SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Osteoblasts and adipocytes originate from common mesenchymal precursors. With aging, there is a decrease in osteoprogenitor cells that parallels an increase of adipocytes in bone marrow. We observed that rabbit serum (RS) induces adipocyte-like differentiation in human osteosarcoma SaOS-2/B10 and MG-63 cell lines, in rat ROS17/2.8 cells, and in mouse calvaria-derived osteoblastic MB1.8 cells, as evidenced by the accumulation of Oil Red O positive lipid vesicles and the decrease in alkaline phosphatase expression. Both SaOS-2/B10 and MG-63 cells, but not ROS17/2.8 nor MB1.8 cells, express significant levels of PPARγ mRNA, a member of the peroxisome proliferator activated receptor (PPAR) family that has been implicated in the control of adipocyte differentiation. However, both ROS17/2.8 and MG-63 cells express significant levels of the adipocyte selective marker, aP2 fatty acid binding mRNA, which can be further increased by RS. These cell types express PPARδ/NUC-1 but not PPARα, indicating that cells that do not express either PPARγ or PPARα are capable of differentiating into adipocyte-like cells. Transfection experiments in COS cells showed that compared with fetal bovine serum (FBS), RS is rich in agents that stimulate PPAR-dependent transcription. The stimulatory activity was ethyl acetate extractable and was 35-fold more abundant in RS than in FBS. Purification and analysis revealed that the major components of this extract are free fatty acids. Furthermore, the same fatty acids, a mixture of palmitic, oleic, and linoleic acids, activate the PPARs and induce adipocyte-like differentiation of both ROS17/2.8 and SaOS-2/B10 cells. These findings suggest that fatty acids or their metabolites can initiate the switch from osteoblasts to adipocyte-like cells.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

During development, mesenchymal stem cells can differentiate into both osteoblasts and adipocytes. It has also been demonstrated that in mature animals, bone marrow–derived stromal cells can differentiate into osteoblastic and adipocytic cells. With aging, the pool of preosteoblastic cells decreases and adipocytes increase, populating the bone marrow.(1) It was proposed that this is due to altered differentiation of common precursor cells. In vitro studies show that calvaria-derived bone cell progenitors can differentiate into adipocytes in response to 1,25-dihydroxyvitamin D3 (1,25(OH)2D3) and dexamethasone.(2) The expression of osteoblastic genes was detected in cells capable of undergoing adipocyte differentiation.(3) Furthermore, it was demonstrated that immortalized mesodermal progenitor cells can differentiate into osteogenic or adipogenic cells in response to treatment with either β-glycerophosphate and ascorbate, plus dexamethasone or dexamethasone plus insulin, respectively.(4-6) Similarly, it was shown that treatment with dexamethasone and 1,25(OH)2D3, at different times during the growth of rat marrow stromal cells, led to an inverse relationship between the differentiation of adipocytic and osteogenic cells.(7)

Recently, the peroxisome proliferator activated receptors (PPARs), a group of receptors that belong to the steroid hormone receptor superfamily, have been implicated in the control of adipocyte differentiation.(8,9) The PPAR family is composed of three receptors, PPARα, NUC-1/PPARδ/PPARβ, and PPARγ.(10-13) Experimental analysis showed that PPARγ is mainly expressed in adipose tissue and the immune system.(9,14) The PPARα receptor has a wider tissue distribution and is found in the liver, heart, kidney, and intestine,(10,14) while PPARδ/NUC-1 is ubiquitously expressed in most cells and tissues.(12,14) These receptors were shown to be activated by free fatty acids and compounds that induce peroxisome proliferation in rodents, such as the hypolipidemic drug clofibrate and structural related analogs.(10-12,14-16) Furthermore, it was shown that the PPARs and their ligands participate in the control of transcription of genes that code for the enzymes of the peroxisomal and microsomal fatty acid oxidation pathways.(9,11,15-17) Recently it was demonstrated that ligands for PPARγ, such as free fatty acids, prostaglandin J2 (PGJ2) and synthetic ligands, can induce the differentiation of preadipocytes into adipose cells.(18-21) Moreover, overexpression of PPARγ and treatment with its cognate ligand stimulates adipose differentiation of cultured fibroblasts.(8) Furthermore, ligands for PPARs were shown to improve glucose tolerance in both human and animal studies.(22-24)

Following the observation that rabbit serum induces adipocytic differentiation in cultured osteosarcoma cells,(25,26) we tried to identify the agents present in rabbit serum that are responsible for this differentiation switch. In this study, we show that compared with FBS, rabbit serum is highly enriched in hydrophobic fat-like material and free fatty acids. These fatty acids purified from RS on the basis of their ability to stimulate PPARs can induce adipocyte-like differentiation of osteoblastic cells like whole rabbit serum.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Chemicals

Various fatty acids and staining reagents were obtained from Sigma Chemical Co. (St. Louis, MO, U.S.A.). Rabbit serum and fetal bovine serum was obtained from GIBCO BRL (Gaithersburg, MD, U.S.A.).

Plasmids

The chimeric receptors were prepared by fusing the amino-terminal portion of the rat glucocorticoid receptor (GR), which includes the DNA binding domain, to the putative ligand binding domain of the nuclear receptors: human NUC-1/PPARδ,(12) mouse PPARα,(10) human PPARγ,(13) and human NER,(27) as described earlier.(12) All the receptors were expressed under the control of SV40-based expression vectors. The reporter gene was the plasmid pJA358 in which the expression of firefly luciferase is under the control of the MMTV promoter.(12)

Cells and tissue culture

SaOS-2/B10 is a human osteosarcoma cell line,(28) ROS17/2.8 is a rat osteosarcoma cell line,(29,30) and MB1.8 osteoblasts are neonatal mouse calvaria derived cells that support in vitro osteoclast formation.(31) Cells were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS) at 37°C and 5% CO2. For the 8-day differentiation assay, cells were seeded at 30,000/well in 12-well plates and allowed to grow to 50% confluence in a standard medium. For the 2-day differentiation assays, cells were seeded at 80,000/well in 12-well plates and allowed to grow to confluency in standard medium. The medium was aspirated and replaced with either 5% FBS or 5% heat inactivated rabbit serum. After 2 or 8 days, the cells were fixed in 10% formalin and were stained with Oil Red O or alkaline phosphatase (ALP). Where indicated, the concentration of rabbit serum was reduced, and a 5% serum total concentration was maintained by the addition of FBS. In some experiments, growth conditions were changed as indicated in the legends.

Adipocyte-like differentiation was measured by Oil Red O staining. Oil Red O stain (0.5 g) was dissolved in 100 ml of isopropanol. The solution is diluted 6:4 with distilled water, allowed to stand for 10 minutes, and then filtered through Whatman #1 paper. Cells are washed with phosphate-buffered saline (PBS) and then fixed in 10% formalin for approximately 10 minutes at room temperature. Cells are stained for approximately 1 h at room temperature then rinsed with distilled water and allowed to air dry.

For ALP staining, cells were washed with PBS and then fixed in 10% formalin for 10 minutes at room temperature, then rinsed again with PBS before staining. One capsule of Fast Blue RR Salt (12 mg) was added to 50 ml of saline buffer. Naphthol AS-MX phosphate, 40 μl/ml, was added to the dissolved Fast Blue in buffered saline and mixed well, 0.5 ml/well (in a 12-well plate) was added and left at room temperature or 37°C until the color appeared.

Fatty acid analysis by microbial identification system

The Microbial Identification System (MIS) consists of a Hewlett-Packard 5890 gas chromatograph (Hewlett-Packard, Palo Alto, CA, U.S.A.), an integrator, and an autosampler. It is a fully automated system that identifies bacteria based on their unique fatty acid profiles. In our study, we used this system to identify the fatty acid profiles of FBS and rabbit serum.

Ligand-dependent transcription assays

Ligand-dependent transcription assays were performed as previously described.(12) The reporter gene, MMTV-luciferase, was coexpressed with plasmids containing cDNA for the hybrid or chimeric steroid hormone receptors in COS-7 cells. After 18 h, the cells were washed and ligands were added. Forty-eight hours later, cell extracts were prepared and assayed for luciferase activity. All compounds were dried and dissolved in ethanol or dimethylsulfoxide (DMSO) and added to the assays at 500- to 1000-fold dilutions of concentrated stock solutions. Data are presented as averages and standard deviation for each point. The assays were done in triplicates or quadruplicates and were repeated at least three times. Luminescence of each sample was measured with the AutoClinilumat (Berthold, Bad Wildbad, Germany).

Ethyl acetate extraction of rabbit serum

Rabbit serum was heat-inactivated by submerging it in a water bath at 56°C for 60 minutes, and a 100 ml aliquot was removed and placed in a 250-ml conical tube. An equal volume (1:1) of ethyl acetate was added to the same tube. The mixture was mixed on an orbit shaker at approximately 1000 rpm for 5 minutes, then centrifuged at 3000 rpm for 20 minutes in a Beckman (Palo Alto, CA, U.S.A.) J2-21 centrifuge for phase separation. The supernatant was collected in prepared tubes and dried down in a rotary dryer. The residue was weighed and compared with that of FBS using the same extraction procedure. The weights of organic solvent extracted material were 140 mg/100 ml for rabbit serum as compared with 4 mg/100 ml for FBS. The rabbit serum extract was 35 times larger when compared with FBS. The residue was dissolved in hexane and loaded onto a silica gel column followed by elution with hexane, chloroform, ethyl-acetate, propanol, and ethanol. The ethanol phase, which contained most of the agonistic activity, was dissolved in ethanol and chromatographed on a Vydac C18, 5 μm, 10 mm × 250 mm column using a Waters Delta Prep 3000 high performance liquid chromatography (HPLC), employing linear gradients of 40% acetonitrile/60% water to 90% acetonitrile/10% water over 60 minutes. The activity of the various fractions was tested after evaporation of the solvents. The material was dissolved in DMSO and added to the ligand screening assays. The final concentration of DMSO did not exceed 0.2% of the cell culture medium.

Structure identification

The structure of the purified compound extracted from rabbit serum that activated the GR/PPARα-chimeric receptor was determined by nuclear magnetic resonance. Proton and carbon nuclear magnetic resonance (NMR) spectra were recorded at room temperature on a Varian VXR-500S NMR spectrometer operating at a proton frequency of 500 MHz and a carbon frequency of 125 MHz. The material present in fraction 5 (Fig. 8B) of the C4 HPLC column was dissolved in 100 atom % D CD3CN (Merck Isotopes, Darmstadt, Germany). Chemical shifts are referenced to internal tetramethylsilane (TMS).

thumbnail image

Figure FIG. 8. Transactivation of PPARα receptor by C18 HPLC fractions. (A) Ethanol fraction eluted from silica gel column (RABB-4, 2 mg) was loaded on Vydac C18 (5 m, 10 mm × 250 mm) HPLC column. The material was eluted by a linear gradient from 80% acetonitrile with 0.1% TFA in water to 20% acetonitrile with 0.1% TFA in water. (B) Samples, 1–6, were pooled (20 mg/ml) as indicated and were tested for the transactivation of GR/PPARα and GR/NER chimeric receptors. (C) Linoleic acid is the active component in fraction 5. Fraction 5 of the C18 column was analyzed by NMR and found to be linoleic acid. An authentic sample of linoleic acid gave an identical NMR spectrum. There was not enough material to analyze fraction 6.

Download figure to PowerPoint

Preparation of RNA and Northern hybridization

Total RNA was extracted by guanidinium isothiocyanate method.(32) Poly(A)+ RNA was isolated from the cells by QuickPrep Micro kit (Pharmacia, Uppsala, Sweden). Total RNA (20 μg/lane) or poly(A)+ RNA (0.35–0.5 μg/lane) was separated on 0.9% agarose gels containing 6.6% formaldehyde and transferred to nylon filter (Boehringer-Mannheim, Indianapolis, IN, U.S.A.). After transfer and ultraviolet fixation (Autolinker, Strategene, La Jolla, CA, U.S.A.), filters were hybridized with cDNA probes in a solution containing 40–50% formamide (Hybrisol I and II, Oncor, Gaithersburg, MD, U.S.A.).

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Effect of rabbit serum on cell differentiation

Using Oil Red O versus alkaline phosphatase staining as criteria for adipocytic versus osteoblastic differentiation we found that in 5% FBS containing medium, none of the cell lines tested stained with Oil Red O (Table 1 and Fig. 1A). Measuring ALP activity by the release of phosphate from fluorescein diphosphate or ALP staining showed high levels of enzyme in osteosarcoma ROS17/2.8 (>95%) and SaOS-2/B10 (>95%) cells with lower levels in MB1.8 cells (approximately 25%; Table 1 and Fig. 2). Only a small fraction of MG-63 showed ALP staining (Table 1). When the cells were switched to a medium containing 5% rabbit serum, all cell lines were positively stained with Oil Red O. The percentage of cells stained were approximately 75% of ROS17/2.8 cells, approximately 50% of the SaOS-2/B10 cells, and the majority of the MG-63 cells (Table 1 and Fig. 1B). Measuring ALP activity in ROS17/2.8, SaOS-2/B10, and MB1.8 cells grown in 5% RS showed a sharp decrease in ALP activity and the percentage of cells stained for ALP by more than 50% (Table 1 and Fig. 2). In addition, a decrease in osteocalcin mRNA levels was found in MB1.8 cells grown in RS (data not shown). Very few MG-63 cells were stained for ALP, and no osteocalcin mRNA expression was found in cells grown in either serum type.

thumbnail image

Figure FIG. 1. Effect of 5% rabbit serum and fatty acid mixture (palmitic acid, oleic acid, and linoleic acid as indicated) on the differentiation of ROS17/2.8 osteoblasts into adipocyte-like cells. Cells were grown for 2 days in the indicated medium and stained with Oil Red O: (A) fetal bovine serum; (B) rabbit serum; (C) fetal bovine serum and 100 μM palmitic acid, 30 μM oleic acid, and 30 μM linoleic acid; (D) fetal bovine serum and 30 μM palmitic acid, 100 μM oleic acid, and 30 μM linoleic acid.

Download figure to PowerPoint

thumbnail image

Figure FIG. 2. Effect of 5% rabbit serum and fatty acid mixture on the expression of ALP in osteoblasts. Human SaOS-2/B10 or ROS17/2.8 cells were grown for 2 days in the indicated medium, and the enzymatic activity of ALP was measured by the release of phosphate from fluorescein diphosphate. Fetal bovine serum (FBS); ▵ rabbit serum (RS); ○ Mix I, FBS and 100 μM palmitic acid, 30 μM oleic acid and 30 μM linoleic acid; ◊ Mix II, FBS and 30 μM palmitic acid, 100 μM oleic acid, and 30 μM linoleic acid.

Download figure to PowerPoint

Table Table 1. Effect of 5% Rabbit Serum on the Differentiation of Osteoblasts into Adipocyte-like Cells
Thumbnail image of

When the cells were grown in rabbit serum alone, the appearance of lipid vesicles were considerably faster. When cells were grown in a medium containing a mixture of 5% FBS and 5% RS, the effect of RS was still dominant and the cells were stained with Oil Red O (data not shown). Furthermore, growing cells in 2% RS and 8% FBS also resulted in cells with Oil Red O–stained vesicles. We concluded, therefore, that RS contains an ingredient that is dominant in the presence of FBS and induces osteoblastic cells to differentiate into adipocyte-like cells.

Expression of adipocytic markers in osteoblastic cells

Since fatty acids and their metabolites are ligands for the members of the PPAR family and are responsible for, at least in part, adipocyte differentiation, we estimated by Northern analysis the expression of the three PPARs in cells grown in FBS or RS. Hybridization experiments revealed significant PPARγ mRNA present in total RNA isolated from SaOS-2/B10 and MG-63 cells (Figs. 3 and 4C). No significant levels of PPARγ were observed in MB1.8 cells (Fig. 3). NUC-1/PPARδ which is expressed in most cells and tissues was found in all these cell types. High expression levels were found in SaOS-2/B10 cells and low levels were found in MB1.8 cells (Fig. 3). Neither cell line expressed significant levels of PPARα. Treatment with RS had no apparent effect on the mRNA expression of any member of the PPAR family in either SaOS-2/B10, MG-63, or MB1.8 cells (Figs. 3 and 4). However, no PPARγ mRNA was detected in RNA isolated from ROS17/2.8 which were grown in either FBS or RS (data not shown).

thumbnail image

Figure FIG. 3. mRNA expression of PPARα, PPARγ, and PPARδ/NUC-1 in human SaOS-2/B10 osteosarcoma cells and mouse calvaria-derived MB1.8 osteoblasts. Cells were grown in medium containing either 5% FBS, 5% FBS + FA (30 μM palmitic acid, 100 μM oleic acid, and 30 μM linoleic acid) or 5% RS. Total RNA was isolated and analyzed (20 μg/lane) by hybridization with the cDNA (ligand binding coding region) of PPARγ, PPARδ/NUC-1, or PPARα.

Download figure to PowerPoint

thumbnail image

Figure FIG. 4. mRNA expression of aP2 and PPARγ in ROS17/2.8, 3T3-L1, and MG-63 cells. (A) Poly(A)+ RNA was isolated from ROS17/2.8 cells and hybridized to the aP2 and actin cDNA probes. The relative increase in aP2 mRNA was determined by Bio-Rad Image Analyzer (Hercules, CA, U.S.A.), and the fold stimulation is indicated in parenthesis. Cells were grown in a medium containing 5% FBS for 2 days, then changed to the indicated serum treatment for an additional 5 days: (lane 1) 5% FBS (1×); (lane 2) 5% FBS + fatty acid mix II (2.1×); (lane 3) 4% FBS + 1% RS (1.1×); (lane 4) 3% FBS + 2% RS (2.8×); (lane 5) 2% FBS + 3% RS (6.5×) or (lane 6) 1% FBS + 4% RS (7.9×). (B) Poly(A)+ RNA was isolated from 3T3-L1 cells, which were grown in a medium containing either 10% FBS or 8% FBS + 2% RS for 8 days and hybridized to the aP2 and GAPDH cDNA probes. (C) Total RNA was isolated from MG-63 cells that were grown in a medium containing either 5% FBS or 5% R for 5 days and hybridized to the PPARγ cDNA probe. The 28S ribosomal RNA is shown as a reference.

Download figure to PowerPoint

The adipocytic potential of these cells can be further demonstrated by the fact that when grown in either FBS or RS, ROS17/2.8 cells express the fatty acid binding protein aP2 (Fig. 4A, lane 1). Treatment of these cells with an increasing concentration of RS resulted in a dose-dependent increase in aP2 mRNA with approximate 8-fold at 4% RS (Fig. 4A, lanes 3–5). Similarly, MG-63 cells that were grown in RS, express significant levels of aP2 gene mRNA that was further increased upon treatment of the cells with ligands for either PPAR or the glucocorticoid receptor (data not shown).

The ability of RS to drive adipocytic differentiation is further demonstrated by experiments with the mouse cell line 3T3-L1. When the cells are grown in FBS, only a small fraction are stained with Oil Red O (data not shown), and very low levels of fatty acid binding protein mRNA, aP2, is expressed (Fig. 4B). When the cells are switched to medium containing RS, a large proportion of the cells differentiate into Oil Red O–stained cells (data not shown), and levels of the adipocytic aP2 mRNA is markedly induced (Fig. 4B). These observations show that RS can drive adipocytic differentiation of the preadipocytic 3T3-L1 cells as well as cells that are at least in part already differentiated into the osteoblastic lineage.

Effect of rabbit serum and FBS on the transcription mediated by PPARα and NUC-1 receptors

To identify the agent present in RS that is responsible for the adipocytic differentiation, we determined that the active ingredient found in RS is heat stable (data not shown). We then tested whether this component could be extracted with hydrophobic solvents. One hundred milliliters of FBS or RS was extracted with ethyl acetate, and the organic solvent was evaporated. The dry weight of the extracted material from FBS and RS was 4 mg and 140 mg, respectively. The RS extracted hydrophobic material had the appearance of fat-like material. Furthermore, the addition of the RS ethyl acetate extractable material to osteoblasts grown in 5% FBS stimulated differentiation into adipocyte-like cells (data not shown).

Since members of the PPAR family have been implicated in the control of lipid metabolism and are activated by fatty acids and their metabolites, we examined whether RS and its extract can activate members of the PPAR family. As previously described, we used ligand-dependent transcription assays in which the expression from the MMTV-luciferase reporter gene is controlled by one of the three GR/PPARs chimeric receptors.(12) As shown in Fig. 5, treatment of the GR/PPARα transfected cells with RS resulted in a 27-fold increase in luciferase expression compared with cells grown in FBS. Similarly, a 5.6-fold increase of transcription by RS was found in COS cells transfected with the GR/NUC-1 chimeric receptor. Furthermore, RS treatment of cells transfected with GR/PPARγ resulted in a 2.7-fold increase in luciferase transcription, which was of the same order of magnitude as stimulation with 1 μM BRL-49653, a PPARγ selective ligand (Fig. 6). Ligand-dependent transcription assays with chimeric receptors composed of the GR DNA-binding domain and the ligand binding domains of two receptors unrelated to PPAR, NER (GR/NER) or progesterone (GR/PR), displayed no response to treatment with RS (Fig. 5). This indicates that the agents present in RS exhibit selective transactivation of the PPARs.

thumbnail image

Figure FIG. 5. Rabbit serum stimulates the transcription mediated by PPARα and NUC-1 receptors. COS cells grown in FBS were cotransfected with MMTV-luciferase reporter gene and either GR/PPARα, GR/NUC-1, GR/NER, or GR/PR chimeric receptors. After transfection, the cells were grown in either FBS or rabbit serum, and the luciferase enzyme activities were measured 2 days later. (Some error bars are not visible due to low standard deviation.)

Download figure to PowerPoint

thumbnail image

Figure FIG. 6. Rabbit serum, BRL-49653, and linoleic acid activate transcription mediated by PPARγ. Cells grown in FBS were transfected with MMTV-luciferase and GR/PPARγ chimeric receptor and treated with 0.1 mg/ml RS extract, 1 μM BRL-49653, and 30 μM linoleic acid. (Some error bars are not visible due to low standard deviation.)

Download figure to PowerPoint

Purification of the active ingredient from rabbit serum

It has been demonstrated that ligands for PPARs induce the differentiation of preadipocytes to adipose cells,(18,19) and that PPARγ is an important regulator of adipocyte differentiation.(8) Therefore, we assumed that the agent(s) in RS that is responsible for adipocytic differentiation may also be responsible for the transactivation of the PPAR, and thus we used the transactivation of the PPARs as assays for the purification of the agent(s) present in RS that are responsible for adipocyte differentiation.

We first examined the ethyl acetate material extracted from RS on the transactivation of the PPARs. Similar to RS, the ethyl acetate extracted material stimulates the transcription controlled by GR/PPARα and not by GR/NER chimeric receptors (Fig. 7A), thus maintaining receptor specificity. Using a solid-phase extraction system, the RS ethyl acetate extract was passed through a silica gel column (70–230 mesh), and the material was eluted with solvents having decreased hydrophobicity: hexane, methylene chloride, ethyl acetate, and ethanol. The fractions were named, RABB-1, -2, -3, and -4, respectively. The solvents were evaporated, and the residues, dissolved in DMSO, were tested in the ligand-dependent transcription assay (Fig. 7B). Fractions RABB-2, -3, and -4 exhibited strong stimulation of GR/PPARα-dependent transcription, fraction RABB-4 showed the highest stimulation at 58-fold. None of the silica gel column fractions stimulated the GR/NER-dependent transcription, indicating that transactivation of PPARα was receptor specific (Fig. 7B).

thumbnail image

Figure FIG. 7. Ethyl acetate extracts and its silica gel fractions selectively stimulate the PPARα-dependent transcription. COS cells grown in FBS were cotransfected MMTV-luciferase reporter gene and with either the GR/PPARα or GR/NER chimeric receptors. After transfection, the cells were grown in FBS supplemented with the indicated fraction as 0.1% of the medium. Luciferase enzyme activities were measured 2 days later. In (A) extract is the ethyl acetate extract of rabbit serum (100 mg/ml). In (B), the ethyl acetate extract was passed on a silica gel column, and fractions were eluted with various solvents. The solvents were evaporated and samples were dissolved in DMSO to 100 mg/ml and added to the cells (hexane, RABB-1; methylene chloride, RABB-2; ethylacetate, RABB-3; and ethanol, RABB-4). (Some error bars are not visible due to low standard deviation.)

Download figure to PowerPoint

To identify the agent that activated PPARα, fraction RABB-4 was further purified on a Vydac HPLC C18 column (Fig. 8A). The 210 nM absorption peaks were tested individually (peaks 1, 3, 4, and 5) or as pools (peaks 2 and 6), as shown in Fig. 8A. Pooled fractions 2 and 6 stimulated transcription mediated by GR/PPARα but not by GR/NER (Fig. 8B). The material in peak 5 was analyzed by NMR spectroscopy and was found to be pure linoleic acid (Fig. 8C). An authentic sample of linoleic acid gave an identical NMR spectrum. The material in peak 6 that activated PPARα was not analyzed due to its impurity and lack of adequate sample volume.

Linoleic acid was shown previously to be a potent activator of both PPARα and NUC-1 receptor.(12) Like RS and 1 μM BRL-49653, 30 μM linoleic acid stimulated transcription mediated by GR/PPARγ to a comparable extent (Fig. 6).

Fatty acid analysis of rabbit and fetal bovine serum by the Microbial Identification System

Since we identified at least one of the agents that may be responsible for the effects of RS, we proceeded to analyze the spectrum of free fatty acids present in FBS and RS by using the MIS chromatographic unit (see Materials and Methods section). Approximately 95% and 97% of the fatty acids present were identified by the MIS in FBS and RS, respectively. As shown in Table 2, and previously observed by ethyl acetate extraction, RS contains 4-fold higher amounts of free fatty acids than FBS, and there were significant differences in the amounts of individual fatty acids found (Table 2). For example, linoleic acid derivatives constitute 7% of total fatty acids in FBS versus 21% in RS. Palmitic acid constitutes 27% of total fatty acids in FBS versus 39% in RS. Interestingly, arachidonic acid derivatives were 8.7% of the total free fatty acids in FBS compared with only 0.67% in RS. Both samples contained relatively small amounts of the lower carbon chain length (C12–C16) saturated fatty acids.

Table Table 2. Fatty Acid Analysis of Rabbit and Fetal Bovine Sera
Thumbnail image of

In FBS, palmitic acid (27%), oleic acid (22%), and steric acid (13%) were the major fatty acid components and accounted for approximately 62% of the identified fatty acids. In RS, palmitic acid (39%), linoleic acid derivatives (21%), and oleic acid (21%) were the major components and accounted for approximately 81% of the identifiable fatty acids.

Phenotypic conversion of osteoblastic into adipocyte-like cells by free fatty acids

We found that compared with FBS, RS is highly enriched in linoleic acid and other free fatty acids; therefore, we tested whether these fatty acids, like RS, can induce adipocytic differentiation in the ROS17/2.8 osteoblastic cell lines. Similar to the finding with RS, growing the cells in FBS and a mixture of fatty acids containing 100 μM palmitic acid, 30 μM oleic acid, and 30 μM linoleic acid (Mix I) resulted in the reduction of ALP levels relative to growing the cells in FBS alone (Fig. 2). The decrease in ALP activity was most pronounced in cells grown in FBS supplemented with a fatty acid mixture (Mix II) containing 30 μM palmitic acid, 100 μM oleic acid, and 30 μM linoleic acid (Fig. 2). This effect was also found in experiments with SaOS-2/B10 cells (data not shown).

Along with a decrease in ALP activity, growing ROS17/2.8 and SaOS-2/B10 cells in FBS and a fatty acid mixture of 100 μM palmitic acid, 30 μM oleic acid and 30 μM linoleic acid (Mix I) induced the development of lipid vesicles stained with Oil Red O in a fraction of cells (Fig. 1C). Growing the cells in FBS supplemented with the fatty acid mixture of 30 μM palmitic acid, 100 μM oleic acid, and 30 μM linoleic acid (Mix II) resulted in a higher level of Oil Red O lipid staining, which was comparable to that induced by 5% RS (Fig. 1D).

In SaOS-2/B10 cells grown in FBS supplemented with the fatty acid mixture of 30 μM palmitic acid, 100 μM oleic acid, and 30 μM linoleic acid (Mix II), there was a significant increase in PPARγ mRNA, while the expression of NUC-1 or PPARα was not changed (Fig. 3). Similar treatment of MB1.8 or Ros17/2.8 cells did not result in changes of PPAR expression. However, treatment of ROS17/2.8 cells with this fatty acid mixture resulted in an increase in aP2 mRNA (Fig. 4, lanes 1 and 2).

To show that osteoblastic cells can undergo differentiation into adipocyte-like cells in the absence of a fatty acid overload from RS, we tested to see if a PPARγ selective ligand, BRL-49653, can induce adipocyte differentiation in SaOS-2/B10 osteoblastic cells. Similar to SaOS-2/B10 osteoblastic cells grown in RS, growing SaOS-2/B10 cells in FBS containing BRL-49653 (1–50 μM) resulted in a dose-dependent adipocyte differentiation as determined by Oil Red O staining (Fig. 9). In contrast, MB1.8 cells that express only trace levels of PPARγ did not respond to the treatment with BRL-49653 (data not shown). Therefore, this indicates that the free fatty acids present in RS can drive adipocytic differentiation of osteoblastic cells.

thumbnail image

Figure FIG. 9. The PPARγ ligand BRL-49653 induces adipocyte differentiation in SaOS-2/B10 osteoblastic cells. Confluent cells were treated with BRL-49653, or 5% RS for 8 days in medium containing 10% serum. Cells were fixed and stained with Oil Red O as described above. The percentage of Oil Red O stained cells was determined by light microscopy, scoring 500 cells from each of three replicate wells.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Osteoblastic cell lines transferred from a medium containing FBS to one containing RS develop adipocyte-like features, such as lipid vesicles which become stained with Oil Red O and a parallel decrease in the activity of the osteoblastic ALP, suggesting a switch from an osteoblastic to an adipocytic-like phenotype. The ability of RS to promote adipocytic differentiation was further demonstrated by the fact that RS induced the late adipocytic differentiation marker gene, aP2, in both 3T3-L1 preadipocytic cells and the osteoblastic ROS17/2.8 cell line. This shows that the agents present in RS are capable of inducing some cells to differentiate into genuine adipocytic cells.

Ligand-dependent transcription assays mediated by the PPARs revealed that RS is rich in components that transactivate the members of the PPAR family. Purification of the RS ingredients responsible for these effects showed that, compared with FBS, RS is highly enriched in free fatty acids that are potent transactivators of the PPARs and can promote the adipocyte-like cell differentiation of osteoblasts. These results agree with the findings that PPAR ligands such as free fatty acids, BRL-49653, WY-14643, 2-bromopalmitate, and their receptors can induce the differentiation of preadipocytes to adipose cells.(8,9,18,19,33)

Comparing the types of fatty acids present in FBS and RS revealed differences in both the type of fatty acids present and their amount. RS contains approximately 35 times more hydrophobic fat-like material than FBS. Although we did not fully analyze the composition of this material, MIS analysis revealed that RS is 4-fold richer in free fatty acids and their derivatives than FBS. The high content of fatty acids and lipid material in RS, compared with FBS, may contribute to the fast accumulation of lipids in the vesicles and shorten the time required for the development of adipocytic-like phenotype. Surprisingly, we found that the fraction of arachidonic acid in FBS is 13-fold higher than in RS. Arachidonic acid is the precursor for prostaglandins, including 15d-PGJ2, which is a ligand for PPARγ and has been implicated in adipocyte differentiation.(20,21) Therefore, one might expect FBS to be more potent than RS in driving adipocytic differentiation. However, the levels of arachidonic acid in FBS may not support the production of metabolites which promote adipocytic differentiation in these cells. From these studies it is clear that fatty acids other than arachidonic acid which are present at high concentrations in RS are capable of driving adipocytic differentiation.

Relative to FBS, the fact that RS is rich in free fatty acids was further confirmed by a direct analysis of the free fatty acid composition of FBS and RS. MIS analysis revealed that palmitic and linoleic acids constitute approximately 39% and 21% of the free fatty acids, respectively, in RS compared with only 27% and 7% in FBS. Indeed, addition of pure synthetic fatty acid mixtures consisting of palmitic, oleic, and linoleic acids to FBS reproduced the induction of lipid vesicles in osteoblasts and partially decreased the expression of ALP. We found that a fatty acid mixture that contains 100 μM oleic acid was more effective than a mixture containing a similar concentration of palmitic acid. The reason for the selective potency of linoleic acid is not clear. However, in our experience and others, unsaturated fatty acids are more potent activators of NUC-1/PPARδ and PPARγ,(11,12,15,34) yet the potential role of oleic acid metabolites cannot be excluded.

Northern analysis of RNA isolated from the various osteoblastic cell lines showed that all expressed some level of PPARδ/NUC-1. This is in agreement with the previous observation that this receptor is expressed in most tissues and cells, including bone.(12,14) As reported in other studies that showed a more restricted distribution of the other two members of the PPAR family, we did not find any significant expression of PPARα in the osteoblastic cell lines. While significant levels of PPARγ were present in SaOS-2/B10 cells and MG-63, only traces of PPARγ were present in MB1.8 or ROS17/2.8 cells. Furthermore, treatment with a PPARγ ligand BRL-49653 did not result in Oil Red O staining of MB1.8 or ROS17/2.8 cell lines, suggesting the lack of PPARγ involvement in these rodent cells. Interestingly, in SaOS-2/B10 cells we observed a significant increase in PPARγ in response to treatment with a mixture of fatty acids, which was shown to be an early phase of adipocytic differentiation.(9) Furthermore, both RS and the fatty acid mixture increased the expression of the adipocyte-specific aP2 fatty acid binding protein in ROS17/2.8 osteoblasts. This shows that these cells have the potential to differentiate into adipocytic-like cells. In addition, like fatty acids, BRL-49653, a PPARγ selective ligand, results in an increase of Oil Red O staining of SaOS-2/B10. This shows that the increase in Oil Red O staining in these cells in not due to fatty acid overload but to an active adipocytic differentiation process.

These observations suggest that fatty acids may induce SaOS-2/B10 or ROS17/2.8 cells to differentiate further on the adipocytic path. Although we found only traces of PPARγ in ROS17/2.8 cells, these cells did express significant levels of the aP2 fatty acid binding protein. This may imply that a cell type that is considered to be a relatively mature osteoblastic cell can still express a late adipocytic marker. However, the expression of adipocytic markers in ROS17/2.8, MG-63, and SaOS-2/B10 cells could result from the fact that these are tumor cells. The fact that osteoblastic cell lines already express some levels of the adipocyte-specific genes may explain the relatively short time it takes for the development of Oil Red O staining in these cells. Although it was demonstrated by overexpressing the PPARs that PPARγ or PPARα, but not NUC-1/PPARδ, is capable of inducing adipocyte differentiation, it is possible that NUC-1/PPARδ may also contribute to this process. It was demonstrated that fatty acids and thiazolidinediones, agents that transactivate PPARs, promote adipocytic differentiation in myoblast cell lines, which express NUC-1/PPARδ but not PPARγ or PPARα, and inhibit the formation of myotubes.(35) Furthermore, it was shown that NUC-1/PPARδ can activate the expression of adipose tissue–specific genes.(24,36,37) Thus, NUC-1/PPARδ may also have the ability to initiate adipocyte differentiation.(33)

In parallel to the development of the adipocyte-like phenotype, we observed that RS and fatty acids inhibit the expression of ALP, which is required for the mineralization process. Thus, with the development of the adipocytic phenotype, these osteoblasts are losing their osteoblastic phenotype. Furthermore, it was reported that treatment of rats with thiazolidinedione (Pioglitazone) for 28 days resulted in an increase of body fat and a decrease of bone mineral density of the tibia.(38) This observation suggests that conversion of the osteoblastic phenotype to an adipocytic one by activators of PPARs in vitro may also occur in vivo. Interestingly, high levels of serum fatty acids (3 mM) and low bone density have been reported in diabetic patients and could be related to the observations described.

In conclusion, this study shows that compared with FBS, RS contains a high concentration of fatty acids. In vitro, these fatty acids suppress ALP expression in osteoblastic cells and induce adipocyte-like feature. This suggests that cells from an osteoblastic lineage may convert to an adipocytic pool of cells, but physiological/pathological significance of this observation remains to be elucidated by in vivo experiments.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

We thank Mr. Joseph C. Vido and Mr. Michael Nuzzolo, Biological Testing Labs, Merck & Co., Inc., for their timely assistance in running our samples on the Microbial Identification System.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES
  • 1
    Gimble JM, Robinson CE, Wu X, Kelly KA 1996 The function of adipocytes in bone marrow stroma: An update Bone 19:421428.
  • 2
    Bellows CG, Wang YH, Heersche JN, Aubin JE 1994 1,25-dihydroxyvitamin D3 stimulates adipocyte differentiation in cultures of fetal rat calvaria cells: Comparison with the effects of dexamethasone Endocrinology 134:22212229.
  • 3
    Dorheim MA, Sullivan M, Dandapani V, Wu X, Hudson J, Segarini PR, Rosen DM, Aulthouse AL, Gimble JM 1993 Osteoblastic gene expression during adipogenesis in hematopoietic supporting murine bone marrow stromal cells J Cell Physiol 154:317328.
  • 4
    Grigoriadis AE, Heersche JN, Aubin JE 1988 Differentiation of muscle, fat, cartilage, and bone from progenitor cells present in a bone-derived clonal cell population: Effect of dexamethasone J Cell Biol 106:21392151.
  • 5
    Grigoriadis AE, Heersche JN, Aubin JE 1990 Continuously growing bipotential and monopotential myogenic, adipogenic, and chondrogenic subclones isolated from the multipotential RCJ 3.1 clonal cell line Dev Biol 142:313318.
  • 6
    Poliard A, Nifuji A, Lamblin D, Plee E, Forest C, Kellermann O 1995 Controlled conversion of an immortalized mesodermal progenitor cell towards osteogenic, chondrogenic, or adipogenic pathways J Cell Biol 130:14611472.
  • 7
    Beresford JN, Bennett JH, Devlin C, Leboy PS, Owen ME 1992 Evidence for an inverse relationship between the differentiation of adipocytic and osteogenic cells in rat marrow stromal cell cultures J Cell Sci 102:341351.
  • 8
    Tontonoz P, Hu E, Spiegelman BM 1994 Stimulation of adipogenesis in fibroblasts by PPAR gamma 2, a lipid-activated transcription factor Cell 79:11471156.
  • 9
    Tontonoz P, Hu E, Graves RA, Budavari AI, Spiegelman BM 1994 mPPAR gamma 2: Tissue-specific regulator of an adipocyte enhancer Genes Dev 8:12241234.
  • 10
    Issemann I, Green S 1990 Activation of a member of the steroid hormone receptor superfamily by peroxisome proliferators Nature 347:645650.
  • 11
    Dreyer C, Krey G, Keller H, Givel F, Helftenbein G, Wahli W 1992 Control of the peroxisomal beta-oxidation pathway by a novel family of nuclear hormone receptors Cell 68:879887.
  • 12
    Schmidt A, Endo N, Rutledge SJ, Vogel R, Shinar D, Rodan GA 1992 Identification of a new member of the steroid hormone receptor superfamily that is activated by a peroxisome proliferator and fatty acids Mol Endocrinol 6:16341641.
  • 13
    Zhu Y, Alvares K, Huang Q, Rao MS, Reddy JK 1993 Cloning of a new member of the peroxisome proliferator-activated receptor gene family from mouse liver J Biol Chem 268:2681726820.
  • 14
    Braissant O, Foufelle F, Scotto C, Dauca M, Wahli W 1996 Differential expression of peroxisome proliferator-activated receptors (PPARs): Tissue distribution of PPAR-alpha, -beta, and -gamma in the adult rat Endocrinology 137:354366.
  • 15
    Dreyer C, Keller H, Mahfoudi A, Laudet V, Krey G, Wahli W 1993 Positive regulation of the peroxisomal beta-oxidation pathway by fatty acids through activation of peroxisome proliferator-activated receptors (PPAR) Biol Cell 77:6776.
  • 16
    Kliewer SA, Forman BM, Blumberg B, Ong ES, Borgmeyer U, Mangelsdorf DJ, Umesono K, Evans RM 1994 Differential expression and activation of a family of murine peroxisome proliferator-activated receptors Proc Natl Acad Sci USA 91:73557359.
  • 17
    Tugwood JD, Issemann I, Anderson RG, Bundell KR, McPheat WL, Green S 1992 The mouse peroxisome proliferator activated receptor recognizes a response element in the 5′ flanking sequence of the rat acyl CoA oxidase gene EMBO J 11:433439.
  • 18
    Amri EZ, Ailhaud G, Grimaldi PA 1994 Fatty acids as signal transducing molecules: Involvement in the differentiation of preadipose to adipose cells J Lipid Res 35:930937.
  • 19
    Chawla A, Schwarz EJ, Dimaculangan DD, Lazar MA 1994 Peroxisome proliferator-activated receptor (PPAR) gamma: Adipose-predominant expression and induction early in adipocyte differentiation Endocrinology 135:798800.
  • 20
    Kliewer SA, Lenhard JM, Willson TM, Patel I, Morris DC, Lehmann JM 1995 A prostaglandin J2 metabolite binds peroxisome proliferator-activated receptor gamma and promotes adipocyte differentiation Cell 83:813819.
  • 21
    Forman BM, Tontonoz P, Chen J, Brun RP, Spiegelman BM, Evans RM 1995 15-Deoxy-delta 12, 14-prostaglandin J2 is a ligand for the adipocyte determination factor PPAR gamma Cell 83:803812.
  • 22
    Bar Tana J, Ben Shoshan S, Blum J, Migron Y, Hertz R, Pill J, Rose Khan G, Witte EC 1989 Synthesis and hypolipidemic and antidiabetogenic activities of beta,beta,beta′,beta′-tetrasubstituted, long-chain dioic acids J Med Chem 32:20722084.
  • 23
    Kobayashi M, Shigeta Y, Hirata Y, Omori Y, Sakamoto N, Nambu S, Baba S 1988 Improvement of glucose tolerance in NIDDM by clofibrate: Randomized double-blind study Diabetes Care 11:495499.
  • 24
    Ibrahimi A, Teboul L, Gaillard D, Amri EZ, Ailhaud G, Young P, Cawthorne MA, Grimaldi PA 1994 Evidence for a common mechanism of action for fatty acids and thiazolidinedione antidiabetic agents on gene expression in preadipose cells Mol Pharmacol 46:10701076.
  • 25
    Hart KA, Campbell JK, Brown E, Johnstone D 1994 Analysis of the adipocytic phenotype of RC2. An immortelized osteosarcoma cell line J Bone Miner Res 9:S231.
  • 26
    Nuttall ME, Olivera DL, Gown M 1994 Control of osteoblast/adipocyte differentiation in MG-63 cells J Bone Miner Res 9:S166.
  • 27
    Shinar DM, Endo N, Rutledge SJ, Vogel R, Rodan GA, Schmidt A 1994 NER, a new member of the gene family encoding the human steroid hormone nuclear receptor Gene 147:273276.
  • 28
    Rodan SB, Fischer MK, Egan JJ, Epstein PM, Rodan GA 1984 The effect of dexamethasone on parathyroid hormone stimulation of adenylate cyclase in ROS 17/2.8 cells Endocrinology 115:951958.
  • 29
    Majeska RJ, Rodan SB, Rodan GA 1980 Parathyroid hormone-responsive clonal cell lines from rat osteosarcoma Endocrinology 107:14941503.
  • 30
    Majeska RJ, Rodan GA 1982 Alkaline phosphatase inhibition by parathyroid hormone and isoproterenol in a clonal rat osteosarcoma cell line: Possible mediation by cyclic AMP Calcif Tissue Int 34:5966.
  • 31
    Schmidt A, Rutledge SJ, Endo N, Opas EE, Tanaka H, Wesolowski G, Leu CT, Huang H, Ramachandaran C, Rodan SB, Rodan GA 1996 Protein tyrosine phosphatase activity regulates osteoclast formation and function: Inhibition by alendronate Proc Natl Acad Sci USA 93:30683073.
  • 32
    Chirgwin JM, Przybla AE, MacDonald RJ, Rutter WJ 1979 Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease Biochemistry 18:52945299.
  • 33
    Mandrup S, Lane MD 1997 Regulating adipogenesis J Biol Chem 272:53675370.
  • 34
    Johnson TE, Holloway MK, Vogel R, Rutledge SJ, Perkins JJ, Rodan GA, Schmidt A, 1997 Structural requirements and cell-type specificity for ligand activation of peroxisome proliferator activated receptors. J Steroid Biochem Mol Biol (in press).
  • 35
    Teboul L, Gaillard D, Staccini L, Inadera H, Amri EZ, Grimaldi PA 1995 Thiazolidinediones and fatty acids convert myogenic cells into adipose-like cells J Biol Chem 270:2818328187.
  • 36
    Brun RP, Tontonoz P, Forman BM, Ellis R, Chen J, Evans RM, Spiegelman BM 1996 Differential activation of adipogenesis by multiple PPAR isoforms Genes Dev 10:974984.
  • 37
    Amri EZ, Bonino F, Ailhaud G, Abumrad NA, Grimaldi PA 1995 Cloning of a protein that mediates transcriptional effects of fatty acids in preadipocytes: Homology to peroxisome proliferator-activated receptors J Biol Chem 270:23672371.
  • 38
    Jennermann C, Tiantafillou J, Cowan D, Pennink BGA, Connolly KM, Morris DC 1995 Effect of thiazolidinedione on bone turnover in the rat J Bone Miner Res 10:S241.