Acute Wnt pathway activation positively regulates leptin gene expression in mature adipocytes

Zong-Lan Chen, Wei-Juan Shao, Fen Xu, Ling Liu, Bei-Si Lin, Xiao- Hong Wei, Zhuo-Lun Song, Huo-Gen Lu, I. George Fantus, Jian-Ping Weng, Tian-Ru Jin
PII: S0898-6568(14)00414-8
DOI: doi: 10.1016/j.cellsig.2014.12.012
Reference: CLS 8361

To appear in: Cellular Signalling

Received date: 9 December 2014
Accepted date: 22 December 2014

Please cite this article as: Zong-Lan Chen, Wei-Juan Shao, Fen Xu, Ling Liu, Bei-Si Lin, Xiao-Hong Wei, Zhuo-Lun Song, Huo-Gen Lu, I. George Fantus, Jian-Ping Weng, Tian-Ru Jin, Acute Wnt pathway activation positively regulates leptin gene expression in mature adipocytes, Cellular Signalling (2014), doi: 10.1016/j.cellsig.2014.12.012

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Acute Wnt pathway activation positively regulates leptin gene expression in mature adipocytes
Zong-Lan Chen1, Wei-Juan Shao2, Fen Xu1, Ling Liu2, Bei-Si Lin1, Xiao-Hong Wei1, Zhuo-Lun Song2, Huo-Gen Lu2,3, I George Fantus2,3, Jian-Ping Weng1,* and Tian-Ru Jin2,*
1Department of Endocrinology and Metabolism, the Third Affiliated Hospital of Sun Yat-sen University; Guangdong Provincial Key Laboratory of Diabetology, Guangzhou, China
2Div. of Advanced Diagnostics, Toronto General Research Institute, University Health Network, Toronto, Ontario, Canada
3Lunenfeld-Tanenbaum Research Institute, Mount Sinai Hospital, Toronto, Ontario, Canada

Short Title: Acute Wnt activation stimulates leptin expression

* Corresponding author:

Dr. Tianru Jin, Rm. 10-354, Toronto Medical Discovery Tower, University Health Network, 101 College St., Toronto, Ontario, Canada, M5G 1L7, Email: [email protected]., or

Abstract

Genome-wide association studies (GWAS) have revealed the implication of several Wnt signaling pathway components, including its effector transcription factor 7-like 2 (TCF7L2) in

diabetes and other metabolic disorders. As TCF7L2 is expressed in adipocytes, we investigated its expression and function in rodent fat tissue and mature adipocytes. We found that TCF7L2 mRNA expression in C57BL/6 mouse epididymal fat tissue was up-regulated by feeding, but down-regulated by intraperitoneal insulin injection. In high-fat diet (HFD) fed mice, db/db mice and Zucker (fa/fa) rats, epididymal fat TCF7L2 mRNA levels were lower than the corresponding controls. Treating rat adipocytes with 100 nM insulin repressed TCF7L2 mRNA and protein levels, associated with the repression of leptin mRNA level. The treatment with 1 nM insulin, however, stimulated TCF7L2 and leptin mRNA levels. This stimulation could be attenuated by iCRT14, an inhibitor of β-catenin/TCF-responsive transcription. Wnt3a stimulated leptin mRNA level, which was also blocked by iCRT14 co-treatment. Utilizing the leptin-expressing cell line HTR8 as a tool, we defined an evolutionarily-conserved CREB binding motif that mediated Wnt3a activation. Although Wnt activation is known to repress the differentiation of 3T3-L1 cells towards mature adipocytes, short term Wnt3a treatment of differentiated 3T3-L1 cells stimulated leptin mRNA levels. Thus, wnt pathway plays a dual function in adipocytes, including the well-known repressive effect on adipogenesis and the stimulation of leptin production in mature adipocytes in response to nutritional status.
Key words: adipocytes, β-catenin, insulin, Leptin, TCF7L2, Wnt3a.

Abbreviations: GWAS, Genome-wide association studies; T2D, type 2 diabetes; SNPs, single nucleotide polymorphisms; TCF7L2, transcription factor 7-like 2; TCF7L2DN, dominant negative TCF7L2; β-cat, β-catenin; Dact1, Dishevelled-binding antagonist of beta-catenin 1; SFRP1, secreted frizzled-related protein 1; TK, thymidine kinase; HFD, high fat diet; PPARγ,

peroxisome proliferator-activated receptor gamma; 6-BnzcAMP, N6-Benzoyladenosine-3′, 5′- cyclic monophosphate sodium; LUC, luciferase; SGBS, Simpson-Golabi-Behmel syndrome; DKK1, Dickkopf 1.

⦁ Introduction

Genome-wide association studies (GWAS) have revealed the implication of Wnt signaling pathway components in type 2 diabetes (T2D) and other metabolic disorders [1, 2]. Extensive investigations have reproducibly revealed that certain single nucleotide polymorphisms (SNPs) of transcription factor 7-like 2 (TCF7L2, also known as TCF-4) [3, 4] are strongly associated with the risk of T2D in different ethnic populations [5-9]. TCF7L2 is abundantly expressed in many metabolic organs, including pancreatic islets, liver, brain, gut and adipocytes. While exploration of the metabolic functions of TCF7L2 in pancreatic β-cells and hepatocytes have generated profound important data but yet controversial conclusions [10-14]; we know relatively little of the metabolic function of TCF7L2 in adipocytes. An earlier investigation revealed that in omental and subcutaneous fat tissue of obese T2D patients, TCF7L2 expression level is significantly lower compared to obese normoglycemic subjects, indicating a potential beneficial role of TCF7L2 in adipocytes [15].
TCF7L2 or other TCF family members (TCF7, TCF7L1 and LEF1) interacts with β-catenin (β-cat), forming the bipartite transcription factor cat/TCF, the key effector of the canonical Wnt signaling pathway [2]. In response to Wnt ligand stimulation, β-cat enters the nucleus, forming the cat/TCF and stimulating Wnt target gene expression. The shuttling of β-cat from cytoplasm into nucleus can also be positively regulated by metabolic hormones including insulin, via stimulating its C-terminal serine residue (S675 and S552) phosphorylation [16, 17]. Thus, the function of TCF members can be influenced by its expression level, as well as by the availability and post-translational modification status of β-cat [16].
Wnt signaling pathway is intensively involved in cellular metabolism [18]. Wnt activation or cat/TCF activity is essential for bone formation [19]. However, in preadipocytes, the inactivation

of Wnt signaling is required to trigger their differentiation towards adipocytes [20-22]. Wnt activation is essential for maintaining preadipocytes in an undifferentiated state via the inhibition of adipogenic transcription factors, including peroxisome proliferator-activated receptor gamma (PPARγ). Functional knockdown of Wnt signaling, either by over-expressing Axin or dominant negative TCF7L2 (TCF7L2DN), stimulates the differentiation of preadipocytes into adipocytes [20]. Dishevelled-binding antagonist of beta-catenin 1 (Dact1) is a known Wnt signaling pathway modulator [23]. Its expression is down-regulated during adipogenesis [24]. In addition, secreted frizzled-related protein 1 (SFRP1), a Wnt antagonist, also promotes adipogenesis [25].
We found previously that in hepatocytes, TCF7L2 level and β-cat S675 phosphorylation can be positively regulated by feeding or by in vitro insulin treatment [14]. Here we show that TCF7L2 as well as leptin mRNA levels in mouse epididymal fat can be activated by re-feeding, while in isolated rat primary mature adipocytes, a low concentration (but not high concentration) of insulin also increased TCF7L2 and leptin mRNA levels. In high fat diet (HFD)-fed mice and two other obese and insulin resistant rodent models, however, TCF7L2 mRNA level is significantly reduced. Interestingly, although Wnt activation is known to repress the differentiation of the immortalized preadipocyte line 3T3-L1, in rat primary adipocytes and differentiated 3T3-L1 cells, short term Wnt3a treatment stimulated the expression of leptin mRNA, which encodes the major peptide hormone leptin by mature adipocytes. Together, this investigation supports the notion that TCF7L2 expression can be nutritionally regulated in rodent adipocytes, and that Wnt signaling pathway exerts dual functions in preadipocytes versus mature adipocytes.

⦁ Materials and methods

⦁ Animals

Male C57BL/6 mice at the age of 7 wks were purchased from the Charles River Laboratories (Montreal, Canada). They were maintained on a 12-h light/dark cycle with free access to food and water. In experiment 1, after acclimatization for 1 week, 12 mice were randomly divided into four groups, with three per group as follows: fasted group (1800h to 0900h), re-fed group (4 h re- feeding after fasting), PBS group (fasted mice with i.p. PBS injection), and insulin group (fasted mice with i.p. insulin injection, 1U/kg body weight). Blood and epididymal fat tissue were collected 30 min after insulin or PBS injection, as we have reported previously [26]. In experiment 2, after acclimatization for 1 week, 5 mice were fed with low-fat diet (LFD, 10% of kcal from fat) and 5 mice with high-fat diet (HFD, 60% of kcal from fat) for 12 wks, followed by epididymal fat tissue collection. Sixteen-week-old male db/db obese mice (n=4) and age- matched heterozygous db/m control lean mice (n=4) were obtained from Model Animal Research Center (Nanjing University, Nanjing, China). Seven-week-old male normal SD rats, twelve-week-old male Zucker fa/fa obese rats, as well as the heterozygous fa/+ control rats were purchased from Charles River Laboratories. All protocols for animal use and euthanasia were in accordance with the NIH guidelines, approved by the Sun Yat-Sen University institutional Animal Care and Use Committee or by the Animal Care Committee of University Health Network.
⦁ Measurement of serum insulin levels and supernatant leptin levels

After overnight fasting, mice used in this study were sacrificed and blood samples were collected. Serum insulin levels were determined using a mouse insulin ELISA kit (Mercodia, Uppsala, Sweden) according to the manufacturer’s instructions. Cell culture supernatants were

harvested after drug treatment and stored at -80℃ until measurements were performed. Supernatant leptin levels were analyzed using a rat leptin ELISA kit (Sigma-Aldrich, Ontario, Canada) according to the manufacturer’s protocols.
⦁ Reagents and cell culture

Recombinant canonical Wnt ligand Wnt3a and non-canonical Wnt ligand Wnt11 were purchased from R&D Systems (Minneapolis, MN, USA). The PKA specific cAMP analogue N6- Benzoyladenosine-3′, 5′-cyclic monophosphate sodium (6-BnzcAMP) and the cAMP promoting agent forskolin were purchased from Sigma-Aldrich (Oakville, Ontario, Canada). iCRT14, an inhibitor of β-catenin/TCF-responsive transcription, was the product of R&D Systems Inc. The human placenta trophoblast cell line HTR8 was a gift of Dr. Chun Peng (York University, Toronto, Canada) and it was cultured in RPMI 1640 medium with10% FBS [27]. The isolation and culture of primary adipocytes from male SD rat epididymal fat tissue has been previously described [28]. With this method, all non-mature adipocytes and non-adipocytes were removed during the isolation procedure [29]. For conducting a given treatment on primary adipocytes, cells were serum starved for at least 3 h.
⦁ Differentiation of 3T3-L1 cells

The differentiation of mouse 3T3-L1 cell line (ZenBio) into mature adipocytes has been previously reported [30]. Briefly, 3T3-L1 cells were cultured in DMEM supplemented with 10% FBS for two days after cells reached 100% confluence. The cells were then incubated in the differentiation medium with an adipogenic cocktail for three days, followed by further differentiation in an insulin-supplemented medium for another two days. Cells at this stage were defined as 3T3-L1D7. These differentiated cells were further cultured in DMEM for another

three days, which were defined as 3T3-L1D10. Undifferentiated or differentiated 3T3-L1 cells were maintained in serum-free medium for 3 h before conducting a given treatment.
⦁ Luciferase (LUC) reporter analysis

The mouse leptin gene 5’ flanking region (740 bp in length) was obtained by PCR against a mouse tail DNA sample from a wild type C57BL/6 mouse, followed by DNA sequencing confirmation and insertion into pGL3 luciferase (LUC) reporter vector (Promega). This fusion gene is defined as Lep-LUC (-740). Two additional Lep-LUC fusion gene constructs were then generated, defined as Lep-LUC (-491WT) and Lep-LUC (-491M), which contain a wild type or a mutated putative CREB motif, respectively. PCR primers for the generation of Lep-LUC fusion gene construct are shown in supplementary Table 1. The TK-LUC fusion gene, in which the expression of the LUC reporter is driven by the HSV thymidine kinase (TK) promoter, has been described in our previous study [31]. Rat primary mature adipocytes were transfected with Lep- LUC (-740) by electroporation as previously reported [28]. Dominant negative TCF7L2 (TCF7L2DN) plasmid [4, 32] or constitutively active β-cat (the S33Y mutant) plasmid [33] was co-transfected with Lep-LUC into the HTR8 cell line with lipofectamine 2000. Transfected cells were cultured for 18 h before being harvested for LUC analysis. For Wnt3a or iCRT14 treatment, cells were further cultured in serum-free medium for 3 h before indicated amount of reagent was added. Method for LUC reporter analyses has been described previously [31].
⦁ Oil Red O staining

At different differentiation stages, 3T3-L1 cells were stained with Oil Red O for lipid content quantification, as previously reported [34]. Briefly, after Oil Red O staining, cells were treated with 1 ml isopropanol for 15 min, followed by absorbance measurement at 510 nm wavelength.
⦁ RT-PCR and quantitative RT-PCR

Total RNA was isolated using Trizol reagent. RT-PCR and Quantitative RT-PCR were conducted as previously described [32]. Relative mRNA levels were calculated using the 2−ΔΔCT method. All PCR primers utilized in this study are included in supplementary Table 1.
⦁ Western blotting and gel shift assay

Methods for protein extraction and Western blotting have been described previously [35, 36]. Antibodies against TCF7L2, β-cat (pS675), β-cat (pS552), CREB, CREB (pS133), AKT, AKT (pS473), β-actin and GAPDH were purchased from Cell Signaling Technology (Danvers, MA), while the β-cat antibody was obtained from Santa Cruz Biotechnology (Dallas, TX). Nuclear protein extraction and gel shift assay have been performed as previously described [37]. Briefly, approximately 5 µg nuclear proteins (unless otherwise stated) were incubated with dCTP (α-32P) labeled double stranded DNA probe (approximately 0.4 ng DNA) at room temperature for 45 min, followed by 10% non-denatured PAGE separation. Gel shift probe sequences are presented in Fig. 6A.
⦁ Statistical analysis

All data are expressed as means ± SE. Results were analyzed by student t-test or ANOVA, as appropriate. Statistical significance was set at P < 0.05.

⦁ Results

⦁ TCF7L2 expression in adipose tissue can be metabolically regulated

Wnt pathway component expression in adipose tissue has been previously demonstrated [38, 39]. We assessed Wnt related gene expression with a custom designed cDNA microarray against epididymal fat from fasted C57BL/6 mice. We assumed that those RNAs of which adjusted CT mean values (n=3) are lower than 32 are abundantly expressed. As shown in supplementary

Table 2, a series of genes, including APC, Axin1, Axin2 , Dact1, certain Frizzled (Frz) receptors, the co-receptors Lrp5 and 6, and the Wnt ligand 2b and 5a are abundantly expressed. In addition to TCF7L2, TCF7 and TCF7L1, but not Lef1, are also abundantly expressed in the fat tissue.
We found previously that re-feeding and in vitro insulin treatment increase hepatic TCF7L2 expression and β-cat Ser675 phosphorylation, suggesting that cat/TCF mediates, at least in part, the function of insulin in the liver [14]. Here we extended these studies to adipose tissue. Firstly, we examined the effect of feeding and i.p. insulin injection on TCF7L2 expression in the epididymal fat tissue of C57BL/6 mice. Feeding increased serum insulin level from ~0.4 ng/ml to ~1.2 ng/ml, while 30 min after insulin injection, serum insulin level increased to ~6.5 ng/ml. As shown in Fig. 1A, fat TCF7L2 mRNA level can be stimulated by feeding but repressed by i.p. insulin injection. The up- and down-regulation of TCF7L2 expression by feeding or insulin injection was positively correlated with the levels of a known Wnt signaling pathway target Axin-2 (Fig. 1B) and leptin, which encodes one of the most important peptide hormones secreted by adipocytes(Fig. 1C). Secondly, we expanded the investigation of adipose tissues to three obese and insulin resistant rodent models. In LFD- and HFD-fed mice, fasting serum insulin levels were 0.5 ng/ml and 1.6 ng/ml, respectively, while in db/m and db/db mice, the fasting serum insulin levels were 0.38 ng/ml and 1.61 ng/ml, respectively. Epididymal fat leptin mRNA levels in these obese models were higher than the corresponding controls (Fig. 1D-F), associated with reduced TCF7L2 mRNA levels (Fig. 1G-I). Thus, the positive correlation between TCF7L2 and leptin mRNA levels was not observed in chronic obese and insulin resistant rodent models.
⦁ TCF7L2 and leptin expression was down-regulated by insulin at a high concentration

Increased TCF7L2 mRNA expression by re-feeding and reduced TCF7L2 expression during chronic insulin resistance prompted us to directly assess the effect of insulin on TCF7L2 and

leptin expression in mature adipocytes in vitro. This way, we are able to avoid the technical limitation of in vivo insulin injection and avoid the “contamination” of other cell lineages in adipose tissue sample collection. Indeed, adipose tissue contains several other cell types, including preadipocytes, and their response to nutritional and hormonal changes on TCF7L2 expression could be different. Rat primary adipocytes were starved for 3 h, followed by 100 nM insulin treatment for 4 or 24 h. We confirmed that 4 h insulin treatment significantly increased Akt phosphorylation (data not shown). Insulin at this concentration significantly repressed leptin mRNA level at both times (Fig. 2A), while its repression on TCF7L2 mRNA was observed at 4 h but not at 24 h (Fig. 2B). Western blotting was then applied for assessing the effect of insulin on TCF7L2 protein level and β-cat S675 phosphorylation (Fig. 2C). Treating adipocytes with 100 nM insulin for 4 or 24 h generated no appreciable effect on β-cat S675 phosphorylation (Fig. 2C and 2D). However, total β-cat level was significantly increased 24 h after insulin treatment (Fig. 2C and 2E). TCF7L2 protein level was significantly repressed by insulin (Fig. 2C and 2F), and this repression correlated with the repression of leptin mRNA level (Fig. 2A). Finally, the level of PPARγ was significantly repressed by 24 h insulin treatment (Fig. 2C and 2G).
⦁ The regulation of leptin and TCF7L2 by insulin in mature adipocytes is dose- dependent and the stimulation requires cat/TCF activity
The lack of stimulation of β-cat S675 phosphorylation by 4 or 24 hr insulin treatment in adipocytes could have been due to the fact that such phosphorylation is a transient event [32]. Thus, we treated rat adipocytes with 1 nM or 100 nM insulin for 5, 15 and 30 min. Membrane permeable cAMP and the cAMP promoting agent forskolin were included as positive controls. With 1 nM insulin, we observed a modest stimulation of β-cat S675 phosphorylation at 5min and 15min, but not at 30 min. With 100 nM insulin, the stimulation was observed at three time points

(Fig. 3A). Interestingly, total β-cat level was increased along with increased β-cat S675 phosphorylation (Fig. 3A). Insulin had no appreciable effect on β-cat S552 phosphorylation, although forskolin did (Fig. 3A).
We then assessed the effect of insulin at three different concentrations on TCF7L2 as well as leptin mRNA expression. At a low concentration of 1 nM, insulin stimulated both TCF7L2 (Fig. 3B) and leptin mRNA levels (Fig. 3C). No significant effect on their mRNA expression was observed when 10 nM of insulin was applied, while repression was observed when the insulin dose was increased to 100 nM. We then found that the stimulation of TCF7L2 and leptin mRNA expression by 1 nM insulin in rat adipocytes can be attenuated by co-treatment with the cat/TCF inhibitor iCRT14 (Fig. 3D and 3E), although its application did not block the stimulatory effect of insulin on TCF7L2 protein levels (Fig. 3F). Together, these observations suggest that at a low dose, insulin positively regulates leptin gene expression via stimulating cat/TCF activity, possibly via increasing TCF7L2 expression and a transient stimulation of β-cat S675 phosphorylation.
⦁ Wnt3a stimulates leptin expression in primary adipocytes

Up- and down-regulated TCF7L2 expression by re-feeding and insulin injection were positively correlated with increased and reduced leptin gene expression, suggesting that TCF7L2 or Wnt signaling positively regulate leptin expression. However, reduced TCF7L2 mRNA in obese animal models was associated with increased leptin expression, suggesting that the potential positive relationship between TCF7L2 or Wnt signaling activation and leptin expression occurs only during acute treatment. Therefore, we directly assessed the effect of Wnt ligand on leptin gene expression in vitro. As shown in Fig. 4A and 4B, Wnt3a (but not Wnt11) treatment increased leptin mRNA level in rat primary adipocytes, associated with increased

leptin hormone secretion (Fig. 4C). Furthermore, the effect of Wnt3a on leptin mRNA expression in mature adipocytes can also be attenuated by iCRT14 co-treatment. iCRT14 on its own also stimulated leptin mRNA level, which could be due to a yet undefined off-target effect (Fig. 4D). Finally, Wnt3a stimulated the activity of Lep-LUC (-740) when transfected into the mature rat adipocytes (Fig. 4E). The activation was not observed when the parental pGL3-LUC was examined (data not shown).
⦁ Wnt3a stimulated leptin promoter activity is dependent on a CREB binding site

As the transfection efficiency for multiple DNAs in primary adipocytes is substantially low, we conducted a set of further verification studies in a leptin-expressing placental cell line HTR8 [40]. Although Wnt3a cannot significantly stimulate Lep-LUC (-740) in this cell line (data not shown), it stimulated the activity of Lep-LUC (-491 WT) (Fig. 5A and 5B). Furthermore, the activation was blocked by iCRT14 co-treatment (Fig. 5B). Such activation was absent when the CREB site mutated Lep-LUC (-491M) was assessed (Fig. 5A and 5C). We then constructed another fusion gene, defined as Lep-TK-LUC, in which one copy of the 31 bp CREB binding site-containing DNA fragment was inserted into the parental TK-LUC fusion gene. As shown in Fig, 5D-E, Wnt3a did not activate the parental TK-LUC, but activated Lep-TK-LUC. This activation was also blocked by iCRT14 co-treatment (Fig. 5E). We hence defined this 31 bp DNA fragment as LepCREB. Finally, in HTR8 cells, co-transfection of TCF7L2DN significantly reduced while co-transfection of the constitutively active β-cat (the S33Y mutant) increased the activity of Lep-LUC (-740) (Fig. 5F and 5G). We hence suggest that leptin is indeed among the direct downstream targets of the canonical Wnt signaling pathway and its key effector cat/TCF.
⦁ Gel shift assay detected the interaction between HTR8 nuclear proteins and LepCREB

We defined the 31 bp DNA fragment as LepCREB and have then located a short fragment within

LepCREB, which shares 87.5% sequence identify with the consensus TCF binding motif (Fig. 6A). Gel shift assay was performed using LepCREB as the probe against nuclear proteins isolated from HTR8 cells. As shown in Fig. 6B, a specific complex was detected. The formation of this complex was attenuated by either 20 fold unlabeled LepCREB probe or unlabeled TCF probe, although the degrees of attenuation differed. Furthermore, the complex was also formed against nuclear proteins of BHK fibroblasts transfected with TCF7L2 (Fig. 6C). By conducting “super- shifting”, we were able to confirm the participation of CREB in the formation of the complex, but not TCF7L2 (Fig. 6D). This could be due to a weak interaction or that the antibody does not work in this “super-shift” assay. We then repeated the gel shift assay using the consensus TCF binding motif as the probe. As shown in Fig. 6E, three specific complexes (C1-C3) were detected, and each of them can be attenuated by either 20 fold of the unlabeled TCF probe or the unlabeled LepCREB probe. The complex C2 was also detected with BHK transfected with TCF7L2, which can be blocked by either the unlabeled TCF or LepCREB probe (Fig. 6E). We then reduced the amount of HTR8 nuclear proteins in the gel shift assay from 5 µg to 2 µg. As shown in Fig. 6F, the formation of the complex was virtually blocked by 20 fold unlabeled LepCREB probe, and the complex can be virtually all “super-shifted” with either 1µl or 2 µl of CREB antibody. An unrelated GATA3 antibody generated no “super-shift”. When nuclear proteins from rat adipocytes were utilized in gel shift assay, we were unable to detect distinct complexes but a smear with the LepCREB probe. However, when CREB antibody was applied, we observed the formation of the “super-shifted” complex (Fig. 6F). Together, we have confirmed the physical interaction between CREB and LepCREB although the interaction between TCF7L2 and LepCREB needs further investigation.
⦁ Wnt3a represses 3T3-L1 differentiation but stimulates leptin expression

The Wnt pathway is known to repress adipogenesis [20, 24]. To confirm the acute stimulatory effect of Wnt3a on leptin expression, we conducted experiments with 3T3-L1 cells (Fig. 7A). Undifferentiated 3T3-L1 cells, or cells that underwent the differentiation for 7 days or 10 days were treated with or without Wnt3a for 24 h, followed by oil Red O staining. In addition, cells during different differentiation stages were also treated with or without Wnt3a for 4 h, followed by RNA extraction and real-time PCR analyses. As shown in Fig. 7B and 7C, 24 h Wnt3a treatment reduced 3T3-L1 differentiation, as lipid content was significantly reduced by the treatment. Four hr Wnt3a treatment reduced the expression of PPARγ (Fig. 7D), which is a key transcription factor that positively regulates adipogenesis [41]. However, leptin mRNA level was significantly increased after 4 h Wnt3a treatment in cells that underwent the differentiation (Fig. 7E). The expression of adiponectin gene, on the other hand, was not affected; although its level was up-regulated during the differentiation procedure (Fig. 7F). The expression of preadipocyte factor-1 (Pref-1) was reduced along with the differentiation procedure. Although Pref-1 expression can be down-regulated by Wnt3a in undifferentiated 3T3-L1 cells; after the differentiation, its expression levels were very low and were not affected by 4 h Wnt3a treatment (Fig. 7G). The expression of Axin-2, a known downstream target of Wnt signaling, was up- regulated by acute Wnt3a treatment (Fig. 7H). Thus, although it is known that Wnt activation represses preadipocytes differentiation; in mature adipocytes, acute Wnt3a treatment stimulates the expression of leptin.

⦁ Discussion

In this study, we assessed the role of 4 hr re-feeding, acute insulin treatment, and Wnt signaling pathway activation on the expression of leptin in mature adipocytes. Our observations suggest that epididymal fat TCF7L2 mRNA expression level can be regulated by nutrients and hormones. It can be up-regulated by re-feeding, or by acute treatment with 1 nM insulin in vitro. This activation was associated with increased leptin mRNA expression. Acute insulin treatment at high concentrations, either in vitro or in vivo via i.p. injection, reduced TCF7L2 expression, and reduced leptin expression. Furthermore, in obese and insulin resistant rodent models with genetic defects of leptin receptors, TCF7L2 level is down-regulated, suggesting that chronic insulin resistance may lead to reduced TCF7L2 expression. These observations are consistent with the report that in omental and subcutaneous fat tissue of obese T2D patients, TCF7L2 expression level is significantly lower compared to obese normoglycemic subjects [15]. We hence support the suggestion that TCF7L2 and Wnt signaling in adipocytes may play a beneficial role. We have also demonstrated in this study that although Wnt pathway activation with 24 h Wnt3a treatment attenuated 3T3-L1 preadipocyte differentiation, in mature adipocytes, acute 4 h Wnt signaling pathway activation resulted in an increase of leptin mRNA level. The recognition of this important dual function of the Wnt signaling pathway provides an explanation why several Wnt ligands, receptors, co-receptors, and effectors are abundantly expressed in mature adipocytes [38].
Following the recognition of TCF7L2 as the top T2D risk gene by GWAS [5], great efforts have been made to functionally explore its metabolic role in pancreatic β-cells and in hepatocytes [11-14, 42-45]. As TCF7L2 is also abundantly expressed in adipocytes, a number of investigations have been conducted on its expression and alternative splicing in human

adipocytes [15, 46-51]. Cauchi et al reported a reduced TCF7L2 expression in subcutaneous and omental fat tissue in T2D patients [15], while Kaminska et al found that TCF7L2 splicing can be regulated by weight loss, involving fatty acid and glucose metabolism [49]. Very recently, the effect of exercise on adipose tissue TCF7L2 gene methylation was also reported [50, 51]. Thus, TCF7L2 falls into the category of genes whose expression can be metabolically regulated. In 2009, Lagathu et al. demonstrated that Dact1, a nutritionally regulated gene, serves as a Wnt signaling pathway modulator in adipocytes and regulates adipogenesis [24]. We show here for the first time that insulin exerts opposite effects on TCF7L2 expression, depending on its concentration. We also show that chronic insulin resistance reduced TCF7L2 expression. Repressed TCF7L2 expression in differentiated human Simpson-Golabi-Behmel syndrome (SGBS) preadipocytes by insulin (5-10 nM) treatment was also reported by Mondal and colleagues [47]. Furthermore, Ahlzen et al found that 1 nM of insulin repressed TCF7L2 mRNA level in cultured human adipocytes regardless of glucose concentration [52]. The discrepancy between the study by Ahlzen et al and our study with 1 nM of insulin could be due to species difference and experimental details. In addition, adipocytes utilized in these two investigations were isolated from different physical locations [52].
More than 14 years ago, Ross et al reported that Wnt signaling pathway activation, likely through the canonical Wnt ligand Wnt10b, maintains preadipocytes in an undifferentiated state via inhibiting adipogenic transcription factors including PPARγ [20]. They have also shown that TCF7L2DN and the Wnt pathway inhibitor Axin can stimulate preadipocytes differentiation. TCF7L2DN and Axin can also trigger the trans-differentiation of myoblasts into adipocytes [20], while the GSK3 inhibitor CHIR99021, which mimics Wnt signaling stimulation, can block adipogenesis [21]. Gustafson and Smith found that another Wnt pathway inhibitor Dickkopf 1

(DKK1) can promote adipogenesis in cells with a low degree of differentiation, and DKK1 and bone morphogenetic protein 4 (BMP4) exert additive effects on differentiating adipocytes [53]. These investigations have defined the fundamental role of Wnt signaling pathway in repressing adipogenesis, a chronic differentiation process [41]. However, in mature adipocytes, Wnt signaling pathway components, including several Wnt ligands, receptors, co-receptors, and effectors are abundantly expressed. The expression of these components in visceral adipose tissue was also shown to be modulated by chronic hypoadiponectinemia [38]. Of interest, Schinner and colleagues demonstrated that human adipocyte-derived Wnt ligands can stimulate β-cell insulin secretion, glucokinase gene transcription and β-cell proliferation [54, 55]. These observations suggest a role of Wnt ligands secreted by mature adipocytes as “endocrine” factors.
As Wnt receptors and co-receptors are also expressed in mature adipocytes, it is reasonable to hypothesize that Wnt ligands produced by adipocytes can function in an “autocrine” manner. We show here that indeed, in mature rat epididymal adipocytes, 4 h Wnt3a treatment stimulated leptin gene expression, leptin hormone secretion and leptin promoter activity. With the utilization of the chemical inhibitor iCRT14, dominant negative TCF7L2 and constitutively active S33Y β-cat, we obtained a battery of data supporting that leptin is among the downstream targets of the Wnt signaling pathway. A previous study by Li et al. has defined a consensus CREB site within the mouse leptin gene promoter [56]. This site is conserved among humans, mice and rats. We demonstrated here that this site mediates the stimulatory effect of Wnt activation, supporting the notion for the existence of crosstalk between Wnt and cAMP-PKA signaling pathways [16]. Further investigations are needed to determine how TCF members and PKA signaling pathway components work together in response to nutritional changes in regulating leptin gene transcript via LepCREB and other yet to be identified cis-elements.

Together, we suggest that Wnt signaling pathway exerts dual function in the adipocytes lineage. In pre-adipocytes, Wnt pathway activation represses adipogenesis; while in mature adipocytes acute Wnt activation stimulates leptin expression.
The Wnt pathway effector cat/TCF not only mediates the effect of Wnt ligands, but also metabolic and other hormones in response to nutritional, environmental and emotional changes [16]. This can be achieved by regulating TCF expression and β-cat S675 and S552 phosphorylation [17, 32, 57]. In this study, we demonstrated for the first time that β-cat S675 phosphorylation in mature adipocytes can be stimulated by insulin, although the stimulation was modest and occurred transiently. Interestingly, total β-cat level was stimulated by insulin, especially after 24 h insulin treatment. This increase is strongly associated with the repression of PPARγ. The positive correlation between S675 β-cat level and total β-cat level was also reported by Gustafson and Smith in 3T3-L1 cells treated with Wnt3a [58]. Another recent study has also utilized total β-cat as an indication of Wnt activation in adipocytes [59]. How the increase in total β-cat contributes to Wnt pathway regulation in adipocytes remains to be further explored.
Mechanistic detail on dosage-dependent effect of insulin on TCF7L2 and leptin expression needs to be further explored. We suggest that the elevation of plasma insulin level after re- feeding leads to increased cat/TCF activity, resulting in the stimulation of leptin gene transcription (Fig. 8A). As high concentration of insulin represses TCF7L2 expression, we see a negative regulation of leptin expression, regardless of the elevation of β-cat level (Fig. 8B). It should be pointed out that chronic insulin resistance leads to repressed TCF7L2 expression and stimulated leptin gene expression. This uncoupling is likely due to a compensatory response of insulin as well as leptin resistance (Fig. 8B).

Leptin is a major hormone produced by adipocytes and its signaling and resistance play important physiological and pathological roles [60]. Following the isolation of leptin cDNA in 1994 [61], mechanisms underlying its transcription have been explored in adipocytes and elsewhere [56, 62-65]. In 3T3-L1 cells, cAMP stimulates Lep-LUC reporter gene expression in the absence of serum [56]. As the cAMP/PKA signaling cascade is able to crosstalk with Wnt pathway via stimulating β-cat S675 phosphorylation [16], it is reasonable to hypothesize that leptin is among the downstream targets of Wnt pathway as well. In this study, with commercially available Wnt3a, we demonstrated for the first time that Wnt activation stimulates leptin mRNA expression, hormone secretion and leptin promoter activity. Whether certain Wnt ligands such as Wnt 2b and Wnt5a, abundantly produced by adipocytes, can serve as “autocrine” factors to regulate leptin production remains to be explored. It is also worth to examine whether the expression of “less abundantly expressed” Wnt ligands can be regulated by nutritional changes in mature adipocytes.
Conflicts of interest

The authors declare no potential conflicts of interest.

Acknowledgments

This work was supported by Canadian Institutes of Health Research (CIHR, MOP-89987 and MOP-97790 to TJ), the National Science Fund for Distinguished Young Scholars of China (No. 81025005 to J.W.), NSFC-CIHR (No. 81261120565 to J.W.), China-Canada Joint Research Initiative (to IGF and TJ), and the National Natural Science Foundation of China (No.81300705 to F.X.). ZC, JW and TJ designed experiments. ZC, WS, LL, FX, BL, XW, ZS, and HL conducted the experiments. ZC, WS, FX, ZS, JW and TJ analyzed the data. ZC, FX, JW and TJ wrote the manuscript. IGF edited the manuscript, and all authors approved the final version.

References

[1] Ip W, Chiang YT, Jin T, Cell Biosci. 2012;2:28. [2] Jin T, Diabetologia. 2008;51:1771-1780.
⦁ Korinek V, Barker N, Moerer P, van Donselaar E, Huls G, Peters PJ, Clevers H, Nat Genet. 1998;19:379-383.
⦁ Yi F, Brubaker PL, Jin T, J Biol Chem. 2005;280:1457-1464.
⦁ Grant SF, Thorleifsson G, Reynisdottir I, Benediktsson R, Manolescu A, Sainz J, Helgason A, Stefansson H, Emilsson V, Helgadottir A, Styrkarsdottir U, Magnusson KP, Walters GB, Palsdottir E, Jonsdottir T, Gudmundsdottir T, Gylfason A, Saemundsdottir J, Wilensky RL, Reilly MP, Rader DJ, Bagger Y, Christiansen C, Gudnason V, Sigurdsson G, Thorsteinsdottir U, Gulcher JR, Kong A, Stefansson K, Nat Genet. 2006;38:320-323.
⦁ Florez JC, Curr Opin Clin Nutr Metab Care. 2007;10:391-396. [7] Grant SF, Diabetes. 2012;61:2657-2658.
⦁ Grant SF, Hakonarson H, Schwartz S, Endocr Rev. 2010;31:183-193.
⦁ Zhao J, Schug J, Li M, Kaestner KH, Grant SF, Diabetologia. 2010;53:2340-2346.
⦁ Shu L, Matveyenko AV, Kerr-Conte J, Cho JH, McIntosh CH, Maedler K, Hum Mol Genet. 2009;18:2388-2399.
⦁ Norton L, Fourcaudot M, Abdul-Ghani MA, Winnier D, Mehta FF, Jenkinson CP, Defronzo RA, Diabetologia. 2011;54:3132-3142.
⦁ Boj SF, van Es JH, Huch M, Li VS, Jose A, Hatzis P, Mokry M, Haegebarth A, van den Born M, Chambon P, Voshol P, Dor Y, Cuppen E, Fillat C, Clevers H, Cell. 2012;151:1595-1607.
⦁ da Silva Xavier G, Mondragon A, Sun G, Chen L, McGinty JA, French PM, Rutter GA, Diabetologia. 2012;55:2667-2676.
⦁ Ip W, Shao W, Chiang YT, Jin T, Am J Physiol Endocrinol Metab. 2012;303:E1166-1176.
⦁ Cauchi S, Meyre D, Dina C, Choquet H, Samson C, Gallina S, Balkau B, Charpentier G, Pattou F, Stetsyuk V, Scharfmann R, Staels B, Fruhbeck G, Froguel P, Diabetes. 2006;55:2903-2908.
⦁ Jin T, George Fantus I, Sun J, Cell Signal. 2008;20:1697-1704.
⦁ Liu Z, Habener JF, J Biol Chem. 2008;283:8723-8735.
⦁ Sethi JK, Vidal-Puig A, Biochem J. 2010;427:1-17.
⦁ Manolagas SC, Almeida M, Molecular endocrinology. 2007;21:2605-2614.
⦁ Ross SE, Hemati N, Longo KA, Bennett CN, Lucas PC, Erickson RL, MacDougald OA, Science. 2000;289:950-953.
⦁ Bennett CN, Ross SE, Longo KA, Bajnok L, Hemati N, Johnson KW, Harrison SD, MacDougald OA, J Biol Chem. 2002;277:30998-31004.
⦁ MacDonald BT, Tamai K, He X, Dev Cell. 2009;17:9-26.
⦁ Cheyette BN, Waxman JS, Miller JR, Takemaru K, Sheldahl LC, Khlebtsova N, Fox EP, Earnest T, Moon RT, Dev Cell. 2002;2:449-461.
⦁ Lagathu C, Christodoulides C, Virtue S, Cawthorn WP, Franzin C, Kimber WA, Nora ED, Campbell M, Medina-Gomez G, Cheyette BN, Vidal-Puig AJ, Sethi JK, Diabetes. 2009;58:609-619.
⦁ Lagathu C, Christodoulides C, Tan CY, Virtue S, Laudes M, Campbell M, Ishikawa K, Ortega F, Tinahones FJ, Fernandez-Real JM, Oresic M, Sethi JK, Vidal-Puig A, Int J Obes (Lond). 2010;34:1695-1705.
⦁ Ip W, Shao W, Chiang YT, Jin T, American journal of physiology. Endocrinology and metabolism. 2012;303:E1166-1176.
⦁ Fu G, Ye G, Nadeem L, Ji L, Manchanda T, Wang Y, Zhao Y, Qiao J, Wang YL, Lye S, Yang BB, Peng C, Hypertension. 2013;61:864-872.
⦁ Beard KM, Lu H, Ho K, Fantus IG, Diabetes. 2006;55:2678-2687.

[29] Rodbell M, J Biol Chem. 1964;239:375-380.
⦁ Zhou Y, Xu F, Deng H, Bi Y, Sun W, Zhao Y, Chen Z, Weng J, PLoS One. 2013;8:e84016.
⦁ Jin T, Drucker DJ, Mol Cell Biol. 1996;16:19-28.
⦁ Shao W, Wang D, Chiang YT, Ip W, Zhu L, Xu F, Columbus J, Belsham DD, Irwin DM, Zhang H, Wen X, Wang Q, Jin T, Diabetes. 2013;62:789-800.
⦁ Ni Z, Anini Y, Fang X, Mills G, Brubaker PL, Jin T, J Biol Chem. 2003;278:1380-1387.
⦁ Putnam K, Batifoulier-Yiannikouris F, Bharadwaj KG, Lewis E, Karounos M, Daugherty A, Cassis LA, Endocrinology. 2012;153:4677-4686.
⦁ Xu F, Burk D, Gao Z, Yin J, Zhang X, Weng J, Ye J, Endocrinology. 2012;153:1706-1716.
⦁ Yu Z, Shao W, Chiang Y, Foltz W, Zhang Z, Ling W, Fantus IG, Jin T, Diabetologia. 2011;54:922-934.
⦁ Sirek AS, Liu L, Naples M, Adeli K, Ng DS, Jin T, Endocrinology. 2009;150:3483-3492.
⦁ Wada N, Hashinaga T, Otabe S, Yuan X, Kurita Y, Kakino S, Ohoki T, Nakayama H, Fukutani T, Tajiri Y, Yamada K, PloS one. 2013;8:e67712.
⦁ Ross SE, Erickson RL, Gerin I, DeRose PM, Bajnok L, Longo KA, Misek DE, Kuick R, Hanash SM, Atkins KB, Andresen SM, Nebb HI, Madsen L, Kristiansen K, MacDougald OA, Mol Cell Biol. 2002;22:5989-5999.
⦁ Munir S, Xu G, Wu Y, Yang B, Lala PK, Peng C, J Biol Chem. 2004;279:31277-31286.
⦁ Li L, Tam L, Liu L, Jin T, Ng DS, Biochem Cell Biol. 2011;89:515-521.
⦁ Lyssenko V, Lupi R, Marchetti P, Del Guerra S, Orho-Melander M, Almgren P, Sjogren M, Ling C, Eriksson KF, Lethagen AL, Mancarella R, Berglund G, Tuomi T, Nilsson P, Del Prato S, Groop L, The Journal of clinical investigation. 2007;117:2155-2163.
⦁ Shu L, Sauter NS, Schulthess FT, Matveyenko AV, Oberholzer J, Maedler K, Diabetes. 2008;57:645- 653.
⦁ da Silva Xavier G, Loder MK, McDonald A, Tarasov AI, Carzaniga R, Kronenberger K, Barg S, Rutter GA, Diabetes. 2009;58:894-905.
⦁ Oh KJ, Park J, Kim SS, Oh H, Choi CS, Koo SH, PLoS Genet. 2012;8:e1002986.
⦁ Prokunina-Olsson L, Welch C, Hansson O, Adhikari N, Scott LJ, Usher N, Tong M, Sprau A, Swift A, Bonnycastle LL, Erdos MR, He Z, Saxena R, Harmon B, Kotova O, Hoffman EP, Altshuler D, Groop L, Boehnke M, Collins FS, Hall JL, Hum Mol Genet. 2009;18:3795-3804.
⦁ Mondal AK, Das SK, Baldini G, Chu WS, Sharma NK, Hackney OG, Zhao J, Grant SF, Elbein SC, J Clin Endocrinol Metab. 2010;95:1450-1457.
⦁ Prokunina-Olsson L, Kaplan LM, Schadt EE, Collins FS, PloS one. 2009;4:e7231.
⦁ Kaminska D, Kuulasmaa T, Venesmaa S, Kakela P, Vaittinen M, Pulkkinen L, Paakkonen M, Gylling H, Laakso M, Pihlajamaki J, Diabetes. 2012;61:2807-2813.
⦁ Ronn T, Volkov P, Davegardh C, Dayeh T, Hall E, Olsson AH, Nilsson E, Tornberg A, Dekker Nitert M, Eriksson KF, Jones HA, Groop L, Ling C, PLoS Genet. 2013;9:e1003572.
⦁ Nilsson E, Jansson PA, Perfilyev A, Volkov P, Pedersen M, Svensson MK, Poulsen P, Ribel-Madsen R, Pedersen NL, Almgren P, Fadista J, Ronn T, Klarlund-Pedersen B, Scheele C, Vaag A, Ling C, Diabetes. 2014.
⦁ Ahlzen M, Johansson LE, Cervin C, Tornqvist H, Groop L, Ridderstrale M, Biochem Biophys Res Commun. 2008;370:49-52.
⦁ Gustafson B, Smith U, Diabetes. 2012;61:1217-1224.
⦁ Schinner S, Ulgen F, Papewalis C, Schott M, Woelk A, Vidal-Puig A, Scherbaum WA, Diabetologia. 2008;51:147-154.
⦁ Schinner S, Horm Metab Res. 2009;41:159-163.
⦁ Li AW, Morash B, Hollenberg AN, Ur E, Wilkinson M, Murphy PR, Mol Cell Endocrinol. 2001;176:57- 65.
⦁ Chiang YA, Shao W, Xu XX, Chernoff J, Jin T, Endocrinology. 2013;154:77-88.

⦁ Gustafson B, Smith U, J Biol Chem. 2010;285:14031-14041.
⦁ Ahn J, Lee H, Kim S, Ha T, Am J Physiol Cell Physiol. 2010;298:C1510-1516.
[60] Zhou Y, Rui L, Front Med. 2013;7:207-222.
⦁ Zhang Y, Proenca R, Maffei M, Barone M, Leopold L, Friedman JM, Nature. 1994;372:425-432.
⦁ Bi S, Gavrilova O, Gong DW, Mason MM, Reitman M, J Biol Chem. 1997;272:30583-30588.
⦁ Mason MM, He Y, Chen H, Quon MJ, Reitman M, Endocrinology. 1998;139:1013-1022.
⦁ He Y, Chen H, Quon MJ, Reitman M, J Biol Chem. 1995;270:28887-28891.
⦁ Morash B, Wilkinson D, Murphy P, Ur E, Wilkinson M, Mol Cell Endocrinol. 2001;185:151-159.

Figure legends

Fig.1. TCF7L2 and leptin mRNA levels in epididymal fat are patho-physiologically regulated. (A-C) C57BL/6 mice were randomly divided into 4 groups: fasted group (overnight), refed group (4 hr re-feeding after overnight fasting), PBS group (fasted mice with PBS injection) and insulin group (fasted mice with insulin injection). Epididymal fat tissue was collected for RT-PCR and quantitative PCR. For insulin and PBS groups, mice were sacrificed 30 min after insulin or PBS injection. After overnight fasting, epididymal fat tissues of three obese rodent models were also harvested for RT-PCR and quantitative PCR. (A) Relative TCF7L2 mRNA levels. (B) Relative Axin-2 mRNA levels. (C) Relative leptin mRNA levels. (D-F) Relative Leptin mRNA levels in epididymal fat tissue of three obese and insulin resistant rodent models. (G-I) Relative TCF7L2 mRNA levels in epididymal fat tissue of three obese and insulin resistant rodent models. LFD and HFD mice (D and G) were sacrificed 12 wks after the corresponding diet feeding (at age of 20 wks). db/db and db/dm mice (E and H) were sacrificed at the age of 17 wks, while fa/fa and fa/+ rats (F and I) were sacrificed at age of 13 wks. PCR primers for TCF7L2 recognize exon 16 - exon 18, which is common for TCF7L2 mRNA that encodes the major isoforms in length of 78 kD and 58 kD. All values are expressed as the mean ± SE. n = 3- 5 per group. *, p﹤0.05 when compared with fasted group (A-C) or a corresponding control group (D-I) respectively. #, p﹤0.05 when compared with the PBS group.
Fig.2. TCF7L2 and leptin mRNA expression is down-regulated by insulin at a high dosage. Rat primary adipocytes were treated with 100 nM insulin for 4 or 24 h. (A-B) Relative leptin and TCF7L2 mRNA levels. (C) Detection of S675 β-cat, β-cat, TCF7L2 and PPARγ by Western blotting. (D-G) Densitometrical analyses of panel C (n=3). All data are expressed as mean ± SE
and repeated at least three times. *, p<0.05 when compared with the corresponding control.

Fig.3. TCF7L2 and leptin expression is up-regulated by 1 nM of insulin and the up- regulation can be blocked by iCRT14. (A) Rat adipocytes were treated with 1 or 100 nM insulin for 5, 15, 30 min, followed by Western blotting with indicated antibody. The numbers on top of β-cat, β-cat (S552) and β-cat (S675) are results of densitometrical analyses, presented as fold (untreated samples designated as 1). cAMP (6-BnzcAMP, 2 nM) and forskolin (10 µM) are known positive controls for β-cat S675 and S552 phosphorylation. (B) and (C) Relative TCF7L2 and leptin mRNA levels in rat adipocytes treated with 1, 10 or 100 nM insulin for 4 h. (D) and
(E) Relative TCF7L2 and leptin mRNA levels in rat adipocytes treated with 1nM insulin, with and without iCRT14 (50 µM) for 4 h. (F) Detection of TCF7L2, p-AKT (S473) and AKT by Western blotting. All data are expressed as mean ± SE and repeated three times. *, p<0.05 when compared with the untreated control. #, p<0.05 when compared with insulin treatment. Panel F is a representative Western blotting image.
Fig.4. Wnt3a stimulates leptin expression and leptin promoter activity in rat mature adipocytes. Canonical Wnt ligand Wnt3a (2.5 nM), non-canonical Wnt ligand Wnt11 (2.5 nM) and the cat/TCF inhibitor iCRT14 (50 µM) were utilized in this study. (A) Leptin mRNA levels in rat adipocytes treated with Wnt3a or Wnt11 for 4 h. (B) Leptin mRNA levels in rat adipocytes treated with Wnt3a for 8 h. (C) Supernatant leptin levels in rat primary adipocytes treated with Wnt3a for 4 and 8 h. (D) Leptin mRNA levels in rat adipocytes treated with or without Wnt3a, in the presence and absence of iCRT14. (E) Lep-LUC (-740) reporter activity in rat adipocytes treated with Wnt3a for 4 and 8 h. All data are expressed as mean ± SE. n > or =3 for each treatment. *, P<0.05 when compared with the untreated control. #, p < 0.05 when compared with wnt3a treatment.

Fig. 5. Wnt3a stimulated leptin gene promoter activity is dependent on a CREB binding motif. (A) Illustration of Lep-LUC (-491 WT) and Lep-LUC (-491 M). (B-C) Wnt3a stimulated Lep-LUC (-491WT) but not Lep-LUC (-491 M). (D-E) Wnt3a did not activate the parental TK- LUC but Lep-TK-LUC. (F) Lep-LUC (-740) reporter activity in HTR8 cells was repressed by TCF7L2DN co-transfection. (G) Lep-LUC (-740) reporter activity in HTR8 cells was increased by S33Y β-cat co-transfection. All data are expressed as mean ±SE and repeated at least three times. *, p<0.05 when compared with the untreated control. #, p<0.05 when compared with the Wnt3a group.
Fig. 6. Detection of binding complexes formation with LepCREB by gel shift. (A) Illustration of the gel shift assay probe LepCREB and TCF. The GATC overhanging sequence was for cloning purpose in making Lep-TK-LUC in Fig. 5 and for the end-labeling with the Klenow enzyme. (B- E) Five µg nuclear proteins from indicated cell type were utilized in the shift assay with 32P- dCTP labeled LepCREB or TCF probe. (F) Two µg nuclear proteins from HTR8 or rat adipocytes (Adip) were utilized in the shift assay with the LepCREB probe. Cmpx., complex; NS, a non- specific complex or a non-specific unlabeled probe; Comp., competitors, unlabelled probe in the amount of 20 fold; S-Cmpx, super-shifted complex. Ab, antibody for super-shifting. C1 and C2, 1 and 2 µl CREB antibody, respectively; T2, 2 µl TCF7L2 antibody, G3, 2 µl unrelated control GATA3 antibody. BHK/TCF, BHK fibroblasts transfected with TCF7L2.
Fig.7. Wnt3a represses 3T3-L1 cell differentiation but stimulates leptin expression. (A) A schematic illustration of the differentiation protocol. (B) Oil Red O staining and (C) Lipid content quantification of 3T3-L1 cells treated with and without Wnt3a for 24 hr during different differentiation stages. (D-H) Relative mRNA levels of PPARγ, leptin, adiponectin, pref-1 and Axin-2 in 3T3-L1 cells treated with and without Wnt3a for 4 hr during different differentiation

stages. All data are expressed as mean ± SE and repeated three times. *, p<0.05 when compared with the corresponding control.
Fig. 8. Illustration of the effect of insulin on Wnt activation and leptin gene expression. (A) Elevated insulin level after feeding stimulates cat/TCF activity, leading to incrased leptin gene transcription. Such activation can also be achieved via canonical Wnt ligand release. This effect can be blocked by iCRT14 and TCF7L2 functional knockdown. (B) Although high dosage of insulin (HDI) can also stimulate β-cat, as it represses TCF7L2 expression, we hence observed the down-regulation of leptin gene transcription. HFD and Lep receptor (LepR) defficiency lead to increased leptin mRNA expression via mechanisms which may not directly involve Wnt signaling pathway.

Highlights

⦁ TCF7L2 and leptin expression is simultaneously regulated by feeding and insulin.

⦁ Insulin stimulates TCF7L2 expression and β-catenin Ser675 phosphorylation.

⦁ Wnt3a-stimulated leptin expression is dependent on β-catenin/TCF activity.

⦁ Wnt3a also stimulates leptin expression in differentiated 3T3-L1 cells.

⦁ The stimulation of Wnt3a on leptin promoter requires a CREB binding site.