Claudia Fischbach1, Jochen Seufert2, Harald Staiger3, Michael Hacker1, Markus Neubauer1, Achim Göpferich1, Torsten Blunk1
1Department of Pharmaceutical Technology,
University of Regensburg, 93040 Regensburg, Germany
2Division of Endocrinology, Metabolism and Molecular Medicine, Medical Poliklinik, University of Würzburg, 97070 Würzburg, Germany
3Department of Endocrinology, Metabolism, and Pathobiochemistry,
Medical Clinic Tübingen, Eberhard-Karls-University, 72076 Tübingen, Germany
Tissue Eng. 2003, in press
Recent in vivo and in vitro studies have demonstrated both promises and current limitations in tissue engineering of fat. Herein, we report the establishment of a well-defined 3-D in vitro model useful for systematic investigations of 3-D adipogenesis. Polyglycolic acid fiber meshes were dynamically seeded with 3T3-L1 preadipocytes; subsequently, cell-polymer constructs were hormonally induced and cultivation under three different conditions was evaluated. Regarding tissue coherence and intracellular lipid content, culture of cell-polymer constructs either dynamically in well plates or in stirred bioreactors yielded similar results which were distinctly improved compared to static conditions in well plates. On the protein and mRNA level, significantly increased expression of genes characteristic for a mature adipose phenotype was demonstrated for constructs dynamically cultured in well plates, as compared to static conditions. Furthermore, investigation of lipolysis under stimulating and inhibiting conditions demonstrated functionality of the dynamically differentiated constructs.
Using dynamic culture conditions, the presented in vitro model system is suggested as a valuable tool serving both fat tissue engineering and basic research by likely facilitating investigations into tissue-inherent features not possible to be appropriately conducted under conventional 2-D culture conditions.
Introduction
The demand for soft tissue equivalents in reconstructive and plastic surgery is continuously increasing [1], even though currently applied surgical techniques fail to produce fully satisfactory results. Clinical approaches to soft tissue reconstruction used in the augmentation of soft-tissue volume for the treatment of congenital deformities, posttraumatic repair, and breast cancer rehabilitation include the autografting of fat pads or injection of adipocyte cell suspensions obtained by liposuction [2-5]. There are, however, major problems associated with these techniques, such as low fat-graft survival and necrosis due to insufficient vascularization followed by progressive resorption of the graft over time [2,3,5-8]. Thus, growing viable and functional adipose tissue constructs by the means of tissue engineering would represent a promising strategy to develop alternative therapeutic approaches aimed at improved predictability, reproducibility, and long-term efficacy, as compared to current autologous fat transplantation procedures.
Over the last couple of years, several groups working in this field have published encouraging data demonstrating de novo tissue generation in vivo. For example, subcutaneous injection of basement membrane supplemented with bFGF as well as local delivery of insulin and IGF-1 by polymeric microspheres resulted in formation of visible fat pads in vivo [9-14].
Furthermore, adipose tissue development is reported for implanted preadipocyte-seeded scaffolds made from various biomaterials [15-18]. Taken together, tissue engineering strategies are useful for fat pad formation by recruitment of endogenous precursor cells as well as by implantation of preadipocytes. However, long-term maintenance of tissue engineered adipose tissue in vivo still remains elusive [16]. Although extensive investigations at the cellular and molecular level could possibly clarify the potential reasons for the failed grafts, the appropriate studies have not yet been performed. Until now, explant characterization has been simply carried out by histological examinations and detection of triglyceride accumulation. Furthermore, RT-PCR has also been recently utilized to analyze the expression of glycerol-3-phosphate dehydrogenase expression [9].
In addition to the above mentioned in vivo studies, the first in vitro investigations were performed by examining three-dimensional (3-D) adhesion behavior of preadipocytes on PLGA scaffolds, freeze-dried collagen scaffolds, and expanded polytetrafluorethylen meshes coated with collagen, albumin, and fibronectin, respectively [15,17,19]. Moreover, adipocytes in three-dimensional cultures were detected by oil red O staining of cells within polymer
scaffolds [15,19] and collagen gels [20]. However, detailed studies of adipogenesis in a tissue engineering context regarding not only lipid accumulation, but also expression of specific differentiation markers, as well as tissue functionality, are still lacking, even though especially this kind of approach may help to explain why previous studies partly failed.
Whereas differentiation control in vivo is complex and contingent on the implantation site as well as the gender and age of the patient [21-23], the development of a standardized in vitro engineered fat construct would facilitate investigations into characteristic tissue properties, e.g. 3-D cell-cell interactions. Thus, basic insights into the events occurring during 3-D adipogenesis could be gained at the molecular and cellular level. Therefore, our intention in this study was not to generate implantable fat, but to establish a 3-D in vitro model exhibiting typical features of white adipose tissue, such as triglyceride biosynthesis, expression of typical fat cell genes, and adipocyte functionality, which is useful for systematic investigations on adipose tissue formation. Such a 3-D model could serve in future studies as a valuable tool to investigate the effects of determinants such as different adipogenic factors, co-cultures, or extracellular matrix components under easily controllable and well-defined conditions. In turn, a thorough understanding of the development of engineered adipose tissue on a molecular and cellular level is likely to prove beneficial for future in vivo approaches to the engineering of fat.
In the current study, in order to establish such a well-defined 3-D adipogenesis model, we generated adipose tissue constructs using the murine preadipose cell line 3T3-L1 [24,25], one of the most frequently utilized and best characterized cell lines for the investigation of adipocyte differentiation in vitro, and commercially available 3-D polymer scaffolds made from polyglycolic acid. To trigger adipose differentiation, the cell-polymer constructs were induced with a hormonal cocktail according to protocols described in the literature. The main goal was to establish suitable conditions for the cultivation of the model. Specifically, static and cultivation in well plates and stirred bioreactors (spinner flasks) were investigated. To assess the properties of the differently cultivated cell-polymer constructs, typical features of adipogenic differentiation were examined. In detail, intracellular lipid accumulation was investigated. Furthermore, expression of typical fat cell genes was analyzed on both the protein and the mRNA level. Finally, lipolysis rate of the 3-D cultivated adipocytes was measured under different conditions to ascertain functionality of the generated tissues.