- Email: [email protected]
PPARγ suppression inhibits adipogenesis but does not promote osteogenesis of human mesenchymal stem cells
The International Journal of Biochemistry & Cell Biology 44 (2012) 377–384
Contents lists available at SciVerse ScienceDirect
The International Journal of Biochemistry & Cell Biology journal homepage: www.elsevier.com/locate/biocel
PPAR␥ suppression inhibits adipogenesis but does not promote osteogenesis of human mesenchymal stem cells Wei-Hua Yu a,1 , Fu-Gui Li a,1 , Xiao-Yong Chen a , Jian-Tao Li a , Yan-Heng Wu b , Li-Hua Huang c , Zhen Wang d , Panlong Li a , Tao Wang a , Bruce T. Lahn a,e , Andy Peng Xiang a,f,g,∗ a Center for Stem Cell Biology and Tissue Engineering, The Key Laboratory for Stem Cells and Tissue Engineering, Ministry of Education, Sun Yat-Sen University, 74# Zhongshan Road 2, Guangzhou, Guangdong 510080, PR China b Medical Imaging Department, The First Affiliated Hospital of Jinan University, No. 613, The West of Huangpu Street, Guangzhou, Guangdong 510630, PR China c Key Laboratory of Modulating Liver to Treat Hyperlipemia SATCM/Level 3 Laboratory of Lipid Metabolism SATCM, Guangdong Pharmaceutical University, Guangzhou, Guangdong 510006, PR China d Institute of Blood Transfusion, Guangzhou Blood Center, Guangzhou, Guangdong 510095, PR China e Department of Human Genetics and Howard Hughes Medical Institute, University of Chicago, 929 E. 57th ST. #W503, Chicago, IL 60637, USA f Department of Biochemistry, Zhongshan Medical School, Sun Yat-sen University, 74# Zhongshan Road 2, Guangzhou, Guangdong 510080, PR China g Cell-gene Therapy Translational Medicine Research Center, The Third Affiliated Hospital, Sun Yat-sen University, Guangzhou 510630, PR China
a r t i c l e
i n f o
Article history: Received 28 June 2011 Received in revised form 14 November 2011 Accepted 15 November 2011 Available online 23 November 2011 Keywords: PPAR␥ RNAi Human mesenchymal stem cells Adipogenesis Osteogenesis
a b s t r a c t Mesenchymal stem cells (MSCs) are the common progenitors of osteoblasts and adipocytes. A reciprocal relationship exists between osteogenesis and adipogenesis in the bone marrow, and the identification of signaling pathways that stimulate MSC osteogenesis at the expense of adipogenesis is of great importance from the viewpoint of developing new therapeutic treatments for bone loss. The adipogenic transcription factor peroxisome proliferator-activated receptor ␥ (PPAR␥) has been reported to play a vital role in modulating mesenchymal lineage allocation within the bone marrow compartment, stimulating adipocyte development at the expense of osteoblast differentiation. Hence, PPAR␥ may be a valuable target for drugs intended to enhance bone mass. However, little direct evidence is available for the role played by PPAR␥ in human mesenchymal lineage allocation. In this study, using human MSCs as an in vitro model, we showed that the two isoforms of PPAR␥, PPAR␥1 and PPAR␥2, were differentially induced during hMSC adipogenesis, whereas only PPAR␥1 was detected during osteogenesis. BADGE and GW9662, two potential antagonists of PPAR␥, as well as lentivirus-mediated knockdown of PPAR␥, inhibited hMSC adipogenesis but did not significantly affect osteogenesis. PPAR␥ knockdown did not significantly influence the expression level of the osteogenic transcription factor Runx2. Together, these results suggest that PPAR␥ is not the master factor regulating mesenchymal lineage determination in human bone marrow. © 2011 Elsevier Ltd. All rights reserved.
1. Introduction An increase in the content of bone marrow (BM) adipose tissue and a decrease in bone volume is observed in patients with
Abbreviations: MSCs, mesenchymal stem cells; BADGE, bisphenol-A-diglycidyl ether; ALP, alkaline phosphatase; LPL, lipoprotein lipase; PPAR␥, peroxisome proliferator-activated receptor ␥; C/EBP␣, CCAAT/enhancer binding protein ␣; AIM, adipocyte induction medium; AMM, adipocyte maintenance medium; GPDH, glycerol-3-phosphate dehydrogenase; FABP4, fatty acid-binding protein 4; ARS, alizarin red S. ∗ Corresponding author at: Center for Stem Cell Biology and Tissue Engineering, The Key Laboratory for Stem Cells and Tissue Engineering, Ministry of Education, Sun Yat-Sen University, 74# Zhongshan Road 2, Guangzhou, Guangdong 510080, PR China. Tel.: +86 20 87335822; fax: +86 20 87335858. E-mail address: [email protected] (A.P. Xiang). 1 These authors made equal contributions to this work. 1357-2725/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.biocel.2011.11.013
age-related osteopenia, as well as in individuals subject to other conditions such as osteoporosis, immobilization, microgravity, ovariectomy, diabetes, or glucocorticoid treatment (Nuttall and Gimble, 2004). Marrow adipocytes share a common MSC origin with bone-forming osteoblasts, as illustrated by the ability of MSCs to differentiate into either lineage (Pittenger et al., 1999). Increasing evidence of transdifferentiation in such cells suggests a large element of plasticity in the mutual transformation of these cell types (Schiller et al., 2001; Wolf et al., 2003). Therefore, any factor controlling the balance between adipogenesis and osteogenesis from BM–MSCs may serve as a target of pharmacological intervention to control bone disorders. The activation of certain transcription factors is essential for cellular commitment to a particular differentiation lineage (Farmer, 2006). PPAR␥ is a member of the nuclear hormone receptor gene superfamily of ligand-activated transcription factors. PPAR␥ is commonly termed the master regulator of adipogenesis and is both
378
W.-H. Yu et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 377–384
necessary and sufficient for adipogenesis; no factor has yet been identified that can induce normal adipogenesis in the absence of PPAR␥. Furthermore, the ectopic expression and activation of the factor is sufficient to induce adipocyte differentiation (Kawai and Rosen, 2010). It has long been hypothesized that PPAR␥ plays a vital role in mesenchymal lineage determination. PPAR␥ haploinsufficient mice showed an increase in bone mass associated with a loss of adipose tissue volume. Primary BM–MSCs from such animals showed enhanced osteogenic but reduced adipogenic differentiation; homozygous PPAR␥-deficient murine ES cells were unable to differentiate into adipocytes but spontaneously differentiated into osteoblasts. Torii reported the differentiation of ES cells into an osteoblastic lineage using siRNA against PPAR␥ without the addition of any osteogenic factors (Yamashita et al., 2006). These findings suggest that PPAR␥ can regulate lineage allocation within the BM, favoring adipocyte development over that of osteoblasts, and that PPAR␥ could serve as a useful target for drugs intended to enhance bone mass. Because of the multiple roles played by PPAR␥ in human physiology (Zieleniak et al., 2008), including the mediation of insulin sensitivity, attempts to design drugs that suppress PPAR␥ expression in BM stromal cells without side effects have not been wholly successful. Intriguingly, although the PPAR␥ antagonist bisphenol-A-diglycidyl ether (BADGE) has been shown to prevent BM adiposity in type 1 diabetic mice, this effect was not associated with any improvement in bone density (Botolin and McCabe, 2006). Only a few reports have appeared about the effects of PPAR␥ antagonists on human mesenchymal lineage allocation. Thus, any benefit of the modulation of PPAR␥ expression as a means of treating bone loss remains uncertain. Given that human BM derived-MSCs are the common progenitors of osteoblasts and adipocytes and that MSCs may be preferential targets of the desired therapy, such cells serve as a promising model to characterize the effects of modulating PPAR␥ expression (Elabd et al., 2008). In the present study, we explored the expression of PPAR␥ during the course of MSC differentiation into adipocytes or osteoblasts. BADGE and GW9662, potential modulators of PPAR␥, and lentivirus-borne PPAR␥ RNAi were used to investigate the involvement of PPAR␥ in mesenchymal lineage determination. The effects of PPAR␥ inhibition on the expression of other MSC-associated transcription factors, Runx2 and C/EBP␣, were also explored. 2. Materials and methods 2.1. Materials Cell culture reagents, including Dulbecco’s modified Eagle’s medium (DMEM), fetal bovine serum (FBS) and trypsin, were obtained from Hyclone (Logan, UT). Dexamethasone, 3-isobutyl1-methylxanthine (IBMX), insulin, indomethacin, ascorbic acid, sodium glycerophosphate, BADGE and GW9662 were purchased from Sigma (St. Louis, MO). Antibodies against PPAR␥ and C/EBP␣ were obtained from Cell Signaling Technologies (Beverly, MA). A monoclonal antibody against GAPDH was supplied by Kangchen (Shanghai, China). 2.2. Cell culture The hMSCs were obtained from Cyagen Biosciences Inc. (Guangzhou, China). The culture medium contained low-glucose DMEM (Hyclone, Logan, UT, USA) supplemented with 10% fetal bovine serum (FBS) (Hyclone), 100 IU/ml penicillin (Hyclone) and 100 g/ml streptomycin (Hyclone). The medium was changed
every 3 days. After confluence was attained, the cells were detached by mild treatment with trypsin (0.25% [w/v], 5 min, 37 ◦ C) and replated at one-third of the confluent density for continued passage.
2.3. Adipogenic differentiation Adipocytic differentiation of human MSCs was performed as previously described (Pittenger et al., 1999). Four-day postconfluent cells were incubated in adipogenesis-inducing medium (AIM) containing 1 M dexamethasone, 0.2 mM indomethacin, 0.5 mM isobutyl-methylxanthine, 0.01 mg/ml insulin and 10% (v/v) FCS in DMEM. After 72 h, the AIM was exchanged for adipogenesismaintenance medium (AMM) containing insulin (0.01 mg/ml) and 10% (v/v) FBS in DMEM for 24 h and then switched back to AIM. This exchange of the two media was repeated three times, after which the cells were maintained in AMM for 1 week. For Oil red O staining, the cells were washed twice with PBS and fixed with 3.7% (v/v) formaldehyde in PBS for 1 h at room temperature. The cells were next stained for 1 h at room temperature with filtered Oil red O solution (0.3% [w/v] Oil red O in 60% [v/v] isopropanol), washed twice with distilled water, visualized under light microscopy and photographed. To extract the incorporated Oil red O, 1 ml isopropanol was added to each well followed by 15 min of shaking at room temperature. After appropriate dilution, the absorbance of triplicate samples was read at 510 nm.
2.4. Osteogenic differentiation Osteogenic differentiation of subconfluent MSC monolayers was induced by 100 nM dexamethasone, 0.05 mM ascorbic acid, 10 mM sodium glycerophosphate and 10% (v/v) FCS, as previously described (Pittenger et al., 1999). For Alizarin Red S (ARS) staining, cells were rinsed with PBS and fixed in ice-cold 70% (v/v) ethanol for 1 h. After a brief wash with water, the cells were stained for 10 min with 40 mM ARS solution (pH 4.2) at room temperature. Next, the cells were rinsed five times with water followed by a 15-min wash with PBS (with rotation) to reduce nonspecific ARS staining. For the calcium assay, cells were washed twice with PBS and incubated overnight in 1 ml of 0.5-N HCl with gentle shaking. After centrifugation at 2000 × g for 10 min, the supernatant was assayed for calcium, following the protocol of the Calcium (CPC) LiquiColor kit (Stanbio Laboratory, USA). The total calcium levels were calculated using a standard curve prepared by employing solutions of known calcium concentration. The protein content was quantified using the DC protein assay from Bio-Rad, employing bovine serum albumin as a standard. The results were expressed as mg calcium/mg cellular protein. To measure alkaline phosphatase (ALP) activity, cells were washed three times with PBS and scraped into a lysis solution containing 10 mM Tris–HCl (pH 7.4), 0.1% (v/v) Triton X-100 and 0.5 mM MgCl2 . Each lysate was sonicated and then centrifuged at 16,000 × g for 10 min at 4 ◦ C. The supernatants were recovered for the assay of ALP activity by the hydrolysis of 5 mM p-nitrophenyl phosphate (for 15 min at 37 ◦ C), followed by colorimetric detection of the product p-nitrophenol at 405 nm. The results were expressed as nanomoles p-nitrophenol produced per mg of cellular protein per min.
2.5. Combined adipogenesis and osteogenesis in culture For this experiment, hMSCs were induced to differentiate into both adipocytes and osteoblasts in the same dish, using a DLP
W.-H. Yu et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 377–384
medium consisting of a mixture of adipogenic and osteogenic medium (1/1, v/v), as described above. 2.6. Real-time PCR Total RNA was isolated using the TRIzol reagent (Life Technologies, Gaithersburg, MD) according to the manufacturer’s instructions. After digestion with DNase I (Fermentas, Hanover, MD), 1 g total RNA was reverse transcribed using a RevertAidTM first-strand complementary DNA synthesis kit (Fermentas). The mRNA levels from the indicated genes were analyzed in triplicate using the QuantiTect SYBR Green PCR Master Mix (Qiagen, Valencia, CA); reactions proceeded in an ABI 7500 real-time PCR instrument (Applied Biosystems, USA) following the manufacturer’s instructions. The mRNA levels were normalized to those of GAPDH (an internal control), and changes in gene expression were calculated as-fold changes (the Ct method). The primers used were as follows: GAPDH (forward), 5 -GAAGGTGAAGGTCGGAGTC-3 ; GAPDH (reverse), 5 -GAAG ATGGTGATGGGATTTC-3 ; PPAR␥ (forward), 5 -CGAGAAGGAGAAGCTGT TGG-3 ; PPAR␥ (reverse), 5 -TCAGCGGGAAGGACTTTATGTATG-3 ; FABP4 (forward), 5 -AGCACCATAACCTTAGATGGGG-3 ; FABP4 (reverse), 5 -CGTGGAAGTGACGCCTTTCA-3 ; GPDH (forward), 5 -AGGAAGACATTGGAGGCAAAAA-3 ; GPDH (reverse), 5 -GCAGCCTGGACCACATCTG-3 ; LPL (forward), 5 -ACAAGAGAGAACCAGACTCCAA-3 ; LPL (reverse), 5 -AGGGTAGTTAAACTCCTCCTCC-3 ; osteocalcin (forward), 5 -CACTCCTCGCCCTATTGGC-3 ; osteocalcin (reverse), 5 -CCCTCCTGCTTGGACACAAAG-3 ; collagen I (forward), 5 -GCCGTGACCTCAAGATGTG-3 ; collagen I (reverse), 5 -GCCGAACCAGACATGCCTC-3 ; Runx2 (forward), 5 -AGATGATGACACTGCCACCTCTG-3 ; Runx2 (reverse), 5 -GGGATGAAATGCTTGGGAACTGC-3 .
2.7. Immunoblotting Cells were washed with cold PBS and directly lysed in Laemmli buffer. Each lysate was sonicated and then centrifuged at 16,000 × g for 10 min at 4 ◦ C. Each supernatant was recovered as a total cell lysate. Equal amounts of protein (20 g) were separated on 8–10% (w/v) SDS-PAGE and then electrotransferred to 0.45-m pore-sized polyvinylidene difluoride membranes (Millipore, Bedford, MA). After the transfer, each membrane was blocked using a solution of 0.1% (v/v) Tween 20/TBS (TBS/T) containing 5% (w/v) non-fat milk powder for 1 h at room temperature and then incubated with appropriate primary antibodies overnight at 4 ◦ C. Specifically bound primary antibodies were detected using peroxidase-coupled secondary antibodies and enhanced chemiluminescence signaling (Cell Signaling Technologies, Beverly, MA).
379
and the non-targeting scrambled shRNA sequence was 5 AACAAGATGAAGAGCACCAA-3 (NTC). All oligonucleotides were synthesized by Sangon (Shanghai, China) and were annealed and subcloned into the pLL3.7 vector. The lentivirus was prepared according to the manufacturer’s protocol (Invitrogen). Briefly, subconfluent 293FT cells were cotransfected with PLL3.7 and a compatible packaging plasmid mixture (Virapower lentiviral packaging system) using Lipofectamine 2000 (Invitrogen). The lentivirus-containing supernatants from the transfected cells were collected 48–72 h later, filtered and ultracentrifuged at 50,000 × g for 90 min. Each enriched virus pellet was resuspended in PBS, aliquoted and stored at −70 ◦ C. Viral titers were determined by the infection of 293FT cells with serial dilutions of vector stocks, followed by fluorescence-activated cell sorter (FACS)-mediated analysis of EGFP-positive cells. The transduction of MSCs proceeded at a multiplicity of infection (MOI) of 10 in the presence of 8 g/ml polybrene (Sigma).
3. Results 3.1. Expression of PPAR during adipogenic and osteogenic differentiation of human MSCs Adipogenesis of MSCs was induced by the addition of dexamethasone, insulin, indomethacin and IBMX. As we have earlier shown, three rounds of induction followed by maintenance treatment increased the numbers of rounded, lipid-laden adipocytes. PPAR␥ is synthesized in two isoforms, PPAR␥1 and the adipocyte-specific PPAR␥2. As the first step in the elucidation of PPAR␥ isoform regulation, we analyzed the temporal expression of the two PPAR␥ isoforms during hMSC differentiation into adipocytes to determine whether one factor was preferentially induced and might thus serve as an early controller of differentiation pathway. Representative immunoblots were shown in Fig. 1A. The induction patterns of the two isoforms of PPAR␥, PPAR␥1 and PPAR␥2, were different. Both isoforms were undetectable in confluent MSCs. Expression of PPAR␥1 was detected 2 days after initiation of adipocyte differentiation, whereas PPAR␥2 appeared later in the differentiation process (on day 6). Interestingly, a switch to adipogenesis-maintenance medium (AMM) transiently inhibited both PPAR␥1 and PPAR␥2 expression and, during the final AMM treatment, expression of both isoforms of PPAR␥ was greatly decreased. In addition, PPAR␥1 was the predominant isoform in hMSC-derived adipocytes. Osteogenesis of hMSCs was induced by the addition of dexamethasone, ascorbic acid and sodium glycerophosphate. We next examined the temporal expression of PPAR␥ during osteogenic differentiation of hMSCs. As shown in Fig. 1B, PPAR␥1 started to increase in level on day 2 after exposure to osteogenic inducers and continued to rise, to attain a maximal level at the end of differentiation. However, PPAR␥2 was not detected at any time during hMSC osteogenic differentiation.
2.8. shRNA transduction The knockdown of endogenous genes was accomplished using short hairpin RNA (shRNA) and PLL3.7 lentiviral vectors carrying the Pol III-dependent U6 promoter (a kind gift of Dr. David L. Garbers, University of Texas Southwestern Medical Center, Dallas, TX). The target sequences were selected with the aid of Dharmacon software. The two sequences targeting PPAR␥ were 5 -GCCCTTCACTACTGTTGAC-3 (PPAR-A) and 5 -GAAGTTCAATGCACTGGAA-3 (PPAR-B); the two targeting sequences for C/EBP␣ were 5 -GCTGGAGCTGACCAGTGACAA3 (CEBP-A) and 5 -GCACGAGACGTCCATCGACAT-3 (CEBP-B);
3.2. BADGE and GW9662, potential antagonists of PPAR, inhibited hMSC adipogenesis but did not promote osteogenesis BADGE, a potential antagonist of PPAR␥, has been reported to inhibit 3T3-L1 adipogenesis (Wright et al., 2000). GW9662 is another potential antagonist of PPAR␥ (David et al., 2007). In the present study, the effects of BADGE and GW9662 on hMSC adipogenesis and osteogenesis were investigated. In adipogenic induction medium, both BADGE and GW9662 greatly inhibited adipogenesis, as shown by Oil red O staining (Fig. 2). No osteogenesis was
380
W.-H. Yu et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 377–384
Fig. 1. Expression of the two isoforms of PPAR␥ during the processes of adipogenesis and osteogenesis. Human MSCs were induced to differentiate into adipocytes (A) or osteoblasts (B) and were harvested at the indicated periods of time. Equal amounts of cell extracts were immunoblotted using antibodies specific for PPAR␥. GAPDH was used as a loading control. Representative data from three separate experiments are shown.
evident, as revealed by results of the alkaline phosphatase (ALP) and calcium assays (data not shown). 3.3. Knockdown of PPAR inhibited adipogenesis of human MSCs but did not promote osteogenesis To further investigate the role of PPAR␥ in adipogenesis and osteogenesis, a lentivirus-based shRNA knockdown system was used. MSCs were transduced with lentivirus containing either PPAR␥-specific shRNA (shPPAR-A and shPPAR-B) or non-target control shRNA (shNTC) and then subjected to adipogenic induction. The lentivirus transduction efficiency of MSCs was at least 90%, as assessed by counting the numbers of cells positive for EGFP, a cassette contained within the lentiviral vector pLL3.7. Fig. 3A and B show the knockdown of PPAR␥ confirmed by immunoblotting and real-time PCR analysis. Compared to MSCs infected with shNTC, shPPAR-expressing MSCs showed greatly impaired lipid accumulation, as revealed by Oil red O staining (Fig. 3C). The knockdown of PPAR␥ did not cause cytoskeletal changes; cells retained the round phenotype of early adipocyte differentiation. The inhibition of adipogenic differentiation by PPAR␥ suppression was further
confirmed by real-time PCR analysis of the expression of the adipogenic genes LPL, FABP4 and GPDH (Fig. 3D). Under conditions conducive to adipogenic induction, PPAR␥ suppression did not cause spontaneous calcium deposition at any time during the entire adipogenic process (data not shown). We found that ALP activity was slightly decreased upon PPAR␥ inactivation, compared with what was seen in non-target controls (Fig. 4A) (p > 0.05). To exclude the possibility that the lack of osteogenesis seen in MSCs deficient in PPAR␥ was caused by the absence of osteogenic stimuli, MSCs transduced with lentivirus containing either PPARspecific shRNA (shPPAR-A and shPPAR-B) or non-target control shRNA (shNTC) were allowed to differentiate in the presence of osteogenic stimuli. As shown in Fig. 4B–D, knockdown of PPAR␥ with PPAR␥-specific shRNAs (shPPAR-A and shPPAR-B) only slightly impaired the osteogenic differentiation of hBMSCs, as demonstrated by ARS staining and the results of the calcium and ALP activity assays (p > 0.05). The effect of PPAR␥ knockdown on the osteogenic differentiation of hBMSCs was further confirmed by real-time PCR analysis of the expression of osteogenic genes, including osteocalcin, collagen I and Runx2 (Fig. 4E). To further confirm and extend the observed effects of PPAR␥ knockdown on MSC commitment to either the osteoblast or adipocyte lineage, we used a dual-lineage-promoting (DLP) medium that allows both cell types to develop in the same dish. We found that, under such conditions, the knockdown of PPAR␥ only slightly inhibited hMSC osteogenesis, as revealed by the results of the calcium and ALP activity assays (p > 0.05; data not shown). 3.4. PPAR suppression inhibited expression of the adipogenic transcription factor C/EBP˛ but not that of the osteogenic transcription factor Runx2
Fig. 2. BADGE and GW9662, the inhibitors of PPAR␥, inhibited adipogenesis. Four days postconfluence, human MSCs were induced to differentiate into adipocytes in the presence of DMSO, 100 nM BADGE or 10 M GW9662. At the end of the culture period, the degree of adipocyte differentiation was determined by Oil red O staining. The lower panel shows the staining of whole dishes. Representative data from three separate experiments are shown. The incorporated Oil red O was extracted with isopropanol, and the quantification was performed as described in Section 2. The results are expressed as the mean ± SEM of three independent experiments (upper panel). A significant difference compared with cells treated with DMSO (p < 0.05) is indicated by *.
It has been reported that PPAR␥ agonists and overexpression of PPAR␥ induce differentiation of BM stromal cells into the adipocyte lineage and negatively regulate osteoblast differentiation by repressing Runx2 (Lecka-Czernik et al., 1999). To investigate the relationship between Runx2 and PPAR␥ expression under conditions of hMSC adipogenesis and osteogenesis, the effect of PPAR␥ knockdown on Runx2 expression was evaluated by Western blotting. As shown in Fig. 4F, Runx2 synthesis was not significantly affected by knockdown of PPAR␥. PPAR␥ and C/EBP␣ are the two transcription factors critical for adipogenesis. As shown in Fig. 5A, the knockdown of PPAR␥ greatly inhibited the expression of C/EBP␣. These two transcription factors have been reported to be mutually regulatory during murine adipogenesis (Rosen and MacDougald, 2006). Thus, to explore the relationship between the transcription factors during hMSC adipogenesis, we constructed lentiviral shRNA specific to C/EBP␣
W.-H. Yu et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 377–384
381
Fig. 3. Lentivirus-mediated PPAR␥ knockdown inhibited hMSC adipogenesis. MSCs were transduced with lentivirus containing either PPAR␥-specific shRNA (shPPAR-A and shPPAR-B) or non-targeting shRNA (shNTC) and then subjected to adipogenic induction. After 7 days of adipocyte induction, the MSCs were harvested and real time PCR was used to analyze the mRNA levels of PPAR␥ (A), LPL, FABP4 and GPDH (D). The mRNA levels were normalized to GAPDH (internal control), and gene expression was presented as-fold changes (Ct method). Representative data from three separate experiments are shown. After seven days of adipocyte induction, cell extracts were immunoblotted using an anti-PPAR␥ antibody. GAPDH was used as a loading control. Representative data from three separate experiments are shown (B). At the end of the culture period, the degree of adipocyte differentiation was determined by Oil red O staining (C). The lower panel shows the staining of whole dishes. Representative data from three separate experiments are shown. The incorporated Oil red O was extracted with isopropanol, and the quantification was performed as described in Section 2. The results are expressed as the mean ± SEM of three independent experiments (upper panel). A significant difference compared with cells infected with shNTC (p < 0.05) is indicated by *.
(shCEBP-A and shCEBP-B) and introduced these shRNA constructs or shNTC into MSCs. The knockdown of C/EBP␣ was confirmed by immunoblotting (Fig. 5D). Downregulation of C/EBP␣ essentially abrogated hMSC adipogenesis compared with that of non-target controls, as revealed by Oil red O staining (Fig. 5B) and real-time PCR analysis of the expression of the adipogenic genes LPL, FABP4 and GPDH (Fig. 5C). Interestingly, we found that the knockdown of C/EBP␣ inhibited PPAR␥2 to a greater extent than PPAR␥1 (Fig. 5D). 4. Discussion A few reports have suggested that PPAR␥ regulates mesenchymal lineage allocation, favoring adipocyte over osteoblast development, and that PPAR␥ may be a useful target of drugs intended to enhance bone mass. However, little direct evidence has been reported on the effect of PPAR␥ antagonists on human mesenchymal lineage allocation. Bone loss induced by Thiazolidinedione (TZD) has not been reversible. In the present study, we showed that BADGE and GW9662, inhibitors of PPAR␥, reduced the extent of adipogenesis, but did not significantly affect osteogenesis. Similar results were obtained upon lentivirus-mediated PPAR␥ knockdown. The fact that PPAR␥ suppression inhibited C/EBP␣ but did not significantly affect expression of the osteogenic transcription factor Runx2 may help to explain the mechanism underlying the choice of differentiation pathway. Although PPAR␥1 expression was detected during osteogenesis, the presence of PPAR␥ RNAi had no obvious effect on osteogenesis under permissive conditions, indicating that PPAR␥1 may not play a critical role in hMSC osteogenesis (Bruedigam et al., 2008).
During adipogenesis, both PPAR␥1 and PPAR␥2 were induced in hMSCs, with PPAR␥1 being the predominant isoform. It is noteworthy that the suppression of C/EBP␣ substantially reduced the level of adipocyte formation to an extent similar to that observed upon PPAR␥ suppression, whereas the knockdown of C/EBP␣ inhibited the activation of PPAR␥2 but not PPAR␥1, suggesting that PPAR␥2 may play the more important role in hMSC adipogenesis. It is thus conceivable that the inhibition of PPAR␥2 reduced adipogenesis but did not affect osteogenesis. Unfortunately, the isoform-specific knockdown of PPAR␥ was not successful in our experiment, due to the great similarity of the sequences of these two isoforms. Further assays using other methods are required to elucidate the respective roles of these two isoforms. Our findings contradict several lines of evidence suggesting that PPAR␥ insufficiency directly promotes the differentiation of osteoblasts. It is possible that a between-species difference is in play. Indeed, there exist important differences between murine MSCs and human MSCs regarding their surface phenotype and immunological properties (Madec et al., 2009). In terms of their differentiation capacity, hMSCs have been observed to have a significant and robust in vitro osteogenic capacity compared with mMSCs. Retinoic acid (RA) is shown to be required for the efficient osteogenesis of mMSCs (Wan et al., 2006). RA is a ligand for nuclear receptors, including retinoic acid receptors (RAR). Retinoic acid can inhibit PPAR␥ expression and activation, presumably by enhancing the heterodimerization of RAR with retinoid X receptor (RXR), thereby making RXR unavailable for dimerization with PPAR␥ (Mandrup and Lane, 1997). High-affinity binding to DNA by PPAR␥ appears to absolutely require dimerization with RXR.
382
W.-H. Yu et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 377–384
Fig. 4. Lentivirus-mediated PPAR␥ knockdown did not significantly affect hMSC osteogenesis. (A) MSCs were transduced with lentivirus containing either PPAR␥-specific shRNA (shPPAR-A and shPPAR-B) or non-targeting shRNA (shNTC) and then subjected to adipogenic induction, and there was no apparent osteogenesis. On the seventh day, the cells were harvested and the alkaline phosphatase activity was determined as described in Section 2. (B–E) The transduced MSCs were subjected to osteogenic induction, and there was no significant difference in osteogenesis compared with cells infected with shNTC. On the seventh day, the cells were harvested and the alkaline phosphatase activity was determined (B). At the end of osteogenic differentiation, the amount of calcium deposition was determined in the treated cells, and the results are expressed as the mean ± SEM of three independent experiments (C). Additionally, at the end of osteogenesis, the formation of bone nodules was assayed by ARS staining. Representative data from three separate experiments are shown (D). Furthermore, at the end of osteogenic differentiation, MSCs were harvested, and real time PCR was used to analyze the mRNA level of osteocalcin, collagen I and Runx2. The mRNA levels were normalized to GAPDH (internal control), and gene expression was presented as-fold changes (Ct method). Representative data from three separate experiments are shown (E). On the seventh day, cells were harvested and the cell extracts were immunoblotted using an anti-Runx2 antibody. GAPDH was used as a loading control. Representative data from three separate experiments are shown (F).
Therefore, the differential response to RA between murine MSCs and hMSCs may be the underlying mechanism for our observation. Notably, although PPAR␥ is required for hMSC adipogenesis, the ectopic expression of PPAR␥ in hMSCs did not induce full adipogenesis (Qian et al., 2010). This finding is not consistent with what has been observed in rodent adipogenesis (Rosen and MacDougald, 2006). Although some indirect insight into PPAR␥ function can be gained from the increased fracture risk associated with the use of synthetic PPAR␥ ligands, the thiazolidinediones (TZDs), the divergent effects of TZDs on human and rodent adipogenesis and osteogenesis suggest that the regulation of skeletal phenotype development cannot be attributed to PPAR␥ alone (Lecka-Czernik et al., 2002). Although some studies of human PPAR␥ gene polymorphisms suggest a role for PPAR␥ in the regulation of bone mass, evidence on the association of a certain PPAR␥ genomic polymorphism with bone status is unconvincing (Cheng et al., 2006; Kim et al., 2007). Recently, accumulating evidence points toward a role for PPAR␥ in osteoclast development (Mbalaviele et al., 2000; Wahli, 2008; Wan et al., 2007), indicating that the phenotype of PPAR␥ knockout mice should be more carefully investigated, as
should the effect of TZDs on bone density. Bone loss after rosiglitazone administration to ovariectomized rats has been shown to result from an increase in bone resorption (Sottile et al., 2004). Few studies have explored the effects of TZDs on hMSC adipogenesis and osteogenesis. Rosiglitazone has been reported to stimulate hMSC adipogenesis, but any effect on osteogenesis remains controversial (Benvenuti et al., 2007; Bruedigam et al., 2010). We could not find any significant effect of rosiglitazone on hMSC osteogenesis (data not shown). One recent report showed that rosiglitazone promoted both adipogenesis and osteogenesis, and rosiglitazone action in the BM may promote osteoblast apoptosis (Bruedigam et al., 2010), thus increasing fracture risk. There is a need to gather more direct evidence on the role played by PPAR␥ in human MSC lineage allocation. Transcription coregulators, such as TAZ, play vital roles in control of the expression and activity of PPAR␥ and Runx2 (Cui et al., 2003). TAZ is the transcriptional co-activator of both factors. The outcome of the interactions of PPAR␥ and Runx2 with TAZ differs substantially in that Runx2 transcription is stimulated, whereas that of PPAR␥ is repressed. Indeed, recent evidence has implicated TAZ as a molecular switch in MSC differentiation (Hong et al., 2005).
W.-H. Yu et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 377–384
383
Fig. 5. Cross-regulation of PPAR␥ and C/EBP␣, the transcription factors of adipogenesis, during the adipocyte differentiation of MSCs. (A) Lentivirus-mediated PPAR␥ knockdown inhibited C/EBP␣. The MSCs were transduced with lentivirus containing either PPAR␥-specific shRNA (shPPAR-A and shPPAR-B) or non-targeting shRNA (shNTC) and then subjected to adipogenic induction. After seven days of adipocyte induction, the cell extracts were immunoblotted using an anti-C/EBP␣ antibody. GAPDH was used as a loading control. Representative data from three separate experiments are shown. (B–D) Lentivirus-mediated C/EBP␣ knockdown inhibited PPAR␥2 and adipogenesis. At the end of culture, the degree of adipocyte differentiation was determined by Oil red O staining (B). The lower panel shows the staining of whole dishes. Representative data from three separate experiments are shown. The incorporated Oil red O was extracted with isopropanol and the quantification was performed as described in Section 2. The results are expressed as the mean ± SEM of three independent experiments (upper panel). A significant difference compared with cells infected with shNTC (p < 0.05) is indicated by *. After seven days of adipocyte induction, MSCs were harvested and real time PCR was used to analyze the mRNA level of LPL, FABP4 and GPDH. The mRNA levels were normalized to GAPDH (internal control), and gene expression was presented as-fold changes (Ct method). Representative data from three separate experiments are shown (C). After seven days of adipocyte induction, cell extracts were immunoblotted using anti-PPAR␥ and anti-C/EBP␣ antibodies. GAPDH was used as a loading control. Representative data from three separate experiments are shown (D).
Additional studies, particularly in vivo, will provide further insights into this process. In conclusion, the knockdown of PPAR␥ inhibited adipogenesis but did not stimulate osteogenesis. To identify the master controls of stem-cell lineage allocation, new model systems and more extensive studies at the genetic, molecular and cellular levels are needed. Acknowledgements This research was supported by the National Natural Science Foundation of China (81000362 and 30928015), National Basic Research Program of China (2009CB522100, 2010CB945400 and 2011CBA01100) the Key Scientific and Technological Projects of Guangdong Province (2007A032100003), the Natural Science Foundation of Guangdong Province (4203004), the Key Scientific and Technological Program of Guangzhou City (Nos. 2008A1E4011-5 and 2010U1-E00551), the Program for New Century Excellent Talents in University and the Fundamental Research Funds for the Central Universities (09ykpy79 and 10ykzd02). The funding agencies had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. References Benvenuti S, Cellai I, Luciani P, Deledda C, Baglioni S, Giuliani C, et al. Rosiglitazone stimulates adipogenesis and decreases osteoblastogenesis in human mesenchymal stem cells. J Endocrinol Invest 2007;30:RC26–30.
Botolin S, McCabe LR. Inhibition of PPARgamma prevents type I diabetic bone marrow adiposity but not bone loss. J Cell Physiol 2006;209:967–76. Bruedigam C, Eijken M, Koedam M, van de Peppel J, Drabek K, Chiba H, et al. A new concept underlying stem cell lineage skewing that explains the detrimental effects of thiazolidinediones on bone. Stem Cell 2010;28:916–27. Bruedigam C, Koedam M, Chiba H, Eijken M, van Leeuwen JP. Evidence for multiple peroxisome proliferator-activated receptor gamma transcripts in bone: fine-tuning by hormonal regulation and mRNA stability. FEBS Lett 2008;582: 1618–24. Cheng S, Afif H, Martel-Pelletier J, Benderdour M, Pelletier JP, Hilal G, et al. Association of polymorphisms in the peroxisome proliferator-activated receptor gamma gene and osteoarthritis of the knee. Ann Rheum Dis 2006;65: 1394–7. Cui CB, Cooper LF, Yang X, Karsenty G, Aukhil I. Transcriptional coactivation of bone-specific transcription factor Cbfa1 by TAZ. Mol Cell Biol 2003;23: 1004–13. David V, Martin A, Lafage-Proust MH, Malaval L, Peyroche S, Jones DB, et al. Mechanical loading down-regulates peroxisome proliferator-activated receptor gamma in bone marrow stromal cells and favors osteoblastogenesis at the expense of adipogenesis. Endocrinology 2007;148:2553–62. Elabd C, Basillais A, Beaupied H, Breuil V, Wagner N, Scheideler M, et al. Oxytocin controls differentiation of human mesenchymal stem cells and reverses osteoporosis. Stem Cell 2008;26:2399–407. Farmer SR. Transcriptional control of adipocyte formation. Cell Metab 2006;4:263–73. Hong JH, Hwang ES, McManus MT, Amsterdam A, Tian Y, Kalmukova R, et al. TAZ, a transcriptional modulator of mesenchymal stem cell differentiation. Science 2005;309:1074–8. Kawai M, Rosen CJ. PPARgamma: a circadian transcription factor in adipogenesis and osteogenesis. Nat Rev Endocrinol 2010;6:629–36. Kim TH, Hong JM, Park EK, Kim SY. Peroxisome proliferator-activated receptorgamma gene polymorphisms are not associated with osteonecrosis of the femoral head in the Korean population. Mol Cell 2007;24:388–93. Lecka-Czernik B, Gubrij I, Moerman EJ, Kajkenova O, Lipschitz DA, Manolagas SC, et al. Inhibition of Osf2/Cbfa1 expression and terminal osteoblast differentiation by PPARgamma2. J Cell Biochem 1999;74:357–71. Lecka-Czernik B, Moerman EJ, Grant DF, Lehmann JM, Manolagas SC, Jilka RL. Divergent effects of selective peroxisome proliferator-activated
384
W.-H. Yu et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 377–384
receptor-gamma 2 ligands on adipocyte versus osteoblast differentiation. Endocrinology 2002;143:2376–84. Madec AM, Mallone R, Afonso G, Abou Mrad E, Mesnier A, Eljaafari A, et al. Mesenchymal stem cells protect NOD mice from diabetes by inducing regulatory T cells. Diabetologia 2009;52:1391–9. Mandrup S, Lane MD. Regulating adipogenesis. J Biol Chem 1997;272:5367–70. Mbalaviele G, Abu-Amer Y, Meng A, Jaiswal R, Beck S, Pittenger MF, et al. Activation of peroxisome proliferator-activated receptor-gamma pathway inhibits osteoclast differentiation. J Biol Chem 2000;275:14388–93. Nuttall ME, Gimble JM. Controlling the balance between osteoblastogenesis and adipogenesis and the consequent therapeutic implications. Curr Opin Pharmacol 2004;4:290–4. Pittenger MF, Mackay AM, Beck SC, Jaiswal RK, Douglas R, Mosca JD, et al. Multilineage potential of adult human mesenchymal stem cells. Science 1999;284:143–7. Qian SW, Li X, Zhang YY, Huang HY, Liu Y, Sun X, et al. Characterization of adipocyte differentiation from human mesenchymal stem cells in bone marrow. BMC Dev Biol 2010;10:47. Rosen ED, MacDougald OA. Adipocyte differentiation from the inside out. Nat Rev Mol Cell Biol 2006;7:885–96. Schiller PC, D’Ippolito G, Brambilla R, Roos BA, Howard GA. Inhibition of gapjunctional communication induces the trans-differentiation of osteoblasts to an adipocytic phenotype in vitro. J Biol Chem 2001;276:14133–8.
Sottile V, Seuwen K, Kneissel M. Enhanced marrow adipogenesis and bone resorption in estrogen-deprived rats treated with the PPARgamma agonist BRL49653 (rosiglitazone). Calcif Tissue Int 2004;75:329–37. Wahli W. PPAR gamma: ally and foe in bone metabolism. Cell Metab 2008;7:188–90. Wan DC, Shi YY, Nacamuli RP, Quarto N, Lyons KM, Longaker MT. Osteogenic differentiation of mouse adipose-derived adult stromal cells requires retinoic acid and bone morphogenetic protein receptor type IB signaling. Proc Natl Acad Sci U S A 2006;103:12335–40. Wan Y, Chong LW, Evans RM. PPAR-gamma regulates osteoclastogenesis in mice. Nat Med 2007;13:1496–503. Wolf NS, Penn PE, Rao D, McKee MD. Intraclonal plasticity for bone, smooth muscle, and adipocyte lineages in bone marrow stroma fibroblastoid cells. Exp Cell Res 2003;290:346–57. Wright HM, Clish CB, Mikami T, Hauser S, Yanagi K, Hiramatsu R, et al. A synthetic antagonist for the peroxisome proliferator-activated receptor gamma inhibits adipocyte differentiation. J Biol Chem 2000;275:1873–7. Yamashita A, Takada T, Nemoto K, Yamamoto G, Torii R. Transient suppression of PPARgamma directed ES cells into an osteoblastic lineage. FEBS Lett 2006;580:4121–5. Zieleniak A, Wojcik M, Wozniak LA. Structure and physiological functions of the human peroxisome proliferator-activated receptor gamma. Arch Immunol Ther Exp (Warsz) 2008;56:331–45.
Contents lists available at SciVerse ScienceDirect
The International Journal of Biochemistry & Cell Biology journal homepage: www.elsevier.com/locate/biocel
PPAR␥ suppression inhibits adipogenesis but does not promote osteogenesis of human mesenchymal stem cells Wei-Hua Yu a,1 , Fu-Gui Li a,1 , Xiao-Yong Chen a , Jian-Tao Li a , Yan-Heng Wu b , Li-Hua Huang c , Zhen Wang d , Panlong Li a , Tao Wang a , Bruce T. Lahn a,e , Andy Peng Xiang a,f,g,∗ a Center for Stem Cell Biology and Tissue Engineering, The Key Laboratory for Stem Cells and Tissue Engineering, Ministry of Education, Sun Yat-Sen University, 74# Zhongshan Road 2, Guangzhou, Guangdong 510080, PR China b Medical Imaging Department, The First Affiliated Hospital of Jinan University, No. 613, The West of Huangpu Street, Guangzhou, Guangdong 510630, PR China c Key Laboratory of Modulating Liver to Treat Hyperlipemia SATCM/Level 3 Laboratory of Lipid Metabolism SATCM, Guangdong Pharmaceutical University, Guangzhou, Guangdong 510006, PR China d Institute of Blood Transfusion, Guangzhou Blood Center, Guangzhou, Guangdong 510095, PR China e Department of Human Genetics and Howard Hughes Medical Institute, University of Chicago, 929 E. 57th ST. #W503, Chicago, IL 60637, USA f Department of Biochemistry, Zhongshan Medical School, Sun Yat-sen University, 74# Zhongshan Road 2, Guangzhou, Guangdong 510080, PR China g Cell-gene Therapy Translational Medicine Research Center, The Third Affiliated Hospital, Sun Yat-sen University, Guangzhou 510630, PR China
a r t i c l e
i n f o
Article history: Received 28 June 2011 Received in revised form 14 November 2011 Accepted 15 November 2011 Available online 23 November 2011 Keywords: PPAR␥ RNAi Human mesenchymal stem cells Adipogenesis Osteogenesis
a b s t r a c t Mesenchymal stem cells (MSCs) are the common progenitors of osteoblasts and adipocytes. A reciprocal relationship exists between osteogenesis and adipogenesis in the bone marrow, and the identification of signaling pathways that stimulate MSC osteogenesis at the expense of adipogenesis is of great importance from the viewpoint of developing new therapeutic treatments for bone loss. The adipogenic transcription factor peroxisome proliferator-activated receptor ␥ (PPAR␥) has been reported to play a vital role in modulating mesenchymal lineage allocation within the bone marrow compartment, stimulating adipocyte development at the expense of osteoblast differentiation. Hence, PPAR␥ may be a valuable target for drugs intended to enhance bone mass. However, little direct evidence is available for the role played by PPAR␥ in human mesenchymal lineage allocation. In this study, using human MSCs as an in vitro model, we showed that the two isoforms of PPAR␥, PPAR␥1 and PPAR␥2, were differentially induced during hMSC adipogenesis, whereas only PPAR␥1 was detected during osteogenesis. BADGE and GW9662, two potential antagonists of PPAR␥, as well as lentivirus-mediated knockdown of PPAR␥, inhibited hMSC adipogenesis but did not significantly affect osteogenesis. PPAR␥ knockdown did not significantly influence the expression level of the osteogenic transcription factor Runx2. Together, these results suggest that PPAR␥ is not the master factor regulating mesenchymal lineage determination in human bone marrow. © 2011 Elsevier Ltd. All rights reserved.
1. Introduction An increase in the content of bone marrow (BM) adipose tissue and a decrease in bone volume is observed in patients with
Abbreviations: MSCs, mesenchymal stem cells; BADGE, bisphenol-A-diglycidyl ether; ALP, alkaline phosphatase; LPL, lipoprotein lipase; PPAR␥, peroxisome proliferator-activated receptor ␥; C/EBP␣, CCAAT/enhancer binding protein ␣; AIM, adipocyte induction medium; AMM, adipocyte maintenance medium; GPDH, glycerol-3-phosphate dehydrogenase; FABP4, fatty acid-binding protein 4; ARS, alizarin red S. ∗ Corresponding author at: Center for Stem Cell Biology and Tissue Engineering, The Key Laboratory for Stem Cells and Tissue Engineering, Ministry of Education, Sun Yat-Sen University, 74# Zhongshan Road 2, Guangzhou, Guangdong 510080, PR China. Tel.: +86 20 87335822; fax: +86 20 87335858. E-mail address: [email protected] (A.P. Xiang). 1 These authors made equal contributions to this work. 1357-2725/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.biocel.2011.11.013
age-related osteopenia, as well as in individuals subject to other conditions such as osteoporosis, immobilization, microgravity, ovariectomy, diabetes, or glucocorticoid treatment (Nuttall and Gimble, 2004). Marrow adipocytes share a common MSC origin with bone-forming osteoblasts, as illustrated by the ability of MSCs to differentiate into either lineage (Pittenger et al., 1999). Increasing evidence of transdifferentiation in such cells suggests a large element of plasticity in the mutual transformation of these cell types (Schiller et al., 2001; Wolf et al., 2003). Therefore, any factor controlling the balance between adipogenesis and osteogenesis from BM–MSCs may serve as a target of pharmacological intervention to control bone disorders. The activation of certain transcription factors is essential for cellular commitment to a particular differentiation lineage (Farmer, 2006). PPAR␥ is a member of the nuclear hormone receptor gene superfamily of ligand-activated transcription factors. PPAR␥ is commonly termed the master regulator of adipogenesis and is both
378
W.-H. Yu et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 377–384
necessary and sufficient for adipogenesis; no factor has yet been identified that can induce normal adipogenesis in the absence of PPAR␥. Furthermore, the ectopic expression and activation of the factor is sufficient to induce adipocyte differentiation (Kawai and Rosen, 2010). It has long been hypothesized that PPAR␥ plays a vital role in mesenchymal lineage determination. PPAR␥ haploinsufficient mice showed an increase in bone mass associated with a loss of adipose tissue volume. Primary BM–MSCs from such animals showed enhanced osteogenic but reduced adipogenic differentiation; homozygous PPAR␥-deficient murine ES cells were unable to differentiate into adipocytes but spontaneously differentiated into osteoblasts. Torii reported the differentiation of ES cells into an osteoblastic lineage using siRNA against PPAR␥ without the addition of any osteogenic factors (Yamashita et al., 2006). These findings suggest that PPAR␥ can regulate lineage allocation within the BM, favoring adipocyte development over that of osteoblasts, and that PPAR␥ could serve as a useful target for drugs intended to enhance bone mass. Because of the multiple roles played by PPAR␥ in human physiology (Zieleniak et al., 2008), including the mediation of insulin sensitivity, attempts to design drugs that suppress PPAR␥ expression in BM stromal cells without side effects have not been wholly successful. Intriguingly, although the PPAR␥ antagonist bisphenol-A-diglycidyl ether (BADGE) has been shown to prevent BM adiposity in type 1 diabetic mice, this effect was not associated with any improvement in bone density (Botolin and McCabe, 2006). Only a few reports have appeared about the effects of PPAR␥ antagonists on human mesenchymal lineage allocation. Thus, any benefit of the modulation of PPAR␥ expression as a means of treating bone loss remains uncertain. Given that human BM derived-MSCs are the common progenitors of osteoblasts and adipocytes and that MSCs may be preferential targets of the desired therapy, such cells serve as a promising model to characterize the effects of modulating PPAR␥ expression (Elabd et al., 2008). In the present study, we explored the expression of PPAR␥ during the course of MSC differentiation into adipocytes or osteoblasts. BADGE and GW9662, potential modulators of PPAR␥, and lentivirus-borne PPAR␥ RNAi were used to investigate the involvement of PPAR␥ in mesenchymal lineage determination. The effects of PPAR␥ inhibition on the expression of other MSC-associated transcription factors, Runx2 and C/EBP␣, were also explored. 2. Materials and methods 2.1. Materials Cell culture reagents, including Dulbecco’s modified Eagle’s medium (DMEM), fetal bovine serum (FBS) and trypsin, were obtained from Hyclone (Logan, UT). Dexamethasone, 3-isobutyl1-methylxanthine (IBMX), insulin, indomethacin, ascorbic acid, sodium glycerophosphate, BADGE and GW9662 were purchased from Sigma (St. Louis, MO). Antibodies against PPAR␥ and C/EBP␣ were obtained from Cell Signaling Technologies (Beverly, MA). A monoclonal antibody against GAPDH was supplied by Kangchen (Shanghai, China). 2.2. Cell culture The hMSCs were obtained from Cyagen Biosciences Inc. (Guangzhou, China). The culture medium contained low-glucose DMEM (Hyclone, Logan, UT, USA) supplemented with 10% fetal bovine serum (FBS) (Hyclone), 100 IU/ml penicillin (Hyclone) and 100 g/ml streptomycin (Hyclone). The medium was changed
every 3 days. After confluence was attained, the cells were detached by mild treatment with trypsin (0.25% [w/v], 5 min, 37 ◦ C) and replated at one-third of the confluent density for continued passage.
2.3. Adipogenic differentiation Adipocytic differentiation of human MSCs was performed as previously described (Pittenger et al., 1999). Four-day postconfluent cells were incubated in adipogenesis-inducing medium (AIM) containing 1 M dexamethasone, 0.2 mM indomethacin, 0.5 mM isobutyl-methylxanthine, 0.01 mg/ml insulin and 10% (v/v) FCS in DMEM. After 72 h, the AIM was exchanged for adipogenesismaintenance medium (AMM) containing insulin (0.01 mg/ml) and 10% (v/v) FBS in DMEM for 24 h and then switched back to AIM. This exchange of the two media was repeated three times, after which the cells were maintained in AMM for 1 week. For Oil red O staining, the cells were washed twice with PBS and fixed with 3.7% (v/v) formaldehyde in PBS for 1 h at room temperature. The cells were next stained for 1 h at room temperature with filtered Oil red O solution (0.3% [w/v] Oil red O in 60% [v/v] isopropanol), washed twice with distilled water, visualized under light microscopy and photographed. To extract the incorporated Oil red O, 1 ml isopropanol was added to each well followed by 15 min of shaking at room temperature. After appropriate dilution, the absorbance of triplicate samples was read at 510 nm.
2.4. Osteogenic differentiation Osteogenic differentiation of subconfluent MSC monolayers was induced by 100 nM dexamethasone, 0.05 mM ascorbic acid, 10 mM sodium glycerophosphate and 10% (v/v) FCS, as previously described (Pittenger et al., 1999). For Alizarin Red S (ARS) staining, cells were rinsed with PBS and fixed in ice-cold 70% (v/v) ethanol for 1 h. After a brief wash with water, the cells were stained for 10 min with 40 mM ARS solution (pH 4.2) at room temperature. Next, the cells were rinsed five times with water followed by a 15-min wash with PBS (with rotation) to reduce nonspecific ARS staining. For the calcium assay, cells were washed twice with PBS and incubated overnight in 1 ml of 0.5-N HCl with gentle shaking. After centrifugation at 2000 × g for 10 min, the supernatant was assayed for calcium, following the protocol of the Calcium (CPC) LiquiColor kit (Stanbio Laboratory, USA). The total calcium levels were calculated using a standard curve prepared by employing solutions of known calcium concentration. The protein content was quantified using the DC protein assay from Bio-Rad, employing bovine serum albumin as a standard. The results were expressed as mg calcium/mg cellular protein. To measure alkaline phosphatase (ALP) activity, cells were washed three times with PBS and scraped into a lysis solution containing 10 mM Tris–HCl (pH 7.4), 0.1% (v/v) Triton X-100 and 0.5 mM MgCl2 . Each lysate was sonicated and then centrifuged at 16,000 × g for 10 min at 4 ◦ C. The supernatants were recovered for the assay of ALP activity by the hydrolysis of 5 mM p-nitrophenyl phosphate (for 15 min at 37 ◦ C), followed by colorimetric detection of the product p-nitrophenol at 405 nm. The results were expressed as nanomoles p-nitrophenol produced per mg of cellular protein per min.
2.5. Combined adipogenesis and osteogenesis in culture For this experiment, hMSCs were induced to differentiate into both adipocytes and osteoblasts in the same dish, using a DLP
W.-H. Yu et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 377–384
medium consisting of a mixture of adipogenic and osteogenic medium (1/1, v/v), as described above. 2.6. Real-time PCR Total RNA was isolated using the TRIzol reagent (Life Technologies, Gaithersburg, MD) according to the manufacturer’s instructions. After digestion with DNase I (Fermentas, Hanover, MD), 1 g total RNA was reverse transcribed using a RevertAidTM first-strand complementary DNA synthesis kit (Fermentas). The mRNA levels from the indicated genes were analyzed in triplicate using the QuantiTect SYBR Green PCR Master Mix (Qiagen, Valencia, CA); reactions proceeded in an ABI 7500 real-time PCR instrument (Applied Biosystems, USA) following the manufacturer’s instructions. The mRNA levels were normalized to those of GAPDH (an internal control), and changes in gene expression were calculated as-fold changes (the Ct method). The primers used were as follows: GAPDH (forward), 5 -GAAGGTGAAGGTCGGAGTC-3 ; GAPDH (reverse), 5 -GAAG ATGGTGATGGGATTTC-3 ; PPAR␥ (forward), 5 -CGAGAAGGAGAAGCTGT TGG-3 ; PPAR␥ (reverse), 5 -TCAGCGGGAAGGACTTTATGTATG-3 ; FABP4 (forward), 5 -AGCACCATAACCTTAGATGGGG-3 ; FABP4 (reverse), 5 -CGTGGAAGTGACGCCTTTCA-3 ; GPDH (forward), 5 -AGGAAGACATTGGAGGCAAAAA-3 ; GPDH (reverse), 5 -GCAGCCTGGACCACATCTG-3 ; LPL (forward), 5 -ACAAGAGAGAACCAGACTCCAA-3 ; LPL (reverse), 5 -AGGGTAGTTAAACTCCTCCTCC-3 ; osteocalcin (forward), 5 -CACTCCTCGCCCTATTGGC-3 ; osteocalcin (reverse), 5 -CCCTCCTGCTTGGACACAAAG-3 ; collagen I (forward), 5 -GCCGTGACCTCAAGATGTG-3 ; collagen I (reverse), 5 -GCCGAACCAGACATGCCTC-3 ; Runx2 (forward), 5 -AGATGATGACACTGCCACCTCTG-3 ; Runx2 (reverse), 5 -GGGATGAAATGCTTGGGAACTGC-3 .
2.7. Immunoblotting Cells were washed with cold PBS and directly lysed in Laemmli buffer. Each lysate was sonicated and then centrifuged at 16,000 × g for 10 min at 4 ◦ C. Each supernatant was recovered as a total cell lysate. Equal amounts of protein (20 g) were separated on 8–10% (w/v) SDS-PAGE and then electrotransferred to 0.45-m pore-sized polyvinylidene difluoride membranes (Millipore, Bedford, MA). After the transfer, each membrane was blocked using a solution of 0.1% (v/v) Tween 20/TBS (TBS/T) containing 5% (w/v) non-fat milk powder for 1 h at room temperature and then incubated with appropriate primary antibodies overnight at 4 ◦ C. Specifically bound primary antibodies were detected using peroxidase-coupled secondary antibodies and enhanced chemiluminescence signaling (Cell Signaling Technologies, Beverly, MA).
379
and the non-targeting scrambled shRNA sequence was 5 AACAAGATGAAGAGCACCAA-3 (NTC). All oligonucleotides were synthesized by Sangon (Shanghai, China) and were annealed and subcloned into the pLL3.7 vector. The lentivirus was prepared according to the manufacturer’s protocol (Invitrogen). Briefly, subconfluent 293FT cells were cotransfected with PLL3.7 and a compatible packaging plasmid mixture (Virapower lentiviral packaging system) using Lipofectamine 2000 (Invitrogen). The lentivirus-containing supernatants from the transfected cells were collected 48–72 h later, filtered and ultracentrifuged at 50,000 × g for 90 min. Each enriched virus pellet was resuspended in PBS, aliquoted and stored at −70 ◦ C. Viral titers were determined by the infection of 293FT cells with serial dilutions of vector stocks, followed by fluorescence-activated cell sorter (FACS)-mediated analysis of EGFP-positive cells. The transduction of MSCs proceeded at a multiplicity of infection (MOI) of 10 in the presence of 8 g/ml polybrene (Sigma).
3. Results 3.1. Expression of PPAR during adipogenic and osteogenic differentiation of human MSCs Adipogenesis of MSCs was induced by the addition of dexamethasone, insulin, indomethacin and IBMX. As we have earlier shown, three rounds of induction followed by maintenance treatment increased the numbers of rounded, lipid-laden adipocytes. PPAR␥ is synthesized in two isoforms, PPAR␥1 and the adipocyte-specific PPAR␥2. As the first step in the elucidation of PPAR␥ isoform regulation, we analyzed the temporal expression of the two PPAR␥ isoforms during hMSC differentiation into adipocytes to determine whether one factor was preferentially induced and might thus serve as an early controller of differentiation pathway. Representative immunoblots were shown in Fig. 1A. The induction patterns of the two isoforms of PPAR␥, PPAR␥1 and PPAR␥2, were different. Both isoforms were undetectable in confluent MSCs. Expression of PPAR␥1 was detected 2 days after initiation of adipocyte differentiation, whereas PPAR␥2 appeared later in the differentiation process (on day 6). Interestingly, a switch to adipogenesis-maintenance medium (AMM) transiently inhibited both PPAR␥1 and PPAR␥2 expression and, during the final AMM treatment, expression of both isoforms of PPAR␥ was greatly decreased. In addition, PPAR␥1 was the predominant isoform in hMSC-derived adipocytes. Osteogenesis of hMSCs was induced by the addition of dexamethasone, ascorbic acid and sodium glycerophosphate. We next examined the temporal expression of PPAR␥ during osteogenic differentiation of hMSCs. As shown in Fig. 1B, PPAR␥1 started to increase in level on day 2 after exposure to osteogenic inducers and continued to rise, to attain a maximal level at the end of differentiation. However, PPAR␥2 was not detected at any time during hMSC osteogenic differentiation.
2.8. shRNA transduction The knockdown of endogenous genes was accomplished using short hairpin RNA (shRNA) and PLL3.7 lentiviral vectors carrying the Pol III-dependent U6 promoter (a kind gift of Dr. David L. Garbers, University of Texas Southwestern Medical Center, Dallas, TX). The target sequences were selected with the aid of Dharmacon software. The two sequences targeting PPAR␥ were 5 -GCCCTTCACTACTGTTGAC-3 (PPAR-A) and 5 -GAAGTTCAATGCACTGGAA-3 (PPAR-B); the two targeting sequences for C/EBP␣ were 5 -GCTGGAGCTGACCAGTGACAA3 (CEBP-A) and 5 -GCACGAGACGTCCATCGACAT-3 (CEBP-B);
3.2. BADGE and GW9662, potential antagonists of PPAR, inhibited hMSC adipogenesis but did not promote osteogenesis BADGE, a potential antagonist of PPAR␥, has been reported to inhibit 3T3-L1 adipogenesis (Wright et al., 2000). GW9662 is another potential antagonist of PPAR␥ (David et al., 2007). In the present study, the effects of BADGE and GW9662 on hMSC adipogenesis and osteogenesis were investigated. In adipogenic induction medium, both BADGE and GW9662 greatly inhibited adipogenesis, as shown by Oil red O staining (Fig. 2). No osteogenesis was
380
W.-H. Yu et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 377–384
Fig. 1. Expression of the two isoforms of PPAR␥ during the processes of adipogenesis and osteogenesis. Human MSCs were induced to differentiate into adipocytes (A) or osteoblasts (B) and were harvested at the indicated periods of time. Equal amounts of cell extracts were immunoblotted using antibodies specific for PPAR␥. GAPDH was used as a loading control. Representative data from three separate experiments are shown.
evident, as revealed by results of the alkaline phosphatase (ALP) and calcium assays (data not shown). 3.3. Knockdown of PPAR inhibited adipogenesis of human MSCs but did not promote osteogenesis To further investigate the role of PPAR␥ in adipogenesis and osteogenesis, a lentivirus-based shRNA knockdown system was used. MSCs were transduced with lentivirus containing either PPAR␥-specific shRNA (shPPAR-A and shPPAR-B) or non-target control shRNA (shNTC) and then subjected to adipogenic induction. The lentivirus transduction efficiency of MSCs was at least 90%, as assessed by counting the numbers of cells positive for EGFP, a cassette contained within the lentiviral vector pLL3.7. Fig. 3A and B show the knockdown of PPAR␥ confirmed by immunoblotting and real-time PCR analysis. Compared to MSCs infected with shNTC, shPPAR-expressing MSCs showed greatly impaired lipid accumulation, as revealed by Oil red O staining (Fig. 3C). The knockdown of PPAR␥ did not cause cytoskeletal changes; cells retained the round phenotype of early adipocyte differentiation. The inhibition of adipogenic differentiation by PPAR␥ suppression was further
confirmed by real-time PCR analysis of the expression of the adipogenic genes LPL, FABP4 and GPDH (Fig. 3D). Under conditions conducive to adipogenic induction, PPAR␥ suppression did not cause spontaneous calcium deposition at any time during the entire adipogenic process (data not shown). We found that ALP activity was slightly decreased upon PPAR␥ inactivation, compared with what was seen in non-target controls (Fig. 4A) (p > 0.05). To exclude the possibility that the lack of osteogenesis seen in MSCs deficient in PPAR␥ was caused by the absence of osteogenic stimuli, MSCs transduced with lentivirus containing either PPARspecific shRNA (shPPAR-A and shPPAR-B) or non-target control shRNA (shNTC) were allowed to differentiate in the presence of osteogenic stimuli. As shown in Fig. 4B–D, knockdown of PPAR␥ with PPAR␥-specific shRNAs (shPPAR-A and shPPAR-B) only slightly impaired the osteogenic differentiation of hBMSCs, as demonstrated by ARS staining and the results of the calcium and ALP activity assays (p > 0.05). The effect of PPAR␥ knockdown on the osteogenic differentiation of hBMSCs was further confirmed by real-time PCR analysis of the expression of osteogenic genes, including osteocalcin, collagen I and Runx2 (Fig. 4E). To further confirm and extend the observed effects of PPAR␥ knockdown on MSC commitment to either the osteoblast or adipocyte lineage, we used a dual-lineage-promoting (DLP) medium that allows both cell types to develop in the same dish. We found that, under such conditions, the knockdown of PPAR␥ only slightly inhibited hMSC osteogenesis, as revealed by the results of the calcium and ALP activity assays (p > 0.05; data not shown). 3.4. PPAR suppression inhibited expression of the adipogenic transcription factor C/EBP˛ but not that of the osteogenic transcription factor Runx2
Fig. 2. BADGE and GW9662, the inhibitors of PPAR␥, inhibited adipogenesis. Four days postconfluence, human MSCs were induced to differentiate into adipocytes in the presence of DMSO, 100 nM BADGE or 10 M GW9662. At the end of the culture period, the degree of adipocyte differentiation was determined by Oil red O staining. The lower panel shows the staining of whole dishes. Representative data from three separate experiments are shown. The incorporated Oil red O was extracted with isopropanol, and the quantification was performed as described in Section 2. The results are expressed as the mean ± SEM of three independent experiments (upper panel). A significant difference compared with cells treated with DMSO (p < 0.05) is indicated by *.
It has been reported that PPAR␥ agonists and overexpression of PPAR␥ induce differentiation of BM stromal cells into the adipocyte lineage and negatively regulate osteoblast differentiation by repressing Runx2 (Lecka-Czernik et al., 1999). To investigate the relationship between Runx2 and PPAR␥ expression under conditions of hMSC adipogenesis and osteogenesis, the effect of PPAR␥ knockdown on Runx2 expression was evaluated by Western blotting. As shown in Fig. 4F, Runx2 synthesis was not significantly affected by knockdown of PPAR␥. PPAR␥ and C/EBP␣ are the two transcription factors critical for adipogenesis. As shown in Fig. 5A, the knockdown of PPAR␥ greatly inhibited the expression of C/EBP␣. These two transcription factors have been reported to be mutually regulatory during murine adipogenesis (Rosen and MacDougald, 2006). Thus, to explore the relationship between the transcription factors during hMSC adipogenesis, we constructed lentiviral shRNA specific to C/EBP␣
W.-H. Yu et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 377–384
381
Fig. 3. Lentivirus-mediated PPAR␥ knockdown inhibited hMSC adipogenesis. MSCs were transduced with lentivirus containing either PPAR␥-specific shRNA (shPPAR-A and shPPAR-B) or non-targeting shRNA (shNTC) and then subjected to adipogenic induction. After 7 days of adipocyte induction, the MSCs were harvested and real time PCR was used to analyze the mRNA levels of PPAR␥ (A), LPL, FABP4 and GPDH (D). The mRNA levels were normalized to GAPDH (internal control), and gene expression was presented as-fold changes (Ct method). Representative data from three separate experiments are shown. After seven days of adipocyte induction, cell extracts were immunoblotted using an anti-PPAR␥ antibody. GAPDH was used as a loading control. Representative data from three separate experiments are shown (B). At the end of the culture period, the degree of adipocyte differentiation was determined by Oil red O staining (C). The lower panel shows the staining of whole dishes. Representative data from three separate experiments are shown. The incorporated Oil red O was extracted with isopropanol, and the quantification was performed as described in Section 2. The results are expressed as the mean ± SEM of three independent experiments (upper panel). A significant difference compared with cells infected with shNTC (p < 0.05) is indicated by *.
(shCEBP-A and shCEBP-B) and introduced these shRNA constructs or shNTC into MSCs. The knockdown of C/EBP␣ was confirmed by immunoblotting (Fig. 5D). Downregulation of C/EBP␣ essentially abrogated hMSC adipogenesis compared with that of non-target controls, as revealed by Oil red O staining (Fig. 5B) and real-time PCR analysis of the expression of the adipogenic genes LPL, FABP4 and GPDH (Fig. 5C). Interestingly, we found that the knockdown of C/EBP␣ inhibited PPAR␥2 to a greater extent than PPAR␥1 (Fig. 5D). 4. Discussion A few reports have suggested that PPAR␥ regulates mesenchymal lineage allocation, favoring adipocyte over osteoblast development, and that PPAR␥ may be a useful target of drugs intended to enhance bone mass. However, little direct evidence has been reported on the effect of PPAR␥ antagonists on human mesenchymal lineage allocation. Bone loss induced by Thiazolidinedione (TZD) has not been reversible. In the present study, we showed that BADGE and GW9662, inhibitors of PPAR␥, reduced the extent of adipogenesis, but did not significantly affect osteogenesis. Similar results were obtained upon lentivirus-mediated PPAR␥ knockdown. The fact that PPAR␥ suppression inhibited C/EBP␣ but did not significantly affect expression of the osteogenic transcription factor Runx2 may help to explain the mechanism underlying the choice of differentiation pathway. Although PPAR␥1 expression was detected during osteogenesis, the presence of PPAR␥ RNAi had no obvious effect on osteogenesis under permissive conditions, indicating that PPAR␥1 may not play a critical role in hMSC osteogenesis (Bruedigam et al., 2008).
During adipogenesis, both PPAR␥1 and PPAR␥2 were induced in hMSCs, with PPAR␥1 being the predominant isoform. It is noteworthy that the suppression of C/EBP␣ substantially reduced the level of adipocyte formation to an extent similar to that observed upon PPAR␥ suppression, whereas the knockdown of C/EBP␣ inhibited the activation of PPAR␥2 but not PPAR␥1, suggesting that PPAR␥2 may play the more important role in hMSC adipogenesis. It is thus conceivable that the inhibition of PPAR␥2 reduced adipogenesis but did not affect osteogenesis. Unfortunately, the isoform-specific knockdown of PPAR␥ was not successful in our experiment, due to the great similarity of the sequences of these two isoforms. Further assays using other methods are required to elucidate the respective roles of these two isoforms. Our findings contradict several lines of evidence suggesting that PPAR␥ insufficiency directly promotes the differentiation of osteoblasts. It is possible that a between-species difference is in play. Indeed, there exist important differences between murine MSCs and human MSCs regarding their surface phenotype and immunological properties (Madec et al., 2009). In terms of their differentiation capacity, hMSCs have been observed to have a significant and robust in vitro osteogenic capacity compared with mMSCs. Retinoic acid (RA) is shown to be required for the efficient osteogenesis of mMSCs (Wan et al., 2006). RA is a ligand for nuclear receptors, including retinoic acid receptors (RAR). Retinoic acid can inhibit PPAR␥ expression and activation, presumably by enhancing the heterodimerization of RAR with retinoid X receptor (RXR), thereby making RXR unavailable for dimerization with PPAR␥ (Mandrup and Lane, 1997). High-affinity binding to DNA by PPAR␥ appears to absolutely require dimerization with RXR.
382
W.-H. Yu et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 377–384
Fig. 4. Lentivirus-mediated PPAR␥ knockdown did not significantly affect hMSC osteogenesis. (A) MSCs were transduced with lentivirus containing either PPAR␥-specific shRNA (shPPAR-A and shPPAR-B) or non-targeting shRNA (shNTC) and then subjected to adipogenic induction, and there was no apparent osteogenesis. On the seventh day, the cells were harvested and the alkaline phosphatase activity was determined as described in Section 2. (B–E) The transduced MSCs were subjected to osteogenic induction, and there was no significant difference in osteogenesis compared with cells infected with shNTC. On the seventh day, the cells were harvested and the alkaline phosphatase activity was determined (B). At the end of osteogenic differentiation, the amount of calcium deposition was determined in the treated cells, and the results are expressed as the mean ± SEM of three independent experiments (C). Additionally, at the end of osteogenesis, the formation of bone nodules was assayed by ARS staining. Representative data from three separate experiments are shown (D). Furthermore, at the end of osteogenic differentiation, MSCs were harvested, and real time PCR was used to analyze the mRNA level of osteocalcin, collagen I and Runx2. The mRNA levels were normalized to GAPDH (internal control), and gene expression was presented as-fold changes (Ct method). Representative data from three separate experiments are shown (E). On the seventh day, cells were harvested and the cell extracts were immunoblotted using an anti-Runx2 antibody. GAPDH was used as a loading control. Representative data from three separate experiments are shown (F).
Therefore, the differential response to RA between murine MSCs and hMSCs may be the underlying mechanism for our observation. Notably, although PPAR␥ is required for hMSC adipogenesis, the ectopic expression of PPAR␥ in hMSCs did not induce full adipogenesis (Qian et al., 2010). This finding is not consistent with what has been observed in rodent adipogenesis (Rosen and MacDougald, 2006). Although some indirect insight into PPAR␥ function can be gained from the increased fracture risk associated with the use of synthetic PPAR␥ ligands, the thiazolidinediones (TZDs), the divergent effects of TZDs on human and rodent adipogenesis and osteogenesis suggest that the regulation of skeletal phenotype development cannot be attributed to PPAR␥ alone (Lecka-Czernik et al., 2002). Although some studies of human PPAR␥ gene polymorphisms suggest a role for PPAR␥ in the regulation of bone mass, evidence on the association of a certain PPAR␥ genomic polymorphism with bone status is unconvincing (Cheng et al., 2006; Kim et al., 2007). Recently, accumulating evidence points toward a role for PPAR␥ in osteoclast development (Mbalaviele et al., 2000; Wahli, 2008; Wan et al., 2007), indicating that the phenotype of PPAR␥ knockout mice should be more carefully investigated, as
should the effect of TZDs on bone density. Bone loss after rosiglitazone administration to ovariectomized rats has been shown to result from an increase in bone resorption (Sottile et al., 2004). Few studies have explored the effects of TZDs on hMSC adipogenesis and osteogenesis. Rosiglitazone has been reported to stimulate hMSC adipogenesis, but any effect on osteogenesis remains controversial (Benvenuti et al., 2007; Bruedigam et al., 2010). We could not find any significant effect of rosiglitazone on hMSC osteogenesis (data not shown). One recent report showed that rosiglitazone promoted both adipogenesis and osteogenesis, and rosiglitazone action in the BM may promote osteoblast apoptosis (Bruedigam et al., 2010), thus increasing fracture risk. There is a need to gather more direct evidence on the role played by PPAR␥ in human MSC lineage allocation. Transcription coregulators, such as TAZ, play vital roles in control of the expression and activity of PPAR␥ and Runx2 (Cui et al., 2003). TAZ is the transcriptional co-activator of both factors. The outcome of the interactions of PPAR␥ and Runx2 with TAZ differs substantially in that Runx2 transcription is stimulated, whereas that of PPAR␥ is repressed. Indeed, recent evidence has implicated TAZ as a molecular switch in MSC differentiation (Hong et al., 2005).
W.-H. Yu et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 377–384
383
Fig. 5. Cross-regulation of PPAR␥ and C/EBP␣, the transcription factors of adipogenesis, during the adipocyte differentiation of MSCs. (A) Lentivirus-mediated PPAR␥ knockdown inhibited C/EBP␣. The MSCs were transduced with lentivirus containing either PPAR␥-specific shRNA (shPPAR-A and shPPAR-B) or non-targeting shRNA (shNTC) and then subjected to adipogenic induction. After seven days of adipocyte induction, the cell extracts were immunoblotted using an anti-C/EBP␣ antibody. GAPDH was used as a loading control. Representative data from three separate experiments are shown. (B–D) Lentivirus-mediated C/EBP␣ knockdown inhibited PPAR␥2 and adipogenesis. At the end of culture, the degree of adipocyte differentiation was determined by Oil red O staining (B). The lower panel shows the staining of whole dishes. Representative data from three separate experiments are shown. The incorporated Oil red O was extracted with isopropanol and the quantification was performed as described in Section 2. The results are expressed as the mean ± SEM of three independent experiments (upper panel). A significant difference compared with cells infected with shNTC (p < 0.05) is indicated by *. After seven days of adipocyte induction, MSCs were harvested and real time PCR was used to analyze the mRNA level of LPL, FABP4 and GPDH. The mRNA levels were normalized to GAPDH (internal control), and gene expression was presented as-fold changes (Ct method). Representative data from three separate experiments are shown (C). After seven days of adipocyte induction, cell extracts were immunoblotted using anti-PPAR␥ and anti-C/EBP␣ antibodies. GAPDH was used as a loading control. Representative data from three separate experiments are shown (D).
Additional studies, particularly in vivo, will provide further insights into this process. In conclusion, the knockdown of PPAR␥ inhibited adipogenesis but did not stimulate osteogenesis. To identify the master controls of stem-cell lineage allocation, new model systems and more extensive studies at the genetic, molecular and cellular levels are needed. Acknowledgements This research was supported by the National Natural Science Foundation of China (81000362 and 30928015), National Basic Research Program of China (2009CB522100, 2010CB945400 and 2011CBA01100) the Key Scientific and Technological Projects of Guangdong Province (2007A032100003), the Natural Science Foundation of Guangdong Province (4203004), the Key Scientific and Technological Program of Guangzhou City (Nos. 2008A1E4011-5 and 2010U1-E00551), the Program for New Century Excellent Talents in University and the Fundamental Research Funds for the Central Universities (09ykpy79 and 10ykzd02). The funding agencies had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. References Benvenuti S, Cellai I, Luciani P, Deledda C, Baglioni S, Giuliani C, et al. Rosiglitazone stimulates adipogenesis and decreases osteoblastogenesis in human mesenchymal stem cells. J Endocrinol Invest 2007;30:RC26–30.
Botolin S, McCabe LR. Inhibition of PPARgamma prevents type I diabetic bone marrow adiposity but not bone loss. J Cell Physiol 2006;209:967–76. Bruedigam C, Eijken M, Koedam M, van de Peppel J, Drabek K, Chiba H, et al. A new concept underlying stem cell lineage skewing that explains the detrimental effects of thiazolidinediones on bone. Stem Cell 2010;28:916–27. Bruedigam C, Koedam M, Chiba H, Eijken M, van Leeuwen JP. Evidence for multiple peroxisome proliferator-activated receptor gamma transcripts in bone: fine-tuning by hormonal regulation and mRNA stability. FEBS Lett 2008;582: 1618–24. Cheng S, Afif H, Martel-Pelletier J, Benderdour M, Pelletier JP, Hilal G, et al. Association of polymorphisms in the peroxisome proliferator-activated receptor gamma gene and osteoarthritis of the knee. Ann Rheum Dis 2006;65: 1394–7. Cui CB, Cooper LF, Yang X, Karsenty G, Aukhil I. Transcriptional coactivation of bone-specific transcription factor Cbfa1 by TAZ. Mol Cell Biol 2003;23: 1004–13. David V, Martin A, Lafage-Proust MH, Malaval L, Peyroche S, Jones DB, et al. Mechanical loading down-regulates peroxisome proliferator-activated receptor gamma in bone marrow stromal cells and favors osteoblastogenesis at the expense of adipogenesis. Endocrinology 2007;148:2553–62. Elabd C, Basillais A, Beaupied H, Breuil V, Wagner N, Scheideler M, et al. Oxytocin controls differentiation of human mesenchymal stem cells and reverses osteoporosis. Stem Cell 2008;26:2399–407. Farmer SR. Transcriptional control of adipocyte formation. Cell Metab 2006;4:263–73. Hong JH, Hwang ES, McManus MT, Amsterdam A, Tian Y, Kalmukova R, et al. TAZ, a transcriptional modulator of mesenchymal stem cell differentiation. Science 2005;309:1074–8. Kawai M, Rosen CJ. PPARgamma: a circadian transcription factor in adipogenesis and osteogenesis. Nat Rev Endocrinol 2010;6:629–36. Kim TH, Hong JM, Park EK, Kim SY. Peroxisome proliferator-activated receptorgamma gene polymorphisms are not associated with osteonecrosis of the femoral head in the Korean population. Mol Cell 2007;24:388–93. Lecka-Czernik B, Gubrij I, Moerman EJ, Kajkenova O, Lipschitz DA, Manolagas SC, et al. Inhibition of Osf2/Cbfa1 expression and terminal osteoblast differentiation by PPARgamma2. J Cell Biochem 1999;74:357–71. Lecka-Czernik B, Moerman EJ, Grant DF, Lehmann JM, Manolagas SC, Jilka RL. Divergent effects of selective peroxisome proliferator-activated
384
W.-H. Yu et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 377–384
receptor-gamma 2 ligands on adipocyte versus osteoblast differentiation. Endocrinology 2002;143:2376–84. Madec AM, Mallone R, Afonso G, Abou Mrad E, Mesnier A, Eljaafari A, et al. Mesenchymal stem cells protect NOD mice from diabetes by inducing regulatory T cells. Diabetologia 2009;52:1391–9. Mandrup S, Lane MD. Regulating adipogenesis. J Biol Chem 1997;272:5367–70. Mbalaviele G, Abu-Amer Y, Meng A, Jaiswal R, Beck S, Pittenger MF, et al. Activation of peroxisome proliferator-activated receptor-gamma pathway inhibits osteoclast differentiation. J Biol Chem 2000;275:14388–93. Nuttall ME, Gimble JM. Controlling the balance between osteoblastogenesis and adipogenesis and the consequent therapeutic implications. Curr Opin Pharmacol 2004;4:290–4. Pittenger MF, Mackay AM, Beck SC, Jaiswal RK, Douglas R, Mosca JD, et al. Multilineage potential of adult human mesenchymal stem cells. Science 1999;284:143–7. Qian SW, Li X, Zhang YY, Huang HY, Liu Y, Sun X, et al. Characterization of adipocyte differentiation from human mesenchymal stem cells in bone marrow. BMC Dev Biol 2010;10:47. Rosen ED, MacDougald OA. Adipocyte differentiation from the inside out. Nat Rev Mol Cell Biol 2006;7:885–96. Schiller PC, D’Ippolito G, Brambilla R, Roos BA, Howard GA. Inhibition of gapjunctional communication induces the trans-differentiation of osteoblasts to an adipocytic phenotype in vitro. J Biol Chem 2001;276:14133–8.
Sottile V, Seuwen K, Kneissel M. Enhanced marrow adipogenesis and bone resorption in estrogen-deprived rats treated with the PPARgamma agonist BRL49653 (rosiglitazone). Calcif Tissue Int 2004;75:329–37. Wahli W. PPAR gamma: ally and foe in bone metabolism. Cell Metab 2008;7:188–90. Wan DC, Shi YY, Nacamuli RP, Quarto N, Lyons KM, Longaker MT. Osteogenic differentiation of mouse adipose-derived adult stromal cells requires retinoic acid and bone morphogenetic protein receptor type IB signaling. Proc Natl Acad Sci U S A 2006;103:12335–40. Wan Y, Chong LW, Evans RM. PPAR-gamma regulates osteoclastogenesis in mice. Nat Med 2007;13:1496–503. Wolf NS, Penn PE, Rao D, McKee MD. Intraclonal plasticity for bone, smooth muscle, and adipocyte lineages in bone marrow stroma fibroblastoid cells. Exp Cell Res 2003;290:346–57. Wright HM, Clish CB, Mikami T, Hauser S, Yanagi K, Hiramatsu R, et al. A synthetic antagonist for the peroxisome proliferator-activated receptor gamma inhibits adipocyte differentiation. J Biol Chem 2000;275:1873–7. Yamashita A, Takada T, Nemoto K, Yamamoto G, Torii R. Transient suppression of PPARgamma directed ES cells into an osteoblastic lineage. FEBS Lett 2006;580:4121–5. Zieleniak A, Wojcik M, Wozniak LA. Structure and physiological functions of the human peroxisome proliferator-activated receptor gamma. Arch Immunol Ther Exp (Warsz) 2008;56:331–45.