Immunophenotype of Human Adipose-Derived Cells: Temporal Changes in Stromal-Associated and Stem Cell–Associated Markers



Adipose tissue represents an abundant and accessible source of multipotent adult stem cells and is used by many investigators for tissue engineering applications; however, not all laboratories use cells at equivalent stages of isolation and passage. We have compared the immunophenotype of freshly isolated human adipose tissue-derived stromal vascular fraction (SVF) cells relative to serial-passaged adipose-derived stem cells (ASCs). The initial SVF cells contained colony-forming unit fibroblasts at a frequency of 1:32. Colony-forming unit adipocytes and osteoblasts were present in the SVF cells at comparable frequencies (1:28 and 1:16, respectively). The immunophenotype of the adipose-derived cells based on flow cytometry changed progressively with adherence and passage. Stromal cell–associated markers (CD13, CD29, CD44, CD63, CD73, CD90, CD166) were initially low on SVF cells and increased significantly with successive passages. The stem cell–associated marker CD34 was at peak levels in the SVF cells and/or early-passage ASCs and remained present, although at reduced levels, throughout the culture period. Aldehyde dehydrogenase and the multidrug-resistance transport protein (ABCG2), both of which have been used to identify and characterize hematopoietic stem cells, are expressed by SVF cells and ASCs at detectable levels. Endothelial cell–associated markers (CD31, CD144 or VE-cadherin, vascular endothelial growth factor receptor 2, von Willebrand factor) were expressed on SVF cells and did not change significantly with serial passage. Thus, the adherence to plastic and subsequent expansion of human adipose-derived cells in fetal bovine serum-supplemented medium selects for a relatively homogeneous cell population, enriching for cells expressing a stromal immunophenotype, compared with the heterogeneity of the crude SVF.


Until recently, bone marrow has been the major tissue source of human adult stem cells. Bone marrow contains hematopoietic stem cells (HSCs), which have served as the prototypic example of an adult stem cell. The defining characteristics of a stem cell, such as surface immunophenotype and the capacity for multipotent differentiation and self renewal, are based primarily on studies of HSCs [1]. Bone marrow also contains stromal progenitor cells, now termed mesenchymal stem cells (MSCs), which have the ability to differentiate along multiple lineage pathways, including adipocytes, chondrocytes, hematopoietic-supporting cells, myocytes, neuronal cells, and osteoblasts [24]. Now, evidence is accumulating to support the existence of stem cells in multiple tissue sites. These include multipotent adult progenitor cells (MAPCs) [5] from the bone marrow, dermal stem cells [68], ear MSCs [9], neural stem cells from the central nervous system [10], hepatic and pancreatic stem cells [11], and stem cells from skeletal muscle [12, 13].

Adipose tissue represents an alternative source of stem cells. Subcutaneous adipose depots are accessible, abundant, and replenishable, thereby providing a potential adult stem cell reservoir for each individual. Many groups working independently have shown that adult stem cells derived from white adipose tissues can differentiate along multiple pathways in vitro, including the adipocyte, chondrocyte, endothelial, epithelial, hematopoietic support, hepatocyte, neuronal, myogenic, and osteoblast lineages [1428]. Thus, adipose-derived cells exhibit potential advantages for tissue engineering applications. However, the cell preparations in different laboratories are not identical. Most groups isolate cells from adipose tissue according to a modification of the methods originally described by Rodbell and colleagues [2931]; tissues are minced, digested with collagenase, and fractionated by differential centrifugation, and the pelleted stromal vascular fraction (SVF) cells are placed in culture. The adherent cell population can then be expanded and used in a variety of assays. Many terms have been used to identify the adherent cells, including adipocyte precursor cells [32], preadipocytes, adipose-derived adult stem (ADAS) cells [14], adipose-derived stromal cells [21], adipose-derived adherent stromal cells [33], processed lipoaspirate cells [24], and adipose-derived stem cells (ASCs) [34]. In accordance with a consensus reached by investigators attending the Second Annual International Fat Applied Technology Society meeting (October 3–5, 2004, Pittsburgh, PA), we will refer to these cells as ASCs from now on. Although some groups have focused their attention exclusively on the minimally processed SVF cell population, others have worked with the expanded plastic adherent ASC subpopulation at various stages of passage. It is likely that the SVF cells and ASCs exhibit different features and properties. This investigation set out to determine the frequency of adherent and lineage progenitor cells in the SVF cell by colony-forming unit (CFU) assays and to define the immunophenotype of human adipose-derived cells at various stages of isolation, purification, and expansion, based on flow cytometric assays using markers associated with endothelial cell, stem cell, and MSC.

Materials and Methods


All materials were obtained from Sigma-Aldrich (St. Louis, or Fisher Scientific (Pittsburgh, unless otherwise noted.

ADAS Cell Isolation and Expansion

All protocols were reviewed and approved by the Pennington Biomedical Research Center Institutional Research Board before the study. Liposuction aspirates from subcutaneous adipose tissue sites were obtained from male and female subjects undergoing elective procedures in local plastic surgical offices. Tissues were washed three to four times with phosphate-buffered saline (PBS) and suspended in an equal volume of PBS supplemented with 1% bovine serum and 0.1% collagenase type I (Worthington Biochemical Corporation, Lakewood, NJ, prewarmed to 37°C. The tissue was placed in a shaking water bath at 37°C with continuous agitation for 60 minutes and centrifuged for 5 minutes at 300 to 500g at room temperature. The supernatant, containing mature adipocytes, was aspirated. The pellet was identified as the SVF cell. Portions of the SVF cells were resuspended in cryopreservation medium (10% dimethylsulfoxide, 10% Dulbecco's modified Eagle's medium [DMEM]/F-12 Ham's, 80% fetal bovine serum [FBS]), frozen at −80°C in an ethanol-jacketed closed container, and subsequently stored in liquid nitrogen. Portions of the SVF were used in CFU assays (see below). The remaining cells of the SVF were suspended and plated immediately in T225 flasks in stromal medium (DMEM/F-12 Ham's, 10% FBS [Hyclone, Logan, UT,], 100 U penicillin/100 μg streptomycin/0.25 μg fungizone) at a density of 0.156 ml of tissue digest/sq cm of surface area for expansion and culture. This initial passage of the primary cell culture was referred to as passage 0 (P0). After the first 48 hours of incubation at 37°C at 5% CO2, the cultures were washed with PBS and maintained in stromal media until they achieved 75%–90% confluence (approximately 6 days in culture). The cells were passaged by trypsin (0.05%) digestion and plated at a density of 5,000 cells/cm2 (passage 1). Cell viability and numbers at the time of passage were determined by trypan blue exclusion and hemacytometer cell counts. Cells were passaged repeatedly after achieving a density of 75%–90% (approximately 6 days in culture) until passage 4.


Confluent cultures of primary ASCs were induced to undergo adipogenesis by replacing the stromal media with adipocyte induction medium composed of DMEM/F-12 with 3% FBS, 33 μM biotin, 17 μM pantothenate, 1 μM bovine insulin, 1 μM dexamethasone, 0.5 mM isobutylmethylxanthine (IBMX), 5 μM rosiglitazone, and 100 U penicillin/100 μg streptomycin/0.25 μg fungizone. After 3 days, media were changed to adipocyte maintenance media that were identical to induction media except for the deletion of both IBMX and rosiglitazone. Cells were maintained in culture for up to 9 days, with 90% of the maintenance media replaced every 3 days. Cultures were rinsed with PBS and fixed in formalin solution, and adipocyte differentiation was determined by staining of neutral lipids with oil red O.


Confluent cultures of primary ASCs were induced to undergo osteogenesis by replacing the stromal medium with osteogenic induction medium composed of DMEM/F-12 Ham's with 10% FBS, 10 mM β-glycerophosphate, 10 nM dexamethasone, 50 μg/ml sodium ascorbate 2-phosphate, and 100 U penicillin/100 μg streptomycin/0.25 μg fungizone. Cultures were fed with fresh osteogenic induction medium every 3 to 4 days for a period of up to 3 weeks. Cultures were rinsed in 0.9% NaCl and fixed in 70% ethanol, and osteogenic differentiation was determined by staining for calcium phosphate with Alizarin red.

Colony-Forming Unit Assays

The frequency of CFUs was determined by limit dilution assay with the assumption that the number of progenitor cells follows a Poisson distribution [3537]. A portion of the SVF equivalent to 25 ml of liposuction tissue aspirate was committed to limit dilution assays to determine the frequency of CFUs. The SVF pellet was suspended in 20 ml of PBS supplemented with 1% bovine serum albumin (BSA) and filtered through an autoclaved metal screen to remove large tissue fragments. A 400-μl portion of the cell suspension was removed to a 2-ml centrifuge tube and centrifuged for 3 minutes at 3,000 rpm at room temperature, and the pellet was resuspended in 400 μl of red cell lysis buffer (Sigma R7757). After a 5-minute lysis period, a 20-μl volume of the lysate was mixed with an equal volume of trypan blue and the number of nucleated cells was determined by hemacytometer count. The remaining cells of the SVF were centrifuged at ×300g for 5 minutes at room temperature, and the resulting pellet was resuspended in stromal medium at a final concentration of 2 × 105 cells/ml. Four 96-well plates were prepared with 100 μl of stromal medium per well. The SVF cell suspension was serially diluted twofold across the 12 columns of each plate, resulting in columns containing from 104 to 4 cells per well. The 96-well plates were incubated at 37°C, 5% CO2, for 9 days. At that time, one of the four plates was committed to a CFU-fibroblast (CFU-F) assay. The plate was rinsed with PBS, fixed in formalin, stained for 20 minutes with 0.1% toluidine blue in formalin, and rinsed with water, and the number of negative wells (i.e., those that did not contain colonies of >20 toluidine blue-positive cells) was determined for each cell concentration. These data were used to determine the number of CFU-fibroblasts (CFU-Fs) based on the Poisson distribution according to the equations Fo = eu and u = −ln Fo, where Fo is the fraction of wells without colonies (negatively staining wells) and u is the average number of precursors or CFUs per well. By solving the equation for the circumstance where the value of u = 1 (or a single CFU unit per well), the fraction of negatively staining wells is determined as Fo = 37%. Thus, under the limit dilution concentration conditions, when three out of eight or 37% of the wells do not contain a colony based on histochemical staining, you will determine the number of SVF cells necessary to obtain a single CFU [3537]. By combining this approach with the knowledge of the total number of cells seeded per well, it is possible to calculate the CFU frequency for any specific lineage by linear regression analysis. The second 96-well plate was committed to a CFU-alkaline phosphatase (CFU-ALP) assay. The plate was rinsed with PBS, fixed in 100% ethanol, incubated for 1 hour in the presence of a solution containing 36 mM sodium metaborate, 0.46 mM 5-bromo-4-chloro-3-indoxyl phosphate, 1.2 mM nitroblue tetrazolium, and 8.3 mM magnesium sulfate (pH 9.3), and rinsed with water, and the number of wells that did not contain colonies > 20 ALP+ cells was determined for each cell concentration; these data were used to determine the number of CFU-ALP according to the above formula. The remaining two 96-well plates were induced to undergo adipogenesis or osteogenesis, respectively, as described above. The CFU-adipocyte (CFU-Ad) was determined by oil red O staining 9 days after induction, whereas the CFU-osteoblast (CFU-O) was determined by Alizarin red staining >14 days after induction. Similar CFU analyses were performed on serial dilutions of ASCs from passages P0 to P4 isolated from a group of n = 4 donors.

Flow Cytometry

Flow cytometry was performed on cells from the SVF as well as from ASCs cultured from passages 0 through 4. All cells were cryopreserved at concentrations of 0.5 to 2.0 million cells per ml in 80% FBS, 10% dimethylsulfoxide, and 10% DMEM/F-12 Ham's medium for periods of ∼1 to 4 months before analysis. Immediately before flow cytometric analysis, individual vials of cells were rapidly thawed in a 37°C water bath (1 to 2 minutes of agitation), resuspended in 5 to 10 ml of medium containing 10% FBS, centrifuged at room temperature, and used for immunostaining. At this time, cell viability for all donors was at least 50% or greater based on trypan blue exclusion. Cells were analyzed for phenotypic markers falling within three general categories (hematopoietic, stromal, and stem cell) as well as aldehyde dehydrogenase (ALDH) activity (Stem Cell Technologies, Vancouver, British Columbia, Canada, The cells were analyzed using both conjugated and unconjugated mouse monoclonal antibodies. Briefly, approximately 4 to 8 × 106 ASCs were acquired from each population. A total of 1 × 106 cells were removed for ALDH analysis, and 1 to 2 × 106 cells were removed for staining with the unconjugated monoclonals. Ten thousand events were acquired per antibody set, and a minimum of 25,000 events was acquired for the ALDH assay on a Becton, Dickinson and Company FACS-Caliber flow cytometer using CELLQuest acquisition software (Becton, Dickinson and Company, Franklin Lakes, NJ, Data analysis was performed using Flow Jo analysis software (Tree Star, Ashland, OR,

Conjugated Monoclonal Antibodies

The cells were washed once in flow wash buffer (1× Dulbecco's phosphate-buffered saline, 0.5% BSA, and 0.1% sodium azide), resuspended in blocking buffer (wash buffer with 25 μg/ml mouse immunoglobulin G), and incubated for 10 minutes on ice. A total of 100 μl of cell suspension (approximately 5 × 105 cells) was aliquoted per tube, and appropriately labeled monoclonal antibodies were added for tricolor analysis (fluorescein isothiocyanate [FITC], phycoerythrin [PE], and antigen-presenting cell). Appropriate isotype control combinations were performed to reflect the monoclonal isotype combinations. Antibodies directed against the following antigens (catalog no.) were purchased from BD Pharmingen (San Diego, unless otherwise indicated and used at the vendor-recommended quantities: CD13 PE (#555394), CD29 FITC (Caltag #CD2901), CD31 FITC (Caltag #MHCD3101), CD34 PE (#348057), CD44 FITC (Cell Sciences #852.601.010), CD49a PE (#559596), CD63 FITC (#557288), CD73 PE (#550257), CD90 FITC (#555595), CD105 PE (Caltag #MHCD10504), CD144 (Chemicon #MAB1989), CD146 PE (#550315), CD166 PE (#559263), ABCG2 FITC (Chemicon #MAB4155F), vascular endothelial growth factor receptor 2 (VEGFr2) (Chemicon #MAB1667), and von Willebrand factor (Chemicon MAB3442). All tubes were incubated on ice and protected from light for 30 minutes. The cells were washed once in wash buffer and fixed in 200 μl of 1% paraformaldehyde.

Unconjugated Monoclonal Antibodies

The cells were washed as stated above, blocked in wash buffer containing 5% goat serum, incubated for 10 minutes, and distributed into 100-μl aliquots. The primary antibodies (CD144, anti-VEGFr2 [KDR], and anti-von Willebrand factor) were added (10 μg/ml), and the cells were incubated for 30 minutes on ice. The cells were washed once in wash buffer and resuspended in wash buffer without serum. Goat anti-mouse PE-conjugated secondary antibody was added (5 μg/ml) to the suspensions containing primary antibody as well as a secondary-only control. The cells were incubated on ice and protected from light for 15 minutes. The cells were then washed in flow wash buffer and fixed with 1% paraformaldehyde.


Cell Yield

Subcutaneous adipose tissue lipoaspirates obtained from a total of 44 donors were processed by collagenase digestion and differential centrifugation. The age (mean ± standard deviation [SD], 41 ± 10; range, 18 to 64), BMI (mean ± SD, 26.1 ± 4.8; range, 19.9 to 39.2), and gender distribution (84% female, 16% male) in the 44 donors were comparable to those reported in previous studies [38, 39]. To assess the frequency of progenitor cells in the adipose tissue, the mean number of nucleated cell numbers present in the SVF was determined as 308,849 per ml of lipoaspirate tissue (Table 1). Based on these calculations, CFU assays were established in 96-well plates by limiting dilution assays to determine the CFU frequency for specific lineage phenotypes based on histochemical staining characteristics and the Poisson distribution as outlined in Materials and Methods (Tables 2, 3). After 9 days in the culture, the number of wells containing cells staining positive for toluidine blue or alkaline phosphatase was used to determine the frequency of CFU-F and CFU-ALP, respectively (Fig. 1). At that time, identical plates were induced to undergo adipogenesis or osteogenesis. The number of wells staining positive for neutral lipids by oil red O or for calcium phosphate by Alizarin red was determined after an additional 9 days or >14 days, respectively. The resulting mean CFU frequencies in the SVF isolated from n = 7 to 12 donors were as follows: CFU-F, 1 per 32 cells; CFU-ALP, 1 per 328 cells; CFU-Ad, 1 per 28 cells; and CFU-Ob, 1 per 16 cells (Table 2). The effect of successive passage on the CFU frequency was determined using a subset of donors (n = 4) (Table 3). The lineage frequencies increased by as much as 10-fold by passage 4, as evidenced by the number of CFU-ALP and CFU-Ad.

After the initial plating, cells were maintained in culture for a mean of 6 days (Table 4) to yield the P0 population. Upon harvest by trypsin digestion, a mean of 247,401 adherent P0 cells (Table 1) were obtained per milliliter of lipoaspirate tissue. These values are comparable to previous studies [38]. Cells were maintained through an additional four successive passages of 6 to 7 days each. During each passage, the cell doubling times ranged between 3.6 and 4.7 days (Table 4).


Flow cytometric analysis was performed on cells cryopreserved after each stage of purification and passage (Table 5); representative flow histograms are shown in Figure 2. The initial SVF cells contained a subset of cells that were positive for a panel of endothelial cell–associated markers, including CD31, CD144 (VE-cadherin), VEGFr2, and von Willebrand factor (Table 5, Fig. 2). The levels of these markers did not change significantly through passage 4 (P4).

Only a subset of the initial SVF cell population expressed stromal cell-associated markers (Table 5, Fig. 3). Less than 1% of the SVFs expressed the activated lymphocyte common adhesion molecule (CD166), whereas 64% of the SVFs expressed the hyaluronate receptor (CD44); the levels of CD29, CD73, CD90, and CD105 were intermediate to these values. With successive passages, the percentage of cells staining positive for each of these markers increased, rising to between 69% (CD166) and 98% (CD44) by passage 4 (P4).

The initial SVF contained a subpopulation of cells positive for stem cell–associated markers. A mean of 60% of the SVFs expressed the HSC-associated marker CD34, a sialomucin and L-selectin ligand [40]. The CD34 levels remained comparable in the P0 population and then declined significantly in successive passages (Fig. 3). The size of the CD34+ population consistently exceeded that of the hematopoietic cell population in each passage based on expression of the pan-hematopoietic marker CD45 [41]. A mean of 31% of the SVFs displayed ABCG2, the multidrug-resistance transporter responsible for the efflux of the Hoescht dye and used in the identification of the side-scatter population of HSCs [42]. Although these levels increased during passage P1 and decreased in subsequent passages, the changes were not statistically significant relative to the SVFs.

High levels of the enzyme ALDH (ALDHbr) have proven to be a novel marker for the identification and isolation of HSCs [4345]. Based on flow cytometric analysis using a fluorescent substrate, the adipose-derived cells contained an ALDHbr sub-population (Table 5, Fig. 4). Although the ALDH levels were low in the SVF cells, the percentage of ALDHbr reached >70% between passages P0 to P4, with mean fluorescent intensities of 114 to 306 (data not shown). The percentage of ALDHbr ASCs fell to 10% when the cells were maintained in culture up to P9 (data not shown).


Previous studies from our laboratory and by other groups have defined the immunophenotype of plastic adherent ASCs at passage 2 or later [23, 33, 38, 46]. The ASCs display a surface protein profile resembling that of bone marrow–derived stromal cells or MSCs [47]. Like MSCs, the ASCs can differentiate along multiple lineage pathways [14, 48]. Indeed, our ring cloning analyses of human ASCs have documented that >50% of the clones expanded through passage 4 are capable of differentiation along two or more lineage-specific pathways [48]. Consequently, adipose tissue presents an accessible, abundant, and alternative source of adult stem cells for potential regenerative medical applications. Studies using bone marrow MSCs isolated from 51 adult human subjects determined that the frequency of CFU-F was approximately 1:10,000 STRO-1+ cells [49]. Since these authors used an enrichment step with the STRO-1 antibody, these are at least three orders of magnitude less than those currently reported for human adipose tissue. Thus, the abundance of CFU-F in adipose tissue is substantially greater than that of bone marrow.

The frequencies of CFU-Ad and CFU-Ob in adipose tissue were comparable to those of the CFU-F; however, the incidence of CFU-ALP was approximately one order of magnitude less frequent. Alkaline phosphatase enzyme activity has been used as a defining characteristic of bone marrow osteoblast progenitors and Westin-Bainton stromal cells [50, 51]. The current study measured alkaline phosphatase activity after 9 days in culture, whereas Alizarin red staining was performed after an additional 14 to 21 days. Since robust alkaline phosphatase staining was associated with multi-tiered cell layers (Fig. 1), it is possible that the frequency of CFU-ALP would have been closer to that of CFU-F and CFU-Ob if it had been assessed after an extended culture period. Progressive passage increased the frequency of CFUs for all lineages examined. By passage 4, CFU-F and CFU-Ob accounted for 33% of the adherent cells whereas CFU-Ad and CFU-ALP represented 16% and 8% of the population, respectively (Table 3). These findings indicate that the adhesion and expansion process enriches for a select functional cell phenotype.

Multiple groups have begun to isolate adipose-derived cells for both in vitro and in vivo applications; however, the degree of consistency between laboratories with respect to the isolation and characterization of the cell population under investigation remains unclear. Recent studies have focused on adipose tissue-derived cells at earlier stages of isolation, using the SVF or adherent cells at early passage number [34, 5254]. These cells displayed markers for the VEGF receptor, Flk-1, CD31, VE-cadherin, von Willebrand factor, and other markers associated with the endothelial cell lineage [34, 5254]. Adipose-derived SVF cells have been used to reconstitute the bone marrow of lethally irradiated mice [55]. The SVF population has been reported to contain progenitors for macrophages and, potentially, other hematopoietic lineages [5658]. Likewise, we find that the SVF cell population includes hematopoietic lineage cells based on their expression of CD11, CD14, CD45, and other markers (data reported in accompanying paper); however, their expression is lost with progressive passage, suggesting that they do not account for the adherent cell population.

The levels of stem cell–associated markers (CD34, ABCG2, ALDHbr) reach their peak in the earliest stages of culture (passages 0/1). Consistent with one of these findings, we have documented the presence of mitochondrial ALDH by tandem mass spectroscopy proteomic analysis of undifferentiated and adipocytes-differentiated human ASCs [59]. The percentage of ASCs that are ALDHbr greatly exceeds the percentage of ALDHbr cells detected in unfractionated bone marrow, which falls at or below 1% of the total cell population [43, 44]. Other groups have used several of these same stem cell–associated markers (CD34, ABCG2) in combination with CD31 and CD144 to characterize and define endothelial progenitor cells in adipose-derived cell populations [34]. It remains to be determined if a subset of antigens or enzyme markers within this panel can be used exclusively to define stem cells derived from adipose tissue in a manner similar to that now used to characterize and isolate HSCs from bone marrow [1].

In the earliest stages of isolation, the cells of the SVF exhibit low levels of stromal-associated markers (CD13, CD29, CD44, CD73, CD90, CD105, CD166). By the later stages of culture (passages 3/4), the cells assume a more homogeneous profile with consistently high levels of stromal markers. Overall, this temporal expression pattern resembles that reported for human bone marrow–derived MSCs [47]. Bone marrow MSCs progressively increased their surface expression of the markers identified as SH2 and SH3, corresponding to endoglin (CD105) and 5′-ecto nucleotidase (CD73), respectively, over 14 days of culture in vitro [47, 60, 61]. By passage 4, five of the stromal markers (CD13, CD29, CD44, CD73, CD90) are consistently present on >90% of the ASC population. Additional stromal markers, such as CD10, may also be of value in demonstrating the homogeneity of this population [38]. These findings are consistent with the current immunophenotypic characterization of the adipose-derived cells at various stages of isolation and expansion.


We have examined cells derived from human adipose tissue based on adherence characteristics and immunophenotype. The initially isolated SVF cells are heterogeneous; however, only 1 of 30 cells actually adheres to plastic and can account for the subsequent expansion of those cells now termed ASCs. The frequency of adipocyte and osteoblast progenitors in the SVF is comparable to that of the adherent cell population. This close correlation between CFU-F, CFU-Ad, and CFU-Ob data are consistent with studies from our laboratory and others demonstrating the presence of bipotent and tripotent clonal cells in human adipose tissue [23]. Classic stromal cell markers (CD13, CD29, CD44, CD73, CD90, CD105, CD166) are present on only 0.8%–54% of the initial SVF cells. By late passage, stromal markers are present on up to 98% of the adipose-derived stem cell population. These temporal changes in expression resemble those reported for human bone marrow MSCs [47]. The human ASCs also express stem cell–associated markers, such as CD34, ABCG2, and aldehyde dehydrogenase; however, it remains to be seen if these proteins can serve as unique identifiers of ASCs in a manner similar to that used for HSCs. Thus, significant changes occur in the adipose-derived cell population as a function of their isolation and culture. In light of recent reports indicating potential transformation of human ASCs after extended passage in culture [62], these findings have implications concerning the potential utility of human adipose tissue as a source of adult stem cells for regenerative medical therapies.

Table Table 1.. Cell yields per milliliter of lipoaspirate tissue
original image
Table Table 2.. Frequency of colony-forming units within the nucleated stromal vascular fraction cell population
original image
Table Table 3.. Frequency of colony-forming units with successive passage (n = 4 donors)
original image
Table Table 4.. Mean cell doubling times and passage lengths
original image
Table Table 5.. Phenotypic characterization of human adipose-derived cells at progressive stages of isolation and passagea
original image
Figure Figure 1..

Colony-forming unit assays. The photomicrographs display staining profiles representative of the following colonies: bottom left, toluidine blue+ colony-forming unit fibroblasts; top right, alkaline phosphatase+ colony-forming unit alkaline phosphatase; top left, oil red O+ colony-forming unit adipocytes; bottom right, Alizarin red+ colony-forming unit osteoblasts.

Figure Figure 2..

Flow cytometry histogram of adipose-derived cells. The flow cytometry histograms for selected hematopoietic, stem cell, and stromal cell markers from a representative donor are displayed at the stromal vascular fraction (SVF) and passage 2 (P2) stages. The percentage of cells staining positive is indicated in the upper right corner of each panel. The blue line indicates the positive staining cells, whereas the red line indicates the isotype-matched monoclonal antibody control.

Figure Figure 3..

Relative change in the immunophenotype of adipose-derived cells as a function of purification and passage. The percentage of positive-staining cells is displayed relative to the isolation stage and passage number. (A): Stromal cell–associated markers CD166, CD73, CD44, and CD29. (Note that the order of the passage numbers is reversed in panel A relative to panel B.) (B): Stem cell–associated markers ABCG2 and CD34.

Figure Figure 4..

Aldehyde dehydrogenase staining of adipose-derived cells as a function of purification and passage.


We thank the Pennington Biomedical Research Foundation for financial support; our colleagues Drs. Barbara Kozak, Ken Eilertsen, Rachel Power, Randy Mynatt, and Farshid Guilak for critical review of the manuscript; and Drs. Elizabeth Clubb and James Wade, their office and nursing staffs, and their patients for donating and providing liposuction waste material to these studies.


J.G. and X.W. have financial interest in Artecel Sciences. S.G., J.G., K.M., J.M., and X.W. have financial interest in Cognate Therapeutics.