The authors state that they have no conflicts of interest.
In vitro, mesenchymal stem cells differentiate to osteoblasts when exposed to bone-inducing medium. However, adipocytes are also formed. We showed that activation of the nuclear protein deacetylase Sirt1 reduces adipocyte formation and promotes osteoblast differentiation.
Introduction: Mesenchymal stem cells (MSCs) can differentiate into osteoblasts, adipocytes, chondrocytes, and myoblasts. It has been suggested that a reciprocal relationship exists between the differentiation of MSCs into osteoblasts and adipocytes. Peroxisome proliferator-activated receptor γ2 (PPARγ2) is a key element for the differentiation into adipocytes. Activation of Sirt1 has recently been shown to decrease adipocyte development from preadipocytes through inhibition of PPARγ2.
Materials and Methods: We used the mouse mesenchymal cell line C3H10T1/2 and primary rat bone marrow cells cultured in osteoblast differentiation medium with or without reagents affecting Sirt1 activity. Adipocyte levels were analyzed by light microscopy and flow cytometry (FACS) after staining with Oil red O and Nile red, respectively. Osteoblast and adipocyte markers were studied with quantitative real-time PCR. Mineralization in cultures of primary rat bone marrow stromal cells was studied by von Kossa and alizarin red staining.
Results: We found that Sirt1 is expressed in the mesenchymal cell line C3H10T1/2. Treatment with the plant polyphenol resveratrol as well as isonicotinamide, both of which activate Sirt1, blocked adipocyte development and increased the expression of osteoblast markers. Nicotinamide, which inhibits Sirt1, increased adipocyte number and increased expression of adipocyte markers. Furthermore, activation of Sirt1 prevented the increase in adipocytes caused by the PPARγ-agonist troglitazone. Finally, activation of Sirt1 in rat primary bone marrow stromal cells increased expression of osteoblast markers and also mineralization.
Conclusions: In this study, we targeted Sirt1 to control adipocyte development during differentiation of MSCs into osteoblasts. The finding that resveratrol and isonicotinamide markedly inhibited adipocyte and promoted osteoblast differentiation may be relevant in the search for new treatment regimens of osteoporosis but also important for the evolving field of cell-based tissue engineering.
Mesenchymal stem cells (MSCs) are multipotent cells able to differentiate into different distinct cell types (i.e., osteoblasts, adipocytes, myoblasts, and chondrocytes).(1) The differentiation into different cell lineages is determined by hormones and different local factors. Studies concerning the isolation, in vitro growth, and differentiation of MSCs are of clinical interest because of their possible use in tissue engineering. In orthopedic surgery, MSCs expanded in cell number and directed to differentiate along the osteoblastic pathway in vitro could be used for reconstruction of bone.(2) Thus, studies about growth and differentiation of MSCs to osteoblasts are important for tissue engineering but may also be relevant for our understanding and treatment of osteoporosis.
Experimental data indicate that a reciprocal relationship exists between differentiation of MSCs into adipocytes and osteoblasts. In age-related osteoporosis, adipose cells are increased in bone marrow.(3,4) Experimental data suggest that mature osteoblasts can transdifferentiate into adipocytes and mature adipocytes into osteoblasts.(5–7) The peroxisome proliferator-activated receptor γ (PPARγ) is important for adipocyte differentiation.(8) PPARγ exists in two isoforms, PPARγ1 and PPARγ2, products of alternative promoter usage and different splicing.(9) The PPARγ1 isoform is expressed at low levels in many cell types, whereas PPARγ2 expression is restricted primarily to adipocytes and has been shown to be the isoform necessary for adipocyte development in mice.(10) Interestingly, PPARγ insufficiency (i.e., reduced expression of PPARγ) enhances osteogenesis through osteoblast formation from MSCs.(11) Ligands for PPARγ are polyunsaturated fatty acids, eicosanoids, and the synthetic antidiabetic thiazolidinedione (TZD) drugs.(12) Because of the reciprocal relationship between adipocyte and osteoblast differentiation, a possible risk for development of osteoporosis could be envisaged on long-term treatment with TZDs. It has been shown, in vitro, that some PPARγ ligands inhibit osteoblast differentiation.(13–15) Also, administration of TZD to mice results in decreased BMC.(16)
Activation of the nuclear NAD-dependent protein deacetylase Sirt1 has recently been shown to result in mobilization of fat in fully differentiated adipocytes and in a decrease in adipocyte formation from pre-adipocytes.(17) This was shown to be caused by Sirt1 docking to the nuclear receptor co-repressors NCoR and SMRT, which in turn inhibited PPARγ. Sirt1 is a nutritional sensor responding to changes in NAD/NADH ratio. Caloric restriction, which results in an increase in NAD/NADH ratio, activates Sirt1, which, interestingly, in several species has been shown to prolong lifespan.(18) The enzymatic activity of Sirt1 is also regulated by specific dietary compounds, such as the plant polyphenol resveratrol (3,5,4-trihydroxystilbene) and nicotinamide, which activates and inhibits Sirt1, respectively.(18)
The classical medium used for differentiation of MSCs to osteoblasts contains β-glycerophosphate, ascorbic acid, and the glucocorticoid dexamethasone. In earlier preliminary studies, using this medium for differentiation of human MSCs as well as for differentiation of a mouse mesenchymal stem cell line, C3H10T1/2, we found that a significant number of adipocytes were developed together with osteoblasts (unpublished data). This stimulated us to use the C3H10T1/2 cell model for studying if Sirt1 can influence adipocyte and osteoblast differentiation from MSCs grown in the above-described osteoblast-inducing medium. We used different compounds that activate or inhibit Sirt1 activity. We found that activation of Sirt1 significantly decreased adipocyte number, and increased expression of osteoblast markers. Inhibition of Sirt1 resulted in increased number of adipocytes and adipocyte markers and decreased expression of osteoblast markers. Furthermore, activation of Sirt1 inhibited the action of the PPARγ agonist troglitazone, indicating that activated Sirt1 inhibits PPARγ in C3H10T1/2 cells.
To further substantiate the above-described findings in C3H10T1/2 cells, rat primary bone marrow stromal cells were exposed to Sirt1-activating or -inhibiting compounds. We found that Sirt1-activating and -inhibiting compounds increased and decreased, respectively, expression of osteoblast markers and mineralization. Learning more about activators of Sirt1 and Sirt1 mechanisms of action during MSC differentiation could be helpful in finding new regimens for treating osteoporosis and also in setting up protocols for optimal growth and differentiation of MSCs to osteoblasts in vitro.
MATERIALS AND METHODS
Eagle's basal medium (BME), FBS, l-glutamine, and gentamicin were from Invitrogen. RNA purification RNeasy kit was obtained from Qiagen, cDNA synthesis Superscript II from Invitrogen, and real-time PCR reagents from Applied Biosystems. 4,4′,4″-[4-propyl-(1H)-pyrazole-1,3,5-triyl]-trisphenol (PPT) was kindly provided by Karo Bio (Stockholm, Sweden). 2,3-bis(4-hydroxyphenyl)-propionitrile (DPN) and 1α,25-dihydroxyvitamin D3 were purchased from Tocris and Calbiochem, respectively. Troglitazone was obtained from Alexis Biochemicals. Estradiol (E2), resveratrol, nicotinamide, and isonicotinamide were purchased from Sigma-Aldrich. Anti-Sirt1 antibody was from Upstate, and β-actin antibody was purchased from Gene Tex.
The murine pluripotent mesenchymal cell line C3H10T1/2 was obtained from American Type Culture Collection. The cells were cultured in BME medium supplemented with 10% FBS, 1× l-glutamine, and 100 μg/ml gentamicin at 37°C in a humidified atmosphere containing 5% CO2. Primary bone marrow stromal cells were collected from 6-week-old Wistar rats and plated in culture dishes. Nonadherent cells were removed after 24 h. Adherent primary rat and C3H10T1/2 cells were seeded in 6-well plates (75,000 cells/well). Osteoblast differentiation was initiated the following day with BME medium supplemented with 50 μg/ml l-ascorbic acid, 10 mM β-glycerophosphate, and 0.1 mM dexamethasone (osteoblast medium [OBM]). To OBM, 50 μM resveratrol, 10 mM nicotinamide, 25 mM isonicotinamide, 0.75 μM troglitazone, 10 nM E2, 10 nM PPT, 10 nM DPN, or 1 nM 1α,25-dihydroxyvitamin D3 [1α,25(OH)2D3] was added.
SDS-PAGE and Western blot
Cells were lysed in lysis buffer (25 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% Nonidet P-40, 5 mM EDTA, 5 mM EGTA, 1 mM vanadate, 20 mM sodium fluoride, 10 mM aprotinin, 10 mM leupeptin, 100 mM phenylmethylfluoride [PMSF], and 5 mM dithiothreitol [DTT]). Protein extract was mixed with Laemmli buffer and boiled for 3 minutes before loading onto SDS-PAGE gels. Proteins were transferred to polyvinylidene fluoride (PVDF) membranes by electroblotting and blocked in 5% milk. The membranes were probed with anti-Sirt1 and anti-β-actin antibodies diluted in blocking solution followed by horseradish peroxidase–conjugated secondary antibody. Proteins were detected with an enhanced chemiluminescence detection system (Amersham Biosciences), followed by exposure to X-ray films.
Adipocyte staining with Oil red O
Cells were washed twice with PBS and fixed with 10% formaldehyde for 45 minutes at room temperature. After washing with distilled water twice and 50% isopropanol once, the cells were stained for 1 h at room temperature with filtered Oil red O/60% isopropanol solution. The cells were washed twice with distilled water and twice with PBS. Adipocytes stained red were recorded by light microscopy.
Adipocyte quantification using Nile red
Adipocytes were quantified with flow cytometry using the lipophilic Nile red fluorescent dye. Briefly, trypsinized cells were washed in PBS once and fixed in 10% formaldehyde for 1 h at 4°C. The cells were stained with 10 μg/ml Nile red for 45 minutes at room temperature. Fluorescent emission between 564 and 604 nm with a band-pass filter was measured using a FACScan flow cytometer (BD Biosciences).
RNA preparation and real-time PCR analysis
Total RNA was isolated from cells grown in 6-well plates with RNeasy (Qiagen). A total of 1 μg RNA was reverse transcribed to cDNA using Superscript II (Invitrogen) according to the manufacturer's protocol. Gene transcript levels of alkaline phosphatase (ALP), collagen 1α1, osteocalcin, runt-related transcription factor 2 (Runx2), RANKL, osteoprotegerin (OPG), PPARγ2, and adipocyte fatty acid binding protein (aP2) were analyzed by real-time PCR (TaqMan gene expression assays) on an ABI7700 Sequence Detection System (Applied Biosystems). The housekeeping gene, GAPDH, was used as an endogenous control.
Staining for mineralization
Rat primary bone marrow stromal cells were cultured for 8 weeks before analysis of mineralization. Both alizarin red and von Kossa staining were used. Briefly, for von Kossa, the cells were washed with PBS and fixed with 4% paraformaldehyde. The cells were rinsed with distilled water and thereafter incubated under UV light in the presence of 5% silver nitrate. After 30 minutes, the cells were washed with distilled water and incubated with 5% sodium thiosulfate. For alizarin red staining, the cells were instead incubated with 2% alizarin red with pH 4.2 for 10 minutes and subsequently washed with distilled water.
Statistical analysis was performed by ANOVA, with significance assigned at the p < 0.05 and p < 0.01 levels.
Formation of adipocytes during osteoblast differentiation of C3H10T1/2
Culture of human or mouse MSCs in vitro in medium containing ascorbic acid, dexamethasone, and β-glycerophosphate promotes differentiation toward osteoblasts.(19) In preliminary studies, when using this medium for osteoblast differentiation of the mouse mesenchymal C3H10T1/2 cell line, we found that adipocytes were also formed. The same had also been found when human adult MSCs were grown in the same medium (O Morales and MKR Samuelsson, personal communication, 2004).
In this study, we initially quantified adipocyte formation in the C3H10T1/2 cell line after 2 weeks of culture in medium containing ascorbic acid, β-glycerophosphate, and dexamethasone (OBM). Adipocytes were detected using Oil red O staining (Fig. 1, left) and quantified by Nile red staining and flow cytometry (FACS) analysis (Fig. 1, middle). After 2 weeks in culture in OBM, 13% of the cells were adipocytes (Fig. 1, right). C3H10T1/2 cell grown in medium without ascorbic acid, β-glycerophosphate, and dexamethasone (BME with 10% FBS) did not differentiate into adipocytes. We also noted that the percentage of adipocytes formed was dependent on the batch of FBS used. In this study, the same batch was used except in the experiment shown in Fig. 2, where a new batch of FBS was used.
Expression of Sirt1 in C3H10T1/2 MSCs
The NAD-dependent protein deacetylase Sirt1 has recently been shown to attenuate development of adipocytes from preadipocytes through inhibition of PPARγ activity.(17) We found it interesting to study if agents known to regulate the activity of Sirt1 could influence the formation of adipocytes during osteoblast differentiation in vitro. We first studied if Sirt1 is present in C3H10T1/2 cells. Whole cell extracts from cells were analyzed by Western blot with anti-Sirt1 antibody. Sirt1 was found to be expressed in the cells before differentiation (day 0) and after differentiation (data not shown).
Effects on adipocyte formation by compounds regulating Sirt1 activity
Because Sirt1 decreases adipocyte formation through inhibition of PPARγ activity, we found it important first to study, in our cell model, if PPARγ activity could be manipulated with compounds earlier known to regulate PPARγ, such as 1α,25(OH)2D3 and troglitazone. 1α,25(OH)2D3 has been shown to downregulate the expression of PPARγ2 and to inhibit adipocyte formation during osteoblast differentiation of MSC.(20) This was also found in our study where 1α,25(OH)2D3 completely inhibited adipocyte development (Fig. 1). The PPARγ2 agonist troglitazone was initially tested at three different concentrations: 0.075, 0.75, and 7.5 μM (data not shown). Interestingly, 0.75 μM troglitazone was most efficient, resulting in 34% adipocytes (Fig. 1). Troglitazone at 7.5 μM resulted in a lower number of adipocytes (data not shown), the reason probably being that high concentrations of troglitazone can lead to downregulation of both PPARγ protein and mRNA.(21)
The compounds used to activate Sirt1 were the plant polyphenol resveratrol and isonicotinamide; to inhibit Sirt1, we used nicotinamide. The mechanism of action of resveratrol is not known in detail but it has been proposed that, as for some other polyphenols with trans-stilbene structure, it induces a conformational change in Sirt1, which lowers Km for both the acetylated substrate and NAD.(18) Nicotinamide, which is cleaved off from NAD by Sirt1, reverses this reaction if present in higher concentrations.(18) Isonicotinamide has recently been described as an antagonist of nicotinamide inhibition of Sirt1.(22)
Nicotinamide, when added to OBM, increased adipocyte number to 21% on day 15 (Fig. 1). Isonicotinamide, on the other hand, significantly attenuated adipocyte formation to 2% (Fig. 1). When resveratrol was added to OBM, adipocyte formation was totally blocked (Fig. 1). Resveratrol, in addition to its inhibitory action on Sirt1, has been shown to act as a phytoestrogen (i.e., binding and activating the estrogen receptor).(23) To exclude that the effect of resveratrol on adipocyte number was caused by its estrogenic character, we added estrogens to OBM. The estrogen receptors α (ERα) and β (ERβ) both bind estradiol (E2). Recently, ERα- and ERβ-specific agonists have been developed. One ERα-specific agonist, PPT, is 410-fold more potent in binding to ERα than ERβ,(24) whereas DPN binds to ERβ with an affinity 72-fold higher than to ERα.(25) When these ER ligands (E2, PPT, DPN) were added to OBM, no significant effect was seen on formation of adipocytes compared with cell culture with OBM only (data not shown). The above findings strongly suggest that Sirt1 influenced adipocyte formation when mesenchymal cells were directed to differentiate into osteoblasts. As described above, Sirt1 was earlier shown to decrease PPARγ activity. To initially study if activated Sirt1 could inhibit PPARγ in our cell model, we cultured cells in OBM in the presence of both the PPARγ agonist troglitazone and the Sirt1 activator resveratrol. As can be seen in Fig. 2, the effect of troglitazone on adipocyte formation was blocked by resveratrol. This indicates that resveratrol inhibited PPARγ.
Analysis of adipocyte and osteoblast markers in C3H10T1/2 cells
To further analyze the influence of Sirt1-affecting compounds on differentiation of C3H10T1/2 cells, we analyzed the expression of markers for adipocyte and osteoblast differentiation with quantitative real-time PCR. Markers chosen for adipocyte differentiation were PPARγ2 and aP2. For osteoblast differentiation, we measured ALP, collagen 1α1, osteocalcin, Runx2, IL-6, OPG, and RANKL. After 2 weeks of culture of cells in OBM, with or without added compounds, RNA was isolated from the cells and marker gene expression analyzed with real-time PCR. Expression was normalized to the expression of GAPDH. The expression of marker genes in the C3H10T1/2 cells cultured in OBM with a compound affecting Sirt1 was compared with gene expression in C3H10T1/2 cells cultured in OBM only.
Cells treated with troglitazone and nicotinamide displayed a general downregulated expression of the osteoblast markers ALP, collagen 1α1, osteocalcin, Runx2, IL-6, OPG, and RANKL compared with cells grown in OBM only (Fig. 3, please note the logarithmic scale on the y-axis). Troglitazone treatment led to a more marked downregulation of Runx2, IL-6, OPG, and RANKL compared with nicotinamide, which correlates to the adipocyte numbers shown in Fig. 1 (i.e., troglitazone and nicotinamide treatment resulted in 34% and 21% adipocytes, respectively). Downregulation of collagen 1α1 and osteocalcin with troglitazone and nicotinamide was similar. Interestingly, ALP was more downregulated in nicotinamide- than troglitazone-treated cells. The reason for this is at present unclear, but the findings show that the marker genes are differently regulated. It is likely that mRNA for these osteoblast markers was not present in adipocytes. Activation of PPARγ2, directly by troglitazone and indirectly through nicotinamide-induced inhibition of Sirt1, obviously resulted in decreased osteoblast differentiation while having a positive effect on the formation of adipocytes.
The effect of troglitazone and nicotinamide on adipocyte markers was also analyzed. Nicotinamide upregulated expression of aP2 and PPARγ2 1.2-fold (Fig. 4, please note the logarithmic scale on y-axis). This upregulation was, however, not significantly different from expression levels in cells only cultured in OBM. Troglitazone upregulated aP2-expression 2.4-fold and downregulated PPARγ2 mRNA 2.7-fold. A likely explanation for the latter effect is, as mentioned above, that troglitazone treatment results in downregulation of PPARγ2 mRNA expression.
Compounds that inhibit PPARγ2 expression or activity resulted in downregulation of the adipocyte markers. The compounds that inhibited formation of adipocytes, resveratrol, isonicotinamide, and 1α,25(OH)2D3, also markedly reduced expression of both adipocyte markers (Fig. 4). With 1α,25(OH)2D3, PPARγ2 mRNA was decreased 6-fold, and for both resveratrol and isonicotinamide, the decrease was 7-fold. The aP2 gene expression was decreased 166-fold with 1α,25(OH)2D3 treatment and 7- and 55-fold with isonicotinamide and resveratrol treatment, respectively. However, their respective effects on expression of osteoblast markers were slightly different (Fig. 5, please note the logarithmic scale on y-axis). They produced a downregulation, 1.5- to 2-fold, of Runx2 expression and increased OPG-expression, 7-, 3.5-, and 8-fold, respectively. With regard to IL-6 expression, resveratrol and 1α,25(OH)2D3 treatment resulted in 1.6- and 2.4-fold increases and were unchanged with isonicotinamide. ALP expression was increased 3-fold with resveratrol but was not affected by isonicotinamide and 1α,25(OH)2D3. Another difference was found for RANKL mRNA expression, which was not affected by isonicotinamide but increased 1.6- and 114-fold after treatment of C3H10T1/2 cells with resveratrol and 1α,25(OH)2D3, respectively.
Effects on rat primary bone marrow stromal cells by compounds regulating Sirt1 activity
To further substantiate the above-described findings with C3H10T1/2 cells, we also studied the effects of Sirt1-activating and -inhibiting compounds on rat primary bone marrow stromal cells. In Fig. 6A can be seen that osteoblast markers ALP and collagen 1α1 decreased in nicotinamide-exposed primary cells, whereas osteocalcin expression was not different from cells cultured in only bone-inducing medium. At present it is not clear why nicotinamide increased expression of Runx2. With isonicotinamide, osteoblast markers ALP, Runx2, and osteocalcin increased, whereas collagen 1α1 was not statistically different from control. Resveratrol increased expression of Runx2 and osteocalcin. The reason for these slightly different effects on osteoblast marker expression is unclear but could be caused by a heterogenous population of rat primary bone marrow stromal cells. Furthermore, we also analyzed the effects of compounds acting on Sirt1 or PPARγ on mineralization in cultures of rat primary cells. Mineralization was detected, by von Kossa and alizarin red, when cells were cultured in bone-inducing medium (Fig. 7). This was prevented with the addition of nicotinamide and troglitazone to medium. Mineralization in isonicotinamide-exposed cells resembled mineralization in bone-inducing medium, whereas 1α,25(OH)2D3 and resveratrol increased mineralization.
The nuclear NAD-dependent protein deacetylase Sirt1 belongs to class III of histone deacetylases and was first described in yeast in which it deacetylates histones, resulting in transcriptional silencing. In eukaryotes, Sirt1 also has nonhistone substrates, including transcription factors (e.g., p53, forkhead transcription factors, NF-κB, and MyoD) and other nuclear proteins [e.g., TAF(I)68 (leading to repression of RNA polymerase I activity and thereby a decrease in protein synthesis) and p300 (leading to repression of p300 acetylase activity)].(18,26) Thus, Sirt1 takes part in epigenetic regulation of genome architecture and gene expression. In general, increased Sirt1 activity results in decreased proliferation, increased differentiation, and decreased apoptosis. The effect on apoptosis is probably caused by Sirt1 deacetylating p53 at specific sites, which results in decreased transcriptional activity of p53.(18) Interestingly, transgenic mice expressing mutated p53 with increased activity exhibited, among other phenotypes, osteoporosis.(27) Sirt1 also indirectly influences the transcriptional activity of the nuclear receptor PPARγ by docking the nuclear receptor co-repressors NCoR and SMRT to PPARγ. This mechanism was shown to be responsible for the ability of Sirt1 to mobilize fat in fully differentiated adipocytes and also attenuate development of adipocytes from pre-adipocytes.(17)
The antidiabetic TZDs are PPARγ ligands that increase insulin sensitivity and reduce plasma glucose and lipid levels. Experimental studies have also shown that TZDs, in vitro and in vivo, inhibit osteoblast differentiation,(13–15) and because of this, it has been put forward that osteoporosis might develop on long-term treatment with TZDs.(16) Because of their ability to enhance adipocyte differentiation, which results in weight gain, efforts are being made to identify new PPARγ ligands.(28) The mechanism by which PPARγ regulates insulin sensitivity is not exactly known. Heterozygous PPARγ (PPARγ+/−) deficient mice having reduced PPARγ expression have been shown to be resistant to high-fat diet-induced obesity and insulin resistance, showing a beneficial effect of reduced PPARγ activity.(29) Furthermore, Rieusset et al.(30) showed that a selective PPARγ antagonist, SR-202, reduced high-fat diet-induced obesity and insulin resistance. These two studies show the complexity of the PPARγ mechanism of action. Interestingly, heterozygous PPARγ (PPARγ+/−) deficient mice were recently shown to have enhanced osteogenesis.(11) Thus, it is possible that inhibition of PPARγ activity could be beneficial in the treatment of both diabetes and osteoporosis.
In preliminary studies, we found that, in addition to inducing differentiation of C3H10T1/2 cells to osteoblasts, BME containing ascorbic acid, β-glycerophosphate, and dexamethasone (OBM) also induced differentiation of a fraction of cells into adipocytes (Fig. 1). Because PPARγ is an important regulator of adipocyte differentiation and its activity can be regulated by Sirt1, we found it interesting to study if activation or inhibition of Sirt1 could influence adipocyte differentiation of C3H10T1/2 mesenchymal stem cells when cultured in osteoblast-inducing medium. To activate Sirt1, we used resveratrol and isonicotinamide, and to inhibit Sirt1, we used nicotinamide. Furthermore, to directly activate PPARγ and increase adipocyte differentiation, we used the TZD troglitazone. We also used 1α,25(OH)2D3 because it is known to inhibit adipogenesis and to stimulate osteoblastogenesis, partly because of its ability to inhibit the expression of PPARγ2.(20) Further studies are needed to identify the underlying molecular mechanisms by which adipocyte formation from C3H10T1/2 cells cultured in OBM is promoted. However, it is likely that PPARγ is activated when C3H10T1/2 cells are cultured in OBM because inhibition of its activity by resveratrol and isonicotinamide or expression by 1α,25(OH)2D3 reduced or inhibited formation of adipocytes. It is not known which ligand(s) might have activated PPARγ under these circumstances. A possible source of PPARγ ligands could be fatty acids present in serum.(31)
Resveratrol and isonicotinamide were shown to be as potent as 1α,25(OH)2D3 in inhibiting adipocyte differentiation from C3H10T1/2 cells. Also, resveratrol and isonicotinamide increased expression of osteoblast markers. Interestingly, the expression pattern for osteoblast markers was different for 1α,25(OH)2D3, resveratrol, and isonicotinamide. In resveratrol-treated cells, ALP expression was higher than in 1α,25(OH)2D3- and isonicotinamide-treated cells. RANKL expression was lower in isonicotinamide- and resveratrol-treated cells. This could be interpreted that, during this treatment period, resveratrol and isonicotinamide did not induce fully differentiated osteoblasts or that fewer cells had reached that stage of differentiation. Further studies comparing 1α,25(OH)2D3, resveratrol, and isonicotinamide are needed to resolve this issue.
Resveratrol has also been shown to act as a phytoestrogen (i.e., activating the estrogen receptor), and because of this, to decrease osteoporosis.(32) To rule out that the effect we found with resveratrol was caused by its estrogenic character, we treated C3H10T1/2 cells with E2, the ERα-specific ligand PPT,(24) or the ERβ-specific ligand DPN,(25) but no effect on adipocyte or osteoblast differentiation was seen. Thus, it is likely that resveratrol was acting on Sirt1. Also, because resveratrol totally blocked the effect of troglitazone on adipocyte formation, it is likely that resveratrol-activated Sirt1 inhibited PPARγ.
We also found that rat primary bone marrow stromal cells responded in a similar way as C3H10T1/2 cells to resveratrol, isonicotinamide, and nicotinamide with regard to expression of osteoblast markers. Furthermore, mineralization was increased with resveratrol and inhibited with nicotinamide. Thus, also in these cells, compounds reported to regulate Sirt1 activity influenced osteoblast differentiation.
To our knowledge, this is the first time the effect of Sirt1 activity and resveratrol, isonicotinamide, and nicotinamide on differentiation of MSCs has been studied. We found that activation of Sirt1 in C3H10T1/2 MSCs decreases adipocyte and increases osteoblast differentiation. It is likely that the decrease in adipocyte differentiation is through inhibition of PPARγ. The increase in osteoblast differentiation could either be caused by inhibition of PPARγ that results in earlier initiation of the osteoblast differentiation program or that Sirt1, in addition to inhibition of PPARγ, also stimulates mechanism(s) regulating osteoblast differentiation. One possible mechanism could be deacetylation of p53. As mentioned above, transgenic mice with activated p53 exhibited osteoporosis.(27) Further studies are needed to resolve this issue.
Sirt1 has been shown to take part in regulation of aging. Caloric restriction and thus activation of Sirt1 results in prolonged lifespan in several model organisms, such as yeast, C. elegans, and Drosophila. Whether Sirt1 also regulates aging in mammals has not been directly shown, but it is clear that caloric restriction prolongs lifespan in mammals. In relation to these findings, it is interesting to note that in age-related osteoporosis, adipose cells are increased in bone marrow.(3,4) It remains to be studied if reduced Sirt1 activity could be one explanation for this.
The results from our study further support the idea, as was suggested earlier,(11) that inhibition of PPARγ could be considered as a means for treating osteoporosis. One way to do this is by using PPARγ antagonists. The finding that resveratrol and isonicotinamide markedly inhibited adipocyte and promoted osteoblast differentiation is interesting and shows that other ways to inhibit PPARγ exist. However, further in vitro and in vivo studies are needed to understand the molecular details of the Sirt1 mechanism of action in MSCs. This study is also important for the evolving field of cell-based tissue engineering. By identifying factors that can control osteoblast differentiation in vitro, better protocols for growing and differentiating mesenchymal cells to be used for bone reconstruction will hopefully be developed.
Dr Marcelo Toro is gratefully acknowledged for technical support with the FACScan equipment and Karo Bio for the kind gift of the PPT compound. This work was supported by the Karolinska Institutet and the Foundation for Strategic Research (SSF).