INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Recently there has been a dramatic increase in the prevalence of obesity attributable to excessive white adipose tissue both in Western countries and in Japan. Because adipocytes play a critical role in energy balance, understanding the molecular mechanisms of adipocyte differentiation may provide clues to strategies for the prevention and treatment of obesity.

The mechanisms of adipocyte differentiation have been extensively studied in preadipocyte culture systems. Characterization of the regulatory regions of adipose-specific genes has led to the identification of key transcription factors in the complex transcriptional cascade that occurs during adipocyte differentiation (36); these factors include peroxisome proliferator-activated receptor  (PPAR) (19, 44), CCAAT/enhancer binding protein (C/EBP) (12, 23, 32, 46, 52, 56), and adipocyte differentiation and determination factor 1 (ADD1)-sterol regulatory element binding protein 1c (SREBP1c) (17, 18, 36, 45).

PPAR is induced before the transcriptional activation of most adipocyte-specific genes, and expression of PPAR has been shown to be sufficient to induce growth arrest and to initiate adipogenesis in exponentially growing fibroblast cell lines, thus demonstrating its critical role in the regulation of adipocyte differentiation (2, 15, 43). In addition, PPAR-deficient cells fail to differentiate into adipocytes, indicating that PPAR plays a pivotal role in adipocyte differentiation (21, 29). Most of the PPAR target genes in adipose tissue including the genes encoding phosphoenolpyruvate carboxykinase (41), lipoprotein lipase (LPL) (33) and adipocyte fatty acid binding protein (A-FABP or aP2) (42) are directly implicated in lipogenic pathways.

C/EBP is the most highly expressed member of C/EBP family in adipose tissue and in liver and has been implicated in the maintenance of the terminally differentiated adipocyte phenotype (4, 9, 25, 56). C/EBP is induced relatively late during adipogenesis in culture, after the induction of PPAR but before the induction of many of the enzymes and proteins characteristic of fully differentiated cells (53), and it transcriptionally activates a large number of adipocyte-specific genes (8). C/EBP-null mice fail to develop white adipose tissue (48), and C/EBP-deficient cells fail to differentiate into adipocytes (51). C/EBP (1, 7, 9, 10, 56) and C/EBP (4, 16, 49) are induced very early and have been shown to play a crucial role in initiating the differentiation of preadipocytes by activating the expression of PPAR (39, 50, 52, 56). An in vitro study of adipocyte differentiation in C/EBP^/ C/EBP^/ embryonic fibroblast (EF) cells shows that the expression of both PPAR and C/EBP is severely reduced (39).

In addition, there is a large amount of literature describing extracellular factors that influence adipogenic potentials. These include insulin-like growth factor 1 (IGF-1) and insulin. However, little is known about the signal transduction pathways by which these hormones regulate the expression of these transcription factors and promote adipogenesis.

IRS-1 and IRS-2 are the two most ubiquitously expressed members of the IRS family of proteins, which can bind signaling proteins with Src-homology-2 domains (SH2 proteins), such as p85 regulatory subunit of phosphatidylinositol 3-kinase (PI 3-kinase), subsequent to the activation of receptors for insulin, IGF-1, and several cytokines (26, 28, 37). IRS-1 plays an important role in the metabolic actions of insulin and IGF-1 mainly in skeletal muscle and adipose tissue, and IRS-2 plays an important role in the metabolic actions of these hormones in the liver (3, 55). The roles of IRS-1 and IRS-2 in adipocyte differentiation, on the other hand, have not been reported.

To investigate the role of IRS-1 and IRS-2 in adipocyte differentiation, we intercrossed mice heterozygous for each of two null alleles (Irs1^+/ [38] and Irs2^+/ [22]) and generated primary EF cells that were wild type, IRS-1 deficient (IRS-1^/), IRS-2 deficient (IRS-2^/), or IRS-1 IRS-2 double deficient (IRS-1^/ IRS-2^/).

We show here that IRS-1 IRS-2 double-deficient cells have no ability to differentiate into adipocytes. We also show that the simultaneous lack of both IRS-1 and IRS-2 dramatically decreases the expression of C/EBP and PPAR at both the mRNA and the protein levels during adipocyte differentiation. Thus, our data provide the first direct evidence for the essential role of IRS-1 and IRS-2 in the expression of adipose-specific transcriptional factors, such as the C/EBP family and PPAR, and adipocyte differentiation.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Materials. The probe for Northern blot and RNase protection assay to 36B4 was a generous gift from Naoya Yahagi (University of Tokyo) (34). The polyclonal antibodies to IRS-1, phosphotyrosine, PPAR, and C/EBP were from Santa Cruz Biotechnology, Inc. Polyclonal antibodies to phospho-mitogen-activated protein kinase (MAPK) and MAPK were from New England Biolabs, Inc. The medium and penicillin-streptomycin solution were purchased from Gibco, Inc. Fetal bovine serum (FBS) was from JRH Biosciences. The nitrocellulose paper used for the immunoblots was from Schleicher & Schuell, Inc. The 3-isobutyl-1-methylxanthine (IBMX) and human recombinant insulin were purchased from Sigma Chemical Co. Dexamethasone (DEX) was purchased from Wako Pure Chemical Industries, Ltd. L-type Wako TG-H (S-R1 and S-R2) for triglyceride measurement was purchased from Wako Pure Chemical Industries, Ltd. BCA Protein Assay Reagent (Pierce) was used for the protein assay.

Preparation of wild-type, IRS-1^/, IRS-2^/, and IRS-1^/ IRS-2^/ EFs. The IRS-1 IRS-2 double-heterozygous mice (IRS-1^+/ IRS-2^+/ mice) were generated by intercrosses of IRS-1-deficient mice (38) and IRS-2-deficient mice (22). To obtain embryos at 13.5 days past coitus, conceptuses that were obtained by fertilization in vitro of ova from IRS-1^+/ IRS-2^+/ female mice with sperm from IRS-1^+/ IRS-2^+/ male mice were implanted into pseudopregnant foster mothers, as previously described (21). The foster mothers were sacrificed at 13.5 days past coitus, the embryos were dissected from the uterus, and the extra-embryonic membranes and viscera were removed. The embryos were cut into small pieces with scissors and soaked for 30 min in 4 ml of 0.25% trypsin-EDTA at room temperature with shaking, and then inactivated with -modified Eagle medium (MEM) supplemented with 10% heat-inactivated FBS, 50 U of penicillin per ml, and 50 µg of streptomycin per ml. The cells were then suspended by pipetting and were plated on two 10-cm dishes per embryo. After 48 h, adherent cells were trypsinized, counted, and replated at a density of 1.2 × 10⁵ cells/cm² in MEM with 10% heat-inactivated FBS, 50 U of penicillin per ml, and 50 µg of streptomycin per ml to induce adipocyte differentiation.

Induction of adipocyte differentiation. The passage number of the EF cells used in these studies was within two passages. Induction of adipocyte differentiation was performed as previously described (39). In brief, cells were cultured on 24- or 6-well or 10-cm plastic dishes and propagated to confluence. Two days later, the medium was replaced with standard differentiation induction medium containing 0.5 mM IBMX, 1 µM DEX, 5 µg of insulin per ml, 10% FBS, 50 U of penicillin per ml, and 50 µg of streptomycin per ml, and the medium was renewed every other day.

Oil-Red O staining and triglyceride-protein assay. Oil-Red O staining solution (0.5% Oil-Red O in isopropyl alcohol solution-distilled water [60:40]) was filtered through the Whatman no. 1 filter paper and, after the cells were washed with phosphate-buffered saline (PBS), they were stained with the filtered staining solution for 30 min at 37°C and then washed with distilled water three times.

Retrovirus-mediated gene transfer. To rescue adipocyte differentiation of EF cells from IRS1^/ IRS2^/, the PPAR2 expression vector for retrovirus-mediated gene transfer was constructed by ligating the BstXI fragment from the pBabe-mPPAR2-puro (kindly provided by Bruce M. Spiegelman) into the BstXI site of pMX-puro (27). Mouse EF cells were infected with equal titers of each recombinant virus as described previously (44), with some modification.

PI 3-kinase assay. PI 3-kinase activity in EF cells was determined in immunoprecipitates with antibodies to phosphotyrosine as described previously (55), with some modification. In brief, EF cells were lysed with lysis buffer (0.25 M sucrose; 5 mM EDTA; 5 mM EGTA; 20 mM Tris-HCl, pH 7.4; 0.2 mM Na₃VO₄; 10 mM NaF; 10 mM sodium PP_i, 1 mM phenylmethylsulfonyl fluoride; 10 µg of leupeptin per ml; 10 µg of aprotinin per ml) containing 1% NP-40. The lysates were cleared by centrifugation at 12,000 rpm in a microcentrifuge at 4°C, and the protein content of the supernatant was determined. Then, 50 µg of the cellular protein was subjected to immunoprecipitation with monoclonal antiphosphotyrosine antibody (PY20; Santa Cruz) by using protein G-Sepharose, and the buffer was changed to PI 3-kinase assay buffer (100 mM sodium chloride; 25 mM Tris-HCl, pH 7.4; 0.5 mM EGTA). The PI 3-kinase activity was then measured as described previously (55).

RNA preparation, Northern blot analysis and RNase protection assay. Total RNA was prepared from EF cells with Trizol (Gibco-BRL) according to the manufacturer's instructions.

Immunoblot analysis. EF cells were lysed in a lysis buffer containing 1% NP-40, and insoluble materials were removed by centrifugation. Proteins in the supernatants were assayed with a BCA protein assay kit (Pierce) and then subjected to immunoblotting analysis.

Histologic analysis. To generate IRS-1^/ IRS-2^/ mice, conceptuses that were obtained by in vitro fertilization of ova from IRS-1^+/ IRS-2^+/ female mice and sperm from IRS-1^/ IRS-2^+/ male mice were implanted into pseudopregnant foster mothers, as previously described (21). At 8 h after birth, neonates were fixed in 10% formaldehyde in PBS. Genomic DNA was prepared from the tail of each neonate, and the genotype was determined by PCR analysis. Neonate was cut into 10-µm sections, and the sections were mounted on silanized slides. The adipose tissue was stained with hematoxylin and eosin. All sections were examined by light microscopy.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Adipocyte differentiation of IRS-1^/ cells is impaired. Wild-type and IRS-1^/ cells were induced to differentiation into adipocytes. Oil-Red O staining was performed 8 days after induction of adipocyte differentiation (Fig. 1A), and the increase in intracellular triglyceride content was measured (Fig. 1B). The ability of IRS-1^/ cells to differentiate into adipocytes was found to be approximately 60% lower than that of wild-type cells. These results were reproduced in six independent IRS-1^/ cells with six independent wild-type cells as controls.

View larger version (51K): [in this window] [in a new window]   FIG. 1.   Ability of wild-type and IRS-1/ cells to differentiate into adipocytes. (A) Oil-Red O staining for fat accumulation in wild-type and IRS-1/ cells at day 8 after induction. Cells were grown to confluence and then induced to differentiation by exposure to IBMX, DEX, and insulin, as described in Materials and Methods. (B) Increase in intracellular triglyceride content from day 0 to day 8 after induction. The assays were performed as described in Materials and Methods. The data represent the means ± the standard errors of the means from four experiments. Wild-type, n = 6; IRS-1/, n = 6; **, P < 0.01.

PI 3-kinase activation during adipocyte differentiation in IRS-1^/ cells was decreased. PI 3-kinase activities during adipocyte differentiation were measured in immunoprecipitates with antibody to phosphotyrosine from lysates of wild-type cells and IRS-1^/ cells (Fig. 2A). PI 3-kinase activity increased during adipocyte differentiation and reached a maximum at 8 days after induction, a result consistent with observations in adipocyte cell lines 3T3-L1 and 3T3-F442A (31). The increase in PI 3-kinase activity of in IRS-1^/ cells during adipocyte differentiation was approximately 50% less than that in the wild-type cells from 0 to 8 days after induction.

View larger version (35K): [in this window] [in a new window]   FIG. 2.   Role of PI 3-kinase in adipocyte differentiation of wild-type and IRS-1/ cells. (A) PI 3-kinase activities during adipocyte differentiation. The antiphosphotyrosine antibody (PY20) immunoprecipitates from 50 µg of protein of cellular lysates of wild-type and IRS-1/ cells at the day indicated after induction into adipocytes were subjected to a PI 3-kinase assay. The autoradiograms of the spots corresponding to PIP are shown in the upper panel. The radioactivity in the spots was measured, and the results are shown in the lower panel, expressed as the ratios of the values for the respective cells to those for wild-type cells at day 0 after induction. The data represent the means ± the standard errors of the means from three independent experiments. **, P < 0.01. (B) Effect of the PI 3-kinase inhibitor LY294002 on adipocyte differentiation of wild-type and IRS-1/ cells. LY294002 (20 µM) was added to adipocyte differentiation medium, and the medium was changed at 36-h intervals. As a control, 0.1% dimethyl sulfoxide was added to the medium, and the medium was changed at the same intervals. At day 8 after induction, the intracellular fat accumulation was assessed by staining with Oil-Red O. (C) Tyrosine-phosphorylated proteins associated with p85 subunit of PI 3-kinase. Total lysates (100 µg of protein each) of cells at day 8 after induction were immunoprecipitated with anti-p85PAN (antibody to the p85 subunit of PI 3-kinase), and the immunoprecipitates were subjected to SDS-PAGE followed by immunoblotting with PY20 conjugated to HRP. (D) Association of IRS-1 with p85 subunit of PI 3-kinase. Total lysates (100 µg of protein each) of cells at day 8 after induction were immunoprecipitated with anti-p85PAN, and the immunoprecipitates were subjected to SDS-PAGE, followed by immunoblotting with anti-IRS-1 antibody.

/

/

/

/

/

/

/

/

View larger version (33K): [in this window] [in a new window]   FIG. 3.   PI 3-kinase and MAPK activities during adipocyte differentiation in wild-type, IRS-1/, IRS-2/, and IRS-1/ IRS-2/ cells. (A) Tyrosine-phosphorylated proteins associated with p85 subunit of PI 3-kinase in cells with four different genotypes, wild-type, IRS-1/, IRS-2/, and IRS-1/ IRS-2/ cells. Total lysates (100 µg of protein each) of cells at day 0 and day 8 after induction were immunoprecipitated with anti-p85PAN (antibody to the p85 subunit of PI 3-kinase), and the immunoprecipitates were subjected to SDS-PAGE, followed by immunoblotting with antiphosphotyrosine antibody (4G10) conjugated to HRP. (B) PI 3-kinase activity associated with tyrosine phosphorylated proteins in wild-type, IRS-1/, IRS-2/, and IRS-1/ IRS-2/ cells. The immunoprecipitates with PY20 from cell lysates (50 µg of protein) were subjected to the PI 3-kinase assay. The autoradiogram of the thin-layer chromatograph is shown. (C) Levels of MAPK protein (upper panel) and phosphorylation of MAPK (lower panel) during adipocyte differentiation. Total lysates (10 µg of protein each) of cells with the four different genotypes were subjected to immunoblotting as described in Materials and Methods using antibody specific to either MAPK or phospho-MAPK.

/

/

/

/

PI 3-kinase and MAPK activities during adipocyte differentiation in wild-type, IRS-1^/, IRS-2^/, and IRS-1^/ IRS-2^/ cells. We performed immunoblotting analysis to identify the tyrosine-phosphorylated proteins associated with the p85 subunit of PI 3-kinase during adipocyte differentiation (Fig. 3A). At day 0 after induction, tyrosine-phosphorylated 170-kDa proteins, which were recognized by anti-IRS-1 antibody (data not shown), were weakly detected in wild-type and IRS-2^/ cells (Fig. 3A; lanes 1 and 3). At 8 days after induction, tyrosine phosphorylation of IRS-1 associated with PI 3-kinase in wild-type and IRS-2^/ cells was markedly increased (Fig. 3A, lanes 5 and 7). In IRS-1^/ cells, tyrosine phosphorylation of the 180-kDa protein associated with the p85 subunit of PI 3-kinase was induced (Fig. 3A, lane 6). This protein was shown to be IRS-2 because it was not detected in IRS-2^/ cells and IRS-1^/ IRS-2^/ cells (Fig. 3A, lanes 7 and 8). In IRS-1^/ IRS-2^/ cells, tyrosine-phosphorylated proteins associated with the p85 subunit of PI 3-kinase were not detected in the 170- to 180-kDa range (Fig. 3A, lane 8).

/

/

/

/

IRS-1^/ IRS-2^/ cells failed to differentiate into adipocytes. The EF cells with each of the four genotypes were induced to differentiation into adipocytes. Oil-Red O staining was performed 8 days after induction to adipocyte differentiation (Fig. 4A). The ability of IRS-1^/ cells to differentiate into adipocytes was again approximately 60% lower than that of wild-type cells. The ability of IRS-2^/ cells to differentiate into adipocytes was approximately 15% lower than that of wild-type cells (Fig. 4B). Surprisingly, IRS-1^/ IRS-2^/ cells were completely unable to differentiate into adipocytes (Fig. 4).

View larger version (43K): [in this window] [in a new window]   FIG. 4.   Adipocyte differentiation of EF cells with the four different genotypes, i.e., wild-type, IRS-1/, IRS-2/, and IRS-1/ IRS-2/ cells. (A) Oil-Red O staining for fat accumulation in cells with the four different genotypes. Wild-type, IRS-1/, IRS-2/, and IRS-1/ IRS-2/ cells were grown to confluence, exposed to IBMX, DEX, and insulin to induce differentiation as described in Materials and Methods and then stained with Oil-Red O at day 8 day after induction. (B) Intracellular triglyceride content during adipocyte differentiation of wild-type, IRS-1/, IRS-2/, and IRS-1/ IRS-2/ cells. The data represent the means ± the standard errors of the means from the analysis of wild type (n = 3), IRS-1/ (n = 4), IRS-2/ (n = 4), and IRS-1/ IRS-2/ (n = 3) cells. *, P < 0.05.

Expression of transcriptional factors for adipocyte differentiation and adipogenic markers. To identify the molecular mechanism for the defective adipocyte differentiation in IRS-1^/ IRS-2^/ cells, we performed Northern blot analysis to investigate gene expression of transcriptional factors for adipocyte differentiation and adipogenic markers (Fig. 5A). The levels of C/EBP expression in IRS-1^/ and IRS-1^/ IRS-2^/ cells were comparable to those of wild-type cells. Somewhat surprisingly, C/EBP expression in IRS-1^/ and IRS-1^/ IRS-2^/ cells was greater than that in wild-type cells. By contrast, the expression of C/EBP in IRS-1^/ IRS-2^/ cells was markedly reduced. The expression of PPAR was also severely decreased in IRS-1^/ IRS-2^/ cells. Although the expression of SREBP1c (ADD1) was significantly reduced 4 and 8 days after induction in IRS-1^/ IRS-2^/ cells, it caught up with that of wild-type cells by 12 days after induction. The expression of LPL and aP2 mRNAs was severely decreased but not abolished in IRS-1^/ IRS-2^/ cells.

View larger version (57K): [in this window] [in a new window]   FIG. 5.   (A) Northern blotting analysis and RNase protection assay of transcriptional factors for adipocyte differentiation and adipogenic markers in wild-type, IRS-1/, IRS-2/, and IRS-1/ IRS-2/ cells during adipocyte differentiation. Total RNAs were extracted from adipocytes at days 0, 2, 4, 8, and 12 after induction, and a 6-µg sample of the total RNA was subjected to electrophoresis and hybridization with probes consisting of 32P-labeled cRNAs encoding C/EBP, C/EBP, C/EBP, and 36B4 (upper panel) and with probes consisting of 32P-labeled cDNAs encoding aP2 and LPL. In an RNase protection assay, 10 µg of the total RNA was hybridized with cRNA encoding PPAR, SREBP1c, and 36B4 (lower panel). 36B4 encodes acidic ribosomal phosphoprotein P0 and was used as a loading control. (B) Effect of the PI 3-kinase inhibitor LY294002 on expression of the mRNAs of PPAR and C/EBP in wild-type cells. LY294002 (20 µM) was added to adipocyte differentiation medium, and the medium was changed at 36-h intervals. As a control, 0.1% dimethyl sulfoxide was added to the medium, and the medium was changed at the same intervals. At day 8 after induction, total RNAs were extracted from wild-type cells, and a 6-µg sample of total RNA was electrophoresed and subjected to Northern blotting with PPAR and C/EBP.

Expression of PPAR and C/EBP was markedly reduced on the protein level in IRS-1^/ IRS-2^/ cells. Immunoblotting analysis was performed to assess the expression of PPAR and C/EBP on the protein level. The expression of C/EBP was markedly reduced on the protein level in IRS-1^/ IRS-2^/ cells at 8 days after induction (Fig. 6A), and the PPAR protein level in IRS-1^/ IRS-2^/ cells was also markedly reduced (Fig. 6B). These results were consistent with those obtained by Northern blotting analysis of the IRS-1^/ IRS-2^/ cells.

View larger version (52K): [in this window] [in a new window]   FIG. 6.   Immunoblot analysis of C/EBP (A) and PPAR (B) in wild-type, IRS-1/, IRS-2/, and IRS-1/ IRS-2/ cells during adipocyte differentiation. Cells were grown to confluence and then exposed to IBMX, DEX, and insulin to induce differentiation. At days 0 and 8 after induction, the cells were lysed, and the lysates (6 µg of protein each) were subjected to SDS-PAGE, followed by immunoblotting with anti-C/EBP antibody (A) and anti-PPAR antibody (B) as described in Materials and Methods. In an immunoblot analysis of PPAR(B), PPAR-deficient cells described in a previous report (21) and the CV-1 cells in which PPAR1 or PPAR2 was overexpressed were also used as control samples.

Retroviral expression of PPAR2 in IRS-1^/ IRS-2^/ cells partially rescued the lack of adipocyte differentiation. If the lack of PPAR expression was responsible for the observed lack of lipid accumulation in IRS-1^/ IRS-2^/ cells, it was thought that forced expression of PPAR might rescue adipocyte differentiation. To determine whether it would, we performed retrovirus-mediated PPAR gene transfer into IRS-1^/ IRS-2^/ cells. The results of Oil-Red O staining showed that forced expression of PPAR2 in IRS-1^/ cells completely rescued the impaired adipocyte differentiation. Forced expression of PPAR2 in IRS-1^/ IRS-2^/ cells, however, rescued it only in part, not to the same level as in mock-vector-transferred wild-type cells (Fig. 7), suggesting that IRS-1 and IRS-2 are required for full adipocyte differentiation in addition to their roles to induce PPAR2.

View larger version (93K): [in this window] [in a new window]   FIG. 7.   Retrovirus-mediated reexpression of PPAR improved the differentiation capacity of IRS-1/ IRS-2/ cells. Wild-type, IRS-1/, or IRS-1/ IRS-2/ cells were transfected with pMX-mPPAR2-puro and mock vector by retrovirus-mediated gene transfer as described in Materials and Methods. At day 8 of differentiation, cells were stained with Oil-Red O to assess the intracellular fat accumulations.

WAT mass of IRS-1^/ IRS-2^/ double-knockout mice was markedly reduced. To address whether the in vitro effects observed in this study are translated into effects on adiposity in vivo, we tried to produce IRS-1^/ IRS-2^/ double-knockout mice. Interestingly, they were carried to term (R. Suzuki et al., unpublished data). Because white adipose tissue (WAT) appears at birth during mouse development, mice at the 8-h postnatal point were examined histologically (Fig. 8). At 8 h after birth, control animals had abundant subcutaneous WAT. In contrast, WAT mass was dramatically reduced, but not abrogated, in the newborn IRS-1^/ IRS-2^/ double-knockout mice. Surprisingly, brown adipose tissue (BAT) mass in IRS-1^/ IRS-2^/ double-knockout mice was almost the same as that in control mice.

View larger version (135K): [in this window] [in a new window]   FIG. 8.   Histological analyses of IRS-1/ IRS-2/ double-knockout mice showed severe reduction in WAT but not in BAT. Transverse sections at the level of the neck were made from ~8-h control (left) and IRS-1/ IRS-2/ (right) mice. The sections were subjected to hematoxylin and eosin staining. Magnifications: ×30 (top) and ×150 (bottom).

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Adipocyte differentiation is an important aspect of energy balance. Although cascades and networks of transcriptional factors during adipocyte differentiation have been investigated rather thoroughly, the signaling mechanism by extracellular hormones and growth factors regulating the cascades and networks of adipose-specific transcriptional factors remains unclear (30). Insulin and IGF-1 have been reported to play important roles in adipocyte differentiation (5, 13, 14). In this study, we investigated the role of IRS-1 and IRS-2, the two major common substrates for insulin and IGF-1 receptor tyrosine kinases, in adipocyte differentiation. Using a gene ablation strategy, we found that the abilities of IRS-1^/ cells and IRS-2^/ cells to differentiate into adipocytes were reduced by 60 and 15%, respectively, compared to that of wild-type cells. These results indicate that even though both IRS-1 and IRS-2 are required for full capacity of adipocyte differentiation, IRS-1 plays a more role in adipocyte differentiation. Moreover, it is possible that IRS-2 and IRS-1 may have rescued the phenotype of IRS-1^/ and IRS-2^/ cells, respectively, in adipocyte differentiation. Therefore, to fully understand the role of IRS-1 and IRS-2 in adipocyte differentiation, we established IRS-1 IRS-2 double-deficient EF cells. Our results clearly show that IRS-1^/ IRS-2^/ cells have no ability to differentiate into adipocytes; providing the first direct evidence that IRS-1 and IRS-2 are absolutely essential to adipocyte differentiation.

Since IRS-1 and IRS-2 are important mediators of insulin and IGF-1 actions, these data may imply that the absence of IRS-1 and IRS-2 impairs hormonal activation of adipogenesis. It should be noted, however, that this study does not address which tyrosine kinase(s) phosphorylate IRS-1 and IRS-2 during adipocyte differentiation. Therefore, it is possible that tyrosine phosphorylation of IRS-1 and IRS-2 by putative upstream tyrosine kinase(s) other than insulin and IGF-1 receptor tyrosine kinases plays a crucial role in in vitro adipocyte differentiation or in vivo adipogenesis. Data from Chaika et al. (6) suggest that the insulin receptor tyrosine kinase was still capable of stimulating adipocyte differentiation even when mutated such that it could no longer phosphorylate IRS proteins.

The present study strongly suggests the important role of tyrosine-phosphorylated IRS-1 and IRS-2 and PI 3-kinase activation by three lines of experimental results: (i) adipocyte differentiation capacity parallels PI 3-kinase activity, which increases during adipocyte differentiation, among the four genotypes; (ii) PI 3-kinase inhibitor completely blocks adipocyte differentiation; and (iii) phosphorylation of MAP kinase is unaltered in any of the four genotypes. Consistent with our results, several studies (31, 40, 47, 54) have reported that PI 3-kinase activity plays a role in adipocyte differentiation by using dominant-negative p85, wortmannin, or LY294002, inhibitors of PI 3-kinase. Moreover, Kohn et al. and Magun et al. have demonstrated the role of activation of PI 3-kinase and one of its downstream mediators, Akt, to induce adipocyte differentiation by constitutive active form of Akt Ser/Thr kinase (20, 24). Taken together with these reports, the present study strongly suggests that PI 3-kinase activity activated through induction of tyrosine phosphorylation of IRS-1 and IRS-2 is essential to adipocyte differentiation.

To further clarify the mechanism by which IRS-1 and IRS-2 fulfill their role in adipocyte differentiation, we investigated the expression of transcriptional factors for adipocyte differentiation and adipogenic markers. The results show that the mRNA expression of C/EBP and PPAR is severely decreased in IRS-1^/ IRS-2^/ cells, although the expression of C/EBP is unaltered and the expression of C/EBP is even slightly increased. It therefore seems possible that IRS-1 and IRS-2 upregulate the expression of C/EBP and PPAR, but the molecular link between IRS-1, IRS-2, and PI 3-kinase and the upregulation of these transcription factors is unclear and requires further investigation. The expression of LPL and aP2, whose promoters have a PPAR response element (PPRE), is parallel to the expression of PPAR. Although the expression of SREBP1c is at least in part dependent on the expression of PPAR, as previously reported (21), SREBP1c has been reported to be also regulated by many other factors (11). At day 12 after induction the expression of SREBP1c in IRS-1^/ IRS-2^/ cells is induced to almost the same amount as that in wild-type cells. It indicates that upstream signaling pathway(s) other than IRS-1 and IRS-2 may be involved in the regulation of the expression of SREBP1c. Nevertheless, normal expression of SREBP1c at day 12 is not sufficient to induce adipocyte differentiation in IRS-1^/ IRS-2^/ cells.

To further study the role of PI 3-kinase in adipocyte differentiation, we examined whether inhibition of PI 3-kinase during adipocyte differentiation affects the mRNA expressions of PPAR and C/EBP in wild-type cells. As shown in Fig. 5B, mRNAs of both PPAR and C/EBP decrease significantly in wild-type cells treated with 20 µM LY294002. It strongly suggests that the inhibition of PI 3-kinase blocks adipogenesis due to severe reduction in expression of both PPAR and C/EBP mRNAs. Our present data appear to differ from those of a previous report (31) that the inhibition of PI 3-kinase in 3T3-F442A cells by expressing dominant-negative p85 does not affect the expression of the transcription factor PPAR at the mRNA level using insulin and 5,8,11,14-eicosatetraynoic acid (ETYA) as inducers for adipocyte differentiation (31). The reasons for the apparent discrepancy between our data and those of Sakaue et al. (31) are unknown at present but could be related to differences in methods such as cell type (primary EF cells versus 3T3-F442A) or inducers for differentiation (IBMX, DEX, and insulin versus ETYA and insulin) or means to inhibit PI 3-kinase (LY294002 versus dominant-negative p85). On the other hand, our present result that PI 3-kinase inhibition using LY294002 decreases PPAR mRNA is consistent with the study by Xia and Serrero (54), in which they demonstrated treatment of 3T3-L1 adipocytes with LY294002 inhibited PPAR expression and adipocyte differentiation induced by IBMX, DEX, and insulin. Identification of signals downstream of IRS-1 and IRS-2 mediating the upregulation of PPAR and C/EBP mRNA expression is an important subject of future research.

Restoration of PPAR to IRS-1^/ IRS-2^/ cells by retrovirus-mediated gene transfer partially rescues the defective adipocyte differentiation. This is consistent with the hypothesis that one of the roles of IRS-1 and IRS-2 in adipocyte differentiation is upregulation of mRNA expression of PPAR. On the other hand, the fact that PPAR is unable to completely rescue the defective adipocyte differentiation suggests that IRS-1 and IRS-2 may stimulate adipocyte differentiation via distinct mechanism(s) in addition to induction of PPAR.

Histological analyses of IRS-1^/ IRS-2^/ double-knockout mice show severe reduction in WAT but not in BAT. These data may provide the evidence that the essential role of IRS-1 and IRS-2 in adipocyte differentiation in vitro can be extended to that in development of WAT in vivo. It should be noted, however, that WAT mass is not completely abrogated in IRS-1^/ IRS-2^/ double-knockout mice. These differences in adipogenesis between in vitro and in vivo have been previously observed in the report of Tanaka et al. on C/EBP and C/EBP knockout mice (39). In in vitro differentiation experiments, the only differentiation stimuli are MIX, DEX, INS and FBS; however, in vivo, there are many other adipogenic factors such as growth hormone, IGF-1, and prostaglandins. With the restricted number of stimuli in vitro, IRS-1 and IRS-2 is absolutely required for the expression of PPAR and C/EBP and for adipocyte differentiation. However, the lack of adipogenesis may be partially rescued with additional stimuli in vivo which are not absolutely dependent upon the presence of IRS-1 and IRS-2.

In conclusion, this study is the first to report that IRS-1 and IRS-2 play an essential role in adipocyte differentiation and adipogenesis. Moreover, we propose that IRS-1 and IRS-2 play crucial roles in adipocyte differentiation through the upregulation of mRNA expression of PPAR and C/EBP subsequent to induction of C/EBP and C/EBP (Fig. 9).

View larger version (16K): [in this window] [in a new window]   FIG. 9.   Schematic model of the roles of IRS-1 and IRS-2 in adipogenesis. IRS-1 and IRS-2 play a crucial role in the upregulation of the expression of C/EBP and PPAR and adipocyte differentiation.