Abstract
Non-union is poorly understood. It is unknown if multipotent cells are present in non-union tissue or whether the activity of such cells is dysfunctional. Clinically, this is important as it may predict the success of novel therapies such as BMP treatments and cell-transplantation. This study aimed to study the characteristics of cell types present in human fracture non-union tissue, in comparison with bone marrow stromal cells (BMSC) from the patient and other healthy patients.
Non-union tissue was harvested (n=8) from long bones. Cells were isolated enzymatically and cultured in monolayer. BMSC were isolated by density gradient centrifugation of iliac crest biopsies. Their phenotype was assessed by FACS analysis for CD34, 45 and 105 markers. Their comparative growth kinetics was examined, as was their osteogenic and adipogenic capacity following extended culture in defined medium. Cell differentiation status was evaluated using alkaline phosphatase, von Kossa and oil-red O staining. Cell senescence was assessed via cell morphology, senescence associated Beta-galactosidase (SA-Beta)-Gal) activity.
Non-union cells grew in monolayer, but showed different morphologies; many non-union cells contained stress filaments (typical of senescent cells) or were of stellate appearance. In addition, significantly more non-union cells were positive for SA-Beta-Gal activity compared to BMSC (P=0.0006). Growth kinetics showed longer doubling times for cells isolated from non-union tissue when compared to BMSC isolated from the patient. Long term culture of non-union cells showed early growth arrest at passages 3–8. FACS analysis showed isolated cells to be CD34/45 negative and CD105 positive. Both non-union cells and BMSC differentiated along osteogenic and adipogenic lineages to varying extents.
Our novel results show that cells from non-union tissue exhibit senescence in culture. Hence, cell senescence is potentially involved in the aetiopathogenesis of non-unions. Whether or not this senescence has arisen through cell division (during failed repair attempts) or via abnormal biomechanical loading warrants further study. The influence of senescent cells on the healing process also requires investigation. Clearly these cells are able to differentiate into osteoblasts in vitro but may have an aberrant influence on union in vivo.
Correspondence should be addressed to Dr Roger Bayston, Division of Orthopaedic and Accident Surgery, Queen’s Medical Centre, Nottingham, NG7 2UH, England.
Acknowledgments: Funding: Institute of Orthopaedics-Oswestry & EU MyJoint(FP6:28861-NEST).
