In mammalians, oxygen plays an important role in energy homeostasis in which oxidative phosphorylation is the most important reaction for energy production. The physiologic oxygen partial pressure in the human body is much lower than the atmosphere, ranging from 1% in cartilage and 1–7% in bone marrow to 10–13% in arteries, lungs, and liver.1, 2 However, in the ex vivo culture system, the oxygen concentration is almost 21% which is much higher than the physiologic oxygen partial pressure in the body. It has been reported that low oxygen pressure in tissue cultures demonstrated positive effects on cell growth3 and, connective tissue cultured in high oxygen conditions promoted bone formation while low oxygen conditions favored cartilage differentiation.4 It has also been reported that reduced oxygen tension enhances proliferation of some cell types. For example, cell proliferation increased for bovine pericytes5 and rat calvarial osteoblasts6 cultured at low oxygen tension. A recent study has reported that hypoxia (1–5% O2) also enhanced the self-renewal of murine neural stem cells7 and embryonic stem (ES) cells8 while some experiments showed that hypoxia induced differentiation.9, 10
Mesenchymal stem cells (MSCs) can be isolated from a variety of somatic tissues.11 These cells have self-renewal capacity and can differentiate into mesodermal lineages. However, the exact influences of hypoxia on differentiation of MSCs still remain inconclusive as there are contradictory reports in the literature.2, 7–10 The purpose of this study was to investigate the effects of hypoxia on human MSCs. We hypothesized that oxygen participates in the intricate balance between cellular proliferation and commitment towards differentiation. In this study, the effects of hypoxia on the proliferation ability, migration ability, cell cycle regulation, stemness gene expression, differentiation potentials, as well as cytokine production of MSCs were examined.
MSC Isolation, Culture Expansion, In Vitro Differentiation, and Hypoxia Treatment
Human bone marrow-derived MSCs were harvested as previously reported (n = 15).11, 12 An approval from the Institutional Review Board as well as the informed consents of the bone marrow donors were obtained prior to the commencement of the study. Single cell-derived, clonally expanded MSCs were tested for their surface immune-phenotype and tri-lineage differentiation potentials as previously described before these cells were used for subsequent experiments.11, 12 Expansion medium consisted of Iscove's modified Dulbecco's medium (IMDM; Gibco, Grand Island, NY) and 10% fetal bovine serum (FBS; Hyclone, Logan, UT) supplemented with 10 ng/ml FGF2, 100 U penicillin, 1,000 U streptomycin, and 2 mM L-glutamine (Gibco). Cells of passage 11–15 were used for the following experiments. Hypoxia (1% O2) was simulated in a water-saturated gas mixture of 1% O2, 5% CO2, and 94% N2 at 37°C. In additional experiments, cells were exposed to hypoxia as described above for different time courses (0, 1, 3, 5, 7 days) and the media were changed twice per week.
Cell Proliferation Assay
Direct Cell Counting
In order to examine the difference of proliferative ability of MSCs between normoxia and hypoxia, cells from passage 9–12 were seeded at 1.5 × 105 or 1.5 × 106 cells/dish density in 100 mm culture dishes with growth medium. Cells were exposed to hypoxia or under normoxia, then harvested and counted at different time courses (1, 3, 5, 7 days).
Bromodeoxyuridine (BrdU) Incorporation Assay
Cell proliferation was estimated by flow cytometry based on BrdU incorporation during the S-phase of cycling cells. MSCs were cultured under normoxia or hypoxia for 7 days. Cultures of MSCs were treated with fluorescein-labeled antibody against BrdU (Cat. No. 559619; BD Biosciences, San Diego, CA) in a biologically effective concentration (10 mM) at day 6. After 24 h, cells were harvested, fixed, and DNA was denaturized. The immune complexes were detected and read by FACScan flow cytometer (BD Biosciences). Data were analyzed by CELL Quest software for determination of the number of BrdU+ cells.
Western Blot Analysis, RNA Isolation, and Quantitative Real-Time PCR
These procedures were performed as previously reported12, 13 and detailed experimental methods were described in the supplementary information.
Cell Migration Assay and Cytokine Array Analysis
Migration assays were performed using the Culture-Insert method (ibidi GmbH, Martinsried, Germany). Human cytokine and growth factor assay was performed using Human Cytokine Antibody Array 1 (Cat #AAH-GF-1, RayBiotech, Norcross, GA) that following the manufacturer's instructions. The detail experimental methods were described in supplementary information.
Statistical analysis was performed using the Statistical Package for Social Science-10 software (SPSS, Inc., Chicago, IL). The data were presented as mean ± SD of the results from three or four independent experiments. Results of cell number as well as stemness gene expressions were analyzed by ANOVA tests with Tukey's Post-Hoc tests. Results of osteogenic, chondrogenic and adipogenic marker gene expressions were analyzed by two-tail, non-paired t tests. A p-value of less than 0.05 was considered statistically significant.
Hypoxia Increased the Proliferation Ability and Cyclin Protein Expression in MSCs
Compared with normoxic controls, there was no significant change in morphology after hypoxic culture (Fig. 1a). The cell number was higher after hypoxic culture for 7 days by direct cell counting (Fig. 1b). From days 0–7, the slop of growth curve in hypoxic group was steeper. That demonstrated the proliferation ability of MSCs in hypoxic group was increased (Fig. 1c). Cell proliferation assay was determined by BrdU incorporation analysis. Under hypoxia, the percentage of BrdU-positive MSCs was higher than normoxia (Fig. 1d). To investigate whether hypoxia affects the expression of cell cycle regulating proteins, MSCs were incubated in hypoxia for different time points (0, 1, 3, 5, 7 days). Results show hypoxia increased the expression of the cell proliferation marker proliferating cell nuclear antigen (PCNA). In addition, cyclin D1 expression was gradually increased with time, and the expression of cyclin-dependent kinases (Cdk) inhibitor p27 was decreased gradually with time (Fig. 1e).
Hypoxia Enhanced the Migration Ability of MSCs
Hypoxia enhanced the migration ability of MSCs as shown in the photomicrographs taken 24 and 36 h after insert removal in which more cells migrated into the scratched area in the hypoxic group (Fig. 2a). Twenty-four hours post removal of the culture insert, the cell-covered area was almost completely full (90% ± 2.89) in the hypoxic group more than normoxic group (56% ± 4.3) (Fig. 2b). Moreover, hypoxia treatment up-regulated the protein level of vimentin as well as fibronectin and N-cadherin as shown by Western blot (Fig. 2c).
Hypoxia Up-Regulated Stemness Gene Expression in MSCs
Real-time RT-PCR analysis showed that hypoxia enhances the important stemness genes such as Oct4, Nanog, Sall4, and Klf4 expression (Fig. 3a–d, respectively). It was also found that hypoxia increases expression of RNA component (TERC) of telomerase in MSCs (Fig. 3e).
Hypoxia Enhanced Osteogenic, Inhibited Adipogenic, and Chondrogenic Potentials of MSCs
During 4 weeks of osteogenic induction, MSCs cultured in hypoxia expressed higher Alk-p (Fig. 4a, upper panel) and mineralization as shown by Von Kossa stain (Fig. 4a, lower panel). MSCs under hypoxia expressed higher levels of osteocalcin and osteopontin than normoxia (Fig. 4b). After 4 weeks of chondrogenic differentiation, the size of pellets cultured under normoxia was larger than those under hypoxia (Fig. 4c, upper panel) and the histological section showed mature chondrocyte morphology in normoxic group (Fig. 4c, lower panel). Under hypoxia, MSCs expressed reduced level of type II collagen (COL2A1), cartilage oligometric protein (COMP) and aggrecan (Fig. 4d) compared to normoxia. Light microscopy revealed increased numbers of lipid vacuoles in cells during the 4-week period of adipogenic induction under normoxic condition. In addition, the number of Oil Red O positive lipid vacuoles was decreased under hypoxia (Fig. 4e). MSCs under hypoxia possessed lower expression levels of adipsin, fatty acid synthase (FASN) and fatty acid binding protein 4 (FABP4) than normoxic controls (Fig. 4f). From the above mentioned findings, MSCs cultured in hypoxia had enhanced osteogenic, but inhibited chondrogenic and adipogenic differentiation.
Hypoxia Induced the Differential Production of Growth Factors in MSCs
The above-mentioned experimental results indicate that hypoxia promotes growth and increases the expression of several cell cycle regulating proteins in MSCs (Fig. 1). In order to explore the mechanisms, cytokine array analysis was performed. MSCs were cultured at normoxia or hypoxia for 7 days. The cultured media were harvested and incubated with cytokine array membranes. Results indicate that hypoxic condition regulates the production of a variety of growth factors on MSCs (Fig. 5).
In this study, we demonstrated that hypoxia increased proliferation potentials and promoted MSCs entry into the cell cycle (Fig. 1). The effects of low oxygen environment on cell growth had been examined by many investigators in different cell models.8, 14, 15 However, the detailed mechanisms of how hypoxia induced proliferation are still unclear. Serum and mitogenic signals are known to activate the cyclin-dependent kinases that are the key regulators of cell cycle progression.16 In view of this proliferation effect, hypoxia regulated the expression of a number of growth factors elucidated from our cytokine array data (Fig. 5). In our model system, it was possible that hypoxia switched on the program of cell cycle through autocrine or paracrine effects. Various mitogens, for example, insulin-like growth factor (IGFs) and VEGF are important for cell proliferation, survival, and maintenance of normal biological motility. It has been reported that VEGF and its receptor play an autocrine or paracrine mitogenic role in MSCs.17 Together, we infer that hypoxia regulated growth factor production, that resulted in mitogenic effects, and thereby induced MSC proliferation.
We also observed that oxygen concentration influenced the physiological motility on MSCs in which hypoxia increased the migration ability and induced mesenchymal gene expression such as vimentin, fibronectin, and N-cadherin (Fig. 2). During embryogenesis, oxygen was a key regulator of ontogeny and regulates the embryonic stem or progenitor cells. In the 1970s, Morriss demonstrated that successful development of the neural fold in vitro was dependent on the creation of culture conditions with low oxygen condition.18 Hypoxia can promote migration of vascular smooth muscle cells through an autocrine pathway.19 Based on our experimental data, hypoxia induced many growth factor and cytokine secretion (Fig. 5) and promoted migration ability of MSCs. These growth factors or cytokines may play crucial roles to regulate cell motility under low oxygen condition.
We also found that hypoxia affects the differentiation potentials of MSCs (Fig. 4). In our experiments, hypoxia promoted differentiation of osteogenesis, but inhibited chondrogenesis and adipogenesis. Rat-derived MSCs cultivated in hypoxic conditions produced more bone than those cultured in normoxic condition when harvested and loaded into porous ceramic cubes and implanted into syngeneic host animals.9 However, contradictory findings were reported by D'Ippolito et al.2 showing that low oxygen tension inhibited osteogenic differentiation of human bone marrow stromal cells. Taken together, the effects of hypoxia on MSCs are cell source- and species-dependent. Furthermore, different procurement and isolation protocols may account for differences in experimental results. In this study, we used single cell-derived, clonally expanded MSCs harvested from human bone marrow aspirates, which have been devoid of committed progenitors. Hypoxia enhanced osteogenesis and reciprocally inhibited adipogenesis in this clearly defined MSC population. In conclusion, we have shown here that hypoxia can influence the balance between proliferation and differentiation of MSCs. Under hypoxia, MSCs proliferate and retain stem cell properties, and may be used to yield great cell numbers for osteoblastic progenies. Hypoxic culture may therefore serve as a useful strategy in the tissue engineering of bones.
This work was supported in part by the UST-UCSD International Center of Excellence in Advanced Bio-engineering sponsored by the Taiwan National Science Council I-RiCE Program under Grant Number: NSC-99-2911-I-009-101. The authors also acknowledge financial support from the Taipei Veterans General Hospital (VGH100E1-010, VGH100C-056, VN100-05, and VGH100D-003-2), Wan Fang Hospital (100swf03) and the National Science Council, Taiwan (NSC99-2120-M-010-001, NSC99-2627-B-010-003, NSC99-3111-B-010-002, NSC98-2314-B-010-001-MY3, NSC 99-2911-I-010-501, and NSC 99-3114-B-002-005). This study was also supported by a grant from the Ministry of Education, Aim for the Top University Plan.