Thy1 (CD90) controls adipogenesis by regulating activity of the Src family kinase, Fyn

Abstract

Worldwide obesity rates are at epidemic levels, and new insight into the regulation of obesity and adipogenesis are required. Thy1 (CD90), a cell surface protein with an enigmatic function, is expressed on subsets of fibroblasts and stem cells. We used a diet-induced obesity model to show that Thy1-null mice gain weight at a faster rate and gain 30% more weight than control C57BL/6 mice. During adipogenesis, Thy1 expression is lost in mouse 3T3-L1 cells. Overexpression of Thy1 blocked adipocyte formation and reduced mRNA and protein expression of an adipocyte marker, fatty acid-binding protein 4, by 5-fold. Although preadipocyte fibroblasts expressed Thy1 mRNA and protein, adipocytes from mouse and human fat tissue had almost undetectable Thy1 levels. Thy1 decreases the activity of the adipogenic transcription factor PPARγ by more than 60% as shown by PPARγ-dependent reporter assays. Using both genetic and pharmacologic approaches, we show Thy1 expression dampens PPARγ by inhibiting the activity of the Src-family kinase, Fyn. Thus, these studies reveal Thy1 blocks adipogenesis and PPARγ by inhibiting Fyn and support the idea that Thy1 is a novel therapeutic target in obesity.—Woeller, C. F., O’Loughlin, C. W., Pollock, S. J., Thatcher, T. H., Feldon, S. E., Phipps, R. P. Thy1 (CD90) controls adipogenesis by regulating activity of the Src family kinase, Fyn.

Obesity, characterized by excessive adipogenesis and/or increases in adipocyte size, has risen dramatically over the last 30 yr. In the United States alone, 60 million people are defined as clinically obese (1). Obesity dramatically increases morbidity and mortality by increasing health problems such as type 2 diabetes, cardiovascular disease, and cancer. Many human tissues can become fatty as a consequence of age, disease, or insult. Examples of this include age-related fatty bone marrow, fatty liver from infection, poor diet or excessive alcohol consumption, and increased fat accumulation leading to proptosis of the eye in thyroid eye disease as a consequence of autoimmunity (2, 3). Thus, the need to understand adipogenesis and adipocyte biology is crucial to worldwide public health.

Adipogenesis is the process by which precursor cells differentiate into mature fat cells, termed adipocytes. The process of adipogenesis is controlled by an intricate network of signaling cascades and transcription factors. The master regulator of adipogenesis is the ligand activated transcription factor peroxisome proliferator-activated receptor γ (PPARγ) (4). Adipogenesis also requires expression and activity of the transcription factors CCAAT/enhancer-binding protein α (C/EBPα) and the retinoid X receptor α (RXRα). RXRα, together with PPARγ, form a heterodimeric complex that is crucial for turning on adipogenic genes such as fatty acid-binding protein 4 (FABP4), adipocyte differentiation-related protein (ADRP), and adiponectin (AdipoQ) (4). Other transcription factors such as STAT5/5A, C/EBPα, C/EBPβ, Klf4, Krox20, and Klf15 are also integral to the adipogenic program (5). Furthermore, PPARγ ligands such as rosiglitazone, a synthetic thiazolidinedione, and 15-deoxy-prostaglandin J2, an endogenous prostaglandin, also promote adipogenesis (6). Although these factors contribute to the regulation of adipogenesis, there remains a large knowledge gap in the understanding of upstream regulators of adipogenesis. There is an urgent need to identify and understand new factors that mediate the adipogenic process.

Mesenchymal stem cells, mesenchymal progenitor cells, fibroblasts, and preadipocyte fibroblasts are key progenitor cells that can become effector cells including myofibroblasts, osteocytes, chondrocytes, and adipocytes (7, 8). Fibroblasts are heterogeneous for expression of the cell surface protein Thy1 (also called CD90) (9, 10). Thy1 is a cell surface glycophosphatidylinositol-anchored glycoprotein belonging to the immunoglobulin superfamily. In addition to being expressed on some fibroblasts, Thy1 is variably expressed on other cells such as mesenchymal stem cells, rodent thymocytes, and T cells (11). Although Thy1 was discovered almost 40 yr ago, its function remains poorly understood. Thy1 plays a role in regulating T cell function during the mouse cutaneous immune response and in the T cell receptor response (12). Thy1 interacts with Fyn, Lyn, and Lck (Src family kinases, SFK) in T cells to mediate signal transduction pathways in the immune response (13). Thy1 also binds to integrin proteins on the surface of certain cells and thus is thought to be involved in cell–cell adhesion (14). The expression of Thy1 on subsets of fibroblasts is important in determining cell fate. Thy1-positive fibroblasts do not readily form adipocytes, unlike Thy1-negative fibroblasts (15). Thy1 may be a marker of certain preadipocytes or may regulate the adipogenic potential of fibroblasts and adipocyte progenitor cells. Regardless, the true function of Thy1 remains a mystery.

Recently, Fyn, an SFK that interacts with Thy1, was shown to be involved in fat tissue metabolism (16, 17). Specifically, Fyn-null mice display reduced free triglycerides, reduced lipid accumulation, and a reduced percentage of adipose tissue mass (17). Fyn-null mice also express 5-fold lower levels of phosphoenolpyruvate carboxykinase (PEPCK), a key adipogenic gene that is induced by PPARγ (18), suggesting that Fyn-null mice have lower PPARγ activity than wild-type animals. Although these studies demonstrate a clear link between Fyn, PPARγ, and adipose tissue, to our knowledge, there is no literature describing a role for Thy1 in adipose tissue accumulation and obesity.

In our study, we asked whether Thy1 plays a fundamental role in controlling adipogenesis. Our novel and interesting findings reveal the seminal importance of Thy1 in regulating adipogenesis in mouse and human cells. Thy1 may offer an important new target in the treatment of obesity or other diseases involving excess fat accumulation.

MATERIALS AND METHODS

Animals

Thy1-null mice (Thy1^−/−) were from Dr. K. Hayakawa (19, 20). Thy1-null or control C57BL/6 mice were housed in colony cages with a 12 h light/12 h dark cycle and fed ad libitum. During the study, 8 wk old male mice were fed with a control diet consisting of 10% kcal/fat (D12450B; Research Diets, New Brunswick, NJ, USA) or a high-fat diet consisting of 60% kcal/fat (D12492; Research Diets). All animal experiments were performed with approval of the University Committee on Animal Resources at the University of Rochester School of Medicine and Dentistry.

Cell culture

3T3-L1 preadipocytes and primary human preadipocyte fibroblasts were acquired and cultured as previously described (21). All media and supplements were purchased from Gibco (Carlsbad, CA, USA). MEFs were isolated as described previously (22). Initial adipogenic medium contained 0.5 mM 3-isobutyl-1-methylxanthine, 0.25 μM dexamethasone, and 1 μg/ml insulin (Sigma-Aldrich, St. Louis, MO, USA) and 2 μM rosiglitazone (Cayman, Ann Arbor, MI, USA) for 2 d. After 2 d, fresh adipogenic medium lacking IBMX was added for 6 to 8 d more. Cells were analyzed for lipid accumulation and adipogenesis by Oil Red O (Sigma-Aldrich) staining and AdipoRed (Cayman) staining as previously described (21).

Isolation of adipocytes from human orbit, eyelid, and abdominal fat tissue

Orbital tissue was obtained from patients undergoing orbital decompression surgery. The tissue was procured following a protocol that was approved by the University of Rochester Research Subjects Review Board, which required informed, written consent to be obtained from all patients. Eyelid and abdominal fat surgical waster was obtained from the Surgical Pathology Clinical Laboratory after unique patient identifiers were removed. All fat tissue was washed in ice-cold PBS and cut into 2 to 3 mm pieces. Vascular and stromal portions were removed. The remaining adipocytes were flash frozen in liquid nitrogen and homogenized by mortar and pestle in a buffer containing 2% SDS and 60 mM Tris, pH 6.8.

Knockdown of gene expression by siRNA

Cells were seeded at 5 to 8 × 10⁴ or 1 × 10⁵ into 24-well or 6-well cell culture plates (Corning, Corning, NY, USA), respectively, and subsequently transiently transfected with 50 to 200 nM of siRNA (Ambion, Foster City, CA, USA) using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) following the manufacturer’s protocol. Cells were incubated with siRNA for 3 to 4 d before being subjected to adipogenic differentiation experiments as described above. At the start of differentiation, cells were transfected again with 100 nM siRNA.

Transient transfection by electroporation

A total of 2 × 10⁶ cells were collected and resuspended in 100 μl of Ingenio electroporation solution (Mirus Bio, Madison, WI, USA) containing appropriate DNA complexes and electroporated with program U-023 on an Amaxa Nucleofector instrument (Lonza, Cologne, Germany) following the instructions from the manufacturer.

Quantitative real-time PCR

Total RNA was isolated using the RNeasy kit (Qiagen, Valencia, CA, USA) following the manufacturer’s protocol, treated with DNase (Qiagen), and assessed for purity by spectroscopy. A total of 100 ng total RNA was used with iScript reverse transcriptase (Bio-Rad, Hercules, CA, USA) to generate cDNA. Quantitative real-time (qPCR) reactions were performed using SYBR Green reagent. Relative gene expression was normalized to reference gene expression as described previously (21). Primers used in this study are listed in Table 1.

TABLE 1.

Primer sets

Gene	Sequence
FABP4	ACAGGAAAGTCAAGAGCACC
	AACTTCAGTCCAGGTCAACG
THY1	ATCTCCTCCCAGAACGTC
	ATCTCTGCACTGGAACTTG
18S rRNA	TGAGAAACGGCTACCACATC
	ACTACGAGCTTTTTAACTGC
ACTB	GCACAGAGCCTCGCCTT
	CCTTGCACATGCCGGAG
Mu Pparγ2	CATGGTTGACACAGAGATGC
	CTTGCATCCTTCACAAGCATG
Mu Cebpa	CTGAGAGCTCCTTGGTCAAG
	GAATCTCCTAGTCCTGGCTTG
Mu Resn	CAAACAAGACTTCAACTCCCTG
	TCTATCCTTGCACACTGGC
Mu Fabp4	ATGTGTGATGCCTTTGTGGGAAC
	TCATGTTGGGCTTGGCCATG
Mu AdipoQ	TGGAGAGAAGGGAGAGAAAGG
	TGAGCGATACACATAAGCGG
Mu Leptin	CTGTGGCTTTGGTCCTATCTG
	TGAAGCCCAGGAATGAAGTC
Mu Plin2 (Adrp)	CACAAATTG CGGTTG CCA AT
	ACTGGC AAC AATCTCGGACGT
Mu Rn18s	GTAACCCGTTGAACCCCATT
	CCATCCAATCGGTAGTAGCG
Mu Actb	GAGGGAAATCGTGCGTGACAT
	ACATCTGCTGGAAGGTGGACA
Mu Gapdh	AGCCTCGTCCCGTAGACAAA
	CCTTGACTGTGCCGTTGAAT
Mu Thy1	CCTTACCCTAGCCAACTTCAC
	AGGATGTGTTCTGAACCAGC

Western blot analysis

Protein was isolated from 1 to 2 × 10⁶ cells and lysed in 60 mM Tris, pH 6.8, 2% SDS, and containing 1× protease inhibitor cocktail (Sigma-Aldrich). Samples were treated with 2× SDS loading buffer containing β-mercaptoethanol before use. Total protein (1 to 10 μg per lane) was subjected to SDS-PAGE. Protein gels were transferred to PVDF membrane (Millipore, Billerica, MA, USA) and probed with antibodies as specified. The Thy1 antibodies used for Western blot analyses in this study were from Sigma-Aldrich (human) and Cell Signaling Technologies (human and mouse; Danvers, MA, USA).

Flow cytometry

Cells were collected by trypsinization and washed in PBS before blocking nonspecific binding with 5% normal mouse serum. Blocked cells were incubated with either anti-mouse Thy1.2-PE conjugated antibody (BD Biosciences, San Jose, CA, USA) or anti-human Thy1-PE conjugated antibody (BD Biosciences). Cells were run on a FACS Canto II flow cytometer (BD Biosciences).

Generation of constitutively active Fyn mutant

The pRK5-Fyn wild-type plasmid was amplified by PCR using KOD polymerase (Novagen, Merck EMD, Darmstadt, Germany) using the following primers: FynY531F 5′-CGACAGAGCCCCAGTTCCAACCTGGTGAAA-AC-3′ and FynY531F- 5′-GTTTTCACCAGGTTGGAACTGGGGCTCTGTCG-3′, where the underlined nucleotides in each primer indicate a codon change from tyrosine to phenylalanine. The mutation was verified by DNA sequencing.

Luciferase reporter assays

A PPAR response element (PPRE) firefly luciferase reporter construct containing 3 copies of a PPRE (PPRE × 3-firefly luciferase) (6) and an SV40 Renilla luciferase construct (Promega, Madison, WI, USA) were introduced into cells by electroporation. Cells were treated with DMSO (vehicle) or 100 nM rosiglitazone for 16 h. Luciferase activity was then measured using Dual-Glo Luciferase Assay buffer (Promega) read on a Varioskan Flash luminescent plate reader (Thermo Fisher Scientific, Waltham, MA, USA). Luciferase readings were normalized to the control, vehicle-treated samples for statistical analysis.

Statistical analysis

Student’s t test and 1-way analysis of variance were used for statistical analysis; P values of 0.05, 0.01, and 0.001 were considered significant.

RESULTS

Thy1-null mice display increased adipogenesis when given a high-fat diet

Previous work from our laboratory demonstrated that Thy1-negative, but not Thy1-positive, human fibroblasts could be driven to adipocytes when stimulated by a PPARγ ligand (15). To test whether or not Thy1 plays a seminal role in adipogenesis and potentially obesity in vivo, male Thy1-null mice (19, 20) and male control C57BL/6 mice were fed a high-fat diet to determine if they had altered weight gain, fat accumulation, or changes in key adipogenic genes. Mice were fed ad libitum for 7 wk on either a high-fat diet (60% kcal/fat) or a control 10% kcal/fat diet and weighed weekly. Thy1-null mice on the high-fat diet gained more weight than control mice (Fig. 1A, B). Tissue and serum were collected at the end of the experiment for analysis of key metabolic markers. Thy1-null mice had greater than 2-fold higher serum levels of resistin, a key factor implicated in obesity and insulin resistance, compared to control mice on a high-fat diet (Fig. 1C). Interestingly, resistin expression was unaffected by genotype when mice were fed a control diet. White adipose tissue (WAT) was analyzed for the expression of key adipogenic genes by qPCR. Thy1-null mice expressed higher levels of PPARγ, C/EBPα, resistin, FABP4, and AdipoQ mRNAs compared to control mice when given a high-fat diet (Fig. 1D). The mRNA expression of leptin, an adipokine involved in regulating obesity (23), was moderately decreased in Thy1-null mice compared to control. Subsequent experiments were performed using female mice (both Thy1-null and control mice) and revealed that Thy1 female mice also gain more weight than control mice when given a high-fat diet (data not shown). As expected, the kinetics of weight gain were different in male and female mice, suggesting that sex specific endocrine factors play a key role in high-fat-diet-induced weight gain.

Figure 1.

Thy1 knockout (KO) mice gain more weight and have increased levels of adipogenic genes compared with control mice on a high-fat diet. A) C57BL/6 mice or Thy1 KO mice were fed a control diet (10% kcal/fat) or a high-fat diet (60% kcal/fat) for 7 wk and weights were measured weekly (n = 5 per group). The graph presents weight gain as percentage of starting weight (100%). B) Thy1 KO mice gain more weight than control mice on a high-fat diet. C) Serum was collected from C57BL/6 mice or Thy1 KO mice and analyzed by ELISA for resistin levels. Thy1 KO mice have dramatically higher serum levels of resistin compared to control mice on a high-fat diet. D) WAT extracts from animals fed a high-fat diet were analyzed for expression of various adipogenic genes by qPCR. On a high-fat diet, Thy1 KO mice express higher mRNA levels of the adipogenic genes PPARγ2, CEBPα, resistin, FABP4, and AdipoQ and lower levels of leptin mRNA. All data are means ± sem. *P < 0.05, **P < 0.01.

Thy1 expression is lost during adipogenesis

The ability of Thy1-null mouse embryonic fibroblasts (MEFs) to undergo preadipocyte differentiation at a greater rate than MEFs with Thy1 was tested. Thy1-null (−/−) and Thy1 heterozygous (+/−) MEF clones from the same litter were subjected to preadipocyte differentiation. After 11 d, cells were analyzed by Western blot test for FABP4, Thy1, and β-tubulin protein levels (Fig. 2A). Culture supernatants were also collected to measure resistin levels. Thy1-null MEFs expressed 2-fold higher levels of FABP4 protein than MEFs expressing Thy1 (Fig. 2A, B). Interestingly, Thy1-null MEFs produced 4-fold more resistin than Thy1 expressing MEFs, suggesting that adipocytes are a source of resistin in our studies (Fig. 2C). Another key finding was that Thy1 levels decreased by 60% during preadipocyte differentiation (Fig. 2A, D).

Figure 2.

Loss of Thy1 increases adipogenesis. A) Thy1, FABP4, and β-tubulin were analyzed by Western blot analysis after Thy1 knockout (KO) (−/−) and Thy1 (+/−) MEF clones (n = 3 each) were induced to differentiate into adipocytes with adipogenic medium (AM). B) FABP4 expression was 2-fold higher in Thy1 KO MEFs compared to Thy1-expressing MEFs. C) Resistin protein levels present in cell culture supernatant from cells treated as in (B). Thy1 KO cells produce 4-fold more resistin than Thy1-expressing cells. D) Thy1 protein expression, as quantified from (A), was reduced in Thy1 expressing MEFs by treatment with adipogenic medium. E) Thy1 mRNA levels were analyzed by qPCR from mouse WAT that was digested and separated into SVF and adipocyte fractions. F) Thy1 protein levels were analyzed by Western blot analysis from mouse SVF and adipocyte fractions. Thy1 is present in SVF but not adipocytes. G) Human Thy1 mRNA levels were measured by qPCR from human orbital stromal fibroblasts, sorted Thy1-positive and -negative fibroblasts, and adipocytes from orbital, eyelid, and visceral fat depots. H) Thy1 protein levels were analyzed by Western blot analysis from orbital stromal fibroblasts and orbital adipocytes. Thy1 is present in human fibroblasts but not in human adipocytes. All data are means ± sem. *P < 0.05, **P < 0.01.

Because Thy1 expression was decreased during preadipocyte differentiation in MEFs, Thy1 expression was analyzed in mature adipocytes from mouse and human fat tissues. WAT from normal mice was digested into a stromal vascular fraction (SVF) and an adipocyte fraction using collagenase as described previously (24). Thy1 mRNA and protein were detected in the SVF, but not in the adipocyte fraction (Fig. 2E, F). Because the SVF contains preadipocyte fibroblasts, monocytes, and lymphocytes, a fraction of SVF cells was cultured to remove nonadherent cells. The remaining adherent cells, which resembled preadipocyte cells, expressed Thy1, thus indicating that preadipocyte cells present in adipose tissue are Thy1 positive.

Thy1 mRNA and protein expression was also analyzed in human fat tissues. Human fat depots were chosen to be representative of WAT depots and to test a broad range of human tissues. Human adipocytes from orbital, eyelid, and visceral fat depots did not express Thy1 mRNA, whereas fibroblasts and SVFs expressed Thy1 mRNA (Fig. 2G). Human orbital adipocytes and the remaining SVF were analyzed for Thy1 protein. Human orbital adipocytes did not contain detectable Thy1 protein, whereas Thy1 was present in the SVF (Fig. 2H).

An investigation of Thy1 function in the well-established mouse 3T3-L1 preadipocyte cell line was also performed (25). Confluent 3T3-L1 cells were treated with a standard adipogenic medium for 8 d (26). Oil Red O staining showed that a majority of cells had differentiated into adipocytes (Fig. 3A). Cells were sampled throughout the experiment to measure protein and mRNA levels of Thy1 and key adipogenic markers. As expected, ADRP and FABP4 mRNA levels were markedly increased during adipogenesis (Fig. 3B). Interestingly, 3T3-L1 preadipocytes expressed Thy1 mRNA (Fig. 3B). However, 1 d after adipogenic treatment, Thy1 mRNA levels decreased to 30% of preadipocyte levels. Thy1 mRNA levels continued to decrease until Thy1 was reduced to less than 2% of original levels by d 7. FABP4 and Thy1 protein levels were also measured during 3T3-L1 preadipocyte differentiation (Fig. 3C). FABP4 protein expression was induced during adipogenesis, whereas Thy1 protein levels were reduced to undetectable levels by d 7.

Figure 3.

Thy1 expression is lost during adipogenesis of 3T3-L1 preadipocytes. A) Oil Red O analysis of 3T3-L1 preadipocytes grown to confluence and induced to differentiate into adipocytes. Cells were analyzed by Oil Red O on d 0, 4, and 8. B) Thy1, FABP4, and ADRP mRNA levels were analyzed by qPCR from 3T3-L1 cells undergoing adipogenesis. Samples of cells were taken on d 0, 1, 3, 5, and 7 of the experiment. C) Thy1, FABP4, and β-tubulin protein levels were analyzed by Western blot analysis from 3T3-L1 cells undergoing adipogenesis. Samples of the cells were taken on d 0, 1, 3, 5, 6, 7, and 8 of the experiment. The experiment was repeated 3 times in duplicate, with 1 representative experiment shown. All data are means ± sem.

Thy1 expression blocks adipogenesis

The ability of ectopic Thy1 expression to impair preadipocyte differentiation of 3T3-L1 cells was tested. Thy1 cDNA was introduced into 3T3-L1 preadipocytes by electroporation. Cells were then induced to differentiate into adipocytes. There was a dramatic reduction in lipid droplet number and size as shown by Oil Red O staining in 3T3-L1 cells with ectopic Thy1 expression (Fig. 4A). Lipid accumulation decreased by approximately 70% in Thy1-expressing cells compared to control (Fig. 4B). Cells ectopically expressing Thy1 show more than a 4-fold reduction in FABP4 mRNA levels and 3-fold reduction in ADRP levels (Fig. 4C). Likewise, FABP4 protein expression was dramatically reduced in cells expressing Thy1 (Fig. 4D).

Figure 4.

Ectopic expression of Thy1 inhibits adipogenesis in 3T3-L1 cells. A) Oil Red O analysis of 3T3-L1 cells either ectopically expressing Thy1 or not that were induced to differentiate into adipocytes. B) Lipid accumulation in 3T3-L1 cells, treated as in (A), was measured using the fluorescent AdipoRed assay. Ectopic expression of Thy1 leads to a dramatic reduction in lipid accumulation. C) FABP4 and ADRP mRNA levels were analyzed by qPCR on adipogenic d 7. Cells ectopically expressing Thy1 express lower levels of the adipogenic genes FABP4 and ADRP. D) Thy1, FABP4, and β-tubulin were analyzed by Western blot analysis on adipogenic d 7. FABP4 expression was markedly inhibited in 3T3-L1 cells ectopically expressing Thy1. The experiment was repeated 3 times in duplicate, with 1 representative experiment shown. All data are means ± sem. *P < 0.05, **P < 0.01.

Because Thy1 expression was lost during adipogenesis of 3T3-L1 cells and forced expression of Thy1 impaired 3T3-L1 adipogenesis, the role of Thy1 in primary human preadipocyte differentiation was evaluated. Control or Thy1-specific siRNA was introduced into human preadipocyte fibroblasts, and the expression of Thy1 was measured by flow cytometry (surface expression) and Western blot analysis (total expression) to verify knockdown (Fig. 5A). Cells were treated with adipogenic medium for 8 d. Samples were visually inspected by microscopy and by Oil Red O staining (Fig. 5B). Thy1 siRNA-treated cells contained more lipid droplets than control. The adipored lipid quantitation assay revealed that lipid accumulation increased by 1.6-fold in Thy1 siRNA-treated cells compared to control cells (Fig. 5C). Thy1 siRNA treatment increased FABP4 mRNA levels by 2.3-fold and FABP4 protein by 1.8-fold (Fig. 5D–F).

Figure 5.

Depletion of Thy1 using siRNA increases adipogenesis and expression of the adipogenic marker FABP4 in human preadipocyte fibroblasts. A) Thy1-specific or control siRNA was introduced into human preadipocytes by lipofection. Total Thy1 levels were analyzed by Western blot analysis, and surface Thy1 levels were measured by flow cytometry. B) Photomicrographs and Oil Red O staining show lipid droplet accumulation in human fibroblasts treated with control or Thy1 siRNA. C) Lipid accumulation was measured using the AdipoRed assay. Cells treated with Thy1 siRNA accumulate more lipid than control cells. D) FABP4 mRNA levels were measured by qPCR. E) FABP4 and β-tubulin expression were analyzed by Western blot analysis. F) Quantification of FABP4 protein levels from Western blot analysis. FABP4 mRNA and protein levels are increased in cells treated with Thy1 siRNA compared to control. Experiments were performed in at least 2 different strains, and results shown are from a representative strain repeated in triplicate. *P < 0.05.

Thy1 inhibits PPARγ activity by blocking Fyn

Next we tested whether Thy1 affects the activity of the key adipogenic transcription factor PPARγ. A PPARγ-dependent firefly luciferase reporter gene (PPREx3-luc) (6), the pSV40-Renilla reporter gene, and either the Thy1 expression plasmid or control plasmid were introduced into 3T3-L1 adipocytes. After verifying Thy1 expression by flow cytometry and Western blot analysis (Fig. 6A), PPARγ activity was measured and normalized to Renilla activity (Fig. 6A, right). There was a 60% reduction in PPARγ activity in cells ectopically expressing Thy1 compared to control. The ability of Thy1 to regulate PPARγ activity in human HEK293FT cells was also tested (Fig. 6B). Because HEK293FT cells do not express endogenous PPARγ protein, a PPARγ expression plasmid was also introduced (27). Importantly, a transfection efficiency of more than 90% was obtained in HEK293FT cells, as demonstrated through a GFP reporter plasmid. Flow cytometry and Western blot analysis revealed robust expression of Thy1 in HEK293FT cells (Fig. 6B, left). Thy1 reduced PPARγ activity by 30%. In the presence of rosiglitazone, a PPARγ ligand, Thy1 expression reduced PPARγ activity by 40% compared to control. Primary human preadipocyte fibroblasts that were transduced with control or Thy1-encoding lentivirus were also tested for PPARγ activity. Thy1 lentivirus transduced cells express 4-fold more Thy1 than control cells (Fig. 6C, left). Thy1 expression reduced the activity of PPARγ by approximately 50% in human fibroblasts (Fig. 6C, right).

Figure 6.

Thy1 inhibits activity of the master adipogenic regulator, PPARγ. A) Thy1 cDNA or a control plasmid along with the PPRE × 3-luc and control SV40-Renilla reporter plasmids were introduced into 3T3-L1 cells via electroporation. Total expression of Thy1 was confirmed by Western blot analysis, and surface expression was confirmed by flow cytometry (left). PPRE-luciferase activity was measured and normalized to Renilla activity (right). Introduction of Thy1 into 3T3-L1 cells reduces PPARγ activity by 60% compared to control. B) HEK293FT cells were treated as in (A) with the addition of a plasmid encoding PPARγ (pcDNA3.1-PPARγ). Cells were treated with vehicle (DMSO) or 100 nM rosiglitazone (Rosi). In both treatments, Thy1 expression reduces PPARγ activity. C) Human preadipocyte fibroblasts were treated as in (A) with DMSO or 100 nM rosiglitazone. D) Reporter plasmids were introduced into human preadipocyte fibroblasts along with a control plasmid or plasmids that express either wild-type Fyn or dominant negative Fyn K299M. Ectopic expression of Fyn results in a 4.5-fold increase in PPARγ transcriptional activity. Expression of Fyn K299M results in a more than 70% reduction in PPARγ transcriptional activity compared to control. E) Reporter constructs were introduced as in (D), along with either a control plasmid (columns 1 and 2) or plasmids expressing Thy1 (columns 3 to 6), Fyn (column 4), or a constitutively active Fyn Y531F (column 5). Thy1 expression reduced PPARγ activity by 50% (column 2 vs. 3), while introduction of Fyn restored approximately 40% of control PPARγ activity (column 2 vs. 4). Introduction of Fyn-Y531F restored PPARγ activity to over 100% of control cells (column 2 vs. 5). Experiments were performed in 2 different strains, and results are from a representative strain repeated in triplicate. Results are normalized firefly luciferase activity ± sem. *P < 0.01, **P < 0.001.

The ability of Thy1 to regulate adipogenesis and PPARγ activity through Fyn was investigated because Thy1 interacts and regulates the activity Fyn in certain cell types (28, 29). A Fyn expression plasmid was introduced into human preadipocyte fibroblasts followed by monitoring PPARγ activity via the luciferase reporter systems. Fyn expression increased the transcriptional activity of PPARγ more than 4-fold (Fig. 6D). To determine whether the increase in PPARγ reporter activity was due to Fyn activity, a dominant-negative Fyn expression plasmid (pRK5-Fyn K299M) that disrupts endogenous SFK activity was used (30). Introduction of Fyn K299M dramatically reduced PPARγ activity in human preadipocytes (Fig. 6D). To determine whether Thy1 prevents Fyn activity from regulating PPARγ, another mutant Fyn expression plasmid (pRK5-Fyn Y531F) was constructed. Fyn Y531F mutant protein is not subject to regulation by Y531 phosphorylation and is thus constitutively active. Thy1, wild-type Fyn or mutant Fyn Y531F was introduced into human preadipocytes and PPARγ activity monitored in the presence of rosiglitazone (Fig. 6E). Wild-type Fyn expression increased PPARγ activity by more than 30%. Remarkably, mutant Fyn Y531F expression increased PPARγ activity more than 2-fold in the presence of Thy1, restoring PPARγ activity (Fig. 6E).

To further pursue the function of endogenous Thy1 and Fyn in adipogenesis, the selective SFK and Fyn inhibitor SU6656 (31) was used. The addition of 10 μM SU6656 to adipogenic medium reduced expression of FABP4 in human preadipocytes to 30% of control (Fig. 7A). SU6656 treatment inhibited lipid accumulation in human fibroblasts (Fig. 7B). To test if Thy1 inhibits adipogenesis in human preadipocytes through Fyn, Thy1 was depleted using Thy1 siRNA followed by preadipocyte differentiation in the presence or absence of SU6656. SU6656 did not have a significant effect on Thy1 mRNA levels, indicating that Fyn does not regulate Thy1 levels. However, as expected, Thy1 mRNA levels decreased by over 85% in Thy1 siRNA-treated cells compared to control siRNA-treated cells (Fig. 7C). Furthermore, depletion of Thy1 increased FABP4 mRNA expression by 2.3-fold compared with control (Fig. 7D). Interestingly, SU6656 attenuated the effects of Thy1 depletion on FABP4 levels (Fig. 7D). SU6656 treatment also attenuated increases in lipid accumulation mediated by Thy1 depletion (Fig. 7E). Taken together, these results show that Thy1 regulates PPARγ activity and preadipocyte differentiation through inhibition of Fyn (Fig. 7E).

Figure 7.

Thy1 regulates adipogenesis by inhibiting the activity of Fyn. A) FABP4 and β-tubulin were analyzed by Western blot analysis from human preadipocyte fibroblasts treated with vehicle, adipogenic medium, or adipogenic medium plus 10 μM SU6656. SU6656 treatment decreased FABP4 protein expression. B) Lipid accumulation from cells treated as in (A) was measured using AdipoRed assay. SU6656 treatment decreased lipid accumulation. C) Thy1-specific or control siRNA was introduced into human fibroblasts by lipofection. Fibroblasts were then treated with an adipogenic medium with or without 10 μM SU6656 for 8 d. Thy1 and FABP4 mRNA expression were analyzed by qPCR. Thy1 mRNA levels in Thy1 siRNA-treated cells were reduced to less than 10% of control siRNA-treated cells. D) FABP4 mRNA expression in cells treated as in (C) was analyzed by qPCR. FABP4 increased 4-fold in Thy1 siRNA treated cells compared to control. SU6656 treatment ablated the effect of Thy1 siRNA on FABP4 expression. E) Lipid accumulation was measured in cells treated as in (C) using AdipoRed assay. Lipid accumulation was increased in cells treated with Thy1 siRNA, and the effect was blunted with SU6656. F) Model describing the role of Thy1 in regulating adipogenesis. Thy1 inhibits the activity of Fyn, which leads to lower PPARγ activity and impaired adipogenesis. *P < 0.01, **P < 0.001.

DISCUSSION

The obesity epidemic is rapidly becoming the most dire health problem in the world as it increases the incidence of type 2 diabetes, hypertension, and cardiovascular disease. A better molecular understanding of obesity and adipogenesis is required to bring about new treatment options. Here, we show that Thy1 plays a fundamental role in regulating mouse and human adipogenesis. Thy1-null mice gain more weight and have increased expression of adipogenic genes compared to control mice. We also used a well-characterized model of adipogenesis, the 3T3-L1 preadipocyte cell line, to show that Thy1 expression inhibits adipogenesis. Interestingly, undifferentiated 3T3-L1 preadipocytes express Thy1, but Thy1 is rapidly lost upon induction of adipogenesis. Fully differentiated 3T3-L1 adipocytes, as well as primary adipocytes isolated directly from mouse and human fat tissue, do not express Thy1. Likewise, MEFs, which normally express Thy1 under standard conditions, lose Thy1 when undergoing adipogenesis. Furthermore, depletion of endogenous Thy1 in human fibroblasts increases their ability undergo adipogenesis. Thus, these data highlight a key new role for Thy1 function in controlling adipogenesis and show that loss of Thy1 expression promotes the adipogenic pathway and obesity.

The role of Thy1 in adipocyte biology and obesity has not been previously recognized. Our discovery that Thy1-null mice have increased adipogenesis and weight gain on a high-fat diet represents a novel and unexpected discovery. Thy1-null mice were originally developed more than 15 yr ago. Because we show that Thy1 is not expressed by mature adipocytes, it is likely that Thy1 expression is lost during preadipocyte differentiation in vivo. This is the case for MEFs, 3T3-L1 cells, and human preadipocyte fibroblasts before adipogenic commitment when they are heterogeneous for Thy1 expression (15).

Our animal studies show that Thy1-null mice have increased weight gain, increased levels of key adipogenic genes, and increased serum resistin levels. Resistin levels are closely associated with obesity, altered adipose tissue metabolism, and resistance to insulin (32). Although the primary source of resistin production appears to be macrophages in human and adipocytes in mice, the association between resistin, diabetes, obesity, and inflammation is strong (33, 34). In our studies, a significant amount of resistin was produced by adipocytes derived from MEF cultures. Importantly, Thy1-null mice fed a high-fat diet have increased serum resistin levels and increased resistin mRNA expression. Thus, our data reveal an additional link between Thy1 levels, obesity, and insulin resistance. Future in vivo studies aimed at reversing the increased weight gain of Thy1 knockout animals by testing various PPARγ and Fyn antagonists such as GW9662 and SU6656, respectively, are needed. Although SU6656 was shown to decrease adiposity in wild-type mice (16), it will be interesting to see if the phenotype of Thy1-null mice on a high-fat diet can be reversed in a dose dependent manner by this Fyn inhibitor. Leptin mRNA expression, however, is decreased in Thy1-null mice compared to control. Leptin, an adipokine, is one of the most important molecules secreted by adipocytes because it serves as an appetite suppressor. Although leptin is normally elevated in obesity, there appears to be leptin resistance pathways that circumvent the negative feedback loop mediated by leptin (35). The fact that obese Thy1-null mice do not have higher levels of leptin suggest that they are still sensitive to leptin, and it may be interesting to determine whether exogenous leptin can attenuate accelerated weight gain of Thy1-null mice. Further studies aimed at investigating a potential link between Thy1 and leptin are needed before this can be resolved.

Thy1 expression in preadipocyte cells appears to follow the expression pattern of Pref-1 (36, 37). Pref-1 is a cell membrane–bound molecule expressed on preadipocytes and mesenchymal stem cells (37). Membrane bound Pref-1 is cleaved to form a soluble mediator that induces MAPK activity, ultimately leading to a block of C/EBPα and C/EBPβ expression that prevents induction of the adipogenic cascade (38). Thy1 also has the potential to be cleaved at its glycophosphatidylinositol anchor by phospholipase enzymes (39, 40). However, membrane cleavage of Thy1 affects its localization and functional properties, and it is thus unclear if cleaved Thy1 could control adipogenesis (41). Recently elevated levels of a soluble form of Thy1 have been associated with scleroderma, a fibrotic disease (42). It may be interesting to determine if soluble Thy1 levels are decreased in obesity and thus serve as a novel biomarker.

Although Thy1 was discovered almost 40 yr ago, its molecular functions remain elusive. Our new results demonstrate a mechanism whereby Thy1 regulates adipogenesis through Fyn. Thy1 can localize with and regulate Fyn and other SFKs in cells such as T cells and fibroblasts. Fyn is an N-myristolated protein that localizes to the plasma membrane and has been implicated in insulin signaling (43). Fyn interacts with numerous signaling molecules and has diverse biologic functions in T cell activation, cell adhesion signaling, and brain function. Fyn involvement in integrin signaling involves activation of Ras (44). In T cells, after T cell receptor activation, Thy1 knockout mice display increased SFK activity compared to wild-type mice, suggesting that Thy1 can inhibit SFKs (13). Furthermore, in MEFs, expression of Thy1 prevents SFK activation by TNF-α (29). These results are consistent with our data showing that Thy1 inhibits Fyn activity to limit PPARγ activity and adipogenesis in human preadipocyte fibroblasts.

Interestingly, a study by Varisco et al. (45) showed that Thy1 could increase PPARγ signaling and lipid accumulation in mouse lung fibroblasts. Lung fibroblasts are undoubtedly distinct from adipose progenitor cells, and thus, while different results were obtained, this may be resolved by noting cell origin, which has been shown to be crucial in phenotype observations. Additionally, while lung fibroblasts can accumulate lipid droplets, these cells are not classified as adipocytes, and lung tissue itself does not get fat. Furthermore, Cohen et al. (28) describe a pathway in which Thy1 promotes Fyn signaling in mouse lung myofibroblasts, which would be expected to increase PPARγ signaling and lipid accumulation in lung fibroblasts. Because Thy1 expression was lost during adipogenesis of 3T3-L1 preadipocytes, and because Thy1 inhibited PPARγ activity in a Fyn-dependent manner in human fibroblasts, our data reveal that Thy1 plays an inhibitory role in Fyn activation during preadipocyte differentiation. Therefore, because Thy1 appears to inhibit SFK activity in preadipocytes, MEFs, and T cells but promotes SFK activity in lung fibroblasts, Thy1 and Fyn cross-talk are regulated in a cell-specific context.

Fyn has been shown to play a role in promoting lipid accumulation and adipogenesis. Fyn-null mice display reduced adipose tissue mass and increased levels of fatty acid oxidation (46). Additionally, treatment of mice with the Fyn inhibitor SU6656 reduced adipose tissue accumulation (47). Our studies using human fibroblasts treated with SU6656 also led to an inhibition of lipid accumulation and adipogenesis. Furthermore, we discovered that SU6656 attenuated the proadipogenic properties of Thy1 siRNA, demonstrating that Thy1 inhibits Fyn to block adipocyte formation. Notably, we observed no changes in Thy1 expression when Fyn was inhibited, suggesting that Fyn does not regulate Thy1 expression.

Our new findings show that Thy1 inhibits PPARγ activity in 3T3-L1 preadipocytes, HEK293FT cells, and human preadipocyte fibroblasts. Because PPARγ is the master regulator of adipogenesis, its expression and activity are subject to numerous levels of regulation. Importantly, we saw no change in PPARγ expression when we modified Thy1 levels in these experiments, suggesting that another form of PPARγ regulation is involved. PPARγ activity is regulated by posttranslational modifications such as sumoylation and phosphorylation, availability of coactivators and ligands, and subcellular localization (4). Because constitutively active Fyn can increase the activity of PPARγ even in the presence of Thy1, this suggests a signal transduction cascade involving phosphorylation. Phosphorylation of PPARγ by extracellular signal–regulated kinase (Erk) 1/2 dramatically reduces the transcription factors activity (48). In T cells, Thy1 signaling increases Erk1/2 activity (49) and thus it is possible that Thy1 promotes Erk1/2 activity in preadipocyte cells to limit PPARγ function. PPARγ can also be phosphorylated at serine 273 by the cyclin-dependent kinase Cdk5. Phosphorylation at serine 273 is associated with obesity and insulin resistance (50). Fyn and Cdk5 are involved in a signal transduction pathway that mediates neuronal guidance (51), and it is interesting to speculate whether a similar complex is involved in insulin resistance and obesity in part through modification of PPARγ. Interestingly, Fyn-null mice have a 5-fold lower level of expression of PEPCK, a PPARγ-dependent gene, compared to wild-type mice (18). These data further suggest a link between Fyn signaling and PPARγ activity.

Because loss of Thy1 appears to be important for preadipocyte differentiation and loss of Thy1 promotes weight gain in mice fed a high-fat diet, it is interesting to speculate how Thy1 expression is regulated at the cellular level in adipocyte precursor cells. There is a knowledge gap in how Thy1 expression is modulated. Reports reveal that Thy1 expression can be affected by epigenetics, specifically through DNA methylation of the Thy1 promoter region (52). It is possible that Thy1 promoter methylation may be altered in obesity or other diseases that involve excess fat accumulation. Thy1 expression is also induced by phorbol dibutryate in endothelial cells (53). Interestingly, phorbol esters are potent inhibitors of adipogenesis and PPARγ activity. It may be that induction of Thy1 expression by phorbol ester compounds is at least in part responsible for the inhibition.

Deletion of Thy1 increases adiposity and adipogenesis in mice fed a high-fat diet. Excess fat accumulation can occur in many tissues. For example, in addition to visceral fat depots associated with obesity, the liver can become fatty as a result of poor diet or excessive alcohol consumption, bone marrow can become fatty with age, and the orbit of the eye can accumulate excess fat as a result of autoimmunity, as in the case of thyroid eye disease (21). Our data demonstrate that Thy1 plays an unexpected yet pivotal role in regulating adipogenesis and obesity. Mouse and human adipocytes from different white fat depots are devoid of Thy1. Furthermore, loss of Thy1 promotes adiposity while increased expression of Thy1 limits adipogenesis. Thus, targeting Thy1 may be important in treating these disorders. Future studies aimed at dissecting the role of Thy1 and Thy1-associated pathways in limiting adipogenesis may lead to new strategies to curb the obesity epidemic and possibly provide clues to treat other disorders of excess fat accumulation.

Glossary

AdipoQ

adiponectin

ADRP

adipocyte differentiation-related protein

C/EBPα

CCAAT/enhancer binding protein α

Erk

extracellular signal–regulated kinase

FABP4

fatty acid-binding protein 4

MEF

mouse embryonic fibroblast

PPARγ

peroxisome proliferator-activated receptor γ

PPRE

PPAR response element

qPCR

quantitative real-time PCR

RXRα

retinoid X receptor α

SFK

Src-family kinase

SVF

stromal vascular fraction

WAT

white adipose tissue