Hedgehog Signaling Alters Adipocyte ... - Wiley Online Library

Download PDF
64 downloads 117 Views 2MB Size Report
Comment
naling activity [7–9]. Several pharmacological modulators of. Hedgehog signaling are available. Cyclopamine is a well-de- scribed inhibitor of the signaling [10], ...
TISSUE-SPECIFIC STEM CELLS Hedgehog Signaling Alters Adipocyte Maturation of Human Mesenchymal Stem Cells CORALIE FONTAINE, WENDY COUSIN, MAGALI PLAISANT, CHRISTIAN DANI, PASCAL PERALDI Institute of Signaling, Biology, Development and Cancer, Universite´ de Nice Sophia-Antipolis, Centre National de la Recherche Scientifique UMR6543, Nice, France Key Words. Mesenchymal stem cells • Differentiation • Hedgehog signaling • Adipocytes

ABSTRACT Human stem cells are powerful tools by which to investigate molecular mechanisms of cell growth and differentiation under normal and pathological conditions. Hedgehog signaling, the dysregulation of which causes several pathologies, such as congenital defects and cancer, is involved in several cell differentiation processes and interferes with adipocyte differentiation of rodent cells. The present study was aimed at investigating the effect of Hedgehog pathway modulation on adipocyte phenotype using different sources of human mesenchymal cells, such as bone marrow stromal cells and human multipotent adipose-derived stem cells. We bring evidence that Hedgehog signaling decreases during human adipocyte differentiation. Inhibition of this pathway is not sufficient to trigger adipogenesis, but activation of Hedgehog pathway alters adipocyte morphology as well as insulin sensitivity. Analysis of glycerol-3-phosphate dehydrogenase activity and expression of adipocyte marker genes indi-

cate that activation of Hedgehog signaling by purmorphamine impairs adipogenesis. In sharp contrast to reports in rodent cells, the maturation process, but not the early steps of human mesenchymal stem cell differentiation, is affected by Hedgehog activation. Hedgehog interferes with adipocyte differentiation by targeting CCAAT enhancer-binding protein ␣ and peroxisome proliferatoractivated receptor (PPAR) ␥2 expression, whereas PPAR␥1 level remains unaffected. Although Hedgehog pathway stimulation does not modify the total number of adipocytes, adipogenesis appears dramatically impaired, with reduced lipid accumulation, a decrease in adipocytespecific markers, and acquisition of an insulin-resistant phenotype. This study indicates that a decrease in Hedgehog signaling is necessary but not sufficient to trigger adipocyte differentiation and unveils a striking difference in the adipocyte differentiation process between rodent and human mesenchymal stem cells. STEM CELLS 2008;26:1037–1046

Disclosure of potential conflicts of interest is found at the end of this article.

INTRODUCTION Alterations in fat mass are frequently associated with modulation of adipocyte differentiation and lead to dysregulation of metabolic pathways, such as cholesterol, fatty acid, and glucose metabolism. This results in the development of dyslipidemia, insulin resistance, diabetes, and obesity. To investigate this phenomenon, various cellular and animal models have been developed. Up to now, preadipocyte cell lines from rodents have been used to gain insights into the molecular mechanisms of adipogenesis and as a tool to develop therapeutic drugs. In regard to human cell models presently available, primary cultures of stroma vascular fraction of adipose tissue are widely used. However, after a few passages, they undergo a dramatic decrease in their ability to differentiate before stopping growth and entering replicative senescence [1]. Recently, we isolated human stem cells derived from adipose tissue. These cells are called human multipotent adipose-derived stem (hMADS) cells, or adipose stem cells according to the new international nomenclature [2]. hMADS cells exhibit a normal karyotype, multipotency at the clonal level, and high self-renewal [3, 4]. They offer a limitless supply of adipocytes to investigate the molecular pathways of adipocyte differentiation and drug metabolism [5].

Adipocyte differentiation is a complex biological process, which is reflected at the cellular level by morphological changes, ability to accumulate cytoplasmic lipid droplets, and acquisition of insulin sensitivity. The nuclear receptor peroxisome proliferator-activated receptor (PPAR) ␥ and the CCAAT enhancer-binding protein (C/EBP) ␣ are the two main key transcription factors that regulate this process [6]. Adipocyte differentiation is highly controlled by stimulatory and inhibitory cues, and modifications in their balance give rise to lipodystrophy or obesity, an epidemic health problem. Hedgehog, the function of which during embryogenesis has been abundantly described, has emerged as a crucial modulator of cell differentiation processes, including adipogenesis. Hedgehog signaling is initiated by the binding of one of the three mammalian ligands: Sonic Hedgehog (Shh), Indian Hedgehog (Ihh), or Desert Hedgehog (Dhh) to the 12-transmembrane protein Patched (Ptch), relieving suppression of the 7-transmembrane protein Smoothened (Smo). In turn, Smo activates an intracellular cascade that results in activation of the Gli2 and Gli3 transcription factors. Gli1 is one of their target genes and has been characterized as a reliable marker of Hedgehog signaling activity [7–9]. Several pharmacological modulators of Hedgehog signaling are available. Cyclopamine is a well-described inhibitor of the signaling [10], and purmorphamine, a

Correspondence: Pascal Peraldi, Ph.D., Centre de Biochimie (UMR 6543 CNRS), Universite´ de Nice-Sophia Antipolis, Faculte des Sciences, 28 avenue Valrose, 06108 Nice Cedex 2, France. Telephone: 33-4-92076437; Fax: 33-4-92076404; e-mail: [email protected] Received November 20, 2007; accepted for publication January 31, 2008; first published online in STEM CELLS EXPRESS February 7, 2008. ©AlphaMed Press 1066-5099/2008/$30.00/0 doi: 10.1634/stemcells.2007-0974

STEM CELLS 2008;26:1037–1046 www.StemCells.com

Hedgehog Alters Adipocyte Differentiation of hMSCs

1038

purine derivative, has been found to activate Hedgehog pathway by directly targeting Smo [11, 12]. Since molecules are developed to modulate Hedgehog signaling in the treatment of various pathologies, such as cancer and alopecia, evaluation of potential side effects of this signaling pathway is of importance [13]. For instance, the observation that Hedgehog signaling is modified in rodent situation of obesity has pointed to a possible physiological implication of this pathway in this syndrome [14]. Antiadipogenic effects of Hedgehog signaling have been evidenced in vitro on murine mesenchymal cells able to differentiate into both adipocyte and osteoblast lineages, such as calvaria cells and C3H10T1/2 and KS483 cell lines [15–17], and in the classic preadipocyte cell line 3T3-L1 [14]. In these cells, activation of Hedgehog signaling favored osteoblastic differentiation at the expense of adipogenesis. However, the precise mechanism of Hedgehog-mediated inhibition of adipocyte differentiation has not been clearly determined in rodents, and no evaluation of the role of the Hedgehog pathway on human adipocyte differentiation has been reported. For this purpose, human bone marrow stromal cells (hBMSCs) [18, 19] and adipocyte-derived stem cells, such as hMADS cells, are invaluable models with which to study human adipogenesis. Together, our results show that Hedgehog signaling activation leads to a decrease in human adipocyte maturation and point out a major difference in the effect of Hedgehog on human and rodent adipocyte differentiation. This suggests that drugs developed to target Hedgehog signaling should be monitored for potential side effects using appropriate human cell models.

MATERIALS

AND

METHODS

Growth and Adipocyte Differentiation of hMADS Cells Establishment and characterization of the multipotency of hMADS cells have been extensively described [4]. In the experiments reported herein, hMADS-1, hMADS-2, and hMADS-3 cells— established from male and female infants—were used. Cells were grown in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal calf serum, 2.5 ng/ml human fibroblast growth factor (hFGF)-2, 60 ␮g/ml penicillin, and 50 ␮g/ml streptomycin. The medium was changed 2 days later in the absence of hFGF-2 until cells reached confluence. At day 2 postconfluence (designated day 0), cells were then induced to differentiate into adipocytes in the presence of DMEM/Ham’s F12 media supplemented with 0.85 ␮M insulin, 10 ␮g/ml transferrin, 0.2 nM triiodothyronine, 1 ␮M dexamethasone (DEX), 100 ␮M isobutyl-methylxanthine (IBMX), and 0.1 ␮M rosiglitazone. Three days later, the medium was changed (DEX and IBMX omitted). The media were then changed every other day, and cells were used at the indicated days. Cyclopamine was dissolved in dimethyl sulfoxide, used at 5 ␮M and added every other day. Purmorphamine was dissolved in ethanol and added in triplicate wells at 2 ␮M (or at the indicated concentration) for different periods of time before cells were collected. Both drugs were found not to have any obvious toxic effect. The specific Hedgehog activator purmorphamine was used instead of bacteria-produced mutant form of Shh (NShh) or conditioned medium from Shh-producing cells (obtained from American Type Culture Collection, Manassas, VA, http://www.atcc.org; CRL-2782). Indeed, recombinant NShh is not covalently modified by cholesterol and palmitate, is 30 times less potent than native Shh, and appears unable to form multimers, prompting some authors to question the physiological relevance of this protein [20, 21]. Conditioned medium from Shh-producing cells was found to exert the same effect on aP2, PPAR␥1, and PPAR␥2 gene expression as purmorphamine. However, conditioned medium also contained undefined molecules interfering with adipocyte differentiation, and reproducibility between batches was loose (supplemental online Fig. 1).

RNA Extraction and Analysis Total RNA was extracted using the TRI-Reagent kit (Euromedex, Souffelweyersheim, France, http://www.euromedex.com) according to the manufacturer’s instructions. Total RNA, digested with DNase I (Promega, Charbonnières-les-bains, France, http://www.promega. com), was subjected to real-time quantitative reverse transcription (RT)-polymerase chain reaction (PCR) analysis. The RT reactions were carried out in the presence of the pd(N)6 random hexamer (Roche Molecular Biochemicals, Meylan, France, http://www. roche-applied-science.com), Moloney murine leukemia virus reverse transcriptase (100 U), supplied buffer, dNTPs (0.5 mM), and RNase inhibitor (10 U) at 37°C for 1 hour in a 20-␮l reaction volume. Real-time quantitative PCRs were performed on an ABI Prism 7000 instrument (Applied Biosystems, Foster City, CA, http://www. appliedbiosystems.com). For each PCR run, a master mix composed of qPCR MasterMix Plus for SYBR Green I (Eurogentec, Seraing, Belgium, http://www.eurogentec.be) was prepared. Primers were designed using Primer Express software (Applied Biosystems; supplemental online Table 1) and validated by testing PCR efficiency using standard curves (85% ⱕ efficiency ⱕ115%). Gene expression was quantified using the comparative CT (threshold cycle) method; TBP was used as reference.

Determination of Cell Size, Lipid Accumulation, GPDH Activity, and Glucose Transport To examine lipid accumulation, cells were fixed with paraformaldehyde (4% in phosphate-buffered saline) and stained with oil red O. Three pictures per single well were realized in triplicate for each condition and were imported into ImageJ (available at http://rsb. info.nih.gov/nih-image). Adipocytes (at least 300 cells for each conditions) were outlined, and the area was calculated with the Analyze tool. The area measurements were imported into Microsoft Excel (Redmond, WA, http://www.microsoft.com), and the mean ⫾ SD of the areas was calculated for control and purmorphaminetreated cells. To evaluate cell morphology, more than 100 single cells were randomly chosen. To determine cell number, three random pictures of each surface were analyzed for each sample. GPDH activity was assayed spectrophotometrically, as previously described [22], on cell homogenates obtained using a 20 mM Tris/HCl buffer, pH 7.5, containing 1 mM EDTA and 1 mM ␤-mercaptoethanol. Assays were carried out in triplicate at day 14 of differentiation. Protein concentrations were determined by the Bradford method, with bovine serum albumin as the standard. Enzyme activity is expressed in mU/mg (i.e., nmol of product formed per minute per mg of protein). Insulin-stimulated glucose transport was determined using nonmetabolizable [3H]2-deoxy-D-glucose. Briefly, adipocytes were starved overnight in DMEM 0.2% bovine serum albumin (BSA) and thereafter incubated in Kreb’s Ringer phosphate medium 0.2% BSA (10 mM HEPES, pH 7.5, 2.5 mM Na2HPO4, 1.25 mM MgSO4, 2 mM CaCl2, 130 mM NaCl, 5 mM KCl) for 4 hours. Human recombinant insulin (10⫺7 M) was added for 45 minutes. During the last 10 minutes, 0.5 ␮Ci/ml 2-deoxy-D-[2,6-3H] glucose (Amersham Biosciences, Piscataway, NJ, http://www.amersham.com) and 0.1 mM 2-deoxy-D-glucose were added. Cells were washed twice in phosphate-buffered saline and lysed in KRPH 1% Triton X-100. Radioactivity was measured using a ␤-scintillation counter. The data were expressed in dpm/minute/mg of protein.

Cell Proliferation Assay hMADS cells were plated into six-well plates (4,500 cells per cm2). After the appropriate time, adherent cells were dissociated in 0.25% trypsin EDTA and counted with a Coulter counter.

Statistical Analysis Statistical differences between groups were analyzed by Student’s t test and are indicated on figures as follow: ⴱ p ⱕ .05, ⴱⴱ p ⱕ .01, and ⴱⴱⴱ p ⱕ .001.

Fontaine, Cousin, Plaisant et al.

1039

Figure 1. The decrease in Hedgehog signaling observed during adipogenesis is neither sufficient nor beneficial to trigger adipocyte differentiation of human mesenchymal stem cells. Cyclopamine (5 ␮M) was added every other day from day 0 to the end of the differentiation (day 14) in the presence or absence of the adipogenic cocktail. Shown is real-time quantitative reverse transcription-polymerase chain reaction (RT-PCR) of Gli1 (A) and Ihh, Dhh, Smo, Ptch, Gli2, and Gli3 (B). (C): Oil red O staining. (D): GPDH activity. (E): Expression of the aP2 gene using real-time quantitative RT-PCR. Data are mean ⫾ SE of triplicates from an experiment representative of three independent experiments or the mean ⫾ SE of three independent experiments (B). Abbreviations: C, control; ctrl, control; D, day 14 of differentiation.

RESULTS Hedgehog Signaling Is Downregulated During Adipocyte Differentiation We studied the role of Hedgehog signaling in human adipocyte differentiation using hMADS cells. These mesenchymal cells present several advantages. They display a normal karyotype and are not transformed, and differentiation takes place in a chemically defined medium, with no addition of serum. So far, they exhibit properties similar, if not identical, to those of human adipocytes [5]. We first investigated the status of Hedgehog signaling during human adipocyte differentiation through Gli1. Indeed, in the absence of Hedgehog signaling, Gli1 is transcriptionally silent, and since Hedgehog is the only known inducer of Gli1, its expression is considered as a marker of pathway activity [7–9, 23–25]. We observed that adipocyte differentiation of hMADS cells was associated with a 90% decrease in Gli1 expression (Fig. 1A), indicating that undifferentiated hMADS cells are endowed with an active Hedgehog signaling pathway that decreases during adipocyte differentiation. Consistently, the expression of Hedgehog receptor Ptch, another Hedgehog-target gene, was also found to be downregulated during adipocyte differentiation (Fig. 1 B). Ihh and Dhh were found to be expressed at low levels, and their expression was not modified during adipocyte differentiation. Shh messenger was also detected, but at levels too low to allow quantification. We then studied other elements of Hedgehog signaling: Smo, Gli2, and Gli3. The expression of Smo was not modified. Interestingly, both Gli2 and Gli3 expression decreased. Since Gli2 is the primary mediator of Hedgehog signaling [7–9, 26], it www.StemCells.com

can be envisioned that a decrease in its expression could be responsible for the decrease in Hedgehog signaling observed during adipocyte differentiation. We then determined whether the decrease in Hedgehog signaling observed during adipocyte differentiation was necessary and/or sufficient for adipocyte differentiation using the potent and specific Hedgehog inhibitor cyclopamine. hMADS cells appear particularly well suited for this purpose, since they differentiated in the absence of serum that could contain undefined molecules affecting Hedgehog signaling. Cells were treated or not treated with an adipogenic cocktail in the presence or absence of cyclopamine. As observed, cyclopamine addition induced a decrease in Gli1 expression, comparable to the one associated with adipocyte differentiation (Fig. 1A). After 14 days, cells were stained with oil red O to visualize lipid accumulation. As illustrated in Figure 1C, cyclopamine did not alter lipid accumulation of hMADS in the presence or absence of the adipogenic cocktail. Accordingly, chronic treatment of cyclopamine during differentiation did not modify either the GPDH activity (Fig. 1D) or aP2 gene expression (Fig. 1E). These data indicate that Hedgehog signaling inhibition is not sufficient to trigger human adipogenesis.

Hedgehog Signaling Impairs Adipogenesis and Lipid Accumulation We then investigated whether a decrease in Hedgehog signaling was necessary for adipocyte differentiation. We first tested the ability of purmorphamine, a potent and highly specific inducer of Hedgehog signaling that directly targets Smo [11, 12, 27, 28], to activate the Hedgehog pathway in hMADS cells. Purmorphamine exhibits an optimal biological effect at 2 ␮M (Fig. 2A), as

1040

Hedgehog Alters Adipocyte Differentiation of hMSCs

Figure 2. Hedgehog signaling is downregulated during differentiation but remains inducible in adipocytes. Expression of Gli1 using real-time quantitative RT-PCR in nondifferentiated cells (A, B) or at the indicated days of adipocyte differentiation (C). Cells were treated with increasing concentrations of purmorphamine for 24 h (A) or for the indicated times at 2 ␮M (B, C). Results are representative of three independent experiments. Abbreviation: h, hour.

observed in other cell types [12]. Kinetics studies demonstrated that Gli1 gene induction is maintained at least 72 hours after purmorphamine treatment (2 ␮M) in hMADS cells (Fig. 2B). Interestingly, the use of purmorphamine during adipogenesis induces Gli1 mRNA levels similar to those present in undifferentiated cells, demonstrating that purmorphamine did not induce a massive overactivation of Hedgehog signaling under these conditions (Fig. 2C). Adipocyte-differentiated cells were fixed and stained with oil red O. As observed in Figure 3A, chronic purmorphamine treatment induced a decrease in the global oil red O staining of hMADS. Microscopic analysis and a count of oil red O-labeled cells in the different conditions indicate that purmorphamine did not modify the total number of adipocytes (Fig. 3B). However, hMADS cells differentiated into adipocytes in the presence of purmorphamine have a reduced cell size compared with control adipocytes (Fig. 3A, 3C). Consistently, adipocytes from purmorphamine-treated cells presented smaller lipid droplets (Fig. 3A, 3D). We then tested the expression of perilipin, the major protein associated with lipid droplets in adipocytes [29, 30], and found that its expression was decreased in purmorphamine-treated cells (Fig. 3E). These morphological changes were accompanied by a 40% reduction of GPDH activity (Fig. 4A), as well as a significant decrease in the induction of the expression of adipocyte-specific markers aP2 (Fig. 4B), adipsin (Fig. 4C), CD36 (Fig. 4D), adiponectin (Fig. 4E), and leptin (Fig. 4F). Similar effects were observed with hBMSCs (supplemental online Fig. 2). Altogether, analysis of triglyceride storage, GPDH activity, and expression of adipocyte markers showed that chronic activation of Hedgehog signaling pathway by purmorphamine impairs normal differentiation, leading to less lipid accumulation, but does not alter adipocyte number.

Hedgehog Signaling Pathway Activation Does Not Inhibit the Early Steps of Adipogenesis To determine the mechanism by which Hedgehog signaling modulates the adipogenic process, we tested whether it could induce the expression of a series of known negative regulators of adipogenesis. GATA-2 [31], GILZ [32], and CHOP-10 [33] gene expression was quantified by real-time quantitative RTPCR, and induction of Gli1 mRNA level was analyzed as a positive response of Hedgehog activation (Fig. 5A). No change in the expression of these antiadipogenic factors was observed when comparing control hMADS cells (day 0) with purmorphamine-treated cells (2 ␮M, 48 hours) (Fig. 5A). Two other factors proposed to drive mesenchymal stem cells toward the osteoblastic at the expense of the adipogenic lineage were analyzed: the morphogen Wnt10b [34] and the transcriptional modulator TAZ [35]. Purmorphamine did not modulate the expression of these factors (Fig. 5A). Consistent with this observation, purmorphamine did not induce gene expression of osteoblastic markers such as alkaline phosphatase or osteoprotegerin (supplemental online Fig. 3). Since Hedgehog signaling has been linked to cell proliferation in a variety of systems [36], we tested this effect on hMADS cells. As shown in Figure 5B, there was no difference in cell proliferation between control and purmorphamine-treated cells in the presence or absence of FGF-2, which is critical for self-renewal of hMADS cells [3]. This indicates that Hedgehog does not interfere with adipocyte differentiation through cell growth modulation. The early steps of adipocyte differentiation take place during the first 3 days, when cells are treated with the complete differentiation medium. To determine whether these early steps are affected by Hedgehog, purmorphamine was added along the differentiation process (days 0 –14), during only the first 3 days

Fontaine, Cousin, Plaisant et al.

1041

Figure 3. Chronic purmorphamine treatment during differentiation decreases adipocyte and lipid droplet size but has no effect on adipocytes cell number. Purmorphamine (2 ␮M) was added every other day from day 0 to completion of differentiation. (A): Oil red O staining. (B): Mean number of cells per cm2. (C, D): Mean size of at least 300 cells per condition (C) and lipid droplets surface area of at least 100 single cells randomly chosen per condition (D) were determined using NIH ImageJ software. (E): Real-time quantitative reverse transcription-polymerase chain reaction of perilipin (mean ⫾ SE of three independent experiments). Abbreviations: Ctrl, control; NS, not significant; PLN, perilipin.

Figure 4. Chronic purmorphamine treatment during adipocyte differentiation decreases GPDH activity and expression of adipocyte marker genes. Purmorphamine (2 ␮M) was added every other day from day 0 to completion of differentiation. (A): GPDH activity. Results are representative of three independent experiments. Each column is the mean value ⫾ SE of triplicate determinations. (B–F): Gene expression analyzed using real-time quantitative RT-PCR. The data presented are means ⫾ SE of at least four independent experiments performed in triplicate. Abbreviations: Ctrl, control; prot, protein.

(days 0 –3), or from day 3 to the end of the experiment (days 3–14). Purmorphamine had no effect on hMADS adipocyte differentiation when added only from day 0 to day 3, as reflected by triglyceride staining (Fig. 5C) and aP2 gene induction (Fig. 5D). Nevertheless, treatment of purmorphamine added chronically since day 3 was sufficient to alter adipocyte differentiation (Fig. 5C, 5D). www.StemCells.com

We assessed the effect of purmorphamine on differentiated adipocytes. A 48-hour purmorphamine treatment of 14-daydifferentiated adipocytes did not interfere with marker expression (supplemental online Fig. 4). Cells were then differentiated for 14 days, and purmorphamine was added for 7 days (days 14 –21). Cells were then compared with adipocytes differentiated without (⫺) or with purmorphamine for 21 days (days

1042

Hedgehog Alters Adipocyte Differentiation of hMSCs

Figure 5. Hedgehog signaling does not inhibit the early step of adipogenesis of human mesenchymal stem cells. (A): hMADS cells (day 0) were treated with purmorphamine (48 hours, 2 ␮M) or vehicle, and gene expression was analyzed using real-time quantitative reverse transcription-polymerase chain reaction (RT-PCR). (B): Counting of hMADS cells grown in the presence or absence of FGF-2 (2 ng/ml) and treated every other day with purmorphamine or vehicle. (C, D): hMADS cells were differentiated for 14 days in the absence (⫺) or presence of purmorphamine from day 0 to day 14 (0 –14), day 0 to day 3 (0 –3), or day 3 to day 14 (3–14). (C): Oil red O staining. (D): Gene expression. (E, F): hMADS were differentiated for 21 days in the absence (⫺) or presence of purmorphamine from day 0 to day 21 (0 –21) or from day 14 to day 21 (14 –21). (E): Oil red O staining. (F): Real-time quantitative RT-PCR presented in fold induction relative to undifferentiated hMADS cells. Results are representative of three (E, F) or four (A–D) independent experiments. Each column is the mean value ⫾ SE of triplicate determinations. Abbreviations: C/EBP␣, CCAAT enhancer-binding protein ␣; ctrl, control; FGF, fibroblast growth factor; hMADS, human multipotent adipose-derived stem; NS, not significant.

Fontaine, Cousin, Plaisant et al.

1043

Figure 6. Glucose transport in response to insulin is inhibited by chronic treatment with purmorphamine during adipogenesis. Purmorphamine (2 ␮M) was added every other day from day 0 to completion of differentiation. (A): [3H]2-Deoxyglucose uptake performed in adipocytes stimulated or not stimulated by insulin. The data presented are mean ⫾ SE of triplicates from an experiment representative of three independent experiments. (B): Glut4 expression analyzed using real-time quantitative reverse transcriptionpolymerase chain reaction. The data presented are mean ⫾ SE of four independent experiments performed in triplicate and are expressed as fold induction relative to undifferentiated hMADS cells (day 0). Abbreviation: Ctrl, control.

Figure 7. Purmorphamine treatment during adipogenesis decreases C/EBP␣ and PPAR␥2 gene induction. Purmorphamine (2 ␮M) was added every other day from day 0 to completion of differentiation. C/EBP␣ (A), PPAR␥ (primers common to both PPAR␥ isoforms) (〉), PPAR␥ coactivators (C), and PPAR␥1 and PPAR␥2 isoform expression (D) were analyzed by real-time quantitative reverse transcription-polymerase chain reaction. The data presented are means ⫾ SE of at least three independent experiments performed in triplicate. Abbreviations: C/EBP␣, CCAAT enhancer-binding protein ␣; ctrl, control; NS, not significant; PPAR, peroxisome proliferator-activated receptor.

0 –21). As observed, whereas increase in Gli1 expression indicated that the Hedgehog signaling pathway was activated, purmorphamine treatment of differentiated cells for 7 days (days 14 –21) did not affect oil red O staining (Fig. 5E) or adipsin, aP2, or C/EBP␣ expression (Fig. 5F). These data indicated that Hedgehog signaling is inefficient to induce dedifferentiation of human adipocytes.

Cells Differentiated in the Presence of Purmorphamine Are Resistant to Insulin We next determined whether the defect in triglyceride accumulation was associated with a decrease in glucose metabolism. hMADS cells were treated under adipogenic conditions in the presence or absence of purmorphamine, and glucose uptake was monitored after 14 days of differentiation. Although basal glucose uptake was not altered by Hedgehog activation pathway, insulin-stimulated glucose uptake was abolished in the presence of purmorphamine during adipogenesis (Fig. 6A). In addition, the insulin-sensitive glucose transporter Glut4 expression dramatically decreased under purmorphamine treatment during adwww.StemCells.com

ipogenesis (Fig. 6B). Thus, the mechanism by which purmorphamine impaired triglyceride accumulation during human adipogenesis may be accounted for, at least by part, by a decrease in insulin-stimulated glucose uptake.

Hedgehog Activation Targets Adipocyte Differentiation Through C/EBP␣ and PPAR␥2 The transcriptional pathway of adipogenesis and insulin sensitivity are regulated by cross-regulation of PPAR␥ and C/EBP␣ [37]. Analysis of these two key transcription factors expression indicated that the induction of C/EBP␣ mRNA level was markedly reduced by Hedgehog activation (Fig. 7A), whereas no significant effect was observed on PPAR␥ expression when primers allowing the detection of total PPAR␥ (i.e., PPAR␥1 and PPAR␥2) were used (Fig. 7B). Nuclear receptors, such as PPAR␥, are known to activate transcription through docking of specific coactivator proteins. Moreover, the relative cellular levels of these cofactors affect the ability of PPAR␥ ligands to induce adipocyte differentiation [38]. Therefore, we assessed whether purmorphamine could

Hedgehog Alters Adipocyte Differentiation of hMSCs

1044

modulate the expression of PPAR␥ coactivators during adipogenesis. As shown in Figure 7C, no change was observed in the expression of the PPAR␥ cofactors tested, such as TRAP220, SRC-1, p300, and PGC1␣. Finally, because PPAR␥1 and PPAR␥2 have distinct patterns of induction during adipocyte differentiation [39, 40] and could exert specific functions in adipocytes [41], it was of interest to examine the expression of each PPAR␥ isoform using specific primers. Interestingly, we found that expression of the most potent proadipogenic isoform, PPAR␥2, decreased under purmorphamine treatment (Fig. 7D), whereas PPAR␥1 levels remained unmodified (Fig. 7D). Similar results were observed using hBMSCs (supplemental online Fig. 5), demonstrating a uniform mechanism of Hedgehog on adipocyte differentiation of human mesenchymal stem cells.

DISCUSSION Although they are more difficult to handle than classic rodent preadipocyte cell lines, human mesenchymal stem cells present several significant advantages for the study of human adipocyte differentiation because they display a normal karyotype, and, in the case of hMADS cells developed in our laboratory, differentiation takes place in a serum-free, chemically defined medium. We show here that human mesenchymal stem cells are endowed with active Hedgehog signaling. Since these cells express several partners of Hedgehog signaling pathway (Ptch, Smo, Gli1, Gli2, and Gli3), it can be envisioned that the expression of Ihh and Dhh is responsible for the maintenance of active signaling in these cells, through an autocrine mechanism. Interestingly, adipocyte differentiation of these cells is associated with a decrease in Hedgehog signaling. This is similar to what has been observed in murine cells [14, 42]. Since we did not detect any modulation in Dhh and Ihh expression during adipocyte differentiation, the decrease in Hedgehog signaling could be associated with a decrease in Gli2 expression (Fig. 1B). Indeed, Gli2 is the primary activator of Hedgehog signaling, and a decrease in its expression in vivo and in vitro leads to a decreased signaling [7–9, 26]. So far, Hedgehog signaling was thought to be regulated primarily through modulation of its ligands. Downregulation of Hedgehog signaling through Gli2 expression could be an original mechanism, specific to adipocyte differentiation. Interestingly, a decrease in Gli2 expression was also observed during adipocyte differentiation of murine cells [14]. Modulation in Hedgehog signaling pathway during adipogenesis suggested that it could play a role in adipocyte differentiation of human mesenchymal stem cells. Experiments performed using cyclopamine indicated that a decrease in this pathway is not sufficient to trigger adipocyte differentiation (Fig. 1). Although debated [14], this is similar to what we observed in murine 3T3-L1 cells [42]. On the other hand, chronic activation of Hedgehog pathway impairs human adipocyte differentiation. Together, this indicates that a decrease in Hedgehog signaling is a necessary but not a sufficient phenomenon for adipocyte differentiation. Under our conditions, purmorphamine maintained a level of Hedgehog signaling similar to the one observed in undifferentiated cells, suggesting that the antiadipogenic effect of Hedgehog can occur without massive overactivation of the pathway. A close examination indicated that the initial events of adipocyte differentiation are not modified by purmorphamine. Indeed, purmorphamine does not modify the total number of adipocytes but interferes with the late steps of differentiation, such as lipid accumulation and the expression of late adipocyte markers. Moreover, activation of the Hedgehog signaling pathway during only the first 3 days did

not affect adipogenesis (Fig. 5C, 5D), indicating that Hedgehog signaling does not interfere with the induction but instead with the maturation of human adipocytes. Interestingly, Hedgehog had no effect on differentiated hMADS (Fig. 5E, 5F). This suggests that the Hedgehog effect is concomitant with the increase in the expression levels of adipocyte master genes, such as PPAR␥ and C/EBP␣ [5]. It is likely that when these factors reach critical levels, the cells become resistant to Hedgehog. These effects of Hedgehog on the maturation of human adipocyte differ from those reported for rodent cells. Indeed, in C3H10T1/2, KS483, and 3T3-L1 cell lines, Hedgehog signaling inhibits the early steps of adipocyte differentiation and diverts cells to an osteoblastic phenotype [14 –17]. This has also been observed in murine primary culture of calvaria [15] and in primary bone marrow stromal cells, using 20(S)-hydroxycholesterol as a Hedgehog activator [43, 44], suggesting that the difference is not specific to a cell model but is indeed a species difference. Suh et al. proposed that Hedgehog-induced adipocyte inhibition of murine cells is mediated through GATA-2 [14], although clear evidence of GATA-2 induction by Hedgehog remained to be strengthened [45]. In human mesenchymal stem cells, purmorphamine did not affect GATA expression and was not found to favor osteoblastic differentiation (supplemental online Fig. 3). Furthermore, it must be stressed that these interspecies differences in Hedgehog signaling on human adipocytes are not specific to the age of the donor or to a given cell line. Indeed, two other hMADS cells, isolated from biopsies from different donors (4 months old and 5 years old), were tested, with similar results (data not shown); in addition, these results were recapitulated in hBMSCs from a 23-year-old patient. Cells differentiated in the presence of purmorphamine exhibited a smaller cell size, with reduced lipid accumulation, and were insulin-resistant. The difference in morphology could be caused, by part, by the decrease in perilipin expression (Fig. 3E). Indeed, perilipin is the major protein associated with lipid droplets in adipocytes [29, 30], and although perilipin does not interfere per se in adipocyte differentiation, adipocytes from perilipin knockout mice have smaller adipocytes than their wild-type counterparts [46, 47]. Interestingly, although it is assumed that large adipocytes found in situation of obesity are less sensitive to insulin than “normal-sized” adipocytes, it has been reported that, in the situation of obesity, insulin resistance is associated with an expanded population of small adipocytes, with a decreased expression in adipocyte-specific markers in human patients [48]. It would be of interest to determine the degree to which immature adipocytes from purmorphamine-treated cells share phenotypic similarities with small adipocytes from these patients. The later stages of adipocyte differentiation are controlled by C/EBP␣ and PPAR␥, two transcription factors positively regulating each other. Expression of C/EBP␣ was found to be downregulated by 70%. Strikingly, although total PPAR␥ was not significantly modified, PPAR␥2 expression decreased by 40%. PPAR␥ is expressed as two isoforms, PPAR␥1 and PPAR␥2, which are derived from the same gene by alternative promoter usage and differential mRNA splicing, so that PPAR␥2 has an additional 30 amino acids at the NH2 terminus. PPAR␥1 can be found in several tissues, including fat, skeletal muscle, heart, and liver, whereas PPAR␥2 is found almost exclusively in adipose tissue. Although it is debated [49], it is generally thought that PPAR␥1 is the predominant PPAR␥ isoform in human adipocytes [50, 51], which could explain why PPAR␥2 expression does not significantly affect the expression of total PPAR␥ (Fig. 7). In vitro, PPAR␥2 is more efficient for promoting adipocyte differentiation; however, the role of PPAR␥1 on adipogenesis remains controversial [52, 53]. It can be noted that in vitro studies were performed using overexpression of PPAR␥ isoforms, which did not exactly match the physiological levels of these proteins. Our data show that maintenance of total

Fontaine, Cousin, Plaisant et al. PPAR␥ levels is not sufficient to maintain full adipocyte differentiation when C/EBP␣ and PPAR␥2 expression are downregulated. By which mechanism does purmorphamine inhibit adipocyte maturation? C/EBP␣ and PPAR␥2 are required for the normal induction of key adipogenic genes, such as aP2, adipsin, CD36, and Glut4. Cells treated with purmorphamine during differentiation display defective expression of these genes and become insulinresistant adipocytes. Adipogenesis of hMADS cells takes place in serum-free defined conditions. Insulin stimulates adipocytes to increase glucose uptake, whose energy is then stored in glycogen and triacylglycerol. The decrease in insulin sensitivity is likely to be one of the causes for the decrease in lipid accumulation. Moreover, since insulin could modulate PPAR␥ transcription in an isoformspecific manner [54, 55], it is tempting to speculate that the decrease in insulin sensitivity could be responsible in part for the downregulation of PPAR␥2 expression. Together, purmorphamine could then trigger a vicious circle: C/EBP␣ and PPAR␥ positively regulate each other and control insulin sensitivity, which in return regulates PPAR␥ levels. Perturbation in one of these elements ultimately would lead to a global decrease in adipocyte maturation. Incidentally, this indicates that therapeutic treatments targeting the Hedgehog signaling pathway could engender insulin resistance.

CONCLUSION Together, our results indicate that a decrease in Hedgehog signaling is necessary but not sufficient for a complete adipo-

REFERENCES 1 2 3 4

5 6 7 8 9 10 11 12 13 14 15

Hauner H, Entenmann G. Regional variation of adipose differentiation in cultured stromal-vascular cells from the abdominal and femoral adipose tissue of obese women. Int J Obes 1991;15:121–126. Horwitz EM, Le Blanc K, Dominici M et al. Clarification of the nomenclature for MSC: The International Society for Cellular Therapy position statement. Cytotherapy 2005;7:393–395. Zaragosi LE, Ailhaud G, Dani C. Autocrine fibroblast growth factor 2 signaling is critical for self-renewal of human multipotent adiposederived stem cells. STEM CELLS 2006;24:2412–2419. Rodriguez AM, Pisani D, Dechesne CA et al. Transplantation of a multipotent cell population from human adipose tissue induces dystrophin expression in the immunocompetent mdx mouse. J Exp Med 2005; 201:1397–1405. Rodriguez AM, Elabd C, Delteil F et al. Adipocyte differentiation of multipotent cells established from human adipose tissue. Biochem Biophys Res Commun 2004;315:255–263. Rosen ED, MacDougald OA. Adipocyte differentiation from the inside out. Nat Rev Mol Cell Biol 2006;7:885– 896. Hooper JE, Scott MP. Communicating with Hedgehogs. Nat Rev Mol Cell Biol 2005;6:306 –317. Ruiz i Altaba A, Mas C, Stecca B. The Gli code: An information nexus regulating cell fate, stemness and cancer. Trends Cell Biol 2007;17:438 – 447. Riobo NA, Manning DR. Pathways of signal transduction employed by vertebrate Hedgehogs. Biochem J 2007;403:369 –379. Taipale J, Chen JK, Cooper MK et al. Effects of oncogenic mutations in Smoothened and Patched can be reversed by cyclopamine. Nature 2000; 406:1005–1009. Wu X, Walker J, Zhang J et al. Purmorphamine induces osteogenesis by activation of the Hedgehog signaling pathway. Chem Biol 2004;11: 1229 –1238. Sinha S, Chen JK. Purmorphamine activates the Hedgehog pathway by targeting Smoothened. Nat Chem Biol 2006;2:29 –30. Kiselyov AS. Targeting the Hedgehog signaling pathway with small molecules. Anticancer Agents Med Chem 2006;6:445– 449. Suh JM, Gao X, McKay J et al. Hedgehog signaling plays a conserved role in inhibiting fat formation. Cell Metab 2006;3:25–34. Spinella-Jaegle S, Rawadi G, Kawai S et al. Sonic Hedgehog increases the commitment of pluripotent mesenchymal cells into the osteoblastic lineage and abolishes adipocytic differentiation. J Cell Sci 2001;114: 2085–2094.

www.StemCells.com

1045

cyte differentiation. Hedgehog signaling inhibits adipocyte maturation of human mesenchymal stem cells, leading to small and insulin-resistant adipocytes without modifying the total level of PPAR␥. Moreover, these results reveal a species-specific effect of Hedgehog signaling activation on adipocyte differentiation and thus highlight the importance of using human mesenchymal stem cells to evaluate side effects of drugs modifying this pathway.

ACKNOWLEDGMENTS We are grateful to Drs. G. Ailhaud, A. Ladoux, S. Giorgetti Peraldi, and Ez-Zoubir Amri for helpful comments. This work was supported by grants from the Centre National de la Recherche Scientifique and the Fondation pour la Recherche Medicale. C.F. and M.P. were supported by the Fondation pour la Recherche Medicale, and W.C. was supported by the Association pour la Recherche sur le Cancer 7696.

DISCLOSURE

POTENTIAL CONFLICTS OF INTEREST

OF

The authors indicate no potential conflicts of interest.

16 Zehentner BK, Leser U, Burtscher H. BMP-2 and sonic Hedgehog have contrary effects on adipocyte-like differentiation of C3H10T1/2 cells. DNA Cell Biol 2000;19:275–281. 17 van der Horst G, Farih-Sips H, Lowik CW et al. Hedgehog stimulates only osteoblastic differentiation of undifferentiated KS483 cells. Bone 2003;33:899 –910. 18 Pittenger MF, Mackay AM, Beck SC et al. Multilineage potential of adult human mesenchymal stem cells. Science 1999;284:143–147. 19 Jaiswal RK, Jaiswal N, Bruder SP et al. Adult human mesenchymal stem cell differentiation to the osteogenic or adipogenic lineage is regulated by mitogen-activated protein kinase. J Biol Chem 2000;275:9645–9652. 20 Taylor FR, Wen D, Garber EA et al. Enhanced potency of human Sonic Hedgehog by hydrophobic modification. Biochemistry 2001; 40:4359 – 4371. 21 Goetz JA, Singh S, Suber LM et al. A highly conserved amino-terminal region of sonic Hedgehog is required for the formation of its freely diffusible multimeric form. J Biol Chem 2006;281:4087– 4093. 22 Wise LS, Green H. Participation of one isozyme of cytosolic glycerophosphate dehydrogenase in the adipose conversion of 3T3 cells. J Biol Chem 1979;254:273–275. 23 Kasper M, Regl G, Frischauf AM et al. GLI transcription factors: Mediators of oncogenic Hedgehog signalling. Eur J Cancer 2006;42: 437– 445. 24 Lee J, Platt KA, Censullo P et al. Gli1 is a target of Sonic Hedgehog that induces ventral neural tube development. Development 1997; 124:2537–2552. 25 Hynes M, Stone DM, Dowd M et al. Control of cell pattern in the neural tube by the zinc finger transcription factor and oncogene Gli-1. Neuron 1997;19:15–26. 26 Thiyagarajan S, Bhatia N, Reagan-Shaw S et al. Role of GLI2 transcription factor in growth and tumorigenicity of prostate cells. Cancer Res 2007;67:10642–10646. 27 Briscoe J. Agonizing Hedgehog. Nat Chem Biol 2006;2:10 –11. 28 Riobo NA, Saucy B, Dilizio C et al. Activation of heterotrimeric G proteins by Smoothened. Proc Natl Acad Sci U S A 2006;103:12607– 12612. 29 Tai ES, Ordovas JM. The role of perilipin in human obesity and insulin resistance. Curr Opin Lipidol 2007;18:152–156. 30 Wolins NE, Brasaemle DL, Bickel PE. A proposed model of fat packaging by exchangeable lipid droplet proteins. FEBS Lett 2006;580: 5484 –5491. 31 Tong Q, Dalgin G, Xu H et al. Function of GATA transcription factors in preadipocyte-adipocyte transition. Science 2000;290:134 –138. 32 Shi X, Shi W, Li Q et al. A glucocorticoid-induced leucine-zipper

Hedgehog Alters Adipocyte Differentiation of hMSCs

1046

33 34 35 36 37 38 39 40

41

42 43 44

protein, GILZ, inhibits adipogenesis of mesenchymal cells. EMBO Rep 2003;4:374 –380. Tang QQ, Lane MD. Role of C/EBP homologous protein (CHOP-10) in the programmed activation of CCAAT/enhancer-binding protein-beta during adipogenesis. Proc Natl Acad Sci U S A 2000;97:12446 –12450. Hu H, Hilton MJ, Tu X et al. Sequential roles of Hedgehog and Wnt signaling in osteoblast development. Development 2005;132:49 – 60. Hong JH, Hwang ES, McManus MT et al. TAZ, a transcriptional modulator of mesenchymal stem cell differentiation. Science 2005;309: 1074 –1078. Agathocleous M, Locker M, Harris WA et al. A general role of Hedgehog in the regulation of proliferation. Cell Cycle 2007;6:156 –159. Wu Z, Rosen ED, Brun R et al. Cross-regulation of C/EBP alpha and PPAR gamma controls the transcriptional pathway of adipogenesis and insulin sensitivity. Mol Cell 1999;3:151–158. Powell E, Kuhn P, Xu W. Nuclear receptor cofactors in PPARgammamediated adipogenesis and adipocyte energy metabolism. PPAR Res 2007;2007:53843. Fajas L, Auboeuf D, Raspe E et al. The organization, promoter analysis, and expression of the human PPARgamma gene. J Biol Chem 1997;272: 18779 –18789. Saladin R, Fajas L, Dana S et al. Differential regulation of peroxisome proliferator activated receptor gamma1 (PPARgamma1) and PPARgamma2 messenger RNA expression in the early stages of adipogenesis. Cell Growth Differ 1999;10:43– 48. Medina-Gomez G, Virtue S, Lelliott C et al. The link between nutritional status and insulin sensitivity is dependent on the adipocyte-specific peroxisome proliferator-activated receptor-gamma2 isoform. Diabetes 2005;54:1706 –1716. Cousin W, Dani C, Peraldi P. Inhibition of the anti-adipogenic Hedgehog signaling pathway by cyclopamine does not trigger adipocyte differentiation. Biochem Biophys Res Commun 2006;349:799 – 803. Kha HT, Basseri B, Shouhed D et al. Oxysterols regulate differentiation of mesenchymal stem cells: Pro-bone and anti-fat. J Bone Miner Res 2004;19:830 – 840. Kim WK, Meliton V, Amantea CM et al. 20(S)-hydroxycholesterol inhibits PPARgamma expression and adipogenic differentiation of bone

45 46 47

48 49

50

51 52 53 54

55

marrow stromal cells through a Hedgehog-dependent mechanism. J Bone Miner Res 2007;22:1711–1719. Cousin W, Fontaine C, Dani C et al. Hedgehog and adipogenesis: Fat and fiction. Biochimie 2007;89:1447–1453. Martinez-Botas J, Anderson JB, Tessier D et al. Absence of perilipin results in leanness and reverses obesity in Lepr(db/db) mice. Nat Genet 2000;26:474 – 479. Tansey JT, Sztalryd C, Gruia-Gray J et al. Perilipin ablation results in a lean mouse with aberrant adipocyte lipolysis, enhanced leptin production, and resistance to diet-induced obesity. Proc Natl Acad Sci U S A 2001;98:6494 – 6499. McLaughlin T, Sherman A, Tsao P et al. Enhanced proportion of small adipose cells in insulin-resistant vs insulin-sensitive obese individuals implicates impaired adipogenesis. Diabetologia 2007;50:1707–1715. Giusti V, Verdumo C, Suter M et al. Expression of peroxisome proliferator-activated receptor-gamma1 and peroxisome proliferator-activated receptor-gamma2 in visceral and subcutaneous adipose tissue of obese women. Diabetes 2003;52:1673–1676. Auboeuf D, Rieusset J, Fajas L et al. Tissue distribution and quantification of the expression of mRNAs of peroxisome proliferator-activated receptors and liver X receptor-alpha in humans: No alteration in adipose tissue of obese and NIDDM patients. Diabetes 1997;46:1319 –1327. Sewter CP, Blows F, Vidal-Puig A et al. Regional differences in the response of human pre-adipocytes to PPARgamma and RXRalpha agonists. Diabetes 2002;51:718 –723. Mueller E, Drori S, Aiyer A et al. Genetic analysis of adipogenesis through peroxisome proliferator-activated receptor gamma isoforms. J Biol Chem 2002;277:41925– 41930. Ren D, Collingwood TN, Rebar EJ et al. PPARgamma knockdown by engineered transcription factors: Exogenous PPARgamma2 but not PPARgamma1 reactivates adipogenesis. Genes Dev 2002;16:27–32. Armoni M, Harel C, Karni S et al. FOXO1 represses peroxisome proliferator-activated receptor-gamma1 and -gamma2 gene promoters in primary adipocytes. A novel paradigm to increase insulin sensitivity. J Biol Chem 2006;281:19881–19891. Rieusset J, Andreelli F, Auboeuf D et al. Insulin acutely regulates the expression of the peroxisome proliferator-activated receptor-gamma in human adipocytes. Diabetes 1999;48:699 –705.

See www.StemCells.com for supplemental material available online.