EXPERIMENTAL Depot-Specific Variation in the Osteogenic and Adipogenic Potential of Human Adipose-Derived Stromal Cells Benjamin Levi, M.D. Aaron W. James, B.A. Jason P. Glotzbach, M.D. Derrick C. Wan, M.D. George W. Commons, M.D. Michael T. Longaker, M.D., M.B.A. Stanford and Los Angeles, Calif. Background: Adipose-derived stromal cells hold promise for use in tissue regen- eration. However, multiple facets of their biology remain unclear. The authors examined the variations in osteogenesis and adipogenesis in adipose-derived stro- mal cells between subcutaneous fat depots and potential molecular causes. Methods: Adipose-derived stromal cells were isolated from human patients from subcutaneous fat depots, including arm, flank, thigh, and abdomen (n = 5 pa- tients). Osteogenic and adipogenic differentiation was performed (alkaline phos- phatase, alizarin red, and oil red O staining, and quantitative real-time polymerase chain reaction). Co-cultures were established to assess the paracrine effect of human adipose-derived stromal cells on mouse osteoblasts. Finally, HOX gene expression was analyzed by quantitative real-time polymerase chain reaction. Results: Subcutaneous fat depots retain markedly different osteogenic and adipo- genic potentials. Osteogenesis was most robust in adipose-derived stromal cells from the flank and thigh, as compared with those from the arm and abdomen (p 0.05 by all markers examined). This was accompanied by elevations of BMP4 and BMPR1B (p 0.05 by all markers examined). The osteogenic advantage of cells from the flank and thigh was again observed when analyzing the paracrine effects of these cells. Conversely, those cells isolated from the flank had a lesser ability to undergo adipogenic differentiation. Adipose-associated HOX genes were less ex- pressed in flank-derived adipose-derived stromal cells. Conclusions: Variations exist between fat depots in terms of adipose-derived stro- mal cell osteogenic and adipogenic differentiation. Differences in HOX expression and bone morphogenetic protein signaling may underlie these observations. This study indicates that the choice of fat depot derivation of adipose-derived stromal cells may be an important one for future efforts in tissue engineering. (Plast. Reconstr. Surg. 126: 822, 2010.) C omposite tissue reconstruction remains a significant challenge in the treatment of pa- tients after traumatic injuries, oncologic re- sections, and congenital anomalies. Synthetic ma- terials and autogenous grafts are currently used to reconstruct skeletal defects, but they are limited by hardware failure, infection, and graft resorption. 1 In addition, bone marrow is less readily available than subcutaneous adipose tissue for the healing of larger defects. 2 Fat transfer for the purposes of soft-tissue augmentation suffers from equal disadvantages. 3 Grafts have been shown to be rela- tively unreliable over time because of resorption. 4 For example, in craniofacial fat transfer, it has been difficult to demonstrate long-term success. 3 It is against the backdrop of the shortcomings of skeletal and soft-tissue reconstruction that our laboratory is interested in the use of adipose-derived stromal cells for both skeletal and soft-tissue regeneration. From the Hagey Pediatric Regenerative Medicine Research Laboratory, Department of Surgery, Plastic and Reconstruc- tive Surgery Division, Stanford University School of Medi- cine, and the Division of Plastic and Reconstructive Surgery, University of California, Los Angeles. Received for publication January 28, 2010; accepted March 5, 2010. The first two authors share equal responsibility in the work presented in this article. Copyright ©2010 by the American Society of Plastic Surgeons DOI: 10.1097/PRS.0b013e3181e5f892 Disclosure: The authors have no financial interest in any of the products, devices, procedures, or any- thing else connected with the article. www.PRSJournal.com 822