Inhibition of Adipocyte Differentiation by Nur77, Nurr1, and Nor1

Nur77 Is Induced by Inflammatory Signals in Adipocytes

Inflammatory signals such as TNFα and lipopolysaccharide (LPS) are known inhibitory factors in adipogenesis (22,23). In addition, TNFα, LPS, and other proinflammatory stimuli have been shown to be potent inducers of Nur77 in macrophages (24,25). We therefore examined whether these inflammatory triggers can induce Nur77 expression in adipocytes. In 3T3-L1 preadipocytes, various inflammatory stimuli were added directly to the culture media, to minimize serum-induced fluctuation of Nur77. As shown in Fig. 2A​2A,, Nur77 was induced approximately 2-fold by TNFα and LPS in 3T3-L1 preadipocytes. To determine whether Nur77 is similarly up-regulated in vivo, wild-type C57Bl/6 mice were injected with PBS or LPS (5 mg/kg). Epididymal white adipose tissue was isolated and flash frozen 1 h after injection. Total RNA preparation from the adipose tissue revealed that, compared with PBS-injected controls, LPS robustly up-regulated Nur77 expression (Fig. 2B​2B).). This finding suggests that one plausible physiological function of Nur77 is to mediate inflammation-induced inhibition of lipid accumulation and adipogenesis.

Open in a separate window

Figure 2

NR4A Expression Is Induced by Inflammatory Stimuli in Vitro and in Vivo

A, Confluent 3T3-L1 preadipocytes were treated with: PBS, mTNFα (25 ng/ml), interferon (IFN)γ (100 U/ml), mIL1β (20 ng/ml), murine epidermal growth factor (mEGF) (10 ng/ml), LPS (1 μg/ml), LTA (100 ng/ml), polyinosinic acid (polyIC) (2 μg/ml), CpG (100 nm). Total RNA was isolated 1 h later. B, Six age-matched nonfasted wild-type male C57Bl/6 mice were injected with PBS or LPS, 5 mg/kg ip. Epididymal white adipose tissue was isolated and flash frozen exactly 1 h after injection. Total RNA was prepared as described in Materials and Methods. *, P < 0.05; **, P < 0.01 vs. PBS. LTA, Lipoteichoic acid; CpG, CpG oligonucleotide.

Fasting Induces Expression of NR4A Receptors

We have previously demonstrated that fasting potently induces hepatic NR4A receptor expression to enhance gluconeogenesis (12). To determine whether NR4A receptors are similarly up-regulated in white adipose tissue in response to fasting, we measured NR4A expression of epididymal fat pads from fasted and fasted/refed wild-type and ob/ob mice. As shown in Fig. 3​3,, Nur77 and NOR1 expression was markedly reduced in the fed state. Nurr1 expression did not appear to be down-regulated by refeeding. Interestingly, ob/ob mice appeared to lack the feeding-regulated suppression of Nur77 and NOR1.

Open in a separate window

Figure 3

NR4A Expression in White Adipose Tissue

Male wild-type (WT) C57Bl/6 and ob/ob mice (n = 8–10) were fasted for 24 h. A subgroup was refed for 8 h. *, P < 0.05; **, P < 0.01, WT vs. ob/ob. ‡, P < 0.05, ‡‡, P < 0.01, fasting vs. refed.

NR4A Overexpression Inhibits Adipogenesis

NR4A receptors have been shown to mediate various metabolic processes in insulin-sensitive tissues such as liver and skeletal muscle (12,13,14,26,27). To determine the function of NR4A receptors in adipocytes, we first tested their roles in differentiation. Using retroviral transduction, each of the three NR4A isotypes was stably overexpressed in 3T3-L1 preadipocytes. As an additional control, we also tested the effect of Nur77–898, an N-terminally truncated dominant-negative receptor (deletion of the first 897 nucleotides of Nur77). Deletion of the N-terminal transactivation domain has previously been shown to antagonize the activity of all three NR4A isotypes (12,17,28). After puromycin selection of large mixed pool of cells (not single clones), each cell line was differentiated in standard differentiation cocktail for 7 d (DMI for 2 d, followed by insulin alone for 5 d). As shown in Fig. 4​4,, expression of each of the NR4A receptors, but not Nur77–898, strongly inhibited adipogenesis of 3T3-L1 preadipocytes (Fig. 4​4,, top row). Similar results were obtained with multiple independent sets of stable lines.

Open in a separate window

Figure 4

NR4A Receptors Inhibit the Differentiation of 3T3 Preadipocytes

Stable cell lines were generated by retroviral transduction as described in Materials and Methods. 3T3-L1 cells were induced to differentiate with DMI alone (top row) or DMI and GW7845 (third row). 3T3-L1 cells expressing both PPARγ and NR4A receptors were differentiated with DMI (bottom row). Induction media was replaced 2 d later with 10% FBS containing insulin, with or without GW7845. 3T3-F442A cell lines were differentiated with 10% FBS containing insulin (second row). Cells were stained with Oil Red O as described in Materials and Methods.

Our finding that NR4A receptors inhibit adipogenesis was reproduced in 3T3-F442A cells (Fig. 4​4,, second row) as well as in C3H10T1/2 cells (data not shown). Even more striking was the observation that treatment of the NR4A-expressing 3T3-L1 preadipocytes with the PPARγ ligand GW7845 (in addition to the standard DMI cocktail) was unable to rescue adipogenesis (Fig. 4​4,, third row). Likewise, NR4A-mediated inhibition of adipogenesis in 3T3-L1 preadipocytes could not be overcome even when PPARγ was ectopically expressed (Fig. 4​4,, bottom row). These data identify NR4A receptors as regulators of adipogenesis. The observation that PPARγ activation is unable to rescue adipogenesis in NR4A-expressing 3T3-L1 preadipocytes suggests that NR4A receptors regulate a PPARγ-independent pathway in adipogenesis (see below).

Examination of the expression of early adipogenic transcription factors revealed that C/EBPβ and C/EBPδ were expressed at lower levels in 3T3-L1 preadipocytes overexpressing Nur77 (Fig. 5A​5A).). We observed an increase in expression of GATA2, a known inhibitor of adipogenesis (29), 24 h after DMI induction in Nur77-expressing cells. There was also a subtle increase in expression of PPARγ 24 h after DMI treatment. Given the minimal expression level of PPARγ at this time point, the change is likely of little biological significance. Expression of Pref-1, another inhibitor of adipogenesis (30,31), was unaffected (Fig. 5B​5B).). The changes in expression of these early adipocyte transcription factors suggest that Nur77 is acting within 24 h of induction of differentiation.

Open in a separate window

Figure 5

Expression of Adipogenic (A) and Antiadipogenic (B) Transcription Factors in 3T3-L1 Cells Overexpressing Nur77

Total RNA was harvested at confluence (time zero) or 24 h after DMI/GW7845 treatment. Each condition represents four to six separate wells. *, P < 0.05; **, P < 0.01 vs. empty vector control. §, P <0.01 vs. empty vector control.

Acute Expression of Nur77 Does Not Cause Dedifferentiation of Adipocytes

Having demonstrated that chronic overexpression of NR4A receptors inhibits the differentiation of preadipocytes, we proceeded to determine whether ectopic Nur77 expression in mature adipocytes would lead to dedifferentiation. 3T3-L1 cells that express the human (h) coxsackie-adenovirus receptor (hCAR; courtesy of B. Spiegelman) were differentiated into mature adipocytes in the presence of standard adipogenic cocktail and GW7845 for 6 d. After differentiation, cells were maintained in 10% fetal bovine serum (FBS)/DMEM and 0.5 μg/ml insulin. Adipocytes were then infected overnight with adenovirus expressing either the green fluorescent protein (GFP) control or Nur77. Cells were harvested on d 5, 10, or 15 after infection for RNA isolation. As shown in Fig. 6​6,, the expression of adipocyte markers such as PPARγ and aP2 was unaffected by acute ectopic expression of Nur77. This finding indicates that although Nur77 inhibits preadipocyte differentiation, it has no apparent effect on maintaining the phenotypic markers of mature adipocytes.

Open in a separate window

Figure 6

Acute Nur77 Expression Does Not Cause Dedifferentiation of 3T3-L1 Adipocytes

Differentiated 3T3-L1 hCAR adipocytes were infected with adenovirus encoding GFP or Nur77 overnight. Total RNA was harvested on d 5, 10, or 15 after infection.

Identification of Nur77-Target Genes that Inhibit Adipogenesis

To identify putative Nur77 targets that may mediate its inhibitory effect on adipogenesis, we performed expression profiling of control (retrovirally transduced with empty vector) or Nur77-expressing 3T3-L1 preadipocytes. Total RNA was harvested at confluence or 24 h after DMI/GW7845 treatment. Genes induced greater than 5-fold by Nur77 are shown in supplemental Table 1 published as supplemental data on The Endocrine Society’s Journals Online web site at http://mend.endojournals.org. Several genes that are induced greater than 2-fold and are involved in adipogenesis are also included in this list. Induction of selected genes relevant to adipogenesis was verified by quantitative real-time PCR (supplemental Table 1). Among the genes shown in supplemental Table 1, gap junction protein, α1 (Gja1), tolloid-like 1 (Tll1), and WNT1-inducible signaling pathway protein 2 (Wisp2) were three particularly interesting genes that were induced by Nur77 in stable 3T3-L1 transformants both at baseline as well as at 24 h after differentiation induction. Gja1 is a member of the connexin family and is the major gap junction protein in the heart involved in synchronous contraction of the myocardium (32,33). It has also been shown to influence mitotic clonal expansion or 3T3-L1 preadipocytes (34). Tll1, an astacin-like metalloprotease with structural similarity to bone morphogenetic protein-1, is crucial for normal cardiac septation in mice (35). Although it is among the most highly induced gene in our microarray, Tll1 was not previously described to have a function in adipose tissue. Wisp2, a known downstream target of Wnt signaling (36), was also a plausible candidate to effect Nur77-driven inhibition of adipogenesis. Wisp2 expression was previously shown to be down-regulated during adipogenic differentiation of human bone marrow-derived mesenchymal stem cells (37).

As shown in Fig. 7A​7A,, we verified with real-time PCR that Gja1, Tll1, and Wisp2 were up-regulated by Nur77 in stable 3T3-L1 transformants. To determine whether the induction in gene expression was responsive to acute as well as chronic overexpression of Nur77, the expression of Gja1, Tll1, and Wisp2 was further analyzed in 3T3-L1 hCAR adipocytes acutely infected with GFP- or Nur77-expressing adenovirus. The data of Fig. 7B​7B confirmed that these genes were robustly up-regulated by Nur77, suggesting that they are likely direct targets of the receptor.

Open in a separate window

Figure 7

Nur77 Induces the Expression of Gja1, Tll1, and Wisp2

A, Total RNA was harvested from vector-control and Nur77-expressing 3T3-L1 cells at confluence (time zero) or 24 h post-DMI/GW7845 treatment. B, 3T3-L1 hCAR adipocytes were infected with Ad-GFP or Ad-Nur77. Total RNA was harvested 3 d after infection. **, P < 0.01 vs. GFP control.

We next determined whether ectopic expression of Gja1, Tll1, and Wisp2 was sufficient to inhibit adipogenesis in 3T3-L1 preadipocytes. The cDNA encoding each of these genes was PCR amplified and inserted into vectors for retroviral transduction of 3T3-L1 cells. After puromycin selection, cells were differentiated using standard adipogenic cocktail for 7 d. Oil red O staining revealed that ectopic expression of Gja1 and Tll1 inhibited adipogenesis, whereas Wisp2 overexpression had no effect (Fig. 8A​8A).). The paucity of lipid droplet accumulation in Gja1- and Tll1-expressing cells was accompanied by diminished expression of adipogenic markers such as aP2, adipsin, lipoprotein lipase, and PPARγ (Fig. 8B​8B).). Similar results were obtained with multiple independent sets of stable lines. Although neither Gja1 nor Tll1 diminished differentiation of 3T3-L1 cells to the same extent as Nur77, their ability to independently inhibit adipogenesis strongly suggests that they contribute to Nur77’s inhibitory activity on adipogenesis.

Open in a separate window

Figure 8

Expression of Gja1 or Tll1 Inhibits Adipogenesis

A, 3T3-L1 preadipocytes were retrovirally transduced with MSCV-vector expressing Nur77, Gja1, Tll1, or Wisp2. Stable transformants were induced to differentiate by DMI. Cells were stained with Oil Red O 7 d after the addition of adipogenic cocktail. The two rows of cells represent two separate experiments. B, Expression of adipocyte markers from 3T3-L1 stable transformants. Total RNA was harvested on d 7 after differentiation. *, P < 0.05; **, P < 0.01 vs. empty vector control. LPL, Lipoprotein lipase.

Cell Culture

We purchased human embryonic kidney (HEK)293T and 3T3-L1 cells from American Type Culture Collection (Manassas, VA). 3T3-F442A cells were a gift from E. Saez. HEK293T cells were maintained in 10% FBS (Omega Scientific, Stamford, CT) in DMEM at 8% CO₂. 3T3-L1 and 3T3-F442A cells were maintained and differentiated as previously described (53). 3T3-L1 hCAR cells (gift from B. Spiegelman) were differentiated as described for 3T3-L1 cells, except that 100 nm GW7845 was added to the culture media during differentiation. For adenovirus infection, cells were cultured with adenovirus at a multiplicity of infection of 250 in 10% FBS/0.5 μg/ml insulin overnight before media refreshment the next day. Oil Red O staining was performed as described (7).

Stable Cell Lines

Full-length cDNA of murine Nur77, Nurr1, NOR1, Nur77–898 (cDNA starts at nucleotide +898), Gja1, Tll1, Wisp2 genes was PCR amplified and cloned into MSCV-pac plasmid. Expression plasmids along with ecotropic helper plasmid were transiently transfected into HEK293T cells using Lipofectamine 2000 reagent (Invitrogen, Carlsbad, CA). Supernatant was sterile filtered 2 d after transfection and added to either 3T3-L1 or 3T3-F442A cells. Puromycin (4 μg/ml) in 10% calf serum was used for selection of stable clones. 3T3-L1 cells overexpressing PPARγ were selected with hygromycin (50 μg/ml).

Quantitative Real-Time PCR

Total RNA was isolated with Trizol (Invitrogen). cDNA was synthesized using reverse transcriptase from Applied Biosystems (Foster City, CA). Real-time PCR was performed as described previously (13). See supplemental Table 2 for a complete list of primer sequences used in real-time PCR.

Microarrays

RNA from control (empty vector) or Nur77-expressing 3T3-L1 cells was harvested at confluence (time zero) or 24 h after DMI/GW7845 (100 nm) treatment. Total RNA prepared by Trizol was further purified through RNEasy columns (QIAGEN, Chatsworth, CA). Each condition was done with replicate arrays, each representing three pooled wells. cRNA preparation and hybridization to Affymetrix Mouse Genome Arrays 430 version 2.0 was performed by the UCLA Microarray Core, and data were analyzed using GeneSpring GX7.3.1. We included only genes with raw data signal greater than 500 for at least one condition for analysis.

Cell Cycle Analysis

Cells (50,000 per well) were plated on gelatin-coated six-well plates. Cells were trypsinized before cell counting and staining. Using the Z1 Coulter Counter (Beckman Coulter, Inc., San Jose, CA), we averaged the mean of cell counts from two independent aliquots per well. Staining of DNA content was performed as previously described (54). Briefly, we resuspended PBS-washed cells in PBS containing 0.1% sodium citrate (wt/vol) (Sigma Chemical Co., St. Louis, MO), 0.3% Triton X-100 (vol/vol) (Sigma), 0.01% propidium iodide (wt/vol) (Calbiochem, La Jolla, CA), and 0.002% ribonuclease A (wt/vol) (Sigma), and incubated the cells for 30 min at 20–25 C. We acquired cell cycle data on a FACS Calibur (BD Biosciences, Palo Alto, CA). Data were analyzed using FloJo Software (Tree Star, Inc., Ashland, OR).

Animals

We purchased age-matched C57Bl/6 wild-type and ob/ob male mice from The Jackson Laboratory (Bar Harbor, ME). Mice were provided with water and standard mouse chow ad libitum and were maintained on 12-h light cycle. PBS or lipopolysaccharide (5 mg/kg) was injected into the peritoneum. Mice were euthanized by isoflurane exactly 1 h after injection. Epididymal adipose tissue was isolated and flash frozen in liquid nitrogen. Total RNA was prepared as described previously (13).

Lipolysis Assay

3T3-L1 hCAR cells were differentiated as described above for 5–7 d. Adipocytes were then infected with adenovirus expressing either GFP or Nur77–898 in 10% FBS/DMEM at a multiplicity of infection of 250, before media refreshment the next day (10% FBS/DMEM). Lipolysis assay was performed as described elsewhere (55).

Immunoblot

3T3-L1 preadipocytes were grown to confluence and treated with 1 mm 8-Br-cAMP. Nuclear extract was prepared as follows: cells were scraped and pelleted in PBS, and washed in cold buffer A (10 mm HEPES, pH 7.9; 10 mm KCl; 0.1 mm EGTA; 0.1 mm EDTA; 1 mm dithiothreitol; 0.5 mm phenylmethylsulfonylfluoride; protease inhibitor) and resuspended in buffer A. After 15 min on ice, Nonidet P-40 was added to the cells to a final concentration of 5.9%. After vigorous vortexing, nuclei were pelleted and resuspended in buffer C (20 mm HEPES, pH 7.9; 420 mm NaCl; 1.5 mm MgCl₂; 0.2 mm EDTA; 25% glycerol; 0.5 mm phenylmethylsulfonylfluoride; 1 mm dithiothreitol; protease inhibitor). Samples were then agitated intermittently, kept on ice for 20 min, and spun down. Nuclear extract (supernatant) was loaded on 4–12% Invitrogen Nupage gel. Nur77 antibody (Cell Signaling Technology, Beverly, MA) was used at 1:1000 dilution and HMG-1 (BD Pharmingen, Franklin Lakes, NJ) at 1:5000.

Statistics

We performed statistical analysis using the Student’s t test, with significance determined a priori at P < 0.05.

We thank K. W. Park and S. Hummasti for discussions. We also thank B. Spiegelman for providing 3T3-L1 hCAR cells and E. Saez for providing 3T3-F442A cells.