Cell Density of Mesenchymal Stem Cells in Chondrogenic and Osteogenic Differentiations

DIFFERENTATIONS *Troken, AJ; +*Mao, JJ (A-NIH) Tissue Engineering Laboratory, University of Illinois at Chicago, Chicago, IL [email protected] INTRODUCTION: Cell density during various stages of the development of synovial joints is generally not well understood. Cell density only accounts for 5% of the total volume of articular cartilage (Stockwell, 1979; Mow et al., 1992; McDevitt et al., 1992). However, it is unclear whether tissueengineered articular cartilage should have a cell density prior to full maturation (Mao, 2005). For this reason, cell density is one of the first pragmatic issues that must be addressed in cell-based tissue engineering approaches. For bone tissue engineering, cell density has potential effects on both mechanical properties and the regeneration outcome (Mow and Hayes, 1997; Spitz and Duboule, 2001). METHODS: Isolation of human mesenchymal stem cells (hMSCs) hMSCs were isolated from donated bone marrow (AllCells, LLC), and cultureexpanded in complete Dulbecco’s Modified Eagle Medium (DMEM), with fresh media change every 3-4 days. Osteogenic differentiation A subpopulation of the hMSCs were differentiated into osteoblasts per our prior methods (Alhadlaq and Mao, 2004; Marion et al., 2005), using a cocktail of 100nM dexamethasone, 50μg/ml L-ascorbic acid-2-phosphate and 10mM beta-glycerophosphate. Chondrogenic differentiation and 3D cell encapsulation in PEGDA Poly(ethylene glycol) based hydrogel (PEG) was used to encapsulate hMSCs, hMSC-derived osteoblasts per our prior method (Alhadlaq et al., 2004; Stosich and Mao, 2005). Upon photopolymerization, the cellhydrogel suspension was fabricated as cylindrical disks (4 x 6 mm: height:dia). Cell densities were 0, 5, 40, and 80 million cells/ml (0 M/ml, 5 M/ml, 40 M/ml, and 80M/ml). A fraction of the hMSC-based constructs was then induced to differentiate in chondrogenic supplemented medium in 3D PEG using 1% 1X ITS+ (Sigma), 1% antibiotic/antimycotic, 100μg/ml sodium pyruvate, 50μg/ml ascorbate, 40μg/ml L-proline, 0.1μM dexamethasone, and 10ng/ml TGF-β3. In vitro 3D incubation – All constructs were further incubated in vitro in corresponding medium (hMSC-based PEG in DMEM; hMSCosteoblast constructs in osteogenic supplemented DMEM; hMSCchondrocytes in chondrogenic supplemented DMEM) for 4 wks. Phenotypic characterization and in vivo implantation The in vitro incubated constructs were processed histologically to observe cell and matrix morphology using H&E staining, von Kossa staining (for mineral deposition), Safranin O/fast green staining (for glycosaminoglycans), and alkaline phosphatase activity. DNA (BioRad) and calcium (Raichem) were quantified. All groups of in vitro incubated constructs were implanted in the dorsum of immunodeficient mice per our prior methods (Alhadlaq et al., 2004; Stosich and Mao, 2005). RESULTS: Histological analysis showed that the control hMSC constructs (without either chondrogenic or osteogenic treatment) stained negatively to both von Kossa (Fig. 1B, C and D) and safranin O (Fig. 2B, C, and D) indicating a lack of mineral deposition or cartilage-related proteoglycans/glycosaminoglycans. HMSC-derived osteoblast constructs (hMSC-Obs) stained positively with von Kossa for mineralization (Fig 1F, G, and H). As cell density increased, mineralization increased qualitatively. However, there was no significant difference in calcium content quantitatively between the 40M/ml and 80M/ml constructs (Fig 3). No alkaline phosphatase enzymatic activity was indicated in the histology for the hMSC control constructs. Minimal alkaline phosphatase activity was suggested with the hMSC-Obs as well. Safranin O stained positively for the chondrogenically induced constructs (Fig 2F, G, and H). Quantification of DNA presented marked increases of DNA content as cell seeding density increased. However, the increases in DNA content were smaller for hMSC-Obs and hMSC-Chs than for the control. CONCLUSIONS: The present data suggest that cell density of encapsulated hMSCs, as well as hMSC-Obs and hMSC-Chs, into PEG is preserved upon in vitro incubation. Thus, cell density may be an important parameter in tissue engineering. The present increases in density have minimal effects on cell viability of encapsulated cells. HMSCs constructs showed no spontaneous osteogenic or chondrogenic differentiation based upon their histology. Von Kossa staining demonstrated retention of the osteoblast phenotype in all hMSC-Ob groups upon harvest. In the highest seeding density of the hMSC-Obs, the von Kossa stain showed smaller quantities of mineralization toward the center of the constructs. Safranin O staining demonstrated chondrogenic activity in hMSC-Ch groups. The larger red area encircling the cells in the 5M/ml group over the other densities implies a stronger proteoglycan synthesis. Also, the highest seeding density shows a distinct line in which the cells toward the center of the construct did not differentiate into the chondrogenic lineage. The calcium content assay presented significant differences between the 5M/ml density and the 40M/ml and 80M/ml densities. Also, within these 40M/ml and 80M/ml groups, significant differences were found between the hMSC-Obs and the hMSCs and hMSC-Chs. The novelty of this ongoing investigation is an optimization of the most effective seeding densities for in vivo tissue engineering to induce mesenchymal stem cells to lead to desired lineages. ACKNOWLEDGEMENTS: This research was supported by NIH grants DE13964, DE15391, and EB02332.