Reduced IRS-2 and GLUT4 expression in PPARγ2-induced adipocytes derived from C/EBPβ and C/EBPδ-deficient mouse embryonic fibroblasts

Adipose tissue plays a central role in the regulation of energy storage and metabolism. For adipocyte differentiation, two transcription factor families,peroxisome proliferator-activated receptor γ (PPARγ) and CCAAT/enhancer-binding proteins (C/EBPs), are known to play critical roles(Fajas et al., 1998; Morrison and Farmer,1999; Rosen et al.,2000). In an in vitro model, some kinds of preadipocyte cell lines(3T3-L1 and 3T3-F422A etc.) and fibroblast cell lines (Swiss 3T3 etc.) could be transformed into mature adipocytes with the aid of an adipogenesis induction cocktail (dexamethasone, insulin, 1-methyl-3-isobutylxanthine and FCS) (Sadowski et al., 1992;Toscani et al., 1990). In these adipocyte differentiation systems, the expression of C/EBPβ and C/EBPδ is upregulated at the early stage, followed by an increase in C/EBPα and PPARγ expression. Gene-targeting studies have supported the importance of these molecules in adipocyte differentiation in vivo. C/EBPα-deficient mice failed to accumulate lipid in hepatocytes and adipocytes (Flodby et al.,1996; Wang et al.,1995). They could not store hepatic glycogen and died soon after birth because of hypoglycemia. In another study, C/EBPβ/δ-double deficient mice showed a reduction in the amount of lipid droplets in brown adipose tissue and the surviving adults showed significantly reduced epidydimal fat pads (Tanaka et al.,1997). Finally, the phenotypes of PPARγ-deficient mice showed that PPARγ is required for adipose tissue development as well as placental and cardiac development (Barak et al., 1999; Kubota et al.,1999; Rosen et al.,1999). Thus, although the C/EBP family proteins and PPARγare known to crossregulate each other's gene expression(Wu et al., 1996;Yeh et al., 1995), these molecules are important for adipocyte differentiation.

In adipose tissue, the ability of cells to respond to insulin and to express genes such as fatty-acid-binding protein (422/aP2), lipoprotein lipase(LPL), adipsin and glucose transporter 4 (GLUT4) is acquired during their differentiation into mature adipocytes(Cornelus et al., 1994). Since C/EBP-binding sites and/or PPAR response elements (PPRE) exist in the promoter regions of some of these genes, C/EBP proteins and PPARγ are involved to a large extent in the transcriptional regulation of these genes(Chaneval et al., 1991;Christy et al., 1989;MacDougald and Lane, 1995;Schoonjans et al., 1996). However, it has remained uncertain whether PPARγ or which C/EBP is involved in the acquisition of these characteristics. It has been recently shown that C/EBPα-deficient mouse embryonic fibroblasts (MEFs) can differentiate into adipocytes after the introduction of PPARγ2 and its subsequent activation (Wu et al.,1999). In these adipocytes lacking C/EBPα, the expression and phosphorylation of insulin receptor (IR) and insulin receptor substrate 1(IRS-1) are impaired and insulin-stimulated glucose uptake is reduced(Wu et al., 1999). These results indicate the essential roles of C/EBPα in the insulin signals of adipocytes. In contrast, it has not yet been demonstrated whether C/EBPβand C/EBPδ are involved in the insulin-responsiveness in adipocytes. To date, it has been difficult to clarify these issues, since C/EBPβ/δ-double deficient MEFs cannot differentiate into mature adipocytes with the aid of a standard adipogenesis induction cocktail(Tanaka et al., 1997). To tackle these issues, we introduced PPARγ2 into C/EBPβ/δ-double deficient MEFs and investigated phenotypic acquisition during adipocyte differentiation.

C/EBPβ^+/- and C/EBPδ^-/- mice were kindly provided by T. Tanaka and S. Akira (Research Institute for Microbiology,Osaka, Japan). C/EBPβ^-/- C/EBPδ^-/-(C/EBPβ/δ-deficient) mice were generated according to the method described previously (Tanaka et al.,1997). Primary MEFs were harvested from 14.5 d.p.c. embryos, and the cells cultured in α-MEM (Gibco BRL, Rockville, MD) supplemented with 10% FCS (Asahi Technoglass, Tokyo, Japan). The cells were plated onto 24-well or 60 mm collagen-coated dishes and passage was discontinued before the cells became confluent.

Mouse PPARγ2 cDNA was kindly provided by B. M. Spiegelman (Harvard Medical School, Cambridge, MA). To generate pAdex-mPPARγ2, mouse PPARγ2 cDNA was inserted into the adenovirus vector (pAdex). The adenovirus was then expanded by a series of infections to HEK293 cells, and twice purified by ultracentrifugation on a caesium chloride-gradient followed by dialysis. The viral titer was determined by plaque assay using HEK293 cells, and calculated as 1.0×10⁹ pfu/ml.

Wild-type and C/EBPβ/δ-deficient MEFs were propagated to confluence. Two days after reaching confluence, the cells were first infected with pAdex or pAdex-mPPARγ2 for 1 hour and then incubated with 3 μM 15-deoxy-Δ^12,14-prostaglandin J₂(15d-PGJ₂) (Cayman Chemical, Ann Arbor, MI) or 10 μM troglitazone (Sankyo, Tokyo, Japan). The media containing either of the drugs were renewed every day or every other day, respectively. Eight days after infection with pAdex or pAdex-mPPARγ2 and stimulation with these drugs,cytoplasmic lipid accumulation was detected with Oil Red O staining(Green and Kehinde, 1974).

Cytoplasmic lipid was isolated with 2-propanol, and the triglyceride content measured with a Triglyceride G-test kit (Wako Junyaku, Osaka, Japan). Whole cellular protein, prepared from another duplicated well with 1% Triton X-100, was measured with the BCA Protein Assay Reagent (Pierce, Rockford, IL). The triglyceride content was then corrected in terms of the protein content.

Total RNA was prepared from MEFs infected with pAdex-PPARγ2 followed by stimulation with troglitazone on the day indicated. Ten μg of RNA was electrophoresed by means of denaturing formaldehyde-agarose gel, and transferred to Hybond-N⁺ nylon membrane (Amersham Pharmacia Biotech, Little Chalfont, UK). The cDNA probes for mouse adipsin, 422/aP2,LPL, GLUT4, C/EBPα, C/EBPβ, C/EBPδ, PPARγ, IR, IRS-1 and IRS-2, were labeled with [α-³²P]dCTP by random primers and using the Megaprime DNA labeling system (Amersham Pharmacia Biotech). Northern blot hybridization was performed with QuickHyb Hybridization Solution(Stratagene, La Jolla, CA). Densitometrical analysis was performed with Immuno Reader FMBIO II (Hitachi, Kanagawa, Japan).

2-deoxyglucose (2-DG)-uptake assay was performed as described before with a minor modification (Fasshauer et al.,2000). Cells were serum-starved for 3 hours in DMEM containing 25 mM glucose and 2 mM glutamine, then incubated with 100 nM insulin at 37°C in serum-starved DMEM without glucose for 30 minutes and finally treated with[³H] 2-DG (0.33 μCi/ml) for an additional 5 minutes.[³H] 2-DG uptake was stopped by three washes with ice-cold PBS. Radioactivity was counted after the cells had been solubilized in 0.1% SDS. Specific [³H] 2-DG uptake was determined by subtraction of nonspecific counts in the presence of 10 μM cytochalasin B from each of the resultant values, which was then corrected in terms of its protein concentration (BCA Protein Assay Reagent).

Cells were harvested on days 0 and 8 during the adipogenic differentiation course and were lysed in 0.5 ml lysis buffer [10 mM Tris-HCl pH 7.6, 5 mM EDTA, 50 mM NaCl, 30 mM sodium pyrophosphate, 50 mM NaF, 0.1 mM Na₃VO₄, 1% Triton X-100, 1 mM PMSF and protease inhibitor cocktail tablet (Roche Molecular Biochemicals, Mannheim, Germany)]. Lysates were clarified by centrifugation at 15,000 g for 10 minutes. 30μg of protein was processed for SDS-PAGE, which was performed on 4-20%gradient gels. The proteins were electrophoretically transferred to Immobilon P (Millipore, Bedford, MA). The blots were blocked with 5% nonfat milk in Tris-buffered saline (TBS, pH 7.4) for 1 hour and then incubated with anti-IR(Santa Cruz Biotechnology, Santa Cruz, CA), anti-IRS-1 (Upstate Biotechnology,Lake Placid, NY), or anti-IRS-2 (Upstate Biotechnology) antibodies in 5%nonfat milk in TBS. They were then washed with TBS and incubated with donkey anti-rabbit IgG conjugated with horseradish peroxidase (1:1000; Amersham Pharmacia Biotech) in 5% nonfat milk in TBS. After washing with TBS, the bound antibodies were visualized by enhanced chemiluminescence (Amersham Pharmacia Biotech) and recorded on X-ray films (Fuji Medical, Tokyo, Japan). The relative amounts of IR, IRS-1 and IRS-2 were determined by densitometry scans.

To investigate whether PPARγ can induce adipocyte differentiation even in the absence of C/EBPβ and C/EBPδ, we infected wild-type MEFs and C/EBPβ/δ-deficient MEFs with mouse PPARγ2-expressing adenovirus vector (pAdex-mPPARγ2) and then stimulated these cells with troglitazone or 15-deoxy-Δ^12,14-prostaglandin J₂(15d-PGJ₂) (Kliewer et al.,1995; Lehman et al.,1995). Eight days after this treatment, both MEFs became round and lipid-laden as shown by Oil Red O staining(Fig. 1). By contrast,wild-type MEFs as well as C/EBPβ/δ-deficient MEFs infected with the empty adenovirus vector (pAdex) failed to demonstrate such morphological change. Intracellular triglyceride contents significantly increased after infection with pAdex-mPPARγ2 and stimulation with troglitazone, and the increases were similar for the adipocytes derived from wild-type MEFs and those from C/EBPβ/δ-deficient MEFs(Fig. 2).

C/EBPβ and C/EBPδ were expressed in wild-type MEFs. By ectopic expression and activation of PPARγ2, C/EBPβ mRNA levels gradually increased, whereas C/EBPδ mRNA levels were soon reduced in the wild-type cells (Fig. 3A,B). The expression of C/EBPα mRNA was detected on day 2 and showed gradually increasing levels (Fig. 3). By contrast, in C/EBPβ/δ-deficient cells the induction of C/EBPαmRNA was lower by ∼30-40% than that in wild-type cells throughout the differentiation course (Fig. 3A,B). Mature adipocytes are known to express proteins involved in fatty acid binding, lipogenesis and lipolysis, and in insulin-sensitive glucose uptake (MacDougald and Lane,1995). Therefore, to define the roles of C/EBPβ and C/EBPδ in such genes' expression during adipocyte differentiation, we compared their expression levels in PPARγ2-induced adipocytes derived from wild-type MEFs and those from C/EBPβ/δ-deficient MEFs. Northern blot analyses demonstrated that fatty-acid-binding protein (422/aP2)and lipoprotein lipase (LPL) mRNA levels started to increase two days after ectopic expression and activation of PPARγ2, reaching a plateau on day 5(Fig. 3A,B). The mRNA levels of wild-type cells and C/EBPβ/δ-deficient cells were not significantly different throughout the experimental course(Fig. 3A). By contrast, the levels of GLUT4 and adipsin mRNA, which also increased during the adipogenic differentiation, were reduced by 40-50% in the C/EBPβ/δ-deficient cells compared with those in the wild-type cells throughout the experiment(Fig. 3A,B).

In the next experiments, we analyzed insulin-stimulated 2-DG uptake in the mature adipocytes. Insulin did not stimulate [³H] 2-DG-uptake in either wild-type MEFs or C/EBPβ/δ-deficient MEFs before the adipocyte differentiation (Fig. 4). In adipocytes derived from wild-type MEFs, insulin significantly enhanced [³H] 2-DG-uptake by a factor of 3.7±0.5 (mean±s.e.m.). In adipocytes derived from C/EBPβ/δ-deficient MEFs, the insulin-stimulated [³H]2-DG-uptake was enhanced 2.1±0.3-fold, which was significantly lower(P<0.005) than that in adipocytes derived from wild-type MEFs(Fig. 4).

Western blot analyses demonstrated that the expression of IR and IRS-2, but not IRS-1, was upregulated during PPARγ2-induced adipogenic differentiation of wild-type MEFs (Fig. 5A). The levels of IR and IRS-1 were similar for the adipocytes derived from wild-type MEFs and those derived from C/EBPβ/δ-deficient MEFs, while the levels of IRS-2 in the C/EBPβ/δ-deficient adipocytes were reduced to approximately 50% of those in the adipocytes derived from wild-type MEFs(Fig. 5A). Northern blot analysis (Fig. 5B) also showed that the mRNA levels for IR and IRS-2, but not IRS-1, increased during PPARγ2-induced adipogenic differentiation of wild-type MEFs. IRS-2 mRNA levels were lower in adipocytes derived from C/EBPβ/δ-deficient MEFs than those in adipocytes derived from wild-type MEFs, whereas the mRNA levels for IR and IRS-1 were similar for the both types of adipocytes.

It has been shown that C/EBPβ/δ-deficient MEFs cannot differentiate into mature adipocytes with the aid of an adipogenesis induction cocktail (Tanaka et al.,1997). In the study presented here, we demonstrated for the first time that the ectopic expression of PPARγ2 and activation with its ligands could induce differentiation of C/EBPβ/δ-deficient MEFs into mature adipocytes, similar to that of wild-type MEFs in terms of cell morphology and accumulation of intracellular triglyceride. Since PPARγis known to be induced via expression of C/EBPβ and C/EBPδ(Wu et al., 1996), our results proved that PPARγ works as a downstream regulator of adipocyte differentiation from C/EBPβ and C/EBPδ.

During adipocyte differentiation, PPARγ and each of the C/EBPs are expressed sequentially. In 3T3-L1 cells treated with the adipogenesis induction cocktail, C/EBPβ and C/EBPδ are expressed in an early phase followed by the induction of C/EBPα and PPARγ(Cao et al., 1991;Wu et al., 1996). We showed that C/EBPα mRNA induction during PPARγ-induced adipocyte differentiation was less extensive in C/EBPβ/δ-deficient cells than that in wild-type cells, indicating that C/EBPβ and C/EBPδ are involved in the induction of C/EBPα during the PPARγ2-induced adipocyte differentiation. This provides direct evidence that C/EBPαexpression is regulated by C/EBPβ and C/EBPδ as well as PPARγ.

In PPARγ-induced adipocytes derived from C/EBPβ/δ-deficient MEFs, the expression of GLUT4 and adipsin mRNA was reduced compared with that in adipocytes derived from wild-type MEFs. By contrast, aP2 and LPL mRNA levels were the same in both adipocytes. Thus,C/EBPβ and C/EBPδ may be involved in the PPARγ2-induced expression of GLUT4 and adipsin, whereas they are not required for the PPARγ2-induced expression of aP2 and LPL. It has also been shown that PPARγ-expressing Swiss 3T3 and BALB/c-3T3 fibroblasts express aP2 and GLUT4 whereas PPARγ-expressing NIH-3T3 cells, despite similar adipocyte morphology and aP2 expression, do not express GLUT4(El-Jack et al., 1999). Together with our results, these findings indicate that such differences may be derived from differences in the expression levels of the various C/EBP molecules in these cells.

We showed that insulin-stimulated [³H] 2-DG-uptake was significantly impaired in C/EBPβ/δ-deficient adipocytes. This suggests the important roles of C/EBPβ and C/EBPδ in the acquisition of insulin-sensitive glucose uptake during adipocyte differentiation. While it is not clear whether PPARγ or which C/EBP is involved in the insulin-stimulated glucose uptake in mature adipocytes, it has been demonstrated in the study using β/δ39 NIH-3T3 fibroblasts expressing C/EBPβ and C/EBPδ but not C/EBPα, that the enhanced expression and activation of PPARγ stimulate synthesis of GLUT4 protein and insulin-responsive glucose uptake(Wu et al., 1998). This finding indicates that PPARγ alone or in combination with C/EBPβand C/EBPδ is capable of activating GLUT4 gene expression. Considering such data together with our results suggests that C/EBPβ and C/EBPδplay important roles in PPARγ-induced GLUT4 expression. The reduced GLUT4 expression may cause the reduction in insulin-stimulated glucose uptake in C/EBPβ/δ-deficient adipocytes.

Recently, C/EBPα has also been shown to be important for insulin-stimulated glucose transport activity, as the result of an analysis of PPARγ-induced adipocytes derived from C/EBPα-deficient MEFs(Wu et al., 1999). In the C/EBPα-deficient adipocytes, gene expression and tyrosine phosphorylation of IR and IRS-1 are reduced, while those of GLUT4 mRNA are increased. Our study demonstrated that, while C/EBPβ/δ-deficient adipocytes express somewhat lower levels of C/EBPα, these adipocytes express lower amounts of IRS-2 but not of IR and IRS-1. Therefore,C/EBPα-deficient and C/EBPβ/δ-deficient adipocytes are different in terms of the expression patterns of GLUT4, IR, IRS-1 and IRS-2. Recently, Fasshauer et al. showed that IRS-2, rather than IRS-1, is critical for insulin-stimulated GLUT4 translocation and glucose uptake in adipocytes(Fasshauer et al., 2000). Thus,the reduced expression of IRS-2 in C/EBPβ/δ-deficient adipocytes also may be responsible for the decrease in the insulin-stimulated glucose uptake.

PPARγ and C/EBP family proteins are expressed at specific times during adipogenesis. Complicated networks exist among these transcriptional factors: C/EBPβ and C/EBPδ turn on the expression of PPARγ(Wu et al., 1996). PPARγupregulates C/EBPα and C/EBPα is necessary to elevate the expression of PPARγ in differentiated adipocytes(Wu et al., 1999), and our study indicates that C/EBPβ and C/EBPδ are also involved in C/EBPα expression. Several gene-targeting studies have demonstrated the biological significance of PPARγ and C/EBP family proteins in adipogenic differentiation (Barak et al.,1999; Kubota et al.,1999; Rosen et al.,1999; Tanaka et al.,1997; Wang et al.,1995; Wu et al.,1999). In PPARγ-induced adipocytes derived from C/EBPα-deficient cells, the levels of fatty acid synthase (FAS),adipsin, LPL and PEPCK decreased whereas those of GLUT4 and aP2 rather increased (Wu et al., 1999). In PPARγ-induced adipocytes deficient of C/EBPβ and C/EBPδ,by contrast, we demonstrated a reduced expression of GLUT4 and adipsin mRNA but normal levels of aP2 and LPL mRNA. Again, C/EBPα deficiency is associated with reduced expression of IR and IRS-1(Wu et al., 1999), whereas C/EBPβ and C/EBPδ deficiencies are associated with reduced IRS-2 expression. Thus, each member of the C/EBP family proteins has a distinct role(s) in the gene regulation during adipogenesis. Activation of PPARγby thiazolidinediones is known to improve insulin sensitivity(Saltiel and Olefsky, 1996). Our results show that C/EBPβ and C/EBPδ are involved in the expression of some insulin-signaling molecules in adipocytes. Since adipose tissue is an important organ for insulin-stimulated glucose transport in the body (Abel et al., 2001), it is also possible that C/EBPβ and C/EBPδ are therapeutic targets for diabetes mellitus.

We thank S. Akira and T. Tanaka (Research Institute for Microbiology, Osaka University, Japan) for providing us with C/EBPβ^+/- mice and C/EBPδ^-/- mice. We also thank B. M. Spiegelman (Harvard Medical School, Cambridge, MA) for providing the mouse PPARγ2 cDNA, and I. Shimomura (Osaka University Graduate School of Medicine, Japan) for providing the mouse IR, IRS-1 and IRS-2 cDNA probes. This work was supported in part by Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan (S.K.) and from the Osaka Foundation for Clinical Immunology (S.K.).