Proteomic Analysis of Primary Cultures of Human Adipose-derived Stem Cells

Adipogenesis plays a critical role in energy metabolism and is a contributing factor to the obesity epidemic. This study examined the proteome of primary cultures of human adipose-derived adult stem (ADAS) cells as an in vitro model of adipogenesis. Protein lysates obtained from four individual donors were compared before and after adipocyte differentiation by two-dimensional gel electrophoresis and tandem mass spectroscopy. Over 170 individual protein features in the undifferentiated adipose-derived adult stem cells were identified. Following adipogenesis, over 40 proteins were up-regulated by ≥2-fold, whereas 13 showed a ≥3-fold reduction. The majority of the modulated proteins belonged to the following functional categories: cytoskeleton, metabolic, redox, protein degradation, and heat shock protein/chaperones. Additional immunoblot analysis documented the induction of four individual heat shock proteins and confirmed the presence of the heat shock protein 27 phosphoserine 82 isoform, as predicted by the proteomic analysis, as well as the crystallin α phosphorylated isoforms. These findings suggest that the heat shock protein family proteome warrants further investigation with respect to the etiology of obesity and type 2 diabetes.

Obesity is a health problem of epidemic proportions. It is estimated that in 2000 over 60% of adults are overweight (BMI 1 Ͼ 25) and that 30% are obese (BMI Ͼ 30); this compares to levels of 46 and 14%, respectively, in 1980. Obesity and increased adiposity are associated clinically with the onset of insulin resistance, dysfunctional glucose sensing and utilization, hypertension, and hypertriglyceridemia, all contrib-uting to the pathologic sequelae of type 2 diabetes. Paradoxically type 2 diabetes also occurs in patients with inherited or acquired forms of lipodystrophy or loss of adipose tissue depots (1,2). Lipodystrophy occurs through defects in genes associated with triglyceride metabolism or as a consequence of antiretroviral therapy in human immunodeficiency viruspositive patients (1,2). Animal models confirm these clinical observations; multiple strains of transgenic mice with a lipodystrophic phenotype exhibit type 2 diabetes (3)(4)(5). Diabetes in these animals responds not to insulin therapy but to transplantation of subcutaneous adipose tissue or to leptin treatment (3)(4)(5). These clinical and experimental observations have led to the hypothesis that a failure in adipocyte differentiation is a critical etiologic factor leading to type 2 diabetes (6 -8). Danforth (6) and others postulate that in obese individuals adipose tissue depots have already committed all of their stem cell reserves to the adipocyte lineage and have lost their capacity to create new adipocytic cells (6 -8). In the face of excess energy balance, both obese and lipodystrophic individuals deposit triglycerides in ectopic sites, such as muscle and liver, thereby contributing to the metabolic dysfunction associated with type 2 diabetes (increased hepatic gluconeogenesis, skeletal muscle insulin resistance, and abnormal pancreatic insulin secretion) (6). As a result of these findings, there has been a renewed interest in adipocyte progenitor cells as therapeutic targets and experimental models for studies of obesity and type 2 diabetes.
Murine cell models, most notably the 3T3-L1 cell line, have been the basis for the majority of studies of adipogenesis at the transcriptional and protein levels. However, there is a growing concern that adipogenesis may differ between human and murine systems. For example, the resistin gene and its secreted protein product were first identified in the 3T3-L1 cells (9). Subsequent in vivo analysis in mice demonstrated an association between resistin levels, obesity, and type 2 diabetes (9). In contrast, clinical studies do not demonstrate a comparable association between serum resistin levels, obesity, and insulin resistance in non-obese and obese human subjects (10). Likewise the regulation of the agouti gene in adipose tissue differs between man and mouse (11). These discrepancies argue for the increased use of human preadipocyte cell models in exploratory research relating to obesity and type 2 diabetes.
Human adipose-derived adult stem (ADAS) cells offer an alternative in vitro model (12,13). These cells can be reproducibly isolated from liposuction aspirates through a procedure involving collagenase digestion, differential centrifugation, and expansion in culture. A single milliliter of tissue yields over 400,000 cells (14). The undifferentiated human ADAS cells express a distinct immunophenotype based on flow cytometric analyses and, following induction, produce additional adipocyte-specific proteins (14 -18). The human ADAS cells display multipotentiality with the capability of differentiating along the adipocyte, chondrocyte, myogenic, neuronal, and osteoblast lineages (14 -29). In the presence of dexamethasone, insulin, isobutylmethylxanthine, and a thiazolidinedione, the undifferentiated human ADAS cells undergo adipogenesis; between 30 and 80% of the cells, based on flow cytometric methods, accumulate lipid vacuoles, which can be stained for neutral lipid with Oil Red O dye (16,17). The current study used two-dimensional electrophoresis/tandem mass spectroscopy to phenotype the proteome of undifferentiated and adipocyte differentiated human ADAS cells. The findings are discussed in the context of proteomic analyses of murine adipocytes and related human cell models.

EXPERIMENTAL PROCEDURES
Liposuction Aspirate Cell Isolation and Culture-The procedures used are modifications of published methods (14,16,17,30). Liposuction aspirates from subcutaneous adipose tissue sites were obtained from male and female subjects (n ϭ 6) undergoing elective procedures in local plastic surgical offices. The mean age and BMI (ϮS.D.) of the subjects were 38 Ϯ 14 years and 30.4 Ϯ 7.1, respectively. Tissues were washed three to four times with phosphatebuffered saline and suspended in an equal volume of PBS supplemented with 1% bovine serum and 0.1% collagenase type I prewarmed to 37°C. The tissue was placed in a shaking water bath at 37°C with continuous agitation for 60 min and centrifuged for 5 min at 300 ϫ g at room temperature. The supernatant was removed, and the pelleted stromal vascular fraction (SVF) was resuspended in Stromal Medium (Dulbecco's modified Eagle's medium/F-12 Ham's, 10% fetal bovine serum, antibiotic/antimycotic) and plated at a density of 0.156 ml of tissue digest/cm 2 of surface area in T225 flasks using Stromal Medium for expansion and culture. This initial passage of the primary cell culture is referred to as "Passage 0." Following the first 48 h of incubation at 37°C at 5% CO 2 , the cultures were washed with PBS and maintained in Stromal Medium until they achieved 100% confluence (mean cell density of ϳ31,000 cells/cm 2 after 4.2 Ϯ 1.5 days in culture). The cells were passaged by trypsin/EDTA digestion and seeded at a density of 30,000 cells/cm 2 ("Passage 1") on 10-cm plates for day 0 protein harvest or on 6-well plates for day 9 adipogenesis protein harvest.
Cell Culture and Adipogenesis-One day after seeding, plates were either harvested for protein (day 0), or the medium was replaced with an Adipogenic Differentiation Medium composed of Dulbecco's modified Eagle's medium/F-12 with 3% fetal bovine serum, 33 M biotin, 17 M pantothenate, 1 M bovine insulin, 1 M dexamethasone, 0.25 mM isobutylmethylxanthine, 5 M rosiglitazone, and 100 units of penicillin, 100 g of streptomycin, 0.25 g of Fungizone. After 3 days, Adipogenic Differentiation Medium was changed to Adipocyte Maintenance Medium, which was identical to the induction medium except for the removal of both isobutylmethylxanthine and rosiglitazone.
Cells were fed every 3rd day and maintained in culture for 9 days prior to protein harvest.
Oil Red O Staining and Quantification-Cells were fixed in 10% formaldehyde/PBS fixative solution for 20 min and then rinsed five times with double distilled H 2 0. 0.3% Oil Red O (Sigma) solution was added to the cells for 2 h on the orbital shaker. The stain was then aspirated, and the cells were washed with double distilled H 2 0 until no residue remained in the culture plates. The plates were then scanned on a flat bed scanner with photo quality resolution, and the images were used to quantitate the percentage of cells stained with Oil Red O using MetaVue software (Universal Imaging Corp.).
Protein Extracts-Cells plated in 10-cm or 6-well plates were washed with ice-cold PBS and lysed directly in 1 ml of Ready Prep Sequential Extraction Reagent 3 prepared according to the manufacturer's (Bio-Rad) instructions. The lysates were sonicated until clear, left at room temperature for 1 h, and centrifuged at 18,000 ϫ g for 10 min at room temperature. Protein extracts were concentrated using Centricon 10 tubes centrifuged at 5000 ϫ g at room temperature. Protein concentrations were determined using the Bio-Rad protein assay reagent and stored at Ϫ80°C prior to use.
Following centrifugation to remove unsolubilized material, samples were rehydrated at ϳ1 mg/gel in DeStreak Reagent (Amersham Biosciences catalog number 17-6003-18, 2-hydroxyethyl disulfide) containing 1% ampholytes, pH 3-10 (Bio-Rad catalog number 163-9094) and were introduced into the dry IPG strips (typically 24 cm, pH 3-10 nonlinear) under conditions of active rehydration (e.g. with a slight voltage applied across the strips). All gels were run in duplicate. Proteins were focused at a maximum 10,000 V for a total of 90,000 V-h. Upon completion of first dimension electrophoresis, the IPG strips were either directly subjected to second dimension SDS-PAGE or frozen at Ϫ80°C for later analysis. For the second dimension, the IPG strips were equilibrated first with 0.375 M Tris-HCl, pH 8.8, 6 M urea, 20% glycerol, 2% SDS, 1% DTT for 15 min followed by a second equilibration with 0.375 M Tris-HCl, pH 8.8, 6 M urea, 20% glycerol, 2% SDS, 2.5% iodoacetamide for 15 min. The strips were rinsed with electrophoresis buffer (25 mM Tris, 190 mM glycine, 0.1% SDS) and then embedded in low melting temperature agarose onto the top of a 25 ϫ 20-cm 12% acrylamide gel. Gels were run at constant current until the bromphenol blue dye front reached the bottom of the gel and stained with Sypro Ruby. The stained gels were scanned with a Molecular Imager FX with data directly imported into PDQuest. For each gel, the relative abundance of each resolved protein feature was quantified by mathematical fitting of Gaussian curves in two dimensions. Data within each were normalized (expressed as a percentage of total spot abundance), and routine statistical analyses available within the software package were used to identify unique spots, absent spots, or spots up-or down-regulated under specified conditions.
Trypsin Digestion-Following electrophoresis, staining, scanning, spot detection, and match set preparation, proteins of interest were selected, and their standard spot numbers were entered into a "Cut List." This Cut List was used by the automated spot cutter to select and excise the protein features in order of least to most abundant from one or more gels. Excised gel plugs were deposited into a 96-well plate and transferred to the MassPrep (Waters/Micromass) station. Proteins within the gel plugs were automatically destained, reduced, alkylated, dehydrated, rehydrated, and digested with trypsin. The resulting peptides were extracted, cleaned up, and then deposited into 96-well plates for analysis.
Q-TOF Analysis-The peptides from each digested spot were separated by capillary liquid chromatography interfaced to an ESI-MS/MS Micromass Q-TOF micromass spectrometer. MassLynx 4.0 software package (Waters) was used to identify individual mass spectrograms. Parameters included calculation of charge states and peaks were deisotoped. The ProteinLynxGlobalServer 1.1 software was used to search Release 43.0 of Swiss-Prot containing 146,720 sequence entries for protein identification using 100 ppm precursor ion and fragment ion mass accuracy; modifications included phosphorylation, oxidation of methionine, and cysteines modified with iodoacetamide; one missed cleavage; and using trypsin. Scores above 100 were generally considered valid identifications, although any identification with a score below 200 was examined carefully to verify that the spectra included a good number of consecutive "y" ions with high mass accuracy. The number of peptides analyzed and the percent coverage of the total amino acid sequence were determined for each protein identified. The data base was checked for redundancy and inspected for single proteins listed under multiple names. The molecular weight and pI of identified proteins were evaluated and verified relative to the electrophoretic mobility of the protein feature on the two-dimensional gel. Proteins were classified into functional categories based on their listed description in the Swiss-Prot data base.
Criteria Used for Analysis-The proteome of the undifferentiated and differentiated human ADAS cells was defined based on the following guidelines: proteins "induced" or "reduced" during adipogenesis displayed both a 98% significance in comparisons between replicate groups and Ͼ2-fold induction (51 features) or Ͼ3-fold reduction (23 features) with adipocyte differentiation.

RESULTS
Adipogenesis-Human ADAS cells isolated from subcutaneous liposuction aspirates of six non-diabetic, healthy donors (mean age ϭ 38 Ϯ 14.2 years, mean BMI ϭ 30.4 Ϯ 7.1) were expanded to Passage 1 and plated at a density of 3 ϫ 10 4 cells/cm 2 in 100-mm plates. Cells from four of the donors (two female and two male) were induced with a combination of adipogenic factors (insulin, isobutylmethylxanthine, dexamethasone, and thiazolidinedione) and differentiated for an additional 9 days in culture. Both female and male subjects were included to avoid biasing this initial analysis of the adipocyte proteome and to focus on those characteristics shared independent of gender. Adipogenesis was accompanied by the increased appearance of lipid vacuoles staining positive with Oil Red O, indicating the presence of neutral lipids. A representative induction is shown in Fig. 1 at the microscopic and macroscopic scale; over 35% of the surface area stained positive with Oil Red O.
Undifferentiated ADAS Proteome-Protein lysates prepared from undifferentiated (n ϭ 6 donors) and adipocyte differentiation (n ϭ 4 donors) human ADAS cells were separated by two-dimensional polyacrylamide electrophoresis and detected with Sypro Ruby (Fig. 2). The number of features identified on each gel ranged from 691 to 795. The undiffer- entiated gels shared 467 features in common as compared with 434 features on the differentiated gels; a total of 288 features were common to all gels. The data from four individual donors was combined to create a "master" map for both the undifferentiated and adipocyte differentiated human ADAS cells (Fig. 2). The functionality and subcellular localization of the 175 proteins identified in the undifferentiated ADAS cells are summarized in Fig. 3. The identity, number of peptides matched, percent coverage, "score," pI, molecular weight, and accession number of each protein are presented in the Supplemental Table 1A. The peptide sequence of those features identified by a single peptide match are presented in Supplemental Table 1B.
Adipocyte Differentiated ADAS Proteome-Comparison between the gels identified proteins that were induced by Ն2-fold or reduced by Ն3-fold following adipocyte differentiation (p Ͻ 0.02, 98% confidence level). Representative protein features are displayed in Fig. 4. The proteins identified as SSP numbers 3101 and 7204 correspond to fatty acid-binding protein (adipocyte) and heat shock protein 20-like protein, respectively, and are both induced with adipogenesis. In contrast, the proteins identified as SSP numbers 3107 and 6521, corresponding to Stathmin and Elfin (as well as LIM and SH3 domain protein 1), respectively, are reduced with adipogenesis. Supplemental Table 2 provides a comprehensive list of those proteins modulated by adipogenesis, including the number of peptides matched, percent coverage, pI, molecular weight, accession number, score, subcellular location, and function. An abbreviated list of these proteins is presented in Table I; those proteins uniquely attributed to differentiated adipocyte human ADAS cells as a result of this study are highlighted in bold fonts. Proteins were categorized into func-tional groups based on their definition within the Swiss-Prot data base. Adipogenesis reduced expression of proteins within selected functional classes, including cytoskeletal/ structural, metabolic, and heat shock protein/chaperonerelated proteins. In addition to these same categories, adipogenesis selectively induced proteins involved in oxidation-reduction and proteasomal degradation and ubiquitination.
Effect of Differentiation on Heat Shock/Chaperone Proteins-Further analysis focused on the heat shock proteins because alterations in their expression has been linked to obesity and diabetes (see "Discussion"). Immunoblots were performed using a panel of antibodies to confirm, validate, and extend the proteomic analysis of the heat shock protein and chaperone family. Control studies documented that each of these antibodies detected an appropriate sized signal in protein lysates prepared from intact human adipose tissue (data not shown). Two of the protein lysates used in the The "Other" location category includes the Golgi, lysosome, plasma membrane, ribosome, and secreted, whereas the "Other" function category includes amyloid-binding protein, cytokine, ion channel, iron-binding protein, signal transduction, and transcription. proteomic analysis described above were examined; both undifferentiated and adipocyte differentiated human ADAS cells from a male and a female donor were examined. Consistent with the proteomic study, the immunoblots demonstrated an induction of crystallin (heat shock protein ␤), HSP20, and HSP27 with adipogenesis by an average of 4-, 5.9-, and 2-fold, respectively (Fig. 5). In addition, HSP60, which was not detected by the mass spectroscopy analysis, displayed a 2.1-fold induction following adipogenesis. In contrast, the relative levels of the heat shock and chaperone proteins HSP47, HSP70, HSP90, and FK506-binding protein showed little or no change following induction of adipogenesis (Fig. 5); each of these proteins had been identified by the proteomic analysis in the undifferentiated ADAS cells and were not changed following adipogenesis.
The mass spectrogram of at least one HSP27 peptide (SSP 3304) suggested that the adipogenesis-induced protein might be phosphorylated on serine residue 82 (Supplemental Table  2A). Immunoblots prepared with the protein extracts prepared from undifferentiated and differentiated cells of all four donors used in the proteomic analysis were probed with antibodies detecting all forms of HSP27 and those specific for the HSP27 phosphoserines 82, 15, and 78. No evidence of the HSP27 phosphoserine 15 or 78 proteins was detected (data not shown); however, the phosphoserine 82 form of HSP27 was induced an average of 5.7-fold, and this exceeded the 1.8fold induction of total HSP27 observed in the same donors (Fig. 6). Similar studies examined the crystallin ␣B phosphorylation status on identical immunoblots (Fig. 7). The serine residues 19, 45, and 59 of the adipogenesis-induced crystallin ␣B proteins each displayed evidence of phosphorylation; upon differentiation, the levels of these phosphoproteins increased by 4.3-, 4.8-, and 3.0-fold, respectively. It should be noted that all donors displayed similar patterns of induction, although the -fold increase varied between individuals (Figs. 6 and 7). DISCUSSION Cells undergoing adipocyte differentiation exhibit well characterized morphologic changes that are reflected in the cell proteome. The current study demonstrates that adipogenesis in human ADAS cells is accompanied by modulation of five major protein categories: cytoskeleton, metabolic, heat shock protein/chaperone, redox, and protein degradation (Table I).
Cytoskeleton-The morphology of the adipocyte is significantly different from that of a fibroblast. In fact, mechanical tension, acting through the actin filament complex, can control the differentiation status of adult stromal stem cells (31). When spread out on a surface, bone marrow-derived adult stem cells formed osteoblasts, whereas, when rounded up, they committed to the adipocyte lineage (31). This process could be manipulated through the RhoA protein, a GTPase affecting the actin cytoskeleton (31). The current study demonstrates that specific proteins regulating actin polymerization (Cofilin2 and destrin) are induced in adipocytes. Previous studies have demonstrated the presence of Cofilin in adipose tissue (32). In addition, there is an induction of cytoskeletal proteins associated with the smooth muscle phenotype (transgelins and myosin light chain alkali). In contrast, individual features identified as Elfin, a protein associated with the formation of actin stress fibers in myoblasts, are both induced and reduced with adipogenesis (33). There is reduced expression of Stathmin, a tubulin polymerization protein whose absence is associated with arrest of the cell cycle and a failure to undergo mitosis (34). The reduction of Stathmin is consistent with the association of adipocyte maturation with cell cycle arrest (35).
Metabolism-The function of the mature adipocyte is to store excess energy in the form of lipid. Consequently adipogenesis in ADAS cells is accompanied by the induction of proteins associated with glycolysis and fatty acid metabolism (glyceraldehyde-3-phosphate dehydrogenase, isocitrate dehydrogenase, phosphoglycerate kinase, and pyruvate dehydrogenase). The metabolic proteins carbonic anhydrase II, fatty acid-binding protein (adipocyte), and glycerol-3-phosphate dehydrogenase were among the first genes found to be up-regulated by adipogenesis in the 3T3-L1 murine preadipocyte model (36,37). Although specific glyceraldehyde-3phosphate dehydrogenase features were induced with adipogenesis (SSP 7503 and 8518), others were reduced (SSP 5005, 6507, 6521, and 8535). Enoyl-CoA dehydrogenase displayed a similar pattern with evidence of both induced and reduced features following cell differentiation. These findings could be related to post-translational modifications.
Redox-Adipogenesis in human ADAS cells is associated with an induction of multiple proteins associated with oxidation/reduction pathways. Although mitochondrial proteins accounted for ϳ8% of the undifferentiated ADAS cell proteome,  Table 2A). Many of these same proteins have been detected in the proteome of mature 3T3-L1 adipocytes (38,39). The mature adipocyte contains an increased number of mitochondria in comparison to fibroblastic cells (40). Recent studies link mitochondrial biogenesis to the etiology of diabetes, and it is postulated that the thiazolidinediones, oral antidiabetic drugs and peroxisome proliferator-activated receptor ␥ ligands, may act in part by regulating mitochondrial formation especially in brown adipose tissue (39,41).
Heat Shock Proteins/Chaperones-The adipogenic induction of crystallin (total and serine phosphoproteins 19, 45, and 59), HSP20, HSP27 (total and serine phosphoprotein 82), and HSP60 in human ADAS cells is intriguing. The heat shock proteins serve as chaperones, controlling protein folding in the endoplasmic reticulum and their subsequent intracellular trafficking (42). There is a growing body of literature linking chaperone-like molecules to adipogenesis, obesity, and diabetes (43)(44)(45)(46). For example, adipogenesis in 3T3-L1 cells is accompanied by increased expression of the chaperone-re-lated immunophilin, FK-binding protein 51 (47). Moreover the nuclear hormone receptors that control adipogenic transcription, the glucocorticoid receptor, and the peroxisome proliferator-activated receptor, are sequestered in the cytosol as a complex with heat shock proteins HSP70 and HSP90 prior to ligand activation (42,48,49). It is interesting to note that clinical studies have linked polymorphisms in HSP70 to an increased risk for obesity and type 2 diabetes (50,51). It is postulated that obesity leads to insulin resistance and diabetes by causing endoplasmic reticulum stress (46). This stress has been found to interfere with the serine/threonine phosphorylation-mediated signal transduction pathway downstream of the insulin receptor (46). Consistent with this is the independent observation that the heat shock protein HSP27 interacts with the insulin-like growth factor receptor 1 and its signal transducer, the serine/threonine kinase protein Akt, which together modulate adipocyte metabolism (52,53). Diabetes alters the metabolism of the chaperone crystallin ␣, increasing its glycation status (43,44). In the lens of the eye, this biochemical change contributes to cataract formation; its effect in adipose tissue, if any, remains to be determined.

TABLE I Functional categories of human ADAS cell proteins modulated during adipogenesis
Bold ϭ unique to current study; italics ϭ identical or related protein identified in 3T3-L1 proteome (38,39,60,61 Phosphorylation of crystallin ␣ alters its subcellular localization and its ability to associate with an adaptor protein of the ubiquitin-protein isopeptide ligase (54). In cardiomyocytes, crystallin ␣ phosphorylation correlates with inhibition of caspase activity and protects the cell from apoptotic events (55). The current study demonstrates that adipogenesis enhances expression of these protective forms of crystallin ␣ in human ADAS cells.
Protein Degradation and Processing-The heat shock proteins/chaperones form complexes that direct translationally modified proteins to the proteasomal pathway for degradation. For example, the adipogenic transcriptional regulator, the peroxisome proliferator-activated receptor ␥, is targeted to the proteasome by ubiquitinylation and sumoylation in 3T3-L1 cells (56 -58). Consistent with this is the current observation that human ADAS cell adipogenesis is associated with induction of a ubiquitin-conjugating enzyme (Table II and  Supplemental Table 2A). In addition, the ubiquitin-like protein SMT3A or SUMO2 is present in ADAS cells (Supplemental Table 2C). These findings suggest that adipogenesis involves selective modifications of the protein processing and degradation pathways.
Contamination-The current study provides a composite profile of the proteome of undifferentiated human ADAS cells obtained from multiple donors. Overall the results confirm and extend previous findings in the literature; however, not all of the proteins identified derive from ADAS cells and may be qualified as contaminants. The most striking examples are hemoglobins A and B, reported in the undifferentiated cell lysates. The presence of hemoglobin reflects the cell isolation procedure. Specifically adipose tissue was digested with collagenase and centrifuged, and the pelleted SVF was placed in culture. The SVF population routinely contains erythrocytes that are washed away after a 48-h period during which the ADAS cells adhere to the plastic surface. The ADAS cells were then passaged once prior to their harvest as cell lysates for the proteomic analysis. Previous flow cytometric studies from our laboratory (14,15) 2 have demonstrated that nucleated hematopoietic cells, expressing the marker CD45, account for less than 1% of the ADAS cell population by Passage 1. Nevertheless it is feasible that some hemoglobin remained bound to the adherent ADAS cell population either as intact erythrocytes or as membrane-bound protein. Alternatively the fetal bovine serum used in the cell culture may have been the source of both the hemoglobin and albumin proteins detected in this analysis.
Comparison to Published Proteomes-The current description of the human ADAS cell proteome in the undifferentiated and differentiated states complements published findings from related murine and human systems (Table II). Proteomic analyses of total cell lysates from murine 3T3-L1 adipocytes have identified between eight and 100 protein features by one-and two-dimensional gel electrophoresis/mass spectroscopy (38,39,59,60). The combined proteome of the undifferentiated and differentiated ADAS cells was identical or similar to Ͼ40% of the proteins identified in these studies. Of note is the fact that some proteins associated exclusively with the differentiated 3T3-L1 adipocytes, such as fatty acid-binding protein (adipocyte), were also detected in the undifferentiated human ADAS cells; nevertheless the protein level increased further upon adipocyte differentiation in the ADAS cells. The fact that the human ADAS cells represent a primary cell population rather than a cell line may account for this discrepancy between systems. Individual publications have focused on the secreted proteins (61) or lipid droplet-associated proteins (59) in 3T3-L1 adipocytes. Because the current study focused on total cell lysates, it is not surprising that only four of the 20 secreted proteins or seven of the 30 lipid droplet proteins were detected in the human ADAS cell proteome (59,61). Further studies examining the secretosome and subcellular fractions of the human ADAS adipocytes are warranted. Proteomic analyses of murine white adipose tissue have identified up to 27 individual protein features that are modified by high fat diet, insulin treatment, or leptin deficiency (62)(63)(64). The ADAS cell proteome included a high percentage of identical or similar proteins (20 -86%).
A single publication has reported the proteome of the caveolae and mitochondrial/nuclear fractions from human adipocytes (65). This work identified ␤-actin, a number of annexins, G protein subunits, F 1 ATPase subunits, and heat shock proteins in common with the current study. The human ADAS cells are known to resemble stromal cells isolated from the bone marrow at the morphologic and differentiation levels (15,66,67). This was reflected at the proteome level where the human ADAS cells displayed Ͼ50% homology with the proteome of human bone marrow stromal cells as well as dermally and synovially derived fibroblasts (68 -71). There is recent evidence that adipocytes and macrophages show similar gene expression profiles; both induce their expression of fatty acid-binding protein (adipocyte) in response to peroxisome proliferator-activated receptor ligands (72). Consistent with this, the ADAS cell proteome displayed between 47 and 64% homology with that of lipopolysaccharide-or phorbol myristic acid-activated human macrophages (73,74).
In conclusion, the current study has characterized the shared proteomic features of undifferentiated and adipocyte differentiated human ADAS cells isolated from both female and male donors. Because gender is known to influence the metabolic characteristics of adipose tissue (75), it will be necessary to compare the ADAS cell proteome between female and male donors. In addition, further proteomic analyses will need to focus on specific subcellular fractions from human ADAS adipocytes. The model can be expanded to explore other lineage pathways because the human ADAS cells are known to differentiate into chondrocytic, osteoblastic, and neuronal phenotypes under appropriate culture conditions (12,13). * This work was supported by the Pennington Biomedical Research Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.