Recent studies have demonstrated that adipose-derived mesenchymal cells (AMCs) offer great promise for cell-based therapies because of their ability to differentiate toward bone, cartilage, and fat. Given that cartilage is an avascular tissue and that mesenchymal cells experience hypoxia during prechondrogenic condensation in endochondral ossification, the goal of this study was to understand the influence of oxygen tension on AMC differentiation into bone and cartilage. In vitro chondrogenesis was induced using a three-dimensional micromass culture model supplemented with TGF-β1. Collagen II production and extracellular matrix proteoglycans were assessed with immunohistochemistry and Alcian blue staining, respectively. Strikingly, micromasses differentiated in reduced oxygen tension (2% O2) showed markedly decreased chondrogenesis. Osteogenesis was induced using osteogenic medium supplemented with retinoic acid or vitamin D and was assessed with alkaline phosphatase activity and mineralization. AMCs differentiated in both 21 and 2% O2 environments. However, osteogenesis was severely diminished in a low-oxygen environment. These data demonstrated that hypoxia strongly inhibits in vitro chondrogenesis and osteogenesis in AMCs.
chondrocytes are continuously challenged by a hypoxic cellular microenvironment as low as 1% oxygen during the process of endochondral ossification (2, 3, 17). Permanent cartilage, which has no blood supply, derives nutrients from the surrounding synovial fluid (15, 22, 24). A recent study has shown that chondrocytes are uniquely adapted to the low-oxygen environment and tend to dedifferentiate in a normoxic environment and redifferentiate in a hypoxic one in vitro (8).
Because of the low capacity of cartilage for repair and the increasing prevalence of cartilage diseases (arthritis), tissue engineering of cartilage has become an important area of research (6, 16). Currently, autologous chondrocyte transplantation is the only cell-based therapy for cartilage pathology. This treatment is relatively ineffective to date, however, because of high donor site morbidity and poor chondrocyte proliferation in culture (4). Adipose-derived mesenchymal cells (AMCs) have been developed and used as alternative sources for cell-based therapies (10). These cells have been shown to differentiate into a variety of lineages, including bone, muscle, cartilage, and fat (9–12, 37). For application of these cells to cartilage regeneration, an understanding of the physiological response of these cells to the cellular environment of a native chondrocyte (hypoxia) is warranted. To date, studies have focused on chondrocyte and mesenchymal cell (MSC) responses to hypoxia have been inconclusive (5, 19, 20, 22).
Osteogenesis, a closely related differentiation pathway, has been shown to be affected by ambient oxygen levels (31). Clinically, after severe bony injury, the vascular supply is disrupted and results in a hypoxic gradient within the wound microenvironment. Poor blood supply can affect the activity of mesenchymal osteoblast precursors during osteogenic repair following injury, including inflammatory cell recruitment and matrix processing, resulting in delayed fracture healing (25). Thus, it is important to understand the response of osteoprogenitors to a hypoxic microenvironment. Osteocytes in vivo, although not experiencing the levels of hypoxia that chondrocytes do, are physiologically exposed to a level of ambient oxygen (4–7%) lower than that of normal atmospheric oxygen (21%) (13, 18, 34). A previous study from our laboratory (31) demonstrated that brief (12 h) exposure to anoxia (0.02% oxygen) inhibited osteogenesis in osteoblasts and human MSCs; however, a similar period of exposure to 2% oxygen did not result in significant changes in osteogenesis. Lennon et al. (19) reported that rat bone marrow MSCs, differentiated in 5% oxygen, exhibited enhanced osteogenesis in ceramic cubes.
The goal of this study was to understand the effect of relatively long-term, low oxygen tension on osteochondrogenic differentiation of AMCs. Our results revealed that both osteogenesis and chondrogenesis of AMCs were significantly inhibited by hypoxic culture conditions, demonstrating that the oxygen microenvironment plays a profound role in AMC differentiation.
MATERIALS AND METHODS
Chemicals and media.
Dulbecco's modified Eagle's medium (DMEM) was purchased from Mediatech (catalog no. 10-017-CV; Herndon, VA). Fetal bovine serum (FBS) was purchased from Gemini Bioproducts (Woodland, CA). Other chemicals were obtained from Sigma-Aldrich (St. Louis, MO).
All experiments were performed using protocols approved by Stanford University Animal Care and Use Committee (IACUC) guidelines. The IACUC Eprotocol/Prism number for our studies is 9999/7373. Three-week-old adult male FVB mice (Charles River Laboratories, Wilmington, MA) were euthanized with CO2. The inguinal fat pads were carefully dissected and washed sequentially in a betadine- and phosphate-buffered saline solution (Fisher Scientific, Fair Lawn, NJ). The fat pads were finely minced and digested with 0.075% collagenase A (Sigma-Aldrich) with vigorous shaking in a 37°C water bath for 30 min. The collagenase A was neutralized with an equal volume of growth medium containing DMEM, 10% FBS, and 1% penicillin-streptomycin. The tissue mixture was passed through a cell strainer (100 μm Falcon) to remove the large, undigested tissue fragments. Subsequently, the cell suspension was centrifuged at 1,000 rpm for 5 min, and the supernatant was discarded. The cell pellet was resuspended in fresh growth medium and plated in tissue culture dishes (∼1 dish per cells from 2 animals). Initially, cells were cultured and expanded in growth medium. Passage 2 cells were used for further experiments.
In vitro chondrogenic differentiation.
The micromass technique was modified from the one described by Ahrens et al. (1). Briefly, passage 2 cells were harvested and resuspended in growth medium at 1 × 107 cells/ml concentration. Droplets (10 μl) were placed in culture dishes and allowed to adhere to each other and the substratum at 37°C for 2 h. Chondrogenic medium was then carefully added around the cell aggregates. Chondrogenic medium consisted of DMEM, 1% FBS, 1% penicillin-streptomycin, 37.5 μg/ml ascorbate-2-phophate, ITS premix (BD Biosciences, Bedford, MA), and 10 ng/ml TGF-β1 (Research Diagnostics, Flanders, NJ). The cells were differentiated in either 21% ambient oxygen or 2% ambient oxygen. By 24 h, the aggregates coalesced and became spherical. The chondrogenic medium was changed every 3 days. For 2% oxygen experiments, fresh chondrogenic medium was preequilibrated to 2% oxygen before each medium change to ensure that micromasses were differentiated in consistent oxygen conditions. Micromasses were harvested for histology and sulfated glycosaminoglycan (sGAG) assay at various time points.
Micromasses differentiated in 21 and 2% ambient oxygen were harvested at days 1, 2, 3, 6, 9, 12, and 15. After a brief wash with phosphate-buffered saline, micromasses were fixed in 4% paraformaldehyde with 4% sucrose for 15 min. Micromass diameters were measured using a Zeiss microscope. At days 3, 6, and 12, micromasses were embedded with optimal cutting temperature compound. Cryosections (10 μm) were mounted onto slides and stained with hematoxylin and eosin and Alcian blue. Immunohistochemistry was performed to assess collagen II expression.
Micromass DNA quantification.
Micromasses at different time points were harvested and digested in a 0.1% papain solution at pH 7. After 3 h of incubation at 65°C, cell pellets were centrifuged and the supernatant was used to perform the sGAG assay. The pellets were resuspended and sonicated, and the DNA content of micromasses was measured using a Quant-iT DNA assay kit (Molecular Probes, Eugene, OR) according to the manufacturer's instructions. Assays were performed in duplicate, and the average DNA quantities from micromasses at days 3, 6, and 12 were calculated.
Alcian blue staining.
The sections were rinsed in 0.1 N HCl (pH 1) for 5 min to equilibrate the pH. To assess proteoglycan content, we placed the sections in a 1% solution of Alcian blue 8-GX (Sigma-Aldrich) in 0.1 N HCl for 30 min and then washed them twice in 0.1 N HCl to remove nonspecific staining. Slides were counterstained with nuclear fast red (Biomeda, Foster City, CA). Photographs were taken with a Zeiss microscope.
Sections were blocked at room temperature for 30 min and incubated with primary antibody at 4°C overnight (anti-collagen II; Santa Cruz Biotechnology, Santa Cruz, CA). After a secondary antibody (Vector Labs, Burlingame, CA) incubation, sections were labeled with ABC reagent (Vector Labs) for 10 min at room temperature. Diaminobenzidine (Vector Labs) was applied to each section, and hematoxylin was used for counterstaining. The negative control was performed by staining with the secondary antibody alone.
Twenty-four micromasses per time point were harvested (12 per replicate). Each replicate was separately digested using a 0.1% papain solution at pH 7. The sGAG quantification was performed on each replicate with a sGAG assay kit (Kamiya Biomedical, Seattle, WA) according to the manufacturer's protocol. All sGAG quantities were normalized by DNA content. Significance (P ≤ 0.05) was assessed using Student's t-test.
In vitro osteogenic differentiation.
After culture expansion, the cells were trypsinized and replated in 12-well culture dishes at a density of 10,000 cells/well. After overnight attachment, the cells were treated with basic osteogenic differentiation medium (ODM: DMEM, 10% FBS, 100 μg/ml ascorbic acid, 10 mM β-glycerophosphate, and 1% penicillin-streptomycin) alone or supplemented with 1 μM retinoic acid or 50 nM vitamin D, each of which has been shown to enhance osteogenic differentiation in AMCs (7, 29, 33). The osteogenic differentiation experiments were performed under 2 and 21% oxygen environments. The osteogenic medium was replenished every 3 days. The osteogenic medium used for 2% oxygen tension was preequilibrated before each medium change to ensure the consistent oxygen gradient. Alkaline phosphatase staining and quantification were assessed at 1 wk. Alizarin red staining and quantification were performed at 3 wk.
Alkaline phosphatase staining.
Medium was removed, and the cells were washed twice with phosphate-buffered saline. The cells were fixed with a 60% acetone-40% citrate working solution for 30 s. Subsequently, the cells were washed with deionized water for 45 s and stained for 30 min with a diazonium salt solution composed of fast violet B (0.024%) and 4% naphthol AS-MX phosphate alkaline solution (Sigma). Afterward, the cells were rinsed with deionized water. Cells positive for alkaline phosphatase were stained red.
Quantitative alkaline phosphatase activity measurement.
Alkaline phosphatase activity also was determined using a biochemical colorimetric assay with the Sigma kit (Document 104) as described by the manufacturer. Briefly, the cells were washed with cold phosphate-buffered saline. The cells were scraped into a radioimmunoprecipitation assay buffer (containing 50 mM Tris·HCl, pH 7.5, 150 mM NaCl, 5% glycerol, 1 mM EDTA, 1% Nonidet P-40, 0.1% SDS, and 0.25% sodium deoxycholate) and centrifuged. The enzymatic alkaline phosphatase activity in the supernatant of the cell lysate was assayed by measuring the p-nitrophenol formed from the enzymatic hydrolysis of p-nitrophenylphosphate, used as the substrate, at 405 nm. Experiments were performed in triplicate wells, and results were calculated as means and standard deviations. Student's t-test was used to assess significance (P ≤ 0.05).
Alizarin red S staining.
At late osteogenic differentiation, calcium deposition also was quantified using Alizarin red S staining. Briefly, cultured cells were washed with cold phosphate-buffered saline, fixed for 15 min in 100% ethanol, and stained with 0.2% Alizarin red S solution (Sigma) for 30 min. Stained cells were washed extensively with deionized water to remove the nonspecific precipitation. Positive red staining represents calcium deposits on the differentiated cells. Photographs were obtained and presented for analysis of late-stage osteogenic differentiation. Matrix mineralization was quantified by extracting the Alizarin red S staining with 100 mM cetypyridinium chloride (Sigma) at room temperature for 3 h. The absorbance of the extracted Alizarin red S stain was measured at 570 nm. Experiments were performed in duplicated wells. Student's t-test was performed to assess statistical significance (P ≤ 0.05).
Low oxygen tension inhibits chondrogenesis of AMCs.
Because chondrocytes proliferate and differentiate in a hypoxic environment and because studies have been mixed regarding hypoxia's effect on in vitro chondrogenesis, we were interested in studying the effects of a low-oxygen environment on chondrogenic differentiation of murine AMCs to manipulate this environment for tissue engineering applications. A three-dimensional (3D) micromass technique with medium supplemented with TGF-β1 was utilized to study in vitro chondrogenic differentiation. After 24 h of culture, the cell aggregates became spheroid micromasses in chondrogenic medium. Concurrently with chondrogenic differentiation, the micromasses accumulated chondrogenic matrix components in the center, causing a reduction of micromass size. To evaluate proliferation and differentiation of micromasses, we calculated average micromass diameter and DNA content at each time point under different oxygen tensions. The sGAG content in the extracellular matrix was stained with Alcian blue in micromass sections at days 3, 6, and 12 of differentiation, and sGAG quantification of whole micromasses was evaluated at days 3, 6, and 12. The presence of collagen II, a specific extracellular matrix component for chondrogenesis, was assessed using immunohistochemistry at day 12 of differentiation.
First, we observed that the micromasses differentiated in 2% oxygen were grossly different in size from those differentiated in 21% oxygen. During the first 3 days, the micromasses differentiated in 2% oxygen condensed similarly to the ones differentiated in 21% oxygen. Subsequently, however, the normoxic micromasses showed progressive reduction in size, whereas the hypoxic micromasses showed less drastic changes. At days 1, 2, 3, 6, 9, 12, and 15, we measured the diameters of the micromasses differentiated in 2 and 21% oxygen environments. All micromass diameters were normalized relative to day 1 micromasses in hypoxia. Results showed that micromasses differentiated in 2% oxygen after day 3 were significantly larger (P ≤ 0.05) than those differentiated in 21% oxygen (Fig. 1A). To assess the proliferation of micromasses at different oxygen tensions, we quantitatively measured the DNA content of micromasses at days 3, 6, and 12. Results showed that the average total DNA quantities were higher in the micromasses grown in 2% oxygen compared with those grown in 21% oxygen, and this difference was statistically significant at 6 days (P ≤ 0.05). During the time of micromass differentiation, results showed that the total DNA quantity decreased as a function of time. These results suggested that the apoptosis occurred within the 3D micromasses while chondrogenesis progressed (Fig. 1B).
Upon sectioning, we observed the morphologies of AMCs in the micromasses differentiated in 2% oxygen: the cells were smaller in size, were more abundant in the center, and had less dense matrix formation (data not shown). In contrast, the extracellular matrix of the micromasses differentiated in 21% oxygen showed progressively more intense Alcian blue staining on days 3, 6, and 12 than those differentiated in 2% oxygen. Hip cartilage and native fat tissue isolated from mice were stained as positive and negative controls, respectively (Fig. 2).
Furthermore, cartilage-specific extracellular matrix protein, type II collagen, was examined using immunohistochemistry to confirm chondrogenesis within the 3D micromass. As expected, at day 12 of differentiation, type II collagen was substantially more abundant in 21% oxygen compared with 2% oxygen (Fig. 3). Control sections incubated with secondary antibody alone showed negative staining microscopically (data not shown). Accordingly, the quantity of sGAG in the whole micromass at days 3, 6, and 12 was measured and normalized by DNA content. Results demonstrated a significant decrease in sGAG content in the micromasses differentiated in 2% oxygen compared with those differentiated in 21% oxygen (P ≤ 0.05) at all time points measured. These data correlated with our Alcian blue staining results (Fig. 4).
Low oxygen tension inhibits osteogenesis of AMCs.
In light of our results regarding chondrogenic differentiation, investigating AMC osteogenic capacity in a lower oxygen environment became more relevant. AMCs were differentiated using three differentiation conditions: a basic osteogenic medium (ODM), ODM plus retinoic acid supplementation, and ODM plus vitamin D acid supplementation. Previous studies demonstrated that ODM supplemented with retinoic acid or vitamin D enhanced osteogenic differentiation of AMCs in vitro (7, 33). Differences in early alkaline phosphatase activity staining and quantification were examined, and late extracellular matrix formation and calcium mineralization were stained with Alizarin red S, an end-point measure of in vitro osteogenic differentiation.
After 1 wk of differentiation, the AMCs differentiated in 21% oxygen showed minimal alkaline phosphatase activity staining in ODM and increased staining with vitamin D or retinoic acid supplementation. AMCs differentiated in 2% oxygen exhibited diminished alkaline phosphatase activity staining compared with cells differentiated in 21% oxygen, regardless of the osteogenic medium used (Fig. 5A). Quantification of alkaline phosphatase activity, which demonstrated the absolute enzymatic quantity per cell, was consistent with the staining results (Fig. 5B).
After 3 wk of differentiation, the AMCs differentiated in 21% oxygen with treatment of ODM showed minimal Alizarin red S staining, whereas a moderate amount of mineralization was seen with vitamin D supplement in ODM. AMCs differentiated with retinoic acid supplement in ODM demonstrated robust mineralization. Data show that 2% oxygen stress significantly impaired the ability of AMCs to undergo terminal osteogenic differentiation with or without vitamin D or retinoic acid (Fig. 6A). Quantitative measurements of Alizarin red S absorbance showed significant differences in mineralization between 2 and 21% oxygen environments in response to retinoic acid or vitamin D. The quantitative analysis correlated with the staining results (Fig. 6B).
The importance of tissue environmental factors in the induction of MSC differentiation has been recently emphasized (14, 30). Among all the nongenetic and environmental factors, perhaps the most important is oxygen tension (30). In particular, the influence of hypoxia in chondrogenesis is a topic of great interest, because chondrocytes grow in a low-oxygen environment (28). AMCs, as an alternative source for cartilage or bone repair, have demonstrated great potential for use in cell-based therapy (9). As an initial step to better understand the functional characteristic of AMCs, we studied the responsiveness of the AMC population to low oxygen tension.
Only a few studies have described the effect of hypoxia on MSCs. Robins et al. (28) described how 1% oxygen upregulated Sox9, the master regulator of chondrogenesis, in monolayer human MSC lines. They described increased Alcian blue staining of monolayer cells, but no specific collagen II assays were performed. Wang et al. (36) reported that adult human bone marrow-derived MSCs encapsulated in alginate beads grown in 5% oxygen had enhanced chondrogenesis and decreased proliferation. Scherer et al. (32) described how 5% oxygen in combination with intermittent hydrostatic pressure promoted chondrogenic gene expression in cells, although no downstream protein assays were examined. In addition to the studies on MSCs, investigations using chondrocytes in hypoxia also reported mixed results regarding proliferation and differentiation (19–22). Most recently, Wang et al. (35) reported that human AMCs cultured in a 5% oxygen environment suspended in alginate beads demonstrated a decrease in growth and an increase of chondrogenic differentiation. The contradictory data can be explained by the variety of species involved, the cell lines versus primary cells used, the ambient oxygen level chosen, the type of culture design (alginate, scaffold, micromass, and monolayer, which would change the type of oxygen gradients to which the cells were exposed), the method of differentiation applied, and the time points evaluated. In our laboratory, the biomaterials used in vitro to aid chondrogenesis, such as alginate beads, might affect AMC differentiation; therefore, we propose that our 3D micromass system not involving biomaterials could account for the differences observed.
Chondrogenesis is a well-orchestrated process driven by chondrogenic progenitors that undergo mesenchymal condensation, proliferation, and chondrocyte differentiation (25). AMCs contain large amounts of MSCs, including chondroprogenitors in different stages of differentiation (38). Certain stimuli by specific factors program progenitor cells to undergo specific lineage differentiation (37). Especially in chondrogenesis, the growth plate and mesenchymal condensation are essential for further chondrocyte differentiation (2). Our results indicated that 3D micromasses differentiated in a low-oxygen environment were larger in diameter than those differentiated at a normal oxygen tension. This result is not necessarily surprising, given the fact that native fat tissue is adapted to much lower oxygen tensions in vivo, and an environment of 2% oxygen may be consistent with physiological conditions (27). However, in addition to the ambient oxygen (normoxia vs. hypoxia), our micromass culture developed hypoxic oxygen and nutrient gradients from the periphery toward the center of the micromass. This gradient is reflected in decreased chondrogenic differentiation of Alcian blue in the center (unpublished data). In addition, our data showing decreased micromass DNA content in 2% oxygen suggested possible apoptosis within the long-term 3D differentiated micromasses. Thus, the long-term 2% oxygen stress may have caused the severe oxygen reduction in the center of hypoxic micromasses.
In addition to chondrogenesis, we also assessed the osteogenic capacity of AMCs under reduced oxygen tension. Unlike human AMCs, murine AMCs do not mineralize to a significant degree in osteogenic medium without retinoic acid or vitamin D supplement (23). Therefore, we assessed the osteogenic potential of AMCs treated with these osteogenic inducers under hypoxia and normoxia. In summary, we observed that oxygen tension strongly affected osteoprogenitor differentiation in AMCs at various conditions. Two percent oxygen tension profoundly influenced osteogenic capacity and mineralization, suggesting that the differentiation of osteoprogenitor cells was inhibited in lower ambient oxygen. In a recent study, transcriptional profiling has revealed that after transient oxygen reduction, Runx2, a critical transcription factor determining bone formation, was downregulated in primary calvarial osteoblasts. Subsequently, hypoxic exposure inhibited bone mineralization in osteoblasts (31). This current study is a first step toward understanding osteoprogenitor behavior in a reduced oxygen microenvironment throughout differentiation. In particular, a growing body of literature demonstrates that hypoxia-inducible factor (HIF)-1α is induced after low-oxygen insult in certain types of cells (26). Therefore, direct association between the elevation of HIF-1α protein and chondro- and osteogenesis in multipotent MSCs will be of great interest. Given that osteoprogenitors are exposed to a hypoxic postinjury milieu and that mesenchymal precursors are the key component of the cascade in bone repair, it is important to elucidate the pathways involved in hypoxic processes for MSCs (20). In addition, the influences of hypoxia in specific lineage transitions within primary culture heterogeneous AMCs must be characterized in future experiments. New insights into the osteoblastic response to hypoxia will further benefit therapeutic applications for wound healing and bone regeneration.
Strategies for efficiently directing multipotent MSCs into their specific lineage progenitor cells have not yet been defined. Nongenetic approaches are attractive because microenvironmental factors, such as oxygen tension, are easier to manipulate. For example, the 21% oxygen tension used as the normoxic condition exceeds the partial pressure of oxygen in most mammalian tissues; it would be interesting to study the effect of other oxygen conditions on AMC lineage commitment. In this study, we found that consistent exposure of a heterogeneous population of AMCs to a reduced oxygen environment inhibited both osteo- and chondroprogenitor lineage differentiation. However, the influence of transient/short exposure to hypoxic environment in AMCs has yet to be elucidated. The proliferative effect of AMCs under lower oxygen tension suggests that the relative physiological environment may provide a suitable condition for MSC proliferation. We hypothesize that transient exposure to reduced oxygen tension on the heterogeneous population of AMCs may inhibit osteoprogenitor cells and may actually promote or enrich other lineage progenitors such as prechondrocytes. Further investigations into the possible effects of hypoxia on the lineage commitment of the heterogeneous AMCs are required.
In conclusion, chondrogenesis and osteogenesis are highly influenced by the oxygen microenvironment of AMCs. Specifically, the in vitro chondrogenesis and osteogenesis models utilizing AMCs will allow us to dissect the multiple functional steps of mesenchymal differentiation during chondro- and osteogenesis.
This study was supported by National Institute of Dental and Craniofacial Research Grants DE-14526 and DE-13194 and an Oak Foundation Grant (to M. T. Longaker).
We greatly thank Asavari Gupte for carefully reading the manuscript.
↵* P. Malladi and Y. Xu contributed equally to this work.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2006 the American Physiological Society