- Email: [email protected]
Insulin-Stimulated Fatty Acid Synthase Gene Expression Does Not Require Increased Sterol Response Element Binding Protein 1 Transcription in Primary Adipocytes
Biochemical and Biophysical Research Communications 291, 439 – 443 (2002) doi:10.1006/bbrc.2002.6467, available online at http://www.idealibrary.com on
Insulin-Stimulated Fatty Acid Synthase Gene Expression Does Not Require Increased Sterol Response Element Binding Protein 1 Transcription in Primary Adipocytes D. Gail Palmer, Guy A. Rutter, and Jeremy M. Tavare´ 1 Department of Biochemistry, School of Medical Sciences, University Walk, Bristol BS8 1TD, United Kingdom
Received January 9, 2002
Sterol response element binding protein 1c (SREBP1c) is a transcription factor that has been implicated in the regulation of expression of key lipogenic genes in hepatocytes, including fatty acid synthase (FAS) and glucokinase. In hepatocytes, insulin stimulates a rapid increase in transcription of SREBP-1c and the appearance of the SREBP-1c protein in the nucleus. SREBP-1 has also been proposed to play an important role in the induction of expression of lipogenic enzymes in adipose tissue in vivo in response to nutritional status. In this paper we have investigated the regulation of the SREBP-1 and FAS genes in adipocytes and find that while an overexpressed constitutively active SREBP-1 mutant is capable of substantially stimulating the FAS promoter, insulin appears to stimulate FAS gene expression in primary adipocytes in the absence of any apparent effect on SREBP-1 transcription. Taken together, our data suggest that insulin does not stimulate FAS gene expression through increasing SREBP-1c transcription in adipose cells. © 2002 Elsevier Science (USA) Key Words: fatty acid synthase; sterol response element binding protein; insulin; adipocyte; insulin.
Fatty acid synthase (FAS) is a key enzyme which catalyses the synthesis of palmitate from acetyl-CoA and malonyl-CoA. In liver and adipose tissues, it is well established that the transcription of the FAS gene is acutely and directly stimulated by insulin. Furthermore, transcription of the FAS gene is repressed by fasting and stimulated by refeeding a high-carbohydrate/low-fat diet [reviewed in (1–3)]. The mechanism by which insulin brings about this effect requires glucose transport or, more specifically, an intracellular Abbreviations used: FAS, fatty acid synthase; SREBP, sterol response element binding protein. 1 To whom correspondence and reprint requests should be addressed. Fax: ⫹44 117 954 6424. E-mail: [email protected].
metabolite of glucose that remains to be formally identified (4). Recently, the effect of insulin on FAS gene expression in the liver has been reported to be dependent on SREBP-1c, a member of the sterol response element binding protein (SREBP) family of transcription factors [reviewed in (5–7)]. The SREBP family comprises three known members; SREBP-1a and SREBP-1c which are derived from a single gene using alternative transcriptional start sites, and SREBP-2 which is encoded for by a distinct gene. SREBP-1c was independently discovered as a transcription factor that controls adipocyte differentiation and, as a result, has also been called ADD1 (8). All the SREBP isoforms are initially synthesized as transmembrane precursor proteins localized in the endoplasmic reticulum and nuclear membranes. Activation occurs by a proteolytic processing of the protein by two proteases acting successively to yield a mature active soluble transcription factor domain that translocates to the nucleus where it can bind to either an E-box (5⬘-CANNTG-3⬘) or sterol response element (SREs) (5⬘-TCACCCCCCAC-3⬘) in its target promoters [reviewed in (5–7)]. The FAS promoter contains both an E-box and an SRE, although it has been proposed that the E-box is more important in the upregulation of FAS promoter activity upon SREBP-1c binding (9). Adenoviral-mediated overexpression of a constitutively-active mutant of SREBP-1c increases the expression of FAS mRNA in rat liver in vivo and in primary cultures of rat hepatocytes; conversely, a dominant-inhibitory SREBP-1c mutant blocks the effect of insulin (10, 11). There is mounting evidence to suggest that the transcription of SREBP-1c is directly stimulated by insulin in the liver in vivo (12) and in primary cultures of hepatocytes (11). Insulin has also been reported to increase the amount of the mature form of SREBP-1c in the nucleus of hepatocytes, although it is not clear
439
0006-291X/02 $35.00 © 2002 Elsevier Science (USA) All rights reserved.
Vol. 291, No. 3, 2002
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
whether this is simply the result of increased net expression of SREBP-1c, or because insulin directly influences the proteolytic processing of the pre-SREBP-1c in the endoplasmic reticulum membrane (13). Furthermore, it has been reported that SREBP-1c is subject to phosphorylation by MAP kinase (14) although the role of MAP kinase in vivo requires further clarification as inhibitors of MAP kinase activation do not affect insulin-stimulated FAS gene expression (13, 15). In contrast to the situation with liver, the mechanism of regulation of SREBP-1c and FAS gene expression in adipose tissue has been less extensively studied. In murine adipose tissue, expression of both FAS and SREBP is decreased by fasting and increased by refeeding (9). In the case of SREBP-1c it is not clear whether this was the result of a direct effect of insulin on the adipocytes, or whether it resulted from alterations in the levels of counter-regulatory hormones in the animals. While SREBP-1c gene expression has been reported to be increased by insulin in cultured 3T3-L1 adipocytes (9) it is also not clear whether this increase can lead to a transactivation of the FAS promoter in these cells. In this study we have directly addressed both of these issues. In adipocytes, we find that a constitutively-active SREBP-1c can indeed transactivate the FAS promoter, although we find no effect of insulin on SREBP-1c expression in adipocytes under conditions where FAS expression is substantially stimulated. MATERIALS AND METHODS Plasmids. We constructed a luciferase reporter plasmid (pFASprom2000.Luc) where the firefly luciferase gene was placed under the control of a 2000-bp fragment of the FAS promoter (⫺2100 to ⫺65). To construct this plasmid we amplified the FAS promoter by PCR and cloned this into the MluI–HindIII sites of the firefly luciferase reporter plasmid pGL3basic (Promega, Madison, WI). pRL.SV40 (Promega) is a plasmid that encodes Renilla luciferase under the control of the constitutive SV40 promoter. A plasmid containing the N-terminal 403 amino acids of the murine SREBP-1c cDNA (SREBP-1c-403), and a plasmid encoding a dominant-negative SREBP-1c-403 possessing an alanine at amino acid 320 in the DNA binding domain (pSREBP-1c-DN), were as described (11). Reporter gene analysis in 3T3-L1 adipocytes. The murine 3T3-L1 fibroblast clone was obtained from ATCC (CCL 92.1; Manassas, U.S.A.). The cells were grown on coverslips, differentiated into adipocytes, and microinjected with plasmid DNA (a mixture of pFASprom-2000.Luc, 100 g/ml; pRL.SV40, 50 g/ml; pSREBP-1c403 or pSREBP-1c-403DN, 50 g/ml) as previously described (16). The cells were serum-starved in DMEM containing 20 mM glucose for 16 h prior to incubation with 100 nM insulin and photon-counting imaging as described (17–19). Briefly, analysis of firefly and Renilla luciferase expression in single microinjected cells was performed using an intensified Hamamatsu photon-counting camera (Hamamatsu, Japan) coupled to a Zeiss Axiovert 100TV microscope. Firefly luciferase activity was measured first by imaging the cells in phosphate-buffered saline for 5 min in the presence of 1 mM luciferin, and then the sum of the firefly and R. reniformis luciferase activities was determined by imaging for an additional 5 min in the
presence of 5 M coelenterazine. The ratio of firefly to Renilla luciferase activities for each cell were subsequently calculated using Argus 50 software (Hamamatsu) and Excel. Isolation and primary culture of rat adipocytes. Rat adipocytes were obtained by collagenase-treatment of epididymal fat pads from male Wistar rats (170 –185 g), as described previously (20). Immediately after isolation, the fat cells (approximately 0.5 ml of packed cell volume) were incubated in a total volume of 2 ml of DMEM at 37°C in a humidified atmosphere (5% CO 2/95% air) for the duration of the experiment (4 or 6 h). Insulin (100 nM) was added to the cells as indicated in the figure legends. Isolation of mRNA and Northern blotting. Total RNA was extracted from the adipocytes using Tri reagent (Sigma, Poole, UK) according to the manufacturer’s instructions. Poly(A) ⫹ RNA was then isolated from equal amounts of total RNA (typically 40 to 60 g per incubation condition) using an mRNA isolation kit (Roche Molecular Biochemicals, Lewes, UK) which separates mRNA from total RNA on biotin-labeled oligo-dT-coated streptavidin beads. The poly(A) ⫹ RNA was then separated by electrophoresis through a 0.9% agarose gel containing 2.2 mM formaldehyde, and transferred to a nylon membrane (Hybond-N; Amersham–Pharmacia Biotech, Amersham, UK) by capillary action for 20 h. For Northern analysis we used either a 32P-labeled 1.8-kb BamHI fragment of the FAS cDNA or a 1.2-kb EcoRI fragment of the SREBP-1 cDNA. 32P-labeling of the probes (25 ng each) was carried out using a random-priming kit (Rediprime labeling kit, Amersham–Pharmacia Biotech). The nylon membrane was prehybridized for 5 min in Perfecthyb Plus hybridization buffer (Sigma) and then with the 32P-labeled cDNA probe for 20 h. Membranes were washed with 2⫻ SSC or 2⫻ SSC/0.1% SDS to remove nonspecific binding. The extent of hybridization of the probes was determined using PhosphorImaging and ImageQuant software (Molecular Dynamics).
RESULTS AND DISCUSSION We have investigated the role of SREBP-1c in regulating FAS gene expression in adipocytes. As cultured 3T3-L1 adipocytes are refractory to standard transfection-based procedures, we introduced reporter genes by microinjection. When coupled with single cell luciferase imaging (19) we were able to measure the activity of the luciferase reporter gene by incubating the cells in the cofactor luciferin and then determining the level of luciferase expression by counting the individual photons produced by the cells using an intensified CCD camera (17, 19). Fully differentiated 3T3-L1 adipocytes were, therefore, microinjected with a firefly luciferase reporter gene under the control of a promoter comprising 2000 bp upstream of the FAS transcriptional start site. To correct for microinjection efficiency and cell viability we also coinjected a second reporter gene, the Renilla luciferase which was under the control of the constitutive SV40 promoter. Thus, firefly luciferase luminescence (Lum F) from single cells can be determined in the presence of luciferin, and then the combined luminescence of the firefly and Renilla luciferases (Lum F⫹R) can be measured during a further incubation with the cofactor for Renilla luciferase, coelenterazine. Finally, the ratio of firefly luciferase to Renilla luciferase expression level is computed
440
Vol. 291, No. 3, 2002
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Constitutively active SREBP-1c-403 upregulates FAS promoter activity in microinjected 3T3-L1 adipocytes. Fully differentiated 3T3-L1 adipocytes were microinjected with pFASprom2000.Luc and pRL.SV40, in the absence (Control) or presence of either pSREBP-1c-403 or pSREBP-1c-DN as indicated. The cells were incubated for 16 prior to photon-counting imaging as described under Materials and Methods. The firefly and Renilla luciferase activities in individual microinjected cells were calculated using Argus 50 software and the data are presented as ratios of firefly luciferase activity/Renilla luciferase activity (means ⫾ SEM for a minimum of 39 cells under each experimental condition).
off-line and is given by Lum F/(Lum F⫹R ⫺ Lum F). The resulting value therefore reflects the activity of the FAS promoter corrected for microinjection efficiency and cell viability. SREBP-1c can be rendered constitutively active by expression of the N-terminal 403 amino acids [i.e., residues 1– 403; SREBP-1c-403; (10)] which constitute the DNA binding and dimerization domains. An inactive (dominant-negative) form of this fragment can be generated by mutating residue 320 to alanine such that DNA binding, but not dimerization, is disrupted (SREBP-1c-DN) (21). As shown in Fig. 1, the activity of the FAS promoter was stimulated approximately fivefold when the constitutively-active SREBP was coexpressed with the FAS reporter gene. In contrast, the SREBP-1c-DN construct was without effect (Fig. 1). Thus elevated levels of active SREBP-1c are clearly capable of transactivating the FAS promoter in these cells, and this effect requires the ability of the SREBP-1c-403 to bind DNA. While others have reported that insulin can increase FAS mRNA expression in 3T3-L1 adipocytes (9, 22), we do not see a significant enough effect of insulin on the activity of the FAS promoter using this technique to allow us to confidently determine the effect of the dominant-negative SREBP-1c on insulin action on this promoter in adipocytes. To examine whether insulin has a direct effect on the expression of the SREBP-1c mRNA we used freshly isolated rat adipocytes. Immediately after isolation, the cells were incubated in the absence or presence of
insulin and then the levels of FAS and SREBP-1c mRNAs were determined by northern blotting. As shown in Fig. 2A, insulin had a significant effect on the expression of FAS transcripts such that after 6 h of treatment this reached four-fold over basal (Fig. 2B). While insulin clearly stimulated FAS expression at both the 4- and 6-h time points, it was without any detectable effect on the level of expression of SREBP-1c mRNA (Figs. 2A and 2B). Theoretically, insulin could regulate nuclear levels of SREBP-1c activity in one or more of three distinct ways: (i) by altering the expression level of SREBP-1c through transcriptional regulation; (ii) by altering the proteolytic processing of endoplasmic reticulumderived SREBP-1c such that the level of the active,
FIG. 2. Insulin stimulates FAS, but not SREBP-1c, gene expression in primary rat adipocytes. In A, primary cultures of adipocytes were prepared by collagenase treatment of epididymal fat pads from Wistar rats. Cells were cultured for 4 or 6 h, as indicated, in DMEM containing 5% BSA in the absence (C) or presence (I) of 100 nM insulin. Poly(A) ⫹ RNA was isolated and analyzed by Northern blotting using 32P-labeled probes for FAS (upper panel) and SREBP-1 (lower panel). In B, the data pooled from four independent experiments, performed as described in the legend to A, are presented for the expression level of FAS (closed circles) or SREBP-1c (open circles) relative to control (*P ⬍ 0.05, **P ⬍ 0.01, or #no significant difference, with respect to the basal level).
441
Vol. 291, No. 3, 2002
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
soluble factor in the nucleus increases through translocation from the cytoplasm; (iii) by a covalent modification (e.g., phosphorylation/dephosphorylation) of SREBP-1c such that it becomes a more efficient transactivator of gene expression. While there is a considerable body of evidence that favors a role for insulinstimulated SREBP-1c gene expression in the induction of FAS gene expression in liver, this is not the case for adipose tissue. While an artificial increase in the level of expression of SREBP-1c (i.e., the active SREBP-1c-403 fragment) can substantially induce the FAS gene promoter in 3T3-L1 cells (Fig. 1), insulin does not increase SREBP-1c transcript expression in freshly isolated adipocytes (Fig. 2). We propose, therefore, that insulinstimulated FAS gene expression in adipose tissue is unlikely to be a consequence of increased SREBP-1c gene expression. These observations have a number of important implications. Firstly, while fasting and refeeding are able to decrease and increase, respectively, the level of SREBP-1c in adipose tissue of mice, this phenomenon may not be due to a direct effect of insulin on the tissue but rather due to alterations in the levels of counterregulatory hormones. This is in contrast to the situation in liver in which insulin has a direct effect on SREBP-1c expression. Secondly, the effect of insulin on FAS gene expression in adipose tissue is well known to require glucose transport (which is substantially stimulated by insulin) and, at least partially, glucose metabolism. The degree to which changes in SREBP-1c activity additionally contribute to the insulin effect on FAS gene expression in this tissue is not known. Our results suggest that changes in transcription are unlikely to play a role. It is more likely, therefore, that insulin either promotes an increase in nuclear SREBP-1c activity due to proteolysis from endoplasmic reticulumderived precursor of SREBP-1c, or transactivates preexisting SREBP-1c via an insulin-induced phosphorylation by MAP kinase or another insulin-stimulated protein kinase. Attempts to investigate the contribution of these possibilities have been hampered by the apparent insolubility of nuclear SREBP-1c in adipocytes such that it remains extremely tightly associated with a chromatin containing fraction and, as a consequence, highly resistant to extraction and isolation (D.G.P., G.A.R., and J.M.T., unpublished observations). Interestingly, it has been reported that insulin increases the amount of nuclear SREBP-1c (13) largely as a result of increased expression of the SREBP1 gene, rather than stimulation of SREBP1 processing. Moreover, increases in the amount of SREBP1c in the nucleus occurred only after the activation of transcription of target genes, including glucokinase (13). Since SREBP1c is indispensable for the expression of a num-
ber of lipogenic genes in the liver (11), it was suggested that a posttranslation modification of the mature nuclear SREBP1c fragment, may also be involved in further stimulating the activity of the factor. The nature of this modification and its importance in adipocytes remain to be tested. In conclusion, we show here that overexpression of SREBP1c is sufficient to activate the FAS promoter, consistent with data in other tissues (13). However, it appears unlikely that activation of SREBP1c transcription is the principal mechanism whereby insulin activates the FAS gene in freshly isolated rat adipocytes. ACKNOWLEDGMENTS We thank Pascal Ferre´ (INSERM, Paris, France) and Kenny Webster and Gavin Welsh (University of Bristol, UK) for useful discussions. This work was supported by grants from the Medical Research Council, the European Union FAIR Programme (CT97-3011), and Diabetes UK.
REFERENCES 1. Ferre´ , P. (1999) Proc. Nutr. Soc. 58, 621– 623. 2. O’Brien, R. M., and Granner, D. K. (1996) Physiol. Rev. 76, 1109 –1161. 3. Sul, H. S., Latasa, M. J., Moon, Y., and Kim, K. H. (2000) J. Nutr. 130, 315S–320S. 4. Girard, J., Ferre´ , P., and Foufelle, F. (1997) Annu. Rev. Nutr. 17, 325–352. 5. Brown, M. S., and Goldstein, J. L. (1999) Proc. Natl. Acad. Sci. USA 96, 11041–11048. 6. Osborne, T. F. (2000) J. Biol. Chem. 275, 32379 –32382. 7. Ferre´ , P., Foretz, M., Azzout-Marniche, A., Be´ card, D., and Foufelle, F. (2001) Biochem. Soc. Trans. 29, 547–552. 8. Tontonoz, P., Kim, J. B., Graves, R. A., and Spiegelman, B. M. (1993) Mol. Cell. Biol. 13, 4753– 4759. 9. Kim, J. B., Sarraf, P., Wright, M., Yao, K. M., Mueller, E., Solanes, G., Lowell, B. B., and Spiegelman, B. M. (1998) J. Clin. Invest. 101, 1–9. 10. Foretz, M., Guichard, C., Ferre´ , P., and Foufelle, F. (1999) Proc. Natl Acad. Sci. USA 96, 12737–12742. 11. Foretz, M., Pacot, C., Dugail, I., Le-Marchand, P., Guichard, C., Le Liepvre, X., Berthelier-Lubrano, C., Spiegelman, B., Kim, J. B., Ferre´ , P., and Foufelle, F. (1999) Mol. Cell. Biol. 19, 3760 –3768. 12. Shimomura, I., Bashmakov, Y., Ikemoto, S., Horton, J. D., Brown, M. S., and Goldstein, J. L. (1999) Proc. Natl Acad. Sci. USA 96, 13656 –13661. 13. Azzout-Marniche, D., Becard, D., Guichard, C., Foretz, M., Ferre´ , P., and Foufelle, F. (2000) Biochem. J. 350, 389 –393. 14. Kotzka, J., Muller-Wieland, D., Koponen, A., Njamen, D., Kremer, L., Roth, G., Munck, M., Knebel, B., and Krone, W. (1998) Biochem. Biophys. Res. Commun. 249, 375–379. 15. Iynedjian, P. B., Roth, R. A., Fleischmann, M., and Gjinovci, A. (2000) Biochem. J. 351, 621– 627. 16. Oatey, P. B., VanWeering, D. H. J., Dobson, S. P., Gould, G. W., and Tavare´ , J. M. (1997) Biochem. J. 327, 637– 642.
442
Vol. 291, No. 3, 2002
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
17. Rutter, G. A., White, M. R., and Tavare´ , J. M. (1995) Curr. Biol. 5, 890 – 899. 18. Kennedy, H. J., Rafiq, I., Pouli, A. E., and Rutter, G. A. (1999) Biochem. J. 342, 275–280. 19. Rutter, G. A., Kennedy, H. J., Wood, C. D., White, M. R., and Tavare´ , J. M. (1998) Chem. Biol. 5, R285–R290.
20. Moule, S. K., Edgell, N. J., Welsh, G. I., Diggle, T. A., Foulstone, E. J., Heesom, K. J., Proud, C. G., and Denton, R. M. (1995) Biochem. J. 311, 595– 601. 21. Kim, J. B., and Spiegelman, B. M. (1996) Genes Dev. 10, 1096–1107. 22. Paulauskis, J. D., and Sul, H. S. (1988) J. Biol. Chem. 263, 7049 –7054.
443
Insulin-Stimulated Fatty Acid Synthase Gene Expression Does Not Require Increased Sterol Response Element Binding Protein 1 Transcription in Primary Adipocytes D. Gail Palmer, Guy A. Rutter, and Jeremy M. Tavare´ 1 Department of Biochemistry, School of Medical Sciences, University Walk, Bristol BS8 1TD, United Kingdom
Received January 9, 2002
Sterol response element binding protein 1c (SREBP1c) is a transcription factor that has been implicated in the regulation of expression of key lipogenic genes in hepatocytes, including fatty acid synthase (FAS) and glucokinase. In hepatocytes, insulin stimulates a rapid increase in transcription of SREBP-1c and the appearance of the SREBP-1c protein in the nucleus. SREBP-1 has also been proposed to play an important role in the induction of expression of lipogenic enzymes in adipose tissue in vivo in response to nutritional status. In this paper we have investigated the regulation of the SREBP-1 and FAS genes in adipocytes and find that while an overexpressed constitutively active SREBP-1 mutant is capable of substantially stimulating the FAS promoter, insulin appears to stimulate FAS gene expression in primary adipocytes in the absence of any apparent effect on SREBP-1 transcription. Taken together, our data suggest that insulin does not stimulate FAS gene expression through increasing SREBP-1c transcription in adipose cells. © 2002 Elsevier Science (USA) Key Words: fatty acid synthase; sterol response element binding protein; insulin; adipocyte; insulin.
Fatty acid synthase (FAS) is a key enzyme which catalyses the synthesis of palmitate from acetyl-CoA and malonyl-CoA. In liver and adipose tissues, it is well established that the transcription of the FAS gene is acutely and directly stimulated by insulin. Furthermore, transcription of the FAS gene is repressed by fasting and stimulated by refeeding a high-carbohydrate/low-fat diet [reviewed in (1–3)]. The mechanism by which insulin brings about this effect requires glucose transport or, more specifically, an intracellular Abbreviations used: FAS, fatty acid synthase; SREBP, sterol response element binding protein. 1 To whom correspondence and reprint requests should be addressed. Fax: ⫹44 117 954 6424. E-mail: [email protected].
metabolite of glucose that remains to be formally identified (4). Recently, the effect of insulin on FAS gene expression in the liver has been reported to be dependent on SREBP-1c, a member of the sterol response element binding protein (SREBP) family of transcription factors [reviewed in (5–7)]. The SREBP family comprises three known members; SREBP-1a and SREBP-1c which are derived from a single gene using alternative transcriptional start sites, and SREBP-2 which is encoded for by a distinct gene. SREBP-1c was independently discovered as a transcription factor that controls adipocyte differentiation and, as a result, has also been called ADD1 (8). All the SREBP isoforms are initially synthesized as transmembrane precursor proteins localized in the endoplasmic reticulum and nuclear membranes. Activation occurs by a proteolytic processing of the protein by two proteases acting successively to yield a mature active soluble transcription factor domain that translocates to the nucleus where it can bind to either an E-box (5⬘-CANNTG-3⬘) or sterol response element (SREs) (5⬘-TCACCCCCCAC-3⬘) in its target promoters [reviewed in (5–7)]. The FAS promoter contains both an E-box and an SRE, although it has been proposed that the E-box is more important in the upregulation of FAS promoter activity upon SREBP-1c binding (9). Adenoviral-mediated overexpression of a constitutively-active mutant of SREBP-1c increases the expression of FAS mRNA in rat liver in vivo and in primary cultures of rat hepatocytes; conversely, a dominant-inhibitory SREBP-1c mutant blocks the effect of insulin (10, 11). There is mounting evidence to suggest that the transcription of SREBP-1c is directly stimulated by insulin in the liver in vivo (12) and in primary cultures of hepatocytes (11). Insulin has also been reported to increase the amount of the mature form of SREBP-1c in the nucleus of hepatocytes, although it is not clear
439
0006-291X/02 $35.00 © 2002 Elsevier Science (USA) All rights reserved.
Vol. 291, No. 3, 2002
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
whether this is simply the result of increased net expression of SREBP-1c, or because insulin directly influences the proteolytic processing of the pre-SREBP-1c in the endoplasmic reticulum membrane (13). Furthermore, it has been reported that SREBP-1c is subject to phosphorylation by MAP kinase (14) although the role of MAP kinase in vivo requires further clarification as inhibitors of MAP kinase activation do not affect insulin-stimulated FAS gene expression (13, 15). In contrast to the situation with liver, the mechanism of regulation of SREBP-1c and FAS gene expression in adipose tissue has been less extensively studied. In murine adipose tissue, expression of both FAS and SREBP is decreased by fasting and increased by refeeding (9). In the case of SREBP-1c it is not clear whether this was the result of a direct effect of insulin on the adipocytes, or whether it resulted from alterations in the levels of counter-regulatory hormones in the animals. While SREBP-1c gene expression has been reported to be increased by insulin in cultured 3T3-L1 adipocytes (9) it is also not clear whether this increase can lead to a transactivation of the FAS promoter in these cells. In this study we have directly addressed both of these issues. In adipocytes, we find that a constitutively-active SREBP-1c can indeed transactivate the FAS promoter, although we find no effect of insulin on SREBP-1c expression in adipocytes under conditions where FAS expression is substantially stimulated. MATERIALS AND METHODS Plasmids. We constructed a luciferase reporter plasmid (pFASprom2000.Luc) where the firefly luciferase gene was placed under the control of a 2000-bp fragment of the FAS promoter (⫺2100 to ⫺65). To construct this plasmid we amplified the FAS promoter by PCR and cloned this into the MluI–HindIII sites of the firefly luciferase reporter plasmid pGL3basic (Promega, Madison, WI). pRL.SV40 (Promega) is a plasmid that encodes Renilla luciferase under the control of the constitutive SV40 promoter. A plasmid containing the N-terminal 403 amino acids of the murine SREBP-1c cDNA (SREBP-1c-403), and a plasmid encoding a dominant-negative SREBP-1c-403 possessing an alanine at amino acid 320 in the DNA binding domain (pSREBP-1c-DN), were as described (11). Reporter gene analysis in 3T3-L1 adipocytes. The murine 3T3-L1 fibroblast clone was obtained from ATCC (CCL 92.1; Manassas, U.S.A.). The cells were grown on coverslips, differentiated into adipocytes, and microinjected with plasmid DNA (a mixture of pFASprom-2000.Luc, 100 g/ml; pRL.SV40, 50 g/ml; pSREBP-1c403 or pSREBP-1c-403DN, 50 g/ml) as previously described (16). The cells were serum-starved in DMEM containing 20 mM glucose for 16 h prior to incubation with 100 nM insulin and photon-counting imaging as described (17–19). Briefly, analysis of firefly and Renilla luciferase expression in single microinjected cells was performed using an intensified Hamamatsu photon-counting camera (Hamamatsu, Japan) coupled to a Zeiss Axiovert 100TV microscope. Firefly luciferase activity was measured first by imaging the cells in phosphate-buffered saline for 5 min in the presence of 1 mM luciferin, and then the sum of the firefly and R. reniformis luciferase activities was determined by imaging for an additional 5 min in the
presence of 5 M coelenterazine. The ratio of firefly to Renilla luciferase activities for each cell were subsequently calculated using Argus 50 software (Hamamatsu) and Excel. Isolation and primary culture of rat adipocytes. Rat adipocytes were obtained by collagenase-treatment of epididymal fat pads from male Wistar rats (170 –185 g), as described previously (20). Immediately after isolation, the fat cells (approximately 0.5 ml of packed cell volume) were incubated in a total volume of 2 ml of DMEM at 37°C in a humidified atmosphere (5% CO 2/95% air) for the duration of the experiment (4 or 6 h). Insulin (100 nM) was added to the cells as indicated in the figure legends. Isolation of mRNA and Northern blotting. Total RNA was extracted from the adipocytes using Tri reagent (Sigma, Poole, UK) according to the manufacturer’s instructions. Poly(A) ⫹ RNA was then isolated from equal amounts of total RNA (typically 40 to 60 g per incubation condition) using an mRNA isolation kit (Roche Molecular Biochemicals, Lewes, UK) which separates mRNA from total RNA on biotin-labeled oligo-dT-coated streptavidin beads. The poly(A) ⫹ RNA was then separated by electrophoresis through a 0.9% agarose gel containing 2.2 mM formaldehyde, and transferred to a nylon membrane (Hybond-N; Amersham–Pharmacia Biotech, Amersham, UK) by capillary action for 20 h. For Northern analysis we used either a 32P-labeled 1.8-kb BamHI fragment of the FAS cDNA or a 1.2-kb EcoRI fragment of the SREBP-1 cDNA. 32P-labeling of the probes (25 ng each) was carried out using a random-priming kit (Rediprime labeling kit, Amersham–Pharmacia Biotech). The nylon membrane was prehybridized for 5 min in Perfecthyb Plus hybridization buffer (Sigma) and then with the 32P-labeled cDNA probe for 20 h. Membranes were washed with 2⫻ SSC or 2⫻ SSC/0.1% SDS to remove nonspecific binding. The extent of hybridization of the probes was determined using PhosphorImaging and ImageQuant software (Molecular Dynamics).
RESULTS AND DISCUSSION We have investigated the role of SREBP-1c in regulating FAS gene expression in adipocytes. As cultured 3T3-L1 adipocytes are refractory to standard transfection-based procedures, we introduced reporter genes by microinjection. When coupled with single cell luciferase imaging (19) we were able to measure the activity of the luciferase reporter gene by incubating the cells in the cofactor luciferin and then determining the level of luciferase expression by counting the individual photons produced by the cells using an intensified CCD camera (17, 19). Fully differentiated 3T3-L1 adipocytes were, therefore, microinjected with a firefly luciferase reporter gene under the control of a promoter comprising 2000 bp upstream of the FAS transcriptional start site. To correct for microinjection efficiency and cell viability we also coinjected a second reporter gene, the Renilla luciferase which was under the control of the constitutive SV40 promoter. Thus, firefly luciferase luminescence (Lum F) from single cells can be determined in the presence of luciferin, and then the combined luminescence of the firefly and Renilla luciferases (Lum F⫹R) can be measured during a further incubation with the cofactor for Renilla luciferase, coelenterazine. Finally, the ratio of firefly luciferase to Renilla luciferase expression level is computed
440
Vol. 291, No. 3, 2002
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Constitutively active SREBP-1c-403 upregulates FAS promoter activity in microinjected 3T3-L1 adipocytes. Fully differentiated 3T3-L1 adipocytes were microinjected with pFASprom2000.Luc and pRL.SV40, in the absence (Control) or presence of either pSREBP-1c-403 or pSREBP-1c-DN as indicated. The cells were incubated for 16 prior to photon-counting imaging as described under Materials and Methods. The firefly and Renilla luciferase activities in individual microinjected cells were calculated using Argus 50 software and the data are presented as ratios of firefly luciferase activity/Renilla luciferase activity (means ⫾ SEM for a minimum of 39 cells under each experimental condition).
off-line and is given by Lum F/(Lum F⫹R ⫺ Lum F). The resulting value therefore reflects the activity of the FAS promoter corrected for microinjection efficiency and cell viability. SREBP-1c can be rendered constitutively active by expression of the N-terminal 403 amino acids [i.e., residues 1– 403; SREBP-1c-403; (10)] which constitute the DNA binding and dimerization domains. An inactive (dominant-negative) form of this fragment can be generated by mutating residue 320 to alanine such that DNA binding, but not dimerization, is disrupted (SREBP-1c-DN) (21). As shown in Fig. 1, the activity of the FAS promoter was stimulated approximately fivefold when the constitutively-active SREBP was coexpressed with the FAS reporter gene. In contrast, the SREBP-1c-DN construct was without effect (Fig. 1). Thus elevated levels of active SREBP-1c are clearly capable of transactivating the FAS promoter in these cells, and this effect requires the ability of the SREBP-1c-403 to bind DNA. While others have reported that insulin can increase FAS mRNA expression in 3T3-L1 adipocytes (9, 22), we do not see a significant enough effect of insulin on the activity of the FAS promoter using this technique to allow us to confidently determine the effect of the dominant-negative SREBP-1c on insulin action on this promoter in adipocytes. To examine whether insulin has a direct effect on the expression of the SREBP-1c mRNA we used freshly isolated rat adipocytes. Immediately after isolation, the cells were incubated in the absence or presence of
insulin and then the levels of FAS and SREBP-1c mRNAs were determined by northern blotting. As shown in Fig. 2A, insulin had a significant effect on the expression of FAS transcripts such that after 6 h of treatment this reached four-fold over basal (Fig. 2B). While insulin clearly stimulated FAS expression at both the 4- and 6-h time points, it was without any detectable effect on the level of expression of SREBP-1c mRNA (Figs. 2A and 2B). Theoretically, insulin could regulate nuclear levels of SREBP-1c activity in one or more of three distinct ways: (i) by altering the expression level of SREBP-1c through transcriptional regulation; (ii) by altering the proteolytic processing of endoplasmic reticulumderived SREBP-1c such that the level of the active,
FIG. 2. Insulin stimulates FAS, but not SREBP-1c, gene expression in primary rat adipocytes. In A, primary cultures of adipocytes were prepared by collagenase treatment of epididymal fat pads from Wistar rats. Cells were cultured for 4 or 6 h, as indicated, in DMEM containing 5% BSA in the absence (C) or presence (I) of 100 nM insulin. Poly(A) ⫹ RNA was isolated and analyzed by Northern blotting using 32P-labeled probes for FAS (upper panel) and SREBP-1 (lower panel). In B, the data pooled from four independent experiments, performed as described in the legend to A, are presented for the expression level of FAS (closed circles) or SREBP-1c (open circles) relative to control (*P ⬍ 0.05, **P ⬍ 0.01, or #no significant difference, with respect to the basal level).
441
Vol. 291, No. 3, 2002
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
soluble factor in the nucleus increases through translocation from the cytoplasm; (iii) by a covalent modification (e.g., phosphorylation/dephosphorylation) of SREBP-1c such that it becomes a more efficient transactivator of gene expression. While there is a considerable body of evidence that favors a role for insulinstimulated SREBP-1c gene expression in the induction of FAS gene expression in liver, this is not the case for adipose tissue. While an artificial increase in the level of expression of SREBP-1c (i.e., the active SREBP-1c-403 fragment) can substantially induce the FAS gene promoter in 3T3-L1 cells (Fig. 1), insulin does not increase SREBP-1c transcript expression in freshly isolated adipocytes (Fig. 2). We propose, therefore, that insulinstimulated FAS gene expression in adipose tissue is unlikely to be a consequence of increased SREBP-1c gene expression. These observations have a number of important implications. Firstly, while fasting and refeeding are able to decrease and increase, respectively, the level of SREBP-1c in adipose tissue of mice, this phenomenon may not be due to a direct effect of insulin on the tissue but rather due to alterations in the levels of counterregulatory hormones. This is in contrast to the situation in liver in which insulin has a direct effect on SREBP-1c expression. Secondly, the effect of insulin on FAS gene expression in adipose tissue is well known to require glucose transport (which is substantially stimulated by insulin) and, at least partially, glucose metabolism. The degree to which changes in SREBP-1c activity additionally contribute to the insulin effect on FAS gene expression in this tissue is not known. Our results suggest that changes in transcription are unlikely to play a role. It is more likely, therefore, that insulin either promotes an increase in nuclear SREBP-1c activity due to proteolysis from endoplasmic reticulumderived precursor of SREBP-1c, or transactivates preexisting SREBP-1c via an insulin-induced phosphorylation by MAP kinase or another insulin-stimulated protein kinase. Attempts to investigate the contribution of these possibilities have been hampered by the apparent insolubility of nuclear SREBP-1c in adipocytes such that it remains extremely tightly associated with a chromatin containing fraction and, as a consequence, highly resistant to extraction and isolation (D.G.P., G.A.R., and J.M.T., unpublished observations). Interestingly, it has been reported that insulin increases the amount of nuclear SREBP-1c (13) largely as a result of increased expression of the SREBP1 gene, rather than stimulation of SREBP1 processing. Moreover, increases in the amount of SREBP1c in the nucleus occurred only after the activation of transcription of target genes, including glucokinase (13). Since SREBP1c is indispensable for the expression of a num-
ber of lipogenic genes in the liver (11), it was suggested that a posttranslation modification of the mature nuclear SREBP1c fragment, may also be involved in further stimulating the activity of the factor. The nature of this modification and its importance in adipocytes remain to be tested. In conclusion, we show here that overexpression of SREBP1c is sufficient to activate the FAS promoter, consistent with data in other tissues (13). However, it appears unlikely that activation of SREBP1c transcription is the principal mechanism whereby insulin activates the FAS gene in freshly isolated rat adipocytes. ACKNOWLEDGMENTS We thank Pascal Ferre´ (INSERM, Paris, France) and Kenny Webster and Gavin Welsh (University of Bristol, UK) for useful discussions. This work was supported by grants from the Medical Research Council, the European Union FAIR Programme (CT97-3011), and Diabetes UK.
REFERENCES 1. Ferre´ , P. (1999) Proc. Nutr. Soc. 58, 621– 623. 2. O’Brien, R. M., and Granner, D. K. (1996) Physiol. Rev. 76, 1109 –1161. 3. Sul, H. S., Latasa, M. J., Moon, Y., and Kim, K. H. (2000) J. Nutr. 130, 315S–320S. 4. Girard, J., Ferre´ , P., and Foufelle, F. (1997) Annu. Rev. Nutr. 17, 325–352. 5. Brown, M. S., and Goldstein, J. L. (1999) Proc. Natl. Acad. Sci. USA 96, 11041–11048. 6. Osborne, T. F. (2000) J. Biol. Chem. 275, 32379 –32382. 7. Ferre´ , P., Foretz, M., Azzout-Marniche, A., Be´ card, D., and Foufelle, F. (2001) Biochem. Soc. Trans. 29, 547–552. 8. Tontonoz, P., Kim, J. B., Graves, R. A., and Spiegelman, B. M. (1993) Mol. Cell. Biol. 13, 4753– 4759. 9. Kim, J. B., Sarraf, P., Wright, M., Yao, K. M., Mueller, E., Solanes, G., Lowell, B. B., and Spiegelman, B. M. (1998) J. Clin. Invest. 101, 1–9. 10. Foretz, M., Guichard, C., Ferre´ , P., and Foufelle, F. (1999) Proc. Natl Acad. Sci. USA 96, 12737–12742. 11. Foretz, M., Pacot, C., Dugail, I., Le-Marchand, P., Guichard, C., Le Liepvre, X., Berthelier-Lubrano, C., Spiegelman, B., Kim, J. B., Ferre´ , P., and Foufelle, F. (1999) Mol. Cell. Biol. 19, 3760 –3768. 12. Shimomura, I., Bashmakov, Y., Ikemoto, S., Horton, J. D., Brown, M. S., and Goldstein, J. L. (1999) Proc. Natl Acad. Sci. USA 96, 13656 –13661. 13. Azzout-Marniche, D., Becard, D., Guichard, C., Foretz, M., Ferre´ , P., and Foufelle, F. (2000) Biochem. J. 350, 389 –393. 14. Kotzka, J., Muller-Wieland, D., Koponen, A., Njamen, D., Kremer, L., Roth, G., Munck, M., Knebel, B., and Krone, W. (1998) Biochem. Biophys. Res. Commun. 249, 375–379. 15. Iynedjian, P. B., Roth, R. A., Fleischmann, M., and Gjinovci, A. (2000) Biochem. J. 351, 621– 627. 16. Oatey, P. B., VanWeering, D. H. J., Dobson, S. P., Gould, G. W., and Tavare´ , J. M. (1997) Biochem. J. 327, 637– 642.
442
Vol. 291, No. 3, 2002
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
17. Rutter, G. A., White, M. R., and Tavare´ , J. M. (1995) Curr. Biol. 5, 890 – 899. 18. Kennedy, H. J., Rafiq, I., Pouli, A. E., and Rutter, G. A. (1999) Biochem. J. 342, 275–280. 19. Rutter, G. A., Kennedy, H. J., Wood, C. D., White, M. R., and Tavare´ , J. M. (1998) Chem. Biol. 5, R285–R290.
20. Moule, S. K., Edgell, N. J., Welsh, G. I., Diggle, T. A., Foulstone, E. J., Heesom, K. J., Proud, C. G., and Denton, R. M. (1995) Biochem. J. 311, 595– 601. 21. Kim, J. B., and Spiegelman, B. M. (1996) Genes Dev. 10, 1096–1107. 22. Paulauskis, J. D., and Sul, H. S. (1988) J. Biol. Chem. 263, 7049 –7054.
443