Regenerative medical techniques will require an abundant source of human adult stem cells that can be readily available at the point of care. The ability to use unmatched allogeneic stem cells will help achieve this goal. Since adipose tissue represents an untapped reservoir of human cells, we have compared the immunogenic properties of freshly isolated, collagenase-digested human adipose tissue-derived stromal vascular fraction cells (SVFs) relative to passaged, plastic-adherent adipose-derived stem cells (ASCs). Parallel studies have shown that adherence to plastic and subsequent expansion of human adipose-derived cells selects for a relatively homogeneous cell population based on immunophenotype. Consistent with these findings, the presence of hematopoietic-associated markers (CD11a, CD14, CD45, CD86, and histocompatible locus antigen-DR [HLA-DR]) detected on the heterogeneous SVF cell population decreased upon subsequent passage of the ASCs. In mixed lymphocyte reactions (MLRs), SVFs, and early passage ASCs stimulated proliferation by allogeneic responder T cells. In contrast, the ASCs beyond passage P1 failed to elicit a response from T cells. Indeed, late passage ASCs actually suppressed the MLR response. Although these results support the feasibility of allogeneic human ASC transplantation, confirmatory in vivo animal studies will be required.
The emerging field of regenerative medicine seeks to combine biomaterials, growth factors, and cells as novel therapeutics to repair damaged tissues and organs. As this specialty grows, there will be a demand for a reliable, safe, and effective source of human adult stem cells to serve in tissue engineering applications. For regulatory purposes, these cells must be defined by quantifiable measures of purity. For practical purposes at the clinical level, these cells should be available as an “off the shelf” product immediately available upon demand at the point of care. The ability to use allogeneic, as opposed to autologous, adult stem cells for transplantation would have a significant positive impact on product development from a commercial standpoint. If this proved to be feasible, a single lot of cells derived from one donor could be transplanted to multiple patients, reducing the costs of both quality control and quality assurance.
We and others have demonstrated that adipose tissue is a potential source of adult stem cells capable of differentiation along multiple lineage pathways, including adipocytes, chondrocytes, endothelial cells, epithelial cells, hematopoietic supporting cells, hepatocytes, myocytes, neuronal-like cells, myocytes, and osteoblasts [1–15]. These cells are isolated by collagenase digestion of the adipose tissue, yielding an initial stromal vascular fraction (SVF) that is further processed by plating and adherence to plastic . Clonal analyses demonstrate that a single adherent human adipose tissue cell can give rise to a daughter cell population capable of multilineage differentiation potential, consistent with the definition of a stem cell . We have quantified the frequency of adherent colony-forming unit cells and found that they represent approximately 3% of the SVF population . These values exceed those of human bone marrow aspirates by at least three orders of magnitude . 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 this adherent cell population as adipose-derived stem cells (ASCs). A single milliliter of liposuction tissue aspirate can generate one-quarter million ASCs within a single passage, and these can expand an additional 64-fold within a 26-day period . Thus, a single liposuction procedure, which routinely generates liters of waste tissue, could potentially yield several billion cells with minimal time in culture.
A number of studies from our laboratory and others have determined that allogeneic bone marrow-derived mesenchymal stem cells or stromal cells (BMSCs) fail to stimulate an immune response based on in vitro assays of T cell function [19–23]. The immunophenotype of BMSCs changes as a function of time in culture , and this may be contributory to their distinct immunogenic properties. We have found that the immunophenotype of human adipose-derived cells changes significantly in a similar manner as a function of adherence to cultureware and expansion . Based on this and other parallels with BMSCs, we set out to determine whether the immunophenotype changes of the human adipose-derived cells correlated with alterations in their immunogenic properties based on in vitro mixed lymphocyte reaction. The results suggest that allogeneic transplantation of human ASCs may prove to be possible.
All protocols were reviewed and approved by the Pennington Biomedical Research Center Institutional Research Board prior to 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 or 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, http://www.worthington-biochem.com) 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 300–500g at room temperature. The supernatant, containing mature adipocytes, was aspirated. The pellet was identified as the SVF. Portions of the SVF were resuspended in cryopreservation medium (10% dimethylsulfoxide, 10% Dulbecco's modified Eagle's medium [DMEM]/Ham's F-12, 80% fetal bovine serum), frozen at −80°C in an ethanol-jacketed closed container, and subsequently stored in liquid nitrogen. Portions of the SVF were used in colony-forming unit assays (see below). The remaining cells of the SVF were suspended and plated immediately in T225 flasks in stromal medium (DMEM/Ham's F-12, 10% fetal bovine serum [HyClone, Logan, UT, http://www.hyclone.com], 100 U of penicillin/100 μg of streptomycin/0.25 μg of Fungizone) at a density of 0.156 ml of tissue digest per cm2 of surface area for expansion and culture. This initial passage of the primary cell culture was referred to as passage 0 (P0). Following 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 per 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.
BMSC Cell Isolation and Expansion
Human bone marrow was purchased from Cambrex (Walkersville, MD, http://www.cambrex.com) or AllCells, LLC (Berkeley, CA). Bone marrow aspirates were collected with heparin and fractionated over a 1.073 g/ml density gradient (Lymphocyte Separation Medium [LSM], Cambrex), and mononuclear cells collected at the interface were plated in HyQ DMEM-Low Glucose (HyClone) containing 10% fetal bovine serum (JRH Biosciences, Lenexa, KS, http://www.jrhbio.com) that was selected based on its ability to support BMSC expansion. Nucleated cells were plated at a density of 30 × 107 cells per T185-cm2 flask. Cells were grown in primary cultures (P0) for 12–17 days with medium changes every 3 or 4 days. When the cells became confluent, the culture was passaged using 0.05% trypsin (Gibco, a subsidiary of Invitrogen, Grand Island, NY, http://www.invitrogen.com) to remove adherent cells and re-plated as P1 cells at 1 × 106 cells per T185-cm2 flask. From this point on, the BMSCs were passaged every 7 days, with one medium change every 3–4 days. At final harvest, BMSC were cryopreserved using a freeze solution containing 10% DMSO (Edwards Life Sciences, Irvine, CA, http://www.edwards.com) and 5% human serum albumin (JRH Biosciences) in plasmalyte (Baxter Health Care, Deerfield, IL, http://www.baxter.com). Expanded BMSCs (P2–P4) represented a homogenous population that was fibroblastic in appearance and negative for hematopoietic markers (CD45, CD14, CD3, major histocompatability class II antigens) and positive for stromal markers (CD13, CD29, CD44, CD90, CD105). BMSCs were multipotent at P2 and P4 as shown by their ability to differentiate along the osteogenic and adipogenic lineages (data not shown).
Flow cytometry was performed as described . Antibodies directed against the following antigens (catalog #) were purchased from BD Pharmingen (San Diego, http://www.bdbiosciences.com/pharmingen) unless otherwise indicated and used at the vendor-recommended quantities: CD11a antigen presenting cell (APC) (550852), CD14 APC (555394), CD40 APC (555591), CD45 FITC (555482), CD54 APC (559771), CD80 FITC (Caltag MHCD8001), CD86 PE (Caltag MHCD8601), HLA-ABC APC (555555), histocompatible locus antigen-DR (HLA-DR) APC (559868). Isotype-matched negative controls were used to define background staining. Analysis gating included ∼1.5% of the negative control fluorescence; thus, populations containing ∼1.5% positive events should be considered negative.
Mixed Lymphocyte Reaction
Human Lymphocyte Populations.
Peripheral blood mononuclear cells (PBMCs) were prepared by centrifugation of leuko-pheresed peripheral blood cells (AllCells; LLC, Emeryville, CA, http://www.allcells.com) over an LSM density gradient. T cells were purified from a portion of the PBMCs by negative selection using magnetic beads. Briefly, PBMCs were treated with a cocktail of monoclonal antibodies (mAbs) (all from Serotec, Inc., Raleigh, NC, http://www.serotec.com) chosen to bind to monocytes (anti-CD14; clone UCHM1), B cells (anti-CD19; clone LT19), natural killer cells (anti-CD56; clone ERIC-1), and cells expressing MHC class II antigens (anti-MHC class II DR; clone HL-39). PBMCs were mixed with magnetic beads coated with anti-mouse IgG antibody (Dynal Corp., Lake Success, NY). Bead-bound cells were removed using a magnet, leaving a population of purified T cells (>90% T cells by flow cytometry using anti-CD3 mAb). Both PBMCs and purified T cells were aliquoted and cryopreserved in liquid nitrogen.
The one-way mixed lymphocyte reaction (MLR) assay was used to determine the immunogenicity of fat-derived cell populations. The MLR was performed in 96-well microtiter plates using Iscove's modified Dulbecco's medium supplemented with sodium pyruvate, nonessential amino acids, antibiotics/antimycotics, 2-mercaptoethanol (all reagents from Gibco, Grand Island, NY, http://www.invitrogen.com), and 5% human AB serum (Pel-Freez, Rogers, AK, http://www.pel-freez.com). Purified T cells derived from two different donors were plated at 2 × 105 cells per donor per well. Different donors were used to maximize the chance that at least one of the T cell populations was a major mismatch to the fat-derived test cells. Stimulator cells used in the assay included autologous PBMCs (baseline response), allogeneic PBMCs (positive-control response), and the test fat-derived cell populations. Stimulator cells were irradiated with 5,000 rads of γ radiation delivered by a cesium irradiator prior to being added to the culture wells at various numbers, typically ranging from 5,000 to 20,000 cells per well. Additional control cultures consisted of T cells plated in medium alone (no stimulator cells). Triplicate cultures were performed for each treatment. The cultures were incubated at 37°C in 5% CO2 for 6 days, pulsed with [3H]thymidine (1 μCi per well; Amersham Biosciences, Piscataway, NJ, http://www.amersham.com) for 16 hours, and the cells were harvested onto glass fiber filter mats using a Skatron 96-well cell harvester (Molecular Devices Corp., Sunnyvale, NY, http://www.moleculardevices.com). Radioactivity incorporated into the dividing T cells deposited on the filters was determined using a scintillation counter (Microbeta Trilux Scintillation and Luminescence Counter; Wallac Inc., Gaithersburg, MD, http://las.perkinelmer.com/).
Three criteria were used in assessing the immunogenicity of cell populations: 1) a statistically significant difference in the T cell-proliferative response (cpm) relative to that induced by autologous PBMCs (p < .05, Student's t test); 2) a stimulation index (S.I.; cpm induced by the test population divided by cpm induced by autologous PBMCs) of at least 3.0; and 3) a difference of at least 750 cpm from the response induced to autologous PBMCs. Test populations that passed all three criteria were considered immunogenic. The rationale for selecting both an S.I. and a minimum Δcpm was to ensure that the proliferative response to the test population was sufficiently higher than the background response to autologous PBMCs, which can be very low (often <100 cpm). Based on our MLR database of hundreds of assays to over a dozen different nonhematopoietic stimulator cells, 750 cpm was chosen as a reasonable threshold since the majority of T cell-proliferative responses exhibiting an S.I. ≥3 also exhibited a δcpm of 750 or greater.
The two-way MLR assay was used to evaluate suppression by adipose-derived cell populations. PBMCs from two different donors were used as the “responder cells” in the MLR. These were mixed in complete culture medium at 2 × 105 cells per donor per well in 96-well microtiter plates. Adipose tissue-derived cells, either SVF cells or ASCs at progressive passages, were added to the MLRs at cell concentrations of 5,000, 10,000, or 20,000 cells per well. Control MLR cultures had no adipose-derived cells added, or human splenic fibroblasts (CRL-7433; American Type Culture Collection, Manassas, VA, http://www.atcc.org) were added at concentrations of 5,000, 10,000, or 20,000 cells per well. Cultures were pulsed with [3H]thymidine on day 6 and harvested for scintillation counting as described above. Previous studies had shown that splenic fibroblasts displayed the least suppressive fibroblastic cell type when added to MLR assays (personal observations on six different fibroblast lines). Therefore, splenic fibroblasts were used in the present experiments to determine nonsuppressive dosing ranges in which to evaluate suppression by test cells. The percentage of suppression was calculated by the following formula: Percentage suppression = (1 − [Test cell + MLR cpm ÷ MLR cpm]) × 100. Statistical significance between control and test cultures was evaluated using Student's t test.
Subcutaneous adipose tissue lipoaspirates were obtained from a donor pool that was predominantly female (84%) with a mean age of 41 and a mean BMI of 26.1 . Samples were processed by collagenase digestion and differential centrifugation. The adipocyte, fibroblast, and osteoblast colony-forming units (CFUs) represented ∼2.5%, 3%, and 8.3% of the initially isolated SVF cells, respectively . With subsequent expansion to passage 4, the CFU frequency for each lineage increased by an additional 3–10-fold . Flow cytometric analysis was performed on cells cryopreserved after each stage of purification and passage. The adipose-derived cells expressed the stem cell-associated markers CD34, ABCG2, and aldehyde dehydrogenase . The expression of the stromal-associated markers CD13, CD29, CD44, CD63, CD73, CD90, and CD166 was initially low on all adipose-derived cells but increased with progressive time in culture . In contrast, the cells expressed the endothelial-associated markers CD31, CD144, or VE-cadherin, VEGF receptor 2, and von Willebrand factor, and their levels did not change significantly with passage .
Further analyses examined the expression of hematopoietic-associated surface antigens (Table 1, Fig. 1). The initial SVF cells freshly isolated from collagenase-digested adipose tissue and the initial adherent P0 ASCs contained a subset of cells that were positive for a panel of hematopoietic markers, including the common leukocyte antigen CD45, the monocyte/macrophage markers CD11a and CD14, the MHC class II DR histocompatibility antigen, and the co-stimulatory molecule CD86 (Table 1). This population disappeared by P1 according to decreased expression for most of the aforementioned markers (Table 1). The presence of cells bearing these surface markers in the freshly isolated adipose tissue-derived cell population is significant, as these cells are potentially immunogenic, could induce an allogeneic-rejection response, and may be a source of proinflammatory cytokines.
Other hematopoietic-associated markers displayed trends similar to markers associated with “stromal cells” . The surface levels of CD40, CD54 (ICAM-1), and MHC class I ABC histocompatibility antigen increased significantly between the SVFs and the P3 ASC populations (Table 1). The range of change varied between 1.3%–66% for CD40 and 67%–92% for HLA-ABC. The high level of class I antigen expression coupled with intermediate to high levels of molecules associated with costimulatory activity (CD40, CD54, CD80) would suggest that these cells could function as antigen presenting cells in the mixed lymphocyte reaction. This was investigated as described below.
One-way MLR assays were performed to assess the immunogenicity of human adipose-derived cells. The proliferation of T cells was measured based on [3H]thymidine incorporation in the presence of increasing doses of irradiated stimulator cells. Autologous and allogeneic PBMCs served as negative and positive stimulator cell controls, respectively. Human SVF cells elicited a dose-dependent MLR response comparable to that of allogeneic PBMCs (Fig. 2). With progressive passage, the human adipose-derived cells elicited a decreased response that fell to undetectable levels by P1. Immunogenicity of adipose-derived cell populations from multiple donors is shown in Table 2. Positive and negative designations for immunogenicity are based on criteria described in Materials and Methods and are shown for the highest cell dose in each experiment, which ranged from 20,000 cells per well (donors 902–917) to 30,000 cells per well (donors 407–611). Based on positive responses for either or both T cell populations, the following populations were immunogenic: SVF cells (four of seven donors), P0 cells (seven of nine donors), and P1 cells (four of seven donors). The remaining passaged cell populations (P2–P4) did not induce T cell proliferation in MLR assays, with the exception of P2 cells from one donor.
The inability of passaged ASCs to stimulate a T cell response may be due to inherent low immunogenicity, to active immunosuppressive mechanisms mediated by the ASCs, or to a combination of both properties. To determine whether the fat-derived cells were immunosuppressive, they were added to MLR cultures at 5,000, 10,000, or 20,000 cells per well. Control MLR cultures either had no cells added or nonsuppressive human splenic fibroblasts were added at the numbers previously described to control for suppression due to cell crowding. As shown in Figure 3, splenic fibroblasts suppressed the MLR cultures only at the highest dose (20,000 cells per well). Using the lower two doses as being valid (no artifactual suppression), significant suppression was mediated by all ASC passages except the SVF population. Percentage suppression of the control MLR response (no cells added) mediated by P0–P4 ASCs ranged from 33%–63%. The results from four donors are summarized in Table 3. Percentage suppression was determined at the lowest dose of cells (5,000 cells per well) since there was no suppression of the MLR at this dose of splenic fibroblasts in any of these experiments. Mean suppression by the SVF population was minimal (10%), whereas suppression by P0–P4 cells averaged 32.0% ± 3.2%. This degree of suppression is significant in view of the low percentage of ASCs in these cultures (1.3%).
It was of interest to compare the suppressive properties of ASCs to BMSCs since BMSCs have similar phenotypic characteristics and differentiation potential as ASCs . Both cell types suppressed the MLR when added at doses of 3,300–10,000 cells per well (Fig. 4). The magnitude of suppression by ASCs exceeded that of BMSCs by up to 13%, but the difference was not statistically significant.
Our studies show that freshly isolated SVF cells can elicit a T cell-proliferative response equivalent to that of allogeneic PBMCs in a mixed lymphocyte reaction. This immunogenic response declined for early passage (P0, P1) ASCs and essentially disappeared for later passage ASCs (P2–P4). The immunogenicity of a cell population in the context of alloreactivity is determined primarily by the presence of APCs within the population. The classic APC is a hematopoietic cell, typically a dendritic cell or macrophage, that expresses MHC class I and class II molecules in addition to costimulatory molecules such as CD80 and CD86. It is important to note that the SVF and P0 populations of adipose-derived cells, which were found to be immunogenic, contained a subpopulation of cells positive for APC-associated markers (CD45, CD11a, CD14, CD86, and MHC class II antigens), whereas P1–P4 populations, which did not contain this subpopulation, were generally not immunogenic.
There are several explanations to account for these observations. First, heterogeneity within the SVF and P0 populations may include the presence of a significant percentage of hematopoietic-derived APCs. The abundance of these hematopoietic-derived cells decline with progressive passage and expansion of the adherent ASCs. Alternatively, the adherent ASCs may themselves express the APC-related surface proteins initially during culture and thereby serve as immunogenic stimuli in the MLR. Certainly, the ASCs express a number of cell surface molecules through passage P4, which can exhibit costimulatory activity, including CD54 , CD40 , CD80, and CD86 . However, in the absence of CD11, CD14, HLA-DR, and other APC-related proteins, these surface antigens may not be sufficient to endow ASCs with APC function. In addition, other mechanisms, such as the ASC's development of active immunosuppression, may override any inherent immunogenicity of the ASCs in culture.
In this study, we have shown that ASCs significantly suppressed T cell proliferation in the MLR. This property was pronounced in P0–P4 cells (mean suppression 32%), but not in the SVF population (mean suppression 10%). There are two alternative explanations for these findings. First, the SVF may contain a subpopulation of cells that possess immunosuppressive function but are not present in sufficient numbers to mediate their effect in an MLR. Alternatively, the ASCs may acquire their immunosuppressive properties only upon adherence and expansion. Theoretically, these two alternatives could be resolved by assessing the immunosuppressive properties in cells purified directly from the SVF population based on their immunophenotype. Although this experiment would be potentially challenging to perform, the results could be important in sorting out issues related to the immunogenicity of the SVF population as described previously (e.g., the presence of contaminating APCs or lack of suppression).
To avoid artifactual interpretation of results (i.e., suppression due to cell crowding), we performed suppression experiments at very high ratios of responding cells in the MLR to the test cells (80:1). Control splenic fibroblasts were not suppressive at this ratio. We compared suppression by ASCs to BMSCs since both cell types have similar phenotypic and functional characteristics and BMSCs have been shown to be immunosuppressive by their ability to inhibit T cell proliferation in MLR assays as well as to mitogenic stimulation [20–23]. Indeed, we found that ASCs and BMSCs exhibited similar magnitude of suppression. Our results confirm and extend those reported recently by Puissant et al. . These investigators examined cells equivalent to our P0 population and found that ASCs stimulated a proliferative response that was significantly less than that elicited by allogeneic PBMCs, similar to the current findings. Likewise, their ASCs displayed a cell concentration-dependent immunosuppressive effect when added to mixed lymphocyte reactions.
Although there is little information on the mechanism of suppression by ASCs, Puissant et al. determined that the immunosuppressive effect of the ASCs was not entirely dependent on direct contact with lymphocytes, consistent with the release of a soluble factor . BMSCs have been reported to elaborate suppressive molecules, including hepatocyte growth factor and transforming growth factor-β , prostaglandins [30, 31], and indoleamine 2,3-dioxygenase . Several different mechanisms have been proposed to account for BMSC-mediated suppression of lymphocyte proliferation. These include division arrest of activated T cells and B cells by inhibition of cyclin D2 expression , induction of regulatory T cells  or APCs , and interference with dendritic cell  and cytotoxic T cell maturation . Further studies are needed to determine whether ASCs mediate suppression by similar mechanisms.
Our immunological data demonstrating that culture-expanded adipose-derived cells do not stimulate but actively suppress alloreactive T cell proliferation suggests that these cells can be transplanted across classical histocompatibility barriers. BMSCs have been reported to survive in immunocompetent allogeneic and xenogeneic recipients for longer than expected periods of time . Whether this is due to their immunosuppressive properties remains to be determined, although there is evidence to suggest that BMSCs can mediate suppression in vivo [19, 34, 35]. Due to the immunogenic nature of the SVF population, it is likely that transplantation of SVF cells will be limited solely to autologous applications, although manipulation of the graft to remove monocytes may diminish immunogenicity of this population. The potential use of allogeneic ASCs as a source of cells for tissue repair or replacement has important implications with respect to the ready availability of adult stem cells for clinical practice and to the practical and commercial aspects of their manufacture and quality assurance. Additional in vivo studies using ASCs from a preclinical animal model will be necessary to determine the feasibility of such an approach.
We have found that the characteristics of cells derived from human adipose tissue change as a function of adhesion and expansion in vitro. The stromal vascular fraction cells, isolated by collagenase digestion and differential centrifugation, were heterogeneous with respect to expression of classical hematopoietic markers. Between 8.1% and 17.6% of these initial cells expressed the monocyte/macrophage and panhematopoietic antigens CD11a, CD14, CD45, CD86, and HLA-DR. After four successive passages, less than 1% of the adherent adipose-derived stem cells expressed CD14, CD45, or CD86, whereas only 3% or fewer of the cells expressed either CD11a or HLA-DR. These changes in immunophenotype correlated with the level of immunogenicity displayed by the human adipose-derived cells in mixed lymphocyte reactions. Although the stromal vascular fraction cells and early passage adipose-derived stem cells (P0/P1) elicited a proliferative response from allogeneic T cells, later passage cells failed to do so. Indeed, the addition of adipose-derived stem cells to mixed lymphocyte reactions suppressed the proliferative response of T cells to allogeneic stimulator cells. Although this work implies that it will be possible to transplant adipose-derived stem cells across traditional histocompatibility barriers, further studies using in vivo animal models will be required to test this hypothesis.
Table Table 1.. Phenotypic characterization of human adipose-derived cells at progressive stages of isolation and passage
Table Table 2.. Immunogenicity of adipose-derived cell populations assessed in the MLR assay against T cells derived from two different donors
Table Table 3.. Percentage suppression of MLR cultures by adipose-derived cell populations from four different donors
J.M.G. and X.W. have financial interests in Artecel Sciences. S.G., J.M.G., L.H., K.M., J.B.M., and X.W. have financial interests in Cognate Therapeutics.
We thank the Pennington Biomedical Research Foundation for financial support; 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.