Saturday, October 24, 2009 - 1:30 PM

Purification of Four Distinct Cell Populations in the Stromal Vascular Fraction of Human Adipose Tissue: Adipogenic Potential and Implications for Soft Tissue Engineering

Han Li, MD, Ludovic Zimmerlin, BS, Kacey Marra, PhD, Vera Donnenberg, PhD, Albert Donnenberg, PhD, and J. Peter Rubin, MD.


Adipose stem cells (ASCs) represent a truly heterogenous population.  An understanding of the functional characteristics of subpopulations will be helpful in planning new ASC cell based therapies for soft tissue reconstruction. The aim of this study was to define four distinct populations within the stromal vascular fraction based on surface marker expression, and evaluate the ability of each cell type to differentiate to mature adipocytes.

Materials & Methods

Subcutaneous whole adipose tissue was obtained by abdominoplasty from human patients. After mechanical and enzymatic dissociation, removal of mature adipocytes by centrifugation and lysis of erythrocytes, the stromal vascular fraction was isolated on a Ficoll/hypaque density gradient and analyzed using a Dako CyAn cytometer and sorted using a Dako MoFlo High Speed Sorter. To distinguish isolated cell populations, we performed an 8-color analysis based on the expression of CD3, CD31, CD34, CD45, CD90, CD117 and CD146.  DAPI staining was performed to exclude apoptotic cells and cell clusters. Cell proliferation was determined by DNA quantification (Cyquant), and cells exposed to adipogenic culture media.  Adipocyte differentiation was assessed by PCR for PPAR Gamma, fatty acid binding protein 4 (FAB4), and Leptin. Lipid accumulation was confirmed with Oil Red-O staining.


Results & Discussion

Using eight-color multiparameter flow cytometry and prototype high throughput parallel processing analytical software (Venturi, Applied Cytometry Systems) four  cell populations were purified and studied. Cells with 2N DNA (DAPI staining, FL6 Log) were selected, and cell clusters (doublet discrimination on FSc), cell debris (FSc x SSc) and cells binding anti-CD3 or CD45 or autofluorescent in the FL1 or FL2 channels were excluded.  Candidate perivascular cells (pericytes), defined as CD146+ CD90dim, were characterized as CD31 and CD34 negative. These comprised 1.45 +/- 0.68% of cells studied.  Two CD31+ endothelial populations were detected and discriminated by CD34 expression and tentatively designated mature endothelial (CD 31+/CD34-: 0.75 +/- 0.29%), and immature endothelial (CD 31+/CD 31-:6.48 +/- 3.7%). CD90+ cells comprised 66% of mature endothelial cell candidates, and 96% of the CD31+ CD34+ population.  Both endothelial populations were heterogeneous with respect to CD146. The CD31-/CD34+ fraction (preadipocyte candidate: 74.26 +/- 11.88%) was also CD90+, but lacked CD146 expression.

Proliferation was greatest in the CD31-/CD34+ group and slowest in the CD146+ CD90dim group.  PPAR gamma mRNA expression was significantly higher in the CD31-/CD34+ group compared with all other populations after exposure to adipgenic medium.  Moreover, the PPAR gamma mRNA expression increased from day 7 to day 14 during culture in adipogenic medium.  A disproportionately high expression of FAB4 mRNA, a downstream target of PPAR gamma and an early marker of adipogenesis,  was also seen in this group.   The highest proportion of positive Oil Red-O staining was noted in this group, as well.


                We have isolated four distinct stromal populations from human adult adipose tissue.  Of these four populations, the CD31-/CD34+ group is the most prevalent and has the greatest potential for adipogenic differentiation.  This cell type appears to hold the most promise for engineering of adipose tissue for reconstructive applications.