Autografts containing bone marrow (BM) are current gold standard in the treatment of critical size bone defects, delayed union and bone nonunion defects. Although reaching unprecedented healing rates in bone reconstruction, the mode of action and cell-cell interactions of bone marrow mononuclear cell (BM-MNC) populations have not yet been described. BM-MNCs consist of a heterogeneous mixture of hematopoetic and non-hematopoetic lineage fractions. Cell culture in a 3D environment is necessary to reflect on the complex mix of these adherend and non-adherend cells in a physiologically relevant context. Therefore, the main aim of this approach was to establish conditions for a stable 3D BM-MNC culture to assess cellular responses on fracture healing strategies.

BM samples were obtained from residual material after surgery with positive ethical vote and informed consent of the patients. BM-MNCs were isolated by density gradient centrifugation, and cellular composition was determined by flow cytometry to obtain unbiased data sets on contained cell populations. Collagen from rat tail and human fibrin was used to facilitate a 3D culture environment for the BM-MNCs over a period of three days. Effects on cellular composition that could improve the regenerative potential of BM-MNCs within the BM autograft were assessed using flow cytometry. Cell-cell-interactions were visualized using confocal microscopy over a period of 24 hours. Cell localization and interaction partners were characterized using immunofluorescence labeled paraffin sectioning.

Main BM-MNC populations like Monocytes, Macrophages, T cells and endothelial progenitor cells were determined and could be conserved in 3D culture over a period of three days. The 3D cultures will be further treated with already clinically available reagents that lead to effects even within a short-term exposure to stimulate angiogenic, osteogenic or immunomodulatory properties. These measures will help to ease the translation from “bench to bedside” into an intraoperative protocol in the end.

Long-term regeneration of cartilage defects treated with tissue engineering constructs often fails because of insufficient integration with the host tissue. We hypothesize that construct integration will be improved when implants actively interact with and integrate into the subchondral bone. Growth and Differentiation Factor 5 (GDF-5) is known to support maturation of chondrocytes and to enhance chondrogenic differentiation and hypertrophy of mesenchymal stromal cells (MSC). Therefore, we investigated whether GDF-5 is capable to stimulate endochondral ossification of MSC in vitro and in vivo and would, thus, be a promising candidate for augmenting fibrin glue in order to support integration of tissue engineering constructs into the subchondral bone plate.

To evaluate the adhesive strength of fibrin glue versus BioGlue^®, a commercially available glue used in vascular surgery, an ex vivo cadaver study was performed and adhesion strength was measured via pull-out testing. MSC were suspended in fibrin glue and cultivated in chondrogenic medium with and without 150 ng/mL GDF-5. After 4 weeks, the formed cartilage was evaluated and half of the constructs were implanted subcutaneously into immunodeficient mice. Endochondral ossification was evaluated after 2 and 4 weeks histologically and by microCT analysis. BioGlue^®and GDF-5-augmented fibrin glue were tested for 4 weeks in a minipig cartilage defect model to assess their orthotopic biocompatibility.

Pull-out testing revealed sufficient adhesive strength of fibrin glue to fix polymeric CellCoTec constructs in 6 mm cartilage defects, however, BioGlue^®showed significantly higher adhesive power. In vitro chondrogenesis of MSC under GDF-5 treatment resulted in equal GAG deposition and COLIIa1 and ACAN gene expression compared to controls. Importantly, significantly increased ALP-activity under treatment with GDF-5 on day 28 indicated enhanced hypertrophic differentiation compared to controls. In vivo, MSC-fibrin constructs pre-cultured with GDF-5 developed a significantly higher bone volume on day 14 and 28 compared to controls. When pre-cultured with GDF-5 constructs showed furthermore a significantly higher bone compactness (bone surface/bone volume coefficient) than controls, and thus revealed a higher maturity of the formed bone at 2 weeks and 4 weeks. Orthotopic biocompatibility testing in minipigs showed good defect filling and no adverse reactions of the subchondral bone plate for defects treated with GDF-5-augmented fibrin glue. Defects treated with BioGlue^®, however, showed considerable subchondral bone lysis.

Thus, BioGlue^®– despite its adhesive strength – should not be used for construct fixation in cartilage defects. GDF-5-augmented fibrin glue is considered promising, because of a combination of the adhesive strength of fibrin with an enhanced osteochondral activity of GDF-5 on MSC. Next step is to perform a large animal study to unravel whether GDF-5 stimulated endochondral ossification can improve scaffold integration in an orthotopic cartilage defect model.