Dr Duque serves as consultant to Procter and Gamble Pharmaceuticals and Merck-Frosst Canada. Dr Rivas states that he has no conflicts of interest.
Published online on July 2, 2007
We committed MSCs to differentiate into either osteoblasts or adipocytes and examined the effect of ALN on both adipogenesis and osteoblastogenesis. ALN inhibited adipogenesis while promoting osteoblast differentiation and activity. Our results reveal a new anabolic effect of ALN in differentiating bone marrow cells.
Introduction: Alendronate (ALN) prevents bone loss in postmenopausal patients through the regulation of osteoclastic activity. However, it has also proven to be effective in older adults where the pathophysiological mechanism is the predominance of adipogenesis over osteoblastogenesis. The aim of this study is to determine the in vitro effect of ALN on both osteoblastogenesis and adipogenesis.
Materials and Methods: Human mesenchymal stem cells (MSCs) were plated at a density of 5 × 105 cells/well in 100-cm 2 dishes containing MSC growth media. After confluence, cells were committed to differentiate adding either adipogenic or osteogenic media with and without 1,25(OH)2D3 (10−8 M) and supplemented with ALN at increasing concentrations (10−9 to 10−7 M). Untreated differentiating MSCs were used as control. Alkaline phosphatase (ALP), oil red O, and Alizarin red staining were performed at timed intervals (weeks 1 and 2). Additionally, levels of expression of both osteogenesis and adipogenesis transcription factors were measured in protein extracts. Finally, the effect of ALN on PPARγ2 nuclear activation complex was assessed.
Results: We found that ALN has a significant and dose-dependent effect on osteoblastogenesis. This effect was not modified by the presence of 1,25(OH)2D3 in the medium. Furthermore, adipogenic differentiation of MSCs was affected by addition of both ALN and 1,25(OH)2D3 to the media as confirmed by phenotype changes and lower number of lipid droplets. Finally, expression of adipogenic transcription factors and PPARγ2 activation were reduced in adipose differentiating MSCs treated with either ALN or ALN + 1,25(OH)2D3.
Conclusions: This study shows a potential anabolic effect of ALN in vitro through the stimulation of osteogenic differentiation of MSCs. Additionally, a previously unknown inhibitory effect of ALN on bone marrow adipogenesis was also found.
Osteoporosis is a common disease in older adults that predispose them to fractures. These fractures not only affect their mobility but also increase mortality in the older population.(1–3) Recent advances in the elucidation of the pathophysiology of osteoporosis in the elderly has guided to a better understanding of its treatment. It is known that additional to the increased bone resorption seen in postmenopausal osteoporosis,(4,5) which is mostly caused by greater osteoclastic activity, an age-related shift in bone marrow cells differentiation with predominant adipogenesis against osteoblastogenesis is the most important mechanism of osteoporosis in older adults.(6–8)
This loss of balance between osteoblastogenesis and adipogenesis determines that aging bones respond differently to the current osteoporosis treatment.(9) Therefore, recent advances in the prevention of osteoporotic fractures have focused on the development of medications with anabolic effect that either promote osteoblastogenesis or inhibit adipogenesis.(6,8,10,11) However, some of the current treatments for osteoporosis with known antiresorptive effect, such as bisphosphonates, are effective in the elderly,(12,13) suggesting that additionally to their capacity of regulating osteoclastic activity, they may have a positive effect on bone formation either through the stimulation of osteoblastogenesis or the inhibition of adipogenesis.
Recent studies have challenged the paradigm that treatment with bisphosphonates lead to a decrease in bone formation. Still et al.(14) have shown that at low concentrations (10−9–10−7 M) both risedronate and alendronate (ALN) increased the formation of fibroblastic colonies in cultures of rat bone marrow. Furthermore, clodronate was found to stimulate osteoblast differentiation of ST2 and MC3T3-E1 cells, increasing their capacity to mineralize.(15) Although these data provide evidence that bisphosphonates may have an anabolic effect through the stimulation of osteoblastogenesis, the effect of bisphosphonates on adipogenesis remains unknown. This study aimed at investigating the effect of a nitrogen-containing bisphosphonate (ALN) on both osteoblastogenesis and adipogenesis using an in vitro model of human mesenchymal stem cell (MSC) differentiation.
MATERIALS AND METHODS
ALN was provided by Merck Pharmaceuticals (Whitehouse Station, NJ, USA). ALN was dissolved in PBS, and the pH was adjusted to 7.4 with 1 N NaOH and filter-sterilized by using a 0.2-μm filter. 1,25(OH)2D3 was purchased from Biomol International (Plymouth Meeting, PA, USA). Other reagents were from Sigma Chemical unless stated otherwise.
In vitro differentiation of MSCs
Human MSCs (BioWhittaker, Walkersville, MD, USA) were induced to differentiate into either osteoblasts or adipocytes as previously described.(16) Briefly, MSCs were plated at a density of 5 × 105 cells/well in 100-cm2 dishes containing MSC growth media (BioWhittaker) with 10% FCS and incubated at 37°C for 24 h. After the cells reached 60% confluence, medium was replaced with MSC growth media or induced to differentiate into adipocytes using adipogenesis induction media (AIM; prepared with DMEM, 4.5 g/liter glucose, 1 μM dexamethasone, 0.2 mM indomethacin, 1.7 μM insulin, 0.5 mM 3-isobutyl-1-methylxanthine, 10% FCS, 0.05 U/ml penicillin, and 0.05 μg/ml streptomycin) for 3 days, incubated 3 days in adipogenesis maintenance medium (DMEM, 4.5 g/liter glucose, 1.7 μM insulin, 10% FBS, 0.05 U/ml penicillin, and 0.05 μg/ml streptomycin), and switched to induction media again to promote adipogenic phenotype as previously described.(16) For osteogenic differentiation, we used osteoblastogenesis induction media (OIM; prepared with MSC growth medium, 10% FCS, 100 nM dexamethasone, 10 mM β glycerol phosphate, and 25μg/ml ascorbic acid). In all experiments, media were changed every 3 days.
Human MSCs were seeded at a density of 4 × 103 cells/well in 6-well plates as previously described.(16) Once confluence was obtained (average being 48 h in culture), MSC growth medium was replaced with either OIM or AIM supplemented with either 1,25(OH)2D3 (10−8 M) alone, increasing concentrations of ALN (10−9 to 10−7 M), or both. In all experiments, untreated differentiating MSCs were used as controls.
Identification of the effect of ALN ± 1,25(OH)2D3 on osteoblasts differentiation and activity
MSCs were plated in 4-cm2 dishes at a density of 4 × 104 cells/dish. At 60% confluence, media were replaced with OIM containing ALN (10−9–10−7 M), 1,25(OH)2D3 (10−8 M), or both. At week 2, media were aspirated, and cells were stained for both alkaline phosphatase (ALP) using TT-blue+ staining and for mineralization using Alizarin red staining as previously described.(16) Relative expression of ALP (blue staining) compared with background was quantified using Sigmascan image analysis software (Systat Software, San Jose, CA, USA). Calcium deposition was also quantified using Alizarin red staining. Briefly, after Alizarin red staining, matrix mineralization was quantified by extracting the Alizarin red staining with 100 mM cetylpyridinium chloride at room temperature for 3 h. The absorbance of the extracted Alizarin red S stain was measured at 570 nm. Six wells were analyzed per experiment. Experiments were performed in triplicate.
Identification of the effect of ALN ± 1,25(OH)2D3 on adipocytes differentiation and activity
MSCs were plated in 4-cm2 dishes at a density of 4 × 104 cells/dish. At 60% confluence, media were replaced with AIM containing either ALN (10−9–10−7 M), 1,25(OH)2D3 (10−8 M), or both. At week 2, media were aspirated, and cells were stained for oil red O and counterstained with hematoxylin. The percentage of adipocytes per field was counted using light microscopy. Differentiated adipocytes were considered those polygonal in shape, with eccentrically located nuclei, considerable cytoplasm, and lipid droplets scattered throughout. Ten fields were counted per well. This experiment was repeated three times.
Western blot analysis
MSCs were treated and induced to differentiate into either osteoblasts or adipocytes as previously described. At week 2, cells were lysed in 20 mM tris-HCl, pH 7.5, 200 mM DTT, 200 mM KCl, 0.5 ml glycerol, and protease inhibitor tablets (Roche Diagnostics Canada, Laval, Canada), freeze-thawed three times in a dry ice-ethanol bath, and centrifuged at 11500 rpm for 15 min to remove insoluble material. Cell lysates were dissolved in SDS electrophoresis buffer (Bio-Rad, Hercules, CA, USA), and proteins were separated on SDS-polyacrylamide gels and subsequently electrotransferred to polyvinylidene difluoride membranes. After membrane blocking with PBS containing 0.1% Tween 20 and 10% nonfat dry milk, membranes were incubated overnight at 4°C using an antibody directed against cbfa-1 (Oncogene, Cambridge, MA, USA) and osteocalcin (Santa Cruz Biotechnology, Santa Cruz, CA, USA) for osteoblastogenesis or CCAAT-element binding protein α (CEBPα), sterol regulatory element binding protein 1 (SREBP-1), and peroxisome proliferator activator γ 2 (PPARγ2) (Santa Cruz Biotechnology) for adipogenesis. The bound antibodies were detected with the corresponding secondary antibodies conjugated with horseradish peroxidase. Blots were developed by enhanced chemiluminescence using Perkin-Elmer reagents (Perkin-Elmer, Boston, MA, USA).
PPARγ activity measurement
DNA-binding PPARγ activity was determined using the ELISA-based PPARγ activation TransAM kit (Active Motif, Rixensart, Belgium) as previously described.(17) The TransAM PPARγ Kit contains a 96-well plate on which an oligonucleotide containing a peroxisome proliferator response element (PPRE; 5′-AACTAGGTCAAAGGTCA-3′) has been immobilized. PPARγ contained in nuclear extract specifically binds to this oligonucleotide. The primary antibody used in the TransAM PPARγ Kit recognizes an accessible epitope on PPARγ protein on DNA binding. Addition of a secondary horseradish peroxidase (HRP)-conjugated antibody provides a sensitive colorimetric readout easily quantified by spectrophotometry (450 nm). To quantify PPARγ activation, 8 μg of nuclear extract was measured using the TransAM PPAR Kit according to the manufacturer's instructions (Active Motif, Carlsbad, CA, USA).
Measurement of cell proliferation
MSCs were seeded at a density of 4 × 102 cells/well in 96-well cluster plates (Falcon, Becton-Dickinson, NJ, USA) and committed to differentiate into either osteoblasts or adipocytes as previously described. Cells were treated with either 1,25(OH)2D3 (10−8 M), ALN (10−9–10−7 M), or both. MTS Formazan assesses mitochondrial function by the ability of viable cells to convert soluble 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTS) into an insoluble dark blue Formazan reaction product measured photometrically as previously described.(18) A stock solution of MTS was dissolved in PBS at a concentration of 5 mg/ml and was added in a 1:10 ratio (MTS/DMEM) to each well incubated at 37°C for 4 h, and the optical density was determined at a wavelength of 570–630 nm on a microplate reader model 3550 (BioRad). In preliminary experiments, the absorbance was found to be directly proportional to the number of cells over a wide range (2 × 102–50 × 103 cells/well). The percent survival was defined as [(experimentalabsorbance − blankabsorbance)/controlabsorbance − blank absorbance)] × 100, where the controlabsorbance is the optical density obtained for 10 × 103 cells/well (number of cells plated at the start of the experiment), and blankabsorbance is the optical density determined in wells containing medium and MTS alone.
All results are expressed as mean ± SE of three replicate determinations, and statistical comparisons are based on one-way ANOVA or Student's t-test. A probability value of p < 0.05 was considered significant.
Effect of ALN on osteoblastogenesis
After 2 wk of treatment with either ALN or ALN + 1,25(OH)2D3 to MSC committed to differentiate into osteoblasts, there was a significantly higher expression of ALP in both ALN- and ALN + 1,25(OH)2D3–treated cells compared with untreated cells (p < 0.01; Figs. 1A–1D). This effect correlates with a higher capacity of mineralization by ALN-treated cells as per Alizarin red staining (Figs. 1E–1G). In contrast, the presence of 1,25(OH)2D3 in the media did not potentate the effect of ALN on mineralization although mineralization remains higher than untreated cells (Figs. 1E–1G).
The dose-dependent effect of ALN on mineralization was assessed looking at osteogenic differentiating cells after 2 wk of treatment. Cells were treated at increasing concentrations of ALN (10−9–10−7 M), stained with Alizarin red and observed under light microscopy (×60). Alizarin red staining was quantified using a colorimetric assay. As shown in Fig. 2, there is a dose-dependent effect of ALN on cell mineralization shown by a higher amount of mineralization at doses of 10−7 (Figs. 2B and 2E) and 10−8 M (Figs. 2C and 2E) compared with a lower concentration of 10−9 M (Figs. 2D and 2E) and untreated cells (Figs. 2A and 2E).
Furthermore, the changes in expression of two major transcription factors for osteoblastogenesis, osteocalcin, and cbfa1 were assessed. Western blot analysis showed that in both ALN (10−8 M)– and ALN + 1,25(OH)2D3 (10−8 M) –treated cells, there was an increase in osteocalcin expression at week 2 of differentiation (Figs. 2F and 2G). In contrast, no changes in osteocalcin expression were seen at week 1 in both treated and untreated cells. Finally, cbfa1 expression was increased at week 1 of treatment with either ALN (10−8 M) or ALN + 1,25(OH)2D3 (Figs. 2F and 2G), with no significant changes seen at week 2.
Effect of ALN on adipogenesis
To determine the effect of ALN either alone or in combination with 1,25(OD)2D3 on adipogenesis, MSCs were induced to differentiate into adipocytes, treated for 2 wk with either ALN (10−8 M) or ALN + 1,25(OH)2D3, and stained with oil red O and counterstained with hematoxylin. There were major changes in the phenotype of differentiating adipocytes in the ALN-treated cells. Treatment-induced changes included bigger nuclei and cytoplasm and marked reduction in the amount and distribution of the lipid droplets. Oil red staining shows that both ALN (Figs. 3B and 3E)– and ALN + 1,25(OH)2D3 (Figs. 3C and 3F)–treated cells showed phenotype changes and a lower amount of fat droplets compared with untreated cells (Figs. 3A and 3D). In addition, this effect was dose dependent, being more remarkable at higher (Figs. 4B and 4C) than lower concentrations (Fig. 4D) compared with nontreated cells (Fig. 4A). Furthermore, the number of MSCs differentiating into adipocytes was significantly reduced by ALN. This effect was dose dependent (Fig. 4E).
Finally, the expression of CEBPα and SREBP-1, two essential transcription factors for adipogenesis,(10,16,19) was determined by Western blot. As shown in Fig. 5A, at week 2 of differentiation, expression of both factors was reduced after treatment with either ALN or ALN + 1,25(OH)2D3. Contrary to SREBP-1, CEBPα expression was lower when cells were treated with ALN with and without 1,25(OH)2D3.
Effect of ALN ± 1,25(OH)2D3 on PPARγ expression and activity
PPRAγ2 is the most essential and better studied transcription factor in adipogenesis.(16) To better characterize the effect of ALN ± 1,25(OH)2D3 on PPARγ2 activity, we quantified nuclear PPARγ2 and PPARγ2 DNA-binding activity. After 2 wk of treatment, ALN decreased both nuclear PPARγ2 protein content (Figs. 5A and 5B) and PPARγ DNA binding activity (Fig. 5C). Finally, combination of ALN and 1,25(OH)2D3 (10−8 M) did not potentiate the reduction in nuclear PPARγ amounts seen after treatment with ALN alone (Figs. 5A and 5B).
Effect of ALN on MSC proliferation
No changes in cell proliferation were found after treatment with either ALN or ALN + 1,25(OH)2D3 (data no shown).
The treatment of osteoporosis has been divided according to their mechanisms of action into antiresorptive and anabolic.(5) Antiresorptive agents include estrogens, selective estrogen-receptor modulators (SERMs), calcitonin, and bisphosphonates.(20) The mechanism of action of these compounds is the inhibition of osteoclastic activity and therefore of bone resorption.(5) In contrast, a second type of agents known as anabolics exert their effect either increasing osteoblast activity, promoting osteoblast survival, or both.(21) Current agents considered as anabolics are PTH,(22) vitamin D,(23) and probably strontium ranelate.(24)
For the specific case of osteoporosis in older adults (both men and women), it is known that, additionally to increased osteoclastic activity, the most important pathophysiological mechanism is a reduction in osteoblast number and activity.(6–8) With aging, osteoblasts are replaced by adipocytes within the bone marrow with significant consequences in osteoblast number, function, and survival.(6,25)
Furthermore, although the effect of anabolic agents on osteoblasts is well known, their effect on preventing bone marrow adipogenesis remains unclear. Recently, Rickard et al.(26) have shown that intermittent doses of PTH inhibit adipocyte differentiation in human bone marrow stromal cells. Similar effect has been reported in 1,25(OH)2D3–treated bone marrow cells both in vitro(27) and in vivo.(28)
In contrast, the effect of antiresorptive agents on the differentiation of bone marrow cells remains to be tested. Estrogens have been found to inhibit adipocyte differentiation in mouse bone marrow stromal cells while enhancing osteoblast differentiation,(29) with an additional inhibitory effect on osteoblast apoptosis.(30) The effect of calcitonin and SERMs on either adipogenesis or osteoblastogenesis has not been tested.
In the case of bisphosphonates, their potential anabolic effect has been partially studied. In general, bisphosphonates have been considered antiresorptive agents with limited effect on the osteoblasts.(31) However, there is evidence suggesting that bisphosphonates may have an anabolic effect that explains the increase in cortical wall thickness seen after treatment with bisphosphonates,(32,33) which may be explained by an increase in localized osteoblastic activity.(14) The anabolic effect of bisphosphonates has been questioned by the fact that they are not internalized by osteoblasts, and therefore, they could not exert a direct effect on osteoblast activity.(31) However, there is evidence to support that there is an effect of bisphosphonates on the osteoblast side not directly on mature osteoblasts but more likely on osteoblast precursors in the bone marrow. First, only nanomolar concentrations of bisphosphonates are needed to stimulate the production of osteoclast inhibitory factor, which is mostly secreted by osteoblast precursors.(34) Second, bisphosphonates have shown to stimulate fibroblastic colony formation by murine and human bone marrow both in vitro and ex vivo cultures.(35) In fact, Im et al.(36) have shown that both risedronate and ALN could increase osteoblast and osteoblast progenitor numbers in primary human trabecular cultures with enhanced expression of bone morphogenic protein-2, type I collagen, and osteocalcin. More recent evidence has shown that ALN induces protein farnesylation in MSCs, facilitating their differentiation into osteoblasts (G Duque, R Akter, D Rivas, unpublished data, 2007). Finally, this effect is not limited to osteoblastogenesis but also through the promotion of osteoblast survival as per their efficacy as anti-apoptotic agents in glucocorticoid-induced osteoporosis.(37)
However, although there is increasing evidence that bisphosphonates have an effect beyond osteoclasts that includes osteoblast differentiation and survival, the question remains whether it also has an effect on bone marrow adipogenesis, therefore facilitating the differentiation of stem cells into the osteoblast lineage.
To answer this question, we assessed the effect of ALN on both osteoblastogenesis and adipogenesis. The doses were selected according to previous reports on the effect of ALN in several models of osteoblastogenesis.(14,35) We found that, in human MSCs, ALN not only induced higher levels of osteoblast differentiation, as per their levels of ALP expression, but also higher osteoblastic activity, as shown by the amount of mineralization seen at week 2, which was not potentiated by the presence of 1,25(OH)2D3. This level of mineralization at week 2 is unexpected because full mineralization is only seen around week 3 of MSC differentiation.(16) This finding correlates with recent evidence on the effect of ALN on mineralization both in vitro(38) and in vivo.(39)
Subsequently, we tested the dose-dependent effect of ALN on mineralization. We found that, in agreement with previous reports,(40) ALN induced mineralization at dosages ranging from 10−7 to 10−9 M, although showing its better effect at a dose of 10−8 M. Finally, the effect of ALN on osteogenic factors was tested using Western blot analysis of osteocalcin and cbfa1 expression. In agreement with the timely order of our model, the expression of cbfa1, a determinant factor for differentiation, was higher at week 1 in ALN-treated cells and similar at week 2 in ALN + 1,25(OH)2D3–treated cells. Furthermore, the expression of osteocalcin was higher at week 2 in both ALN and ALN + 1,25(OH)2D3–treated cells, which may explain the early mineralization.
Subsequently, we were interested in the effect of ALN on bone marrow adipogenesis. We found that ALN inhibited adipogenesis in a dose-dependent manner. Additionally, ALN-treated cells showed major changes in their phenotype and a severe limitation in their capacity of producing fat droplets. Similar to the effect seen in the osteoblastogenesis model, combination of ALN and 1,25(OH)2D3 did not potentiate the effect of ALN alone. Additionally, the levels of expression of the major transcription factors required for adipogenesis were reduced after treatment with either ALN or ALN + 1,25(OH)2D3. Finally, from a mechanistic approach, we tested the effect of ALN on the PPARγ2 nuclear complex activity. We found that both ALN and ALN + 1,25(OH)2D3 reduced the activity of the PPARγ2 complex when added to the media. This evidence suggests that the inhibitory effect of ALN on adipogenesis is not only explained by its effect on protein transcription for PPARγ2 but also through limiting its capacity to interact with the nuclear complex. Interestingly, this effect is not potentiated by addition of 1,25(OH)2D3 to the media.
Taken together, our data suggest that, in addition to its antiresorptive effect, ALN has also an anabolic effect not only through the induction of osteoblastogenesis but also through the inhibition of adipogenesis. Furthermore, the addition of 1,25(OH)2D3 did not potentiate these effects. In conclusion, this evidence supports the notion that bisphosphonates have a local effect as bone-forming agents that could explain the gain in bone mass seen in patients treated with these compounds. Additionally, intermittent treatments with anabolics (i.e., PTH) and ALN, which have shown to be effective in preventing fractures, may be related to a persistent anabolic effect of ALN that has not been completely identified; therefore, further studies should be pursued to test this effect in vivo.
This work was supported by an operating grant from the Canadian Institutes of Health Research and a medical school grant from Merck Co. USA. Dr Duque holds a career award from the Fonds de la Recherche en Santé du Quebec.