (7). Using the FirstStrand cDNA synthesis kit (Pharmacia Biotech) the RNA was reverse transcribed following the directions of the manufacturer. A cDNA fragment corresponding to the amino acids 1-460 of SREBP1a, corresponding to transcriptional active domain, was amplified by 0006-291X/98 $25.00
375
Copyright q 1998 by Academic Press All rights of reproduction in any form reserved.
AID
BBRC 9161
/
695b$$$481
08-01-98 15:47:26
bbrcg
AP: BBRC
Vol. 249, No. 2, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Role of ADD1/SREBP-1c in hormonal activation of LDLR promoter activity. HepG2/pcDNA3 cells [A] and SREBP-1(-) cells without [B] or with ectopically expressed ADD1/SREBP-1c-NT (0.05 mg/well) [C] were transiently transfected with LDLR promoter construct phLDL4 (1 mg/well). Before harvesting as described under experimental procedures cells were incubated with insulin (1 x 1007 M) or PDGF (3.3 x 1009 M) for 4 h. Transfection efficiency was monitored and normalized by cotransfecting cells with pSV b-galactosidase vector (1 mg/ well). All transfection experiments were carried out in triplicate. The results are given as means ({S.D.) of five independent experiments, each done in triplicate.
PCR from this single stranded cDNA using the sequence-specific 5*-primer (GGGGATCCCCATGGACGAGCCACCCTTCA) and 3*primer (GGAATTCTCAGTC AGGCTCCGAGTCACTGCCA), which contain additional bases (underlined sequence) to generate a BamHI and a EcoRI restriction site, respectively. This fragment was inserted in frame to glutathione-S-transferase into pGEX 3X (Pharmacia Biotech). The transcriptional active domain of ADD1/SREBP-1c (aa 1-436), the splice variant of SREBP-1a, was amplified by PCR with BamHI/HgaI digested SREBP-1a-NT clone as matrix using sequence-specific 5*-primer (CGGGATCCCGATGGATTGCACTTTCGAAGACATGCTTCAGCTTATC) with BamHI restriction site (underlined sequence) and 3*-primer identical to 3*-primer of SREBP-1a. To obtain ADD1/SREBP-1c-NT in pGEX 3X vector the SREBP-1aNT was replaced by BamHI/SmaI digested PCR product. The ADD1/ SREBP-1c-NT expression vector was constructed by ligating the Nterminal domain of ADD1/SREBP-1c as a 1308 bp BamHI/EcoRI fragment into the BamHI and EcoRI site of pcDNA3 (Invitrogen). The sequence of the constructs was confirmed by using a model 373A DNA sequencer (Applied Biosystem Inc.). Transient transfection assay. HepG2 cells were maintained in RPMI 1640 medium (Sigma) supplemented with 10% fetal calf serum (GibcoBRL) and antibiotics. Mock-transfected (pcDNA3) HepG2 cells and HepG2 cells with reduced intracellular level (õ10%) of SREBP1 protein, named SREBP-1(-) cells, generated by antisense technique as described previously (7), were cultured in RPMI 1640 medium supplemented with 10% fetal calf serum (GibcoBRL), antibiotics and 500 mg/ml G418. Before electroporation, cells were released by trypsinization, washed and suspended in Opti-MEM (GibcoBRL) supplemented with 10% FCS. Cell suspensions (2 x 105 cells/well) were mixed with vectors as indicated in figure legends. Samples were transferred to an electroporation cuvette (interelectrode distance 0,4 cm) and pulsed for 18 msec in GenePulser II (Bio-Rad). Before seeding on 6 well plates (Costar) cell suspension was diluted with RPMI 1640 with 10% (v/v) FCS and antibiotics. If external induction experiments were carried out, cells were serum-starved on day two for 16
h. Before harvesting, cells were incubated with insulin (1007 M) or PDGF (3.3 x 1009 M) for 4 h. For endogenous induction using expression vectors for dominant active MEKK1 or activated MEK1, cells were harvested 16 h after electroporation without further treatment. Luciferase activity was measured according to the supplier’s instructions (Promega). Transfection efficiency was monitored by cotransfection of b-galactosidase expression vector pSV-bGal (Promega). bGalactosidase activity was determined by galactolight assay (Tropix). The data represent relative luciferase activity as x-fold induction of either endogenous or exogenous stimulation relative to unstimulated cells as indicated in the legends to the figures. Protein kinase assay. The glutathione-S-transferase (GST)ADD1/SREBP-1c-NT fusion protein was expressed in E. coli strain BL21 and purified according to the manufacturer’s recommendations (Pharmacia Biotech). Protein phosphorylation by MAP kinases ERK1 or ERK2 (Upstate Biotechnology) was performed with 10 mg GSTADD1/SREBP-1c-NT fusion protein or 0.5 mg MBP and activated GST-MAP kinase ERK1 (100 ng/assay) or ERK2 (100 ng/assay) fusion proteins in kinase buffer (5 mM MOPS (pH 7.2), 6.25 mM bglycerophosphate, 1.25 mM EGTA, 0.25 mM sodium-orthovanadate, 0.25 mM DTT). The reaction was initiated by addition of 50 mM [g-32P]ATP (10 Ci/mmol) in a final volume of 40 ml kinase buffer. The reaction was terminated after 15 min at 257C by addition of Laemmli sample buffer. The phosphorylation of ADD1/SREBP-1c-NT was examined after sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) by autoradiography. Protein phosphatase assay. GST-ADD1/SREBP-1c-NT fusion protein (1 mg), phosphorylated by activated ERK2, was suspended in AP buffer (50 mM Tris/HCl, pH 8.0, 0.1 mM EDTA). The reaction was initiated by addition of 0-0.5 U alkaline phosphatase (Boehringer Mannheim) in a final volume of 50 ml AP buffer. The reaction was terminated after 60 min at 257C by addition of Laemmli sample buffer. The dephosphorylation of ADD1/SREBP-1c-NT was examined after SDS-PAGE by autoradiography.
376
AID
BBRC 9161
/
695b$$$481
08-01-98 15:47:26
bbrcg
AP: BBRC
Vol. 249, No. 2, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 2. MEKK1 and MEK1 enhances transcriptional activation of LDLR gene by ADD1/SREBP-1c.HepG2 cells were transiently transfected with LDLR promoter construct phLDL4 (1 mg/well) along with expression vectors (0.05 mg/well) without (black bars) or containing (gray bars) [A] ADD1/SREBP-1c-NT and dominant active MEKK1 or [B] ADD1/SREBP-1c-NT and activated MEK1, respectively. Conditions are described under experimental procedures. Transfection efficiency was monitored and normalized by cotransfecting cells with pSV bgalactosidase vector (1mg/well). All transfection experiments were carried out in triplicate. The results are given as a mean ({S.D.) of five independent experiments, each done in triplicate.
RESULTS To evaluate the role of ADD1/SREBP-1c in signal transduction of insulin stable HepG2 cell lines were generated using anti-sense RNA technique being deficient in protein levels (õ10%) of SREBP-1 (7). Transient transfection experiments using this cell line showed that the effects of insulin and PDGF on LDLR promoter activity using a construct (phLDL4) containing an intact SREBP-1c binding sterol-regulatory cis element sre-1 were abolished in SREBP-1 deficient cells compared to mock-transfected control cell lines, see Fig. 1. Accordingly, overexpression of the transcriptional active N-terminal region of ADD1/SREBP-1c in SREBP-1-deficient cells led to a complete reconstitution of insulin and PDGF action on LDLR promoter activity. One major pathway linking receptor associated tyrosine kinases, like the receptors of insulin and PDGF, to the transcriptional machinery of various genes is the intracellular MAP kinase cascade. In accordance to that, we could show (Fig. 2) that overexpression of constitutive active MEKK1 or MEK1, which are upstream activators of MAP kinases, stimulated LDLR promoter activity in HepG2 cells several-fold. Ectopic expression of the transcriptional active region of ADD1/ SREBP-1c (ADD1/SREBP-1c-NT) had a stimulatory action on LDLR promoter activity per se and a synergistic effect when coexpressed with MEKK1 or MEK1.
To study, whether SREBP-1c might be a direct substrate of MAP kinases we produced and purified recombinant GST-fusion proteins of SREBP-1c. Incubation of ADD1/SREBP-1c-NT fusion protein with activated recombinant ERK1 or ERK2 showed significant phosphorylation, in vitro (Fig. 3). In contrast to that, the C-terminal region was not phosphorylated (data not
FIG. 3. Phosphorylation of ADD1/SREBP-1c by MAP kinases ERK1 and ERK2 in vitro. Bacterial-synthesized glutathione-S-transferase (GST)-ADD1/SREBP-1c-NT fusion protein was treated with activated GST-MAP kinase ERK1 or ERK2 fusion proteins and separated by SDS-PAGE followed by autoradiography. A representative result of five independent experiments are shown.
377
AID
BBRC 9161
/
695b$$$481
08-01-98 15:47:26
bbrcg
AP: BBRC
Vol. 249, No. 2, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 4. Dephosphorylation of ADD1/SREBP-1c by phosphatase in vitro. (GST)-ADD1/SREBP-1c-NT fusion protein phosphorylated by active GST-MAP kinase ERK2 fusion protein was treated with increasing amounts of alkaline phosphatase (AP), followed by SDSPAGE and autoradiography. A representative result of four independent experiments are shown.
shown). It is interesting to note, that phosphorylation by ERK2 appeared to be more prominent when compared to ERK1. Phosphoamino acid analyses revealed phosphorylation predominantly on serine and slightly on threonine residues (data not shown). This phosphorylation was reversible by incubating phosphorylated ADD1/SREBP-1c-NT with increasing concentrations of alkaline phosphatase (Fig. 4). The autoradiography of the SDS-PAGE showed two bands indicating possibly more than one phosphorylation site.
DISCUSSION The promoter of LDLR gene contains a sterol regulatory element (sre-1/ATCACCCCAC), which is regulated by the intracellular content of sterols (8). This DNA sequence is the target of three transcription factors, named SREBP-1a, SREBP-1c, and SREBP-2 (9,10). Some essential features of the structure-function relationship of SREBPs led them belong to the family of basic helix-loop-helix leucine zipper (bHLH-LZ) transcription factors. The basic domain of these proteins regulates DNA binding to a consensus sequence referred to as E-box motif (CANNTG). This latter motif is found in promoters of many genes. Most bHLH-LZ proteins contain an arginine in the basic region which restricts protein-DNA interaction to this E-box motif. However, SREBPs uniquely contain a tyrosine at this position instead of an arginine. Thereby, SREBPs might have a dual specificity, i.e. not only for E-box but also for sre-1-like (consensus sequence: Py-CA-Py; Py: pyrimidine) elements (11, 12). Sterol-responsive sre-1-like cis elements seem to exist in promoters of many more genes coding for enzymes involved not only in cholesterol metabolism, but also
in triglyceride synthesis and possibly others (for review, see 3, 4, 13). Furthermore, there are some data indicating that SREBPs might not only mediate the effects of intracellular cholesterol levels on gene transcription, but are also integrating signalling cascades activated by various stimuli (7, 14). Interestingly, SREBP-1c is the human homologue to ADD1, which has been cloned by Spiegelman and colleagues in an attempt to identify transcription factors linked to adipocyte differentiation (15). Expression of wild-type and dominant negative forms of ADD1 in preadipocytes and nonadipogenic cells have shown that ADD1/SREBP-1c plays a pivotal role in fat cell differentiation and gene expression (5, 15). Further experiments provided evidence, that ADD1/SREBP-1c promotes PPAR-g2-mediated transcriptional activity. Using a stable SREBP-1 deficient HepG2 cell line this study shows that the transcription factor ADD1/ SREBP-1c integrates the signalling of insulin as well as PDGF receptor associated kinases on the LDLR promoter. The LDLR promoter was used, because it contains the classical SREBPs binding cis element sre-1 (8). SREBP-1a and SREBP-1c are splice variants coded by a single gene (16). One major difference between SREBP-1a and SREBP-1c is an alternative exon 1. The antisense construct used to generate the SREBP-1 deficient cells is directed to exon 2 thereby affecting both splice variants. However, the reconstitution experiments (Fig. 1) were performed with a construct specific for the N-terminal domain of ADD1/SREBP-1c showing that the hormonal effects can be restored completely. This might indicate the existence of a regulatory mechanism affecting ADD1/SREBP-1c activity independent of intracellular sterol levels. Recently, the group of Spiegelman also provided evidence that ADD1/SREBP1c may be modulated by a signalling mechanism related to insulin (14). Since not only insulin but also PDGF action was affected by intracellular levels of ADD1/SREBP-1c we focussed on a common intracellular signalling pathway like MAP kinase cascade. Ectopic expression of upstream activators of MAP kinases like MEKK1 and MEK1 stimulates LDLR promoter activity. Coexpression of ADD1/SREBP-1c has a synergistic effect (Fig. 2). Since several transcription factors appear to be direct substrates of MAP kinases (17), we wanted to prove the hypothesis that ADD1/SREBP-1c might be a direct substrate of the MAP kinases ERK1 and ERK2. Our experiments show, that ADD1/SREBP1c is phosphorylated by ERK1 and ERK2, in vitro, the phosphorylation being reversible by alkaline phosphatase treatment. The fact that only the transcriptional active N-terminal domain is phosphorylated but not the C-terminal sequence involved in precursor cleavage supports the hypothesis that insulin regulates ADD1/ SREBP-1c mediated gene transcription by stimulating the activity rather than cleavage rate at the endoplasmatic reticulum of ADD1/SREBP-1c.
378
AID
BBRC 9161
/
695b$$$481
08-01-98 15:47:26
bbrcg
AP: BBRC
Vol. 249, No. 2, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Taken together, this study shows that ADD1/ SREBP-1c activity is not only regulated by intracellular sterol concentrations, but also by hormones acting possibly via phosphorylation. Accordingly, ADD1/ SREBP-1c might be a unique convergence point integrating metabolic and endocrine signals to gene regulatory events. ACKNOWLEDGMENTS We thank the German Ministry of Education and Research (BMBF/01 KS 9502) for current support. G.R. was supported by Graduiertenkolleg ‘‘Molekularbiologische Grundlagen Pathophysiologischer Vorga¨nge’’ of the German Research Foundation (DFG).
REFERENCES 1. Spiegelman, B. M. (1997) Eur. J. Med. Res. 2, 257–264. 2. Spiegelman, B. M. (1998) Diabetes 47, 507–514.
3. Brown, M. S., and Goldstein, J. L: (1997) Cell 89, 331–340. 4. Spiegelman, B. M., and Flier, J. S. (1996) Cell 5, 377–389. 5. Kim, J. B., and Spiegelman, M. B. (1996) Genes Dev. 10, 1096– 1107. 6. Shimano, H., et al. (1997) J. Clin. Invest. 29, 846–854. 7. Streicher R., et al. (1996) J. Biol. Chem. 271, 7128–7133. 8. Smith, J. R., et al. (1990) J. Biol. Chem. 265, 2306–2310. 9. Wang, X., et al. (1993) J. Biol. Chem. 268, 14497–14504. 10. Briggs, R. B., et al. (1993) J. Biol. Chem. 268, 14490–14496. 11. Kim, J. B., et al. (1995) Mol. Cell. Biol. 15, 2582–2588. 12. Magan˜a, M. M., and Osborne, T. F. (1996) J. Biol. Chem. 271, 32689–32694. 13. Mu¨ller-Wieland, D., et al. (1997) Curr. Opin. Lipidol. 8, 348– 353. 14. Kim, B., et al. (1998) J. Clin. Invest. 1, 1–9. 15. Tontonoz, P., et al. (1993) Mol. Cell Biol. 13, 4753–4759. 16. Hua, X., et al. (1995) Genomics 25, 667–673. 17. Whitmarsh, A. J., and Davis, R. J. (1996) J. Mol. Med. 74, 589– 607.
379
AID
BBRC 9161
/
695b$$$481
08-01-98 15:47:26
bbrcg
AP: BBRC