Chibby Promotes Adipocyte Differentiation through Inhibition of {beta}-Catenin Signaling

The canonical Wnt/ß-catenin signaling pathway playsdiverse roles in embryonic development and disease. Activationof this pathway, likely by Wnt-10b, has been shown to inhibitadipogenesis in cultured 3T3-L1 preadipocytes and in mice. Here,we report that the ß-catenin antagonist Chibby (Cby)is required for adipocyte differentiation. Cby is expressedin adipose tissue in mice, and Cby protein levels increase duringadipogenic differentiation of 3T3-L1 cells. Ectopic expressionof Cby induces spontaneous differentiation of these cells intomature adipocytes to an extent similar to that of dominant-negativeTcf-4. In contrast, depletion of Cby by RNA interference potentlyblocks adipogenesis of 3T3-L1 and mouse embryonic stem cells.In support of this, embryonic fibroblasts obtained from Cby-deficientembryos display attenuated differentiation to the adipogeniclineage. Mechanistically, Cby promotes adipocyte differentiation,in part by inhibiting ß-catenin, since gain or lossof function of Cby influences ß-catenin signalingin 3T3-L1 cells. Our results therefore establish Cby as a novelproadipogenic factor required for adipocyte differentiation.

INTRODUCTION

Top

INTRODUCTION

Adipose tissue plays critical roles in the regulation of energyhomeostasis by storing and releasing fuel as a reservoir andby secreting a number of hormones and cytokines as an endocrineorgan (37). Excess body fat, or obesity, is a major public healthproblem, particularly in industrialized countries, increasingthe risk of diabetes, cardiovascular diseases and several typesof cancers (28, 37). Conversely, lipoatrophy, the lack of adiposetissue, is also associated with diabetes and a number of othermetabolic abnormalities (32). Hence, understanding the signalingpathways that govern adipocyte differentiation is necessaryto develop comprehensive therapeutic strategies for the preventionand treatment of these disorders.

Adipogenesis involves the formation of preadipocytes from mesenchymalprogenitor cells and their differentiation into adipocytes (29,33, 34). The cellular and molecular mechanisms of adipocytedifferentiation have been extensively studied using preadipocyteculture systems, such as 3T3-L1 and 3T3-F442A cell lines (10,31, 33, 34). In response to hormonal stimuli of adipogenesis,two transcription factors, CCAAT/enhancer-binding protein ß(C/EBPß) and C/EBP ${delta}$ , are rapidly and transiently induced.These proteins then stimulate expression of the key adipogenictranscription factors, C/EBP ${alpha}$ and peroxisome proliferator-activatedreceptor ${gamma}$ (PPAR ${gamma}$ ), which act synergistically to induce expressionof various adipocyte-specific genes.

Intracellular signaling by the Wnt family of secreted glycoproteinsregulates cell proliferation, differentiation, and polaritythroughout vertebrate embryonic development (27, 44, 46). ß-Cateninplays a pivotal role as a transcriptional coactivator in thecanonical Wnt pathway (39). In the absence of Wnt signaling,cytoplasmic ß-catenin becomes phosphorylated by glycogensynthase kinase-3ß (GSK-3ß) in a complexcontaining Axin and the tumor suppressor Adenomatous polyposiscoli and is targeted for ubiquitin-mediated proteasomal degradation(23, 30). Wnt binding to the seven transmembrane Frizzled receptorsand the low-density lipoprotein receptor-related protein (LRP)coreceptors, LRP5 and LRP6, leads to inhibition of GSK-3ßactivity, resulting in stabilization of cytoplasmic ß-catenin.Subsequently, stabilized ß-catenin translocates intothe nucleus and binds to the T-cell factor/lymphoid enhancerfactor (Tcf/Lef) family of transcription factors, leading toactivation of target genes (1, 2, 6, 26, 43).

The Wnt/ß-catenin pathway has been shown to inhibitadipogenesis (3, 5, 7, 36). Wnt signaling maintains preadipocytesin an undifferentiated state through inhibition of C/EBP ${alpha}$ andPPAR ${gamma}$ . Wnt-10b is a good candidate for the endogenous signalbecause it is expressed in dividing and confluent preadipocytes,and its expression decreases during differentiation into matureadipocytes (5, 36). Forced expression of Wnt-10b or Wnt-1 in3T3-L1 cells stabilizes free cytosolic ß-catenin andblocks adipogenesis. In contrast, overexpression of negativeregulators of the pathway, such as Axin or dominant-negativeTcf-4 (dnTcf-4), which binds to Tcf/Lef consensus binding sitesbut lacks the N-terminal ß-catenin-binding domain,results in spontaneous adipogenesis. Moreover, transgenic miceoverexpressing Wnt-10b in adipose tissues show a 50% reductionin total body fat and resist diet-induced obesity (22). Inductionof PPAR ${gamma}$ appears to further suppress canonical Wnt signalingby stimulating degradation of ß-catenin by the proteasome(20, 21, 25).

We previously reported a ß-catenin-associated antagonist,Chibby (Cby) (11, 38). Cby is a 14.5-kDa nuclear protein evolutionarilyconserved from fly to human. Cby physically interacts with theC-terminal activation domain of ß-catenin and repressesß-catenin-mediated transcriptional activation. Mechanistically,Cby competes with Tcf/Lef transcription factors to bind to ß-catenin.Depletion of Drosophila Cby by RNA interference (RNAi) resultsin segment polarity defects that mimic the gain-of-functionphenotype of a Drosophila Wnt homolog, wingless (wg), highlightingthe biological importance of Cby's function (11, 38, 42).

In this report, we demonstrate that Cby is required for adipocytedifferentiation. Cby protein levels increase during differentiationof 3T3-L1 preadipocyte cells. Intriguingly, ectopic expressionof Cby in 3T3-L1 cells causes differentiation into mature adipocyteseven without hormonal stimulation. In contrast, depletion ofCby by RNAi profoundly suppresses adipogenesis of 3T3-L1 andmouse embryonic stem (ES) cells. Mouse embryonic fibroblasts(MEFs) lacking Cby display a reduced ability to differentiateinto the adipogenic lineage. Additionally, in 3T3-L1 cells,loss of Cby results in upregulation of ß-catenin transcriptionalactivity, implying that Cby promotes adipogenesis, at leastin part, by inhibiting ß-catenin signaling. In supportof this mechanism, overexpression of stabilized ß-cateninpartially suppresses Cby-induced adipogenesis, and furthermore,forced expression of dnTcf-4 rescues adipocyte differentiationof Cby-RNAi cells. Our findings suggest that Cby is a novelproadipogenic factor required for adipocyte differentiation.

MATERIALS AND METHODS

Top

MATERIALS AND METHODS

Cell culture and adipocyte differentiation. 293T cells were purchased from ATCC and maintained in Dulbecco'smodified Eagle's medium (DMEM) (Invitrogen) with 10% fetal bovineserum (FBS) and 100 units/ml penicillin-streptomycin.

Mouse 3T3-L1 preadipocytes (a kind gift from Jeffrey Pessin)were maintained in growth media containing DMEM supplementedwith 10% calf serum (HyClone), 1 mM L-glutamine, and 100 units/mlpenicillin-streptomycin at 37°C in 8% CO₂. 3T3-L1 cellswere induced to differentiate into mature adipocytes as describedpreviously (36). Briefly, 2 days after reaching confluence (dayzero), cells were placed in differentiation media consistingof DMEM, 10% FBS, and the differentiation cocktail of 0.5 mMmethylisobutylxanthine, 0.25 µM dexamethasone, and 1 µg/mlof insulin (MDI). Thereafter, the medium was replenished every3 days.

Mouse ES cells were maintained in an undifferentiated stateon gelatin-coated culture plates in Knock-out DMEM (Invitrogen)containing 10% Knockout-Serum Replacement (Invitrogen), 1% FBS(Atlanta Biologicals), 2 mM L-glutamine, 100 units/ml penicillin-streptomycin,25 mM HEPES (Invitrogen), 300 µM monothioglycerol (Sigma),and 1,000 units/ml recombinant mouse leukemia inhibitory factor(Chemicon). The Cby knockdown ES and control cell lines wereestablished by transfecting pSuppressor Retro-Cby short hairpinRNA (shRNA) or empty vector into R1 ES cells and selecting for14 days with G418. Their differentiation into adipocytes wasinduced by the hanging-drop method essentially as describedpreviously (9). Embryoid bodies were treated with retinoic acidprior to switching them to differentiation media.

To study adipogenesis of MEFs, wild-type and Cby^–/–MEFs were prepared from embryonic day 12.5 (E12.5) embryos ofCby^+/– crosses as described previously (24, 41). The generationand characterization of Cby knockout mice will be describedelsewhere. MEFs were maintained in DMEM with 10% FBS and 100units/ml penicillin-streptomycin and induced to undergo adipogenicdifferentiation 1 day postconfluence with MDI as described abovefor 3T3-L1 cells.

Retroviral constructs and infection. For ectopic expression, human Cby and dnTcf-4 cDNAs were amplifiedby PCR and subcloned into the retroviral vector pLXIN (neomycinresistant; Clontech). A stabilized form of human ß-catenin(S33Y) and dnTcf-4 were subcloned into the retroviral vectorpQCXIP (puromycin resistant; Clontech). A detailed descriptionof the constructs is available upon request. The retroviralpackaging cell line RetroPack PT67 (Clontech) was transfectedwith retroviral expression vectors using Lipofectamine 2000(Invitrogen). After 48 h of transfection, the medium containingretroviruses was collected, filtered, treated with Polybrene,and used to infect 3T3-L1 fibroblasts. The infected cells wereselected with either 500 µg/ml G418 (Invitrogen) aloneor in combination with 2.5 µg/ml puromycin (Invitrogen)and maintained in growth media.

RNAi. For knockdown of Cby expression in 3T3-L1 and mouse ES cells,two complementary oligonucleotides for mouse Cby shRNA or CbyshRNA- ${Delta}$ C harboring a single-base deletion in the targeted sequencewere annealed and subcloned into the SalI and XbaI sites ofthe pSuppressor Retro vector (Imgenex) containing the U6 promoterand a neomycin resistance cassette. The target site correspondsto nucleotides 218 to 238 of the mouse Cby gene, with the Aof the ATG start codon assigned as position 1. The sequencesof the oligonucleotides were as follows: Cby shRNA-top, 5'-TCGAGTGGCAGACTCCGTGATTAGTTTCAAGAGAACTAATCACGGAGTCTGCCACTTTTT-3';Cby shRNAi bottom, 5'-CTAGAAAAAGTGGCAGACTCCGTGATTAGTTCTCTTGAAACTAATCACGGAGTCTGCCAC-3'(the underlined G is the deleted single nucleotide within thereverse primer). The C deleted in Cby shRNA- ${Delta}$ C is underlined.All plasmids were verified by sequencing them. In order to examinethe efficiency of RNAi, these constructs were transiently transfectedinto 293T cells in combination with a mouse Cby expression construct(38), and 24 h later, the cells were lysed and subjected toWestern blotting for Cby as described below.

Western blot analyses. Monolayer cells or mouse epididymal fat pads were rinsed withphosphate-buffered saline (PBS), harvested in ice-cold lysisbuffer (20 mM Tris-HCl [pH 8.0], 135 mM NaCl, 1.5 mM MgCl₂,1 mM EGTA, 1% Triton X-100, and 10% glycerol), and incubatedfor 20 min on ice with intermittent vortexing. After centrifugationfor 30 min at 14,000 x g at 4°C, the supernatants were collected,and the protein concentrations of the extracts were determinedby the Bradford assay (Bio-Rad). Equal amounts of protein wereresolved by 4 to 20% gradient sodium dodecyl sulfate-polyacrylamidegel electrophoresis (Invitrogen), transferred onto a nitrocellulosemembrane (Bio-Rad), and immunoblotted with the following antibodies.Rabbit polyclonal anti-Cby antibodies were previously described(38). Anti-C/EBP ${alpha}$ and anti-PPAR ${gamma}$ antibodies were purchased fromSanta Cruz. Anti-ß-actin and anti-aP2 antibodies werefrom Abcam and ProSci, respectively. Anti-ß-cateninwas purchased from BD Transduction Laboratories, and anti-greenfluorescent protein antibody was from Invitrogen. dnTcf-4 wastagged with the Flag epitope at the N terminus and detectedwith an anti-Flag M2 antibody (Sigma). Horseradish peroxidase-conjugatedsecondary antibodies were purchased from Jackson ImmunoresearchLaboratories. Membranes were visualized using the SuperSignalWest Pico Chemiluminescent Substrate kit (Pierce).

Tcf/Lef reporter assays. 3T3-L1 preadipocytes were transiently transfected using LipofectaminePlus (Invitrogen). Luciferase assays were performed using theDual-Luciferase Reporter Assay System (Promega) as describedpreviously (38, 40). Lucifearse activities were measured 48h after transfection and normalized for transfection efficiencyusing the internal Renilla luciferase control.

Oil Red-O staining. Detection of neutral lipids with Oil Red-O was performed asdescribed previously (36). Briefly, cells were washed twicewith PBS, fixed with 4% paraformaldehyde for 30 min, and thenstained for 2 h at room temperature with Oil Red-O solution(0.5% Oil Red-O from Sigma in isopropanol). The cells were washedtwice with PBS and visualized with an inverted microscope.

RNA extraction and RT-PCR. Total RNA was purified from epidydimal white adipose tissue(WAT), interscapular brown adipose tissue (BAT), and liver of2-month-old C57BL/6J mice with RNeasy spin columns (QIAGEN),and synthesis of cDNA was performed with oligo(dT) primers usingthe ThermoScript reverse transcription (RT)-PCR System (Invitrogen)according to the manufacturer's instructions. The primer sequenceswere as follows: Cby (forward, 5'-CGTTTCCTCACTGAGTTAGG-3', andreverse 5'-TAGTCTGCTAATCTGACGGG-3'); GAPDH (glyceraldehyde-3-phosphatedehydrogenase) (forward, 5'-ACCACAGTCCATGCCATCAC-3', and reverse,5'-TCCACCACCCTGTTGCTGTA-3').

Total RNA from ES cells was extracted using the RNA aqueouskit (Ambion), and cDNA was synthesized as described above. Amplificationswere performed using primers for C/EBP ${alpha}$ (forward, 5'-GGACAAGAACAGCAACGAGTACC-3',and reverse, 5'-GGCGGTCATTGTCACTGGTC-3'), PPAR ${gamma}$ (forward, 5'-CTCCTGTTGACCCAGAGCAT-3',and reverse, 5'-ACCCTTGCATCCTTCACAAG-3'), aP2 (forward, 5'-GATGCCTTTGTGGGAACCTGG-3',and reverse, 5'-TTCATCGAATTCCACGCCCAG-3'), and ß-actin(forward, 5'-ATGGATGACGATATCGCTG-3', and reverse, 5'-ATGAGGTAGTCTGTCAGGT-3').

RESULTS

Top

RESULTS

Expression of Cby in differentiating 3T3-L1 cells and murine adipose tissue. Mouse 3T3-L1 preadipocytes are a well-characterized in vitromodel of adipocyte differentiation that can reliably differentiateinto mature fat cells upon exposure to MDI in the presence ofFBS (12, 13). In order to investigate a potential role of Cbyin adipogenesis, we first analyzed Cby expression levels duringdifferentiation of 3T3-L1 preadipocytes into mature adipocytes.Interestingly, Western blot analysis demonstrated that Cby proteinlevels increased in a fashion similar to that of the adipocytedifferentiation marker aP2 (adipocyte fatty acid-binding protein)and peaked at day 8 following hormonal stimulation (Fig. 1A).It is worthy of note that Cby protein levels inversely correlatewith those of cytosolic ß-catenin, and this is consistentwith the previous notion that Wnt signaling needs to be suppressedfor adipogeneis to occur (5, 36). In adult mice, relativelyhigh levels of Cby mRNA expression were seen in both WAT andBAT (Fig. 1B). In agreement with this finding, Cby protein wasdetected in epidydimal fat tissue from wild-type, but not Cby-null,animals (Fig. 1C).

View larger version (22K):
[in this window]
[in a new window]

FIG. 1. Cby expression in cultured 3T3-L1 cells and murine adipose tissue. (A) Cby protein levels increase during differentiation of 3T3-L1 preadipocytes. Whole-cell lysates were extracted at the indicated days of differentiation and subjected to immunoblot analyses with antibodies against Cby, ß-catenin, and the adipocyte differentiation marker aP2. ß-Actin was used as a loading control. (B) Cby is expressed in WAT and BAT. Total RNA was prepared from epidydimal WAT, interscapular BAT, and liver of 2-month-old C57BL/6J mice and used to synthesize cDNAs for RT-PCR for Cby. By Northern blot analysis of adult mouse tissues, liver was found to be one of the tissues in which Cby is highly expressed (data not shown), and it served as a positive control. GAPDH was used as a loading control. (C) Cby protein expression in WAT. Western blot analysis was performed on whole-cell lysates prepared from wild-type (WT) or Cby-deficient (Cby^–/–) epidydimal WAT using anti-Cby antibodies. The stars denote nonspecific bands detected in both wild-type and Cby^–/– samples.

Ectopic expression of Cby induces spontaneous differentiation of 3T3-L1 preadipocyte cells. The above-mentioned expression results suggested that Cby mightbe involved in adipocyte differentiation. It has been demonstratedthat forced expression of negative regulators of the Wnt/ß-cateninpathway, such as Axin or dnTcf-4 in 3T3-L1 cells, leads to spontaneousadipocyte differentiation in the absence of hormonal stimulation(36). Since Cby represses ß-catenin-mediated transcriptionalactivation, overexpression of Cby in preadipocytes might similarlyresult in spontaneous adipogenesis.

To test this hypothesis, 3T3-L1 cells were infected with a retroviruscarrying human Cby or dnTcf-4 as a positive control, or a controlretrovirus, and maintained without MDI treatment. As shown inFig. 2A, the efficient expression of Cby in preadipocytes wasconfirmed by immunoblotting using anti-Cby antibodies. Two weekspostconfluence, control cells failed to differentiate and maintainedfibroblast morphology (Fig. 2B). In accordance with the previousfinding (36), ectopic expression of dnTcf-4 induced spontaneousdifferentiation of these cells as assessed by Oil Red-O staining.Cby-expressing cells also underwent spontaneous adipogenesisto an extent similar to that of dnTcf-4 (Fig. 2B).

View larger version (41K):
[in this window]
[in a new window]

FIG. 2. Forced expression of Cby induces spontaneous differentiation of 3T3-L1 cells without hormonal stimulation. Preadipocytes were infected with a retrovirus vector alone (pLXIN) or retroviruses carrying dnTcf-4 or Cby, selected with G418, and maintained in growth media without the standard differentiation cocktail, MDI. (A) Total cell lysates were prepared from confluent cells for Western blot analyses using antibodies for Cby and ß-actin. (B) Cells were stained with Oil Red-O to visualize lipid droplets 14 days postconfluence. (C) Cby-induced adipogenesis is accompanied by expression of the adipogenic markers C/EBP ${alpha}$ and aP2. Total cell lysates were prepared 14 days postconfluence for Western blot analyses using the indicated antibodies.

To gain insights into the molecular mechanisms underlying theseeffects, we performed Western blot analyses for adipogenic markers(Fig. 2C). As expected, overexpression of Cby induced expressionof C/EBP ${alpha}$ and aP2. These data argue that forced expression ofCby in 3T3-L1 cells induces spontaneous differentiation, mostlikely by inhibiting endogenous ß-catenin.

Cby antagonizes ß-catenin signaling in 3T3-L1 cells. If our hypothesis were correct, providing excess ß-cateninmight overcome the effect of Cby overexpression, leading toa blockade of adipogenesis. In an attempt to examine this, controlor Cby-overexpressing 3T3-L1 cells were reinfected with a controlretrovirus or a retrovirus expressing ß-catenin stabilizedby a missense mutation of tyrosine for serine at codon 33 inthe N-terminal GSK-3ß phosphorylation site (ß-cateninS33Y) (18). Cells were then induced to differentiate in thepresence of MDI. Overexpression of ß-catenin and Cbywas verified by Western blot analyses (Fig. 3A). Consistentwith a prior report (36), ectopic expression of ß-cateninS33Y potently inhibited differentiation (Fig. 3B). Interestingly,Cby overexpression further facilitated adipocyte differentiationof these cells compared to the control retrovirus-infected cells.Under these conditions, ß-catenin S33Y partially suppressedCby-induced adipocyte differentiation, whereas it exhibitedno effect on that of dnTcf-4-expressing cells. The extent ofdifferentiation was further confirmed by Western blot analysisfor aP2 at day 8 (Fig. 3A). These data suggest that Cby promotesadipocyte differentiation, at least in part by inhibiting ß-cateninsignaling.

View larger version (47K):
[in this window]
[in a new window]

FIG. 3. Overexpression of ß-catenin S33Y partially suppresses adipocyte differentiation of Cby-expressing 3T3-L1 cells. Control (pLXIN), Cby- or dnTcf-4-expressing cells were reinfected with a retrovirus vector alone (pQCXIP) or retroviruses carrying stabilized ß-catenin S33Y and were coselected with G418 and puromycin. Two days postconfluence (day zero), the cells were induced to differentiate into mature adipocytes with MDI. (A) Whole-cell lysates were prepared at day 0 or 8 for Western blot analyses for ß-catenin, Cby, aP2, and ß-actin. (B) Cells were also stained with Oil Red-O to visualize lipid droplets 8 days postinduction.

To directly assess whether Cby negatively regulates ß-cateninsignaling in 3T3-L1 preadipocytes, we performed Tcf/Lef luciferasereporter assays (19). Expression of stabilized ß-cateninS33Y stimulated the activity of TOPFLASH containing three optimalTcf/Lef binding sites (Fig. 4B). Coexpression of Cby repressedß-catenin-mediated transcriptional activation. Inorder to test if silencing of endogenous Cby influences TOPFLASHactivity in 3T3-L1 cells, we constructed an RNAi vector expressingshRNAs for mouse Cby. We first examined the efficiency of CbyRNAi by transient transfection into 293T cells. These cellswere transfected with Cby shRNA or Cby shRNA- ${Delta}$ C harboring a single-basedeletion near the center of the Cby shRNA, together with a smallamount of mouse Cby expression plasmid, followed by immunoblottingfor Cby.

View larger version (20K):
[in this window]
[in a new window]

FIG. 4. Cby antagonizes ß-catenin signaling in 3T3-L1 cells. (A) The efficiency of Cby RNAi was tested in 293T cells by transient transfection. 293T cells were transfected with 0.1 µg of a mouse Cby plasmid (mCby) (38) and increasing amounts (0.25, 0.5, or 1 µg) of an expression vector for Cby shRNA or Cby shRNA- ${Delta}$ C, harboring a single-base deletion near the center of the targeted sequence, as indicated. Whole-cell lysates were analyzed by immunoblotting for Cby. A green fluorescent protein expression plasmid (0.1 µg) was cotransfectd as a control. (B) Tcf/Lef luciferase reporter assays in 3T3-L1 cells. 3T3-L1 preadipocytes were transfectd with 0.25 µg of TOPFLASH or the mutant FOPFLASH luciferase reporter and expression vectors for stabilized ß-catenin S33Y (0.125 µg), human Cby (hCby) (2.5 µg) (38), Cby shRNA (0.5 µg), and Cby shRNA- ${Delta}$ C (0.5 µg) in combinations as indicated, and relative luciferase activities were measured.

As shown in Fig. 4A, Cby shRNA markedly reduced Cby proteinlevels in a dose-dependent fashion. Importantly, Cby shRNA- ${Delta}$ Chad no major effect on Cby expression. In 3T3-L1 cells, expressionof Cby shRNA resulted in a modest but consistent increase inTOPFLASH activity, whereas Cby shRNA- ${Delta}$ C showed only a minor effect(Fig. 4B). As a control, cotransfection of Cby shRNA with ahuman Cby expression construct lacking the mouse Cby shRNA targetsequence rescued TOPFLASH activity. These results suggest thatCby represses ß-catenin-dependent transcription in3T3-L1 cells.

Depletion of Cby attenuates adipocyte differentiation of 3T3-L1 and mouse ES cells. The above-mentioned data indicate that Cby positively regulatesadipocyte differentiation. If so, one might hypothesize thatknockdown of endogenous Cby would impair the differentiationprocess. To test this hypothesis, we infected 3T3-L1 preadipocyteswith a retrovirus expressing Cby shRNA, Cby shRNA- ${Delta}$ C, or a controlvirus, and the infected cells were selected with G418 and allowedto proliferate to confluence. We observed no major differencesin morphology, viability, or growth rate between Cby shRNA andcontrol retrovirus-infected preadipocyte cells. Two days later(day zero), these stable cells were induced to differentiateinto mature adipocytes in the presence of MDI. Six days postinduction,total cell extracts were prepared and subjected to Western blotanalysis for Cby. As shown in Fig. 5B, endogenous Cby proteinwas substantially decreased by Cby RNAi. In parallel, fat accumulationwas visualized by staining lipids with Oil Red-O (Fig. 5A).3T3-L1 preadipocytes infected with a control retrovirus underwentefficient differentiation, as expected. In marked contrast,Cby RNAi almost completely blocked the accumulation of lipiddroplets, concomitant with reduced expression levels of adipogenicmarkers (Fig. 5B). A single-nucleotide deletion in the Cby shRNAsequence totally abolished these effects, confirming the specificityof Cby RNAi (Fig. 5A and B).

View larger version (84K):
[in this window]
[in a new window]

FIG. 5. Cby RNAi blocks the adipogenic conversion of 3T3-L1 cells, and dnTcf-4 rescues adipogenesis of Cby-RNAi cells. (A and B) 3T3-L1 cells were infected with a retrovirus vector alone (pSuppressor Retro) or a retrovirus expressing Cby shRNA or Cby shRNA- ${Delta}$ C, selected with G418, and induced to differentiate using MDI. Six days later, the cells were stained with Oil Red-O to visualize the degree of lipid accumulation (A). Whole-cell extracts were subjected to immunoblot analyses with antibodies against the indicated proteins (B). (C and D) Cby shRNA-expressing cells were reinfected with a retrovirus vector alone (pQCXIP) or retroviruses carrying dnTcf-4, coselected with G418 and puromycin, and maintained in growth media without MDI for 14 days after confluence. The cells were then stained with Oil Red-O and photographed at two different magnifications (C). Total cell lysates were prepared for Western blot analyses using the indicated antibodies (D).

If loss of Cby by RNAi blocks adipogenesis via elevation ofß-catenin activity, expression of dnTcf-4 could rescueadipogenic differentiation of Cby-RNAi cells. In an effort totest this possibility, the above-mentioned Cby shRNA-expressing3T3-L1 cells were reinfected with a control retrovirus or aretrovirus expressing dnTcf-4 and maintained in the absenceof MDI. As shown in Fig. 5C, expression of dnTcf-4 indeed rescuedadipocyte differentiation of Cby-RNAi cells, with a concomitantinduction of aP2 (Fig. 5D). These observations further supportour model, in which Cby facilitates adipogenesis by suppressingWnt/ß-catenin signaling.

ES cells are known to differentiate into a variety of cell typesin culture, including adipocytes (9, 17). To determine if Cbyis involved in adipogenesis in pluripotent ES cells, Cby shRNAwas stably expressed to suppress endogenous Cby expression inmouse R1 ES cells. Cby was expressed both in undifferentiatedES cells and during their in vitro differentiation, and CbyRNAi efficiently reduced Cby protein levels similar to thoseobserved in 3T3-L1 cells expressing Cby shRNA (data not shown).Reduction of Cby did not exhibit any adverse effects on maintenanceand early differentiation of ES cells (data not shown).

These stable lines were then differentiated by the standardhanging-drop method, and embryoid bodies were treated with retinoicacid and allowed to terminally differentiate into the adipogeniclineage in the presence of adipogenic hormones (9). Silencingof Cby expression significantly impaired adipocyte differentiation,while vector-transfected control cells efficiently differentiatedinto mature adipocytes under these conditions (Fig. 6A). Asexpected, expression of adipogenic markers was severely reducedin Cby knockdown cells, but not in control cells (Fig. 6B).Taken together, these results clearly demonstrate that Cby isrequired for adipocyte differentiation.

View larger version (46K):
[in this window]
[in a new window]

FIG. 6. Cby RNAi attenuates adipocyte differentiation of mouse ES cells. Mouse ES cells were stably transfected with an empty vector or Cby shRNA, selected by G418, and induced to differentiate into the adipogenic lineage for 20 days according to standard procedures. (A) Cells were stained with Oil Red-O. (B) RNA was purified and used to synthesize cDNAs for RT-PCR for adipogenic markers. ß-Actin was used as a loading control.

Cby^–/– embryonic fibroblasts display impaired adipocyte differentiation. To further investigate the importance of Cby in adipogenesis,primary mouse embryonic fibroblasts (MEFs) were prepared fromwild-type and Cby^–/– embryos at E12.5, culturedto confluence, and then treated with MDI to induce adipocytedifferentiation. As shown in Fig. 7A, Cby protein is not presentin Cby^–/– MEFs. Under these conditions, wild-typeMEFs differentiated into adipocytes and accumulated lipid droplets(Fig. 7B). In contrast, Cby^–/– MEFs exhibited reducedadipocyte differentiation, as verified by a decrease in aP2levels (Fig. 7A). These observations strongly support the notionthat Cby plays a proadipogenic role during adipocyte differentiation.

View larger version (52K):
[in this window]
[in a new window]

FIG. 7. Cby^–/– MEFs differentiate poorly into the adipogenic lineage. MEFs were prepared from wild-type (WT) or Cby-deficient (Cby^–/–) E12.5 embryos. (A) Whole-cell lysates were extracted from confluent cells and subjected to Western blotting with the indicated antibodies. (B) MEFs were induced to differentiate into adipocytes for 10 days with MDI, followed by Oil Red-O staining to visualize lipid droplets.

DISCUSSION

Top

DISCUSSION

Recent studies established that Wnt/ß-catenin signalsinhibit adipogenesis (3, 5, 7, 25, 35). Wnt-10b is likely tomediate this effect, since it is highly expressed in preadipocytes,but its expression declines upon adipogenic differentiation(5, 36). Wnt signaling appears to exert its effect through regulationof mitotic clonal expansion (35). However, the precise mechanismsthrough which Wnts block adipogenesis remain largely unknown.

Cby is an evolutionarily conserved nuclear protein that antagonizesß-catenin signaling (38). We have previously shownthat Cby binds to the C-terminal portion of ß-cateninand competes with Tcf/Lef transcription factors to repress genetranscription. In this study, we have uncovered a novel roleof Cby in adipocyte differentiation. Cby protein levels increaseduring differentiation of 3T3-L1 preadipocytes. Overexpressionof Cby facilitates adipogenesis of these cells, whereas depletionof Cby by RNAi inhibits adipocyte differentiation of 3T3-L1and mouse ES cells. Furthermore, MEFs lacking Cby show a diminishedability to differentiate into the adipogenic lineage. Our gain-and loss-of-function studies clearly indicate that Cby is anovel proadipogenic factor required for adipocyte differentiation.

It is worth noting that silencing of Cby by RNAi in 3T3-L1 orES cells does not affect cell morphology or the rate of proliferation.This implies that Cby is not essential for normal growth. Rather,Cby is crucial for the adipocyte differentiation process perse. In addition, Cby RNAi profoundly decreases induction ofC/EBP ${alpha}$ and PPAR ${gamma}$ (Fig. 5B and 6B), demonstrating that Cby functionsupstream of these key adipogenic transcription factors. Thisis in agreement with the prior report that activation of Wntsignaling abrogates expression of C/EBP ${alpha}$ and PPAR ${gamma}$ (36).

In 3T3-L1 cells, ectopic expression of Cby represses ß-catenin-mediatedtranscriptional activation, while loss of Cby by RNAi leadsto elevation of ß-catenin activity (Fig. 4). In supportof this, forced expression of stabilized ß-cateninpartially suppresses Cby-induced adipocyte differentiation (Fig.3), and moreover, expression of dnTcf-4 rescues adipogenesisof Cby shRNA-expressing cells (Fig. 5C and D), suggesting thatCby promotes adipocyte differentiation in part through inhibitingß-catenin. This is consistent with the previous findingthat the differentiation of preadipocytes into adipocytes requiressuppression of the Wnt/ß-catenin pathway (3, 25, 36).In this regard, Wnt-5b has been recently reported to promoteadipogenesis by antagonizing the Wnt/ß-catenin pathway(15, 16).

During differentiation of 3T3-L1 preadipocytes, nuclear ß-cateninlevels decrease; however, it remains present at basal levelsin mature adipocytes (25). Cby protein levels inversely correlatewith ß-catenin levels (Fig. 1A). Thus, it is plausiblethat Cby binds to nuclear ß-catenin and repressesits activity in adipocytes. This may ensure terminal adipocytedifferentiation by preventing ectopic transcriptional activationby ß-catenin. This repressive activity of Cby is criticalduring Drosophila embryogenesis, since depletion of Cby by RNAileads to ectopic activation of Wingless (Wg) target genes, evenin a wg-null background, leading to embryonic lethality (38).

Little is known about the in vivo functions of the Wnt/ß-cateninpathway in adipose tissue development. In contrast to in vitroobservations and gain-of-function studies in mice, Wnt-10b knockoutmice have no obvious defects in adipose tissue development (4,45). However, it is possible that other Wnt family members compensatefor the loss of Wnt-10b in the in vivo setting. Therefore, loss-of-functionexperiments are needed to rigorously document the involvementof ß-catenin signaling in adipogenesis in vivo. Inmice, Cby is expressed throughout embryonic development andin multiple adult tissues. Cby-null mice display severe digestive/absorptionproblems, and the vast majority die by weaning age, most likelydue to starvation (K.-I. Takemaru, V. Voronina, and R. T. Moon,unpublished data). Although they manifest reduced adiposityin early postnatal life, it is difficult to assess whether itis a primary or secondary effect due to the nutritional problems.Hence, further studies will be needed to define potential deficitsin adipocyte development in Cby-null animals.

In summary, our study provides compelling evidence that Cbyis a novel proadipogenic factor required for adipocyte differentiation.More recently, mutations in the Wnt-10b gene have been describedin human obesity (8), and LRP5 polymorphisms have been shownto be significantly associated with obesity phenotypes (14).These observations raise the possibility that manipulating Wntsignaling, such as blocking Cby's activity, may eventually providemeans to develop new therapies for obesity and its associateddisorders, such as diabetes.

ACKNOWLEDGMENTS

We thank Jeffrey Pessin for reagents and helpful discussions.

This work was supported by a Carol M. Baldwin Breast CancerResearch Award and NIH R01 DK073191 to K.-I.T. and by the AmericanHeart Association with a grant-in-aid to N.T. and a predoctoralFellowship to A.M.S. R.T.M. is an investigator of Howard HughesMedical Institute.

FOOTNOTES

* Corresponding author. Mailing address: SUNY at Stony Brook, Department of Pharmacology, BST 7-169 Nicolls Rd., Stony Brook, NY 11794-8651. Phone: (631) 444-7976. Fax: (631) 444-3218. E-mail: takemaru{at}pharm.sunysb.edu

Published ahead of print on 2 April 2007.

REFERENCES

Top