Introduction

The incidence of obesity in Western society is rapidly increasing (1, 2). An increase in adipose mass is associated with a shorter life span, type 2 diabetes, sensitivity toward septic shock, hypertension, and cancer (3, 4, 5, 6, 7, 8). Overfeeding studies with monozygotic twins have predicted the presence of important genetic determinants in weight regulation (9). In contrast to obesity of the subcutaneous compartment, obesity of the omental compartment is associated with a higher risk for obesity-related illnesses (10, 11, 12, 13, 14). Although genetically determined, the mechanism behind the preference to store energy in the omental or the subcutaneous compartment is poorly understood (15, 16, 17, 18, 19).

An increase in adipose mass may be the result of the recruitment of new adipocytes (adipogenesis) or an increase in the volume of existing adipocytes or an attenuated apoptosis (20, 21, 22, 23, 24). Within the stromal vascular cell (SVC)¹ fraction of the adipose tissue, there are cells present that possess the capability of becoming adipocytes in vitro and in vivo (25, 26, 27, 28). The high adipogenic potential of these SVCs could be an important determinant in fat mass regulation and obesity-associated disorders (29, 30, 31).

Studies in rats and humans indicate that the ability of preadipocytes to become adipocyte declines with aging (25, 32). In addition, preadipocytes from different fat stores have been described to have different intrinsic rates of adipogenesis and apoptosis (23, 33). However, there is no consensus on these issues (34, 35).

In contrast to studies on functional aspects of preadipocytes, no study has focused on the assessment of preadipocyte number in adipose tissue, most likely because the isolation of preadipocytes, first described by Hauner et al., is intricate and error prone (25). Here, experimental data are presented that indicate that the assessment of preadipocyte number and function is strongly dependent on the isolation and differentiation protocol used. Despite our standardized isolation and differentiation protocol, large differences in preadipocyte number were observed in obese to severely obese subjects.

Research Methods and Procedures

Cell Culture

For isolation of SVCs, subcutaneous and omental adipose tissues were obtained during gastric restriction surgery. All participants in the study gave permission in writing, and the study was approved by the local ethics committee of the Academic Hospital Maastricht. Subject inclusion criteria were as follows: BMI > 40 kg/m² and, in the case of severe comorbid conditions, BMI > 35 kg/m². Subject exclusion criteria were as follows: severe cardiopulmonary pathology (American Society of Anesthesiologists class 3); history of bariatric surgery; manifest psychopathology; <18 years or >60 years of age; and >50 years of age when subject underwent previous upper-abdominal surgery. All women in the study had regular menstrual periods. In Figure 1, the distribution of BMI and age in obese to severely obese subjects is presented. Adipose tissue was collected in phosphate-buffered saline (NBPI, Oss, The Netherlands) and cut into 3 mm 3 mm pieces with scissors. The 3 mm 3 mm pieces were further processed with a scalpel. Next, the pieces of adipose tissue were digested in Dulbecco's modified Eagle's medium (DMEM)-high glucose (Invitrogen Life Technologies, Carlsbad, CA) containing 4% bovine serum albumin (ICN, Aurora, OH) and 2 mg/mL collagenase II (C6685; Sigma, Steinheim, Germany). No relation was observed between lots (99,103,107) of collagenase and recovery of SVCs. Adipose tissue (1 to 2.5 grams) was digested in 5 mL of the above-described solution. Digestion was performed at 37 °C on a shaking platform (200 rpm) for 1 to 3.5 hours. Next, 5 mL of digest was transferred to a 5-mL syringe and gently pressed over a 500-m sterile pore-size disposable nylon mesh (ITK Diagnostics, Uithoorn, The Netherlands). This procedure was repeated until the complete digest was filtered. Depending on the origin of the tissue and the digestion time with collagenase, more than one filter was used. SVCs were separated from adipose cells by centrifugation (1 minute, 170g) in a standard table centrifuge. Adipose cells were removed, and the SVCs were precipitated by centrifugation (5 minutes, 350g; Hettich, Tuttlingen, Germany). Red blood cells were lysed by resuspending the cell pellet in 10 mL of red cell lysing solution (154 mM NH₄CL, 10 mM KHCO₃, and 0.1 mM EDTA). After 5 minutes, SVCs were spun down (5 minutes, 350g) and resuspended in DMEM/F12 containing 10% fetal calf serum (FCS) (Bodinko, Alkmaar, The Netherlands), 2 mM glutamine (Invitrogen Life Technologies), 100 IU/mL penicillin, and 100 g/mL streptomycin (Roche, Mannheim, Germany); this mixture is referred to as complete medium (CM). SVCs were counted using an electronic counter (Beckman Coulter, Fullerton, CA), and viability was assessed with tryptan blue exclusion. SVCs (1 10⁶/mL) were serial diluted (1 to 1024) in CM. Two hundred microliters of diluted SVC suspension was plated in two 96-well tissue culture dishes. After 48 hours, one 96-well plate was used to determine the number of cells attached to the 96-well dish, whereas the other plates were used for differentiation studies.

Figure 1.

Characteristics of the study population. Twenty-seven obese individuals (18 women and 9 men) were selected for this study. (A) Distribution of BMI (44 10). (B) Distribution of age (40 9).

Full figure and legend (67K)

Determination of the Number of Plastic Adherent SVCs (aSVCs)

Two hundred microliters of culture medium was aspired from the wells, and the cells were fixed in 100 L of DMEM/F12 with 3.7% formaldehyde (Merck Research Laboratories, Rahway, NJ) for 10 minutes at room temperature (RT). The fixative was discarded, and aSVCs were washed once with MilliQ (Millipore, Billerica, MA) and stained with 100 L of Giemsa (0.25% Giemsa in 4 mM Na-phosphate buffer, pH 7.0). The working solution was filtered with a 0.22-m filter. The cells were stained overnight at RT. Next, aSVCs were washed once with MilliQ, and the number of aSVCs was determined by counting eight entire wells per serial dilution. The number of aSVCs was determined in three serial dilutions by counting the entire well. Within the omental aSVC fraction endothelial, cells were present as determined by -CD31 staining (36, 37). The number of endothelial cells never exceeded 0.4% of the cells present. No endothelial cells were observed in the subcutaneous aSVC fraction.

Induction of Adipogenesis and the Determination of Fat Accumulation

Further serial diluted aSVCs (see above, under "Cell Culture") were used for induction of adipogenesis. After 48 hours, the medium was replaced by differentiation medium consisting of CM (instead of 10% FCS, the concentration of FCS varied in some experiments) plus 15 mM NaHCO₃ (Merck), 15 mM HEPES (Merck), 33 M biotin (ICN), 17 M panthothenate (ICN), 200 pM T3 (ICN), 1 M dexamethasone (ICN), 500 nM insulin (Roche), 4 g/mL transferrin (Invitrogen Life Technologies), and 10 M carbacyclin (cPGI₂) (Biomol, Plymouth Meeting, PA) in the presence 2% FCS. In some experiments, cPGI₂ was replaced by Wy14643 (Biomol), ciglitazone (Biomol), L783983, or L805645 (a kind gift of Dr. J. Berger, Merck) (38). The cells were cultured for 20 days with culture medium replaced every 4 days. After 20 days, the triacylglycerol and Oil Red O extraction was performed.

The Extraction of Oil Red O from aSVC Cultures

Cells were fixed in 100 L of 3.7% formaldehyde in DMEM/F12 for 10 minutes at RT. The cells were washed once with 200 L of H₂0 and once with 200 L of 70% ethanol followed by an incubation with 50 L of Oil Red O solution for 30 minutes at RT (39). Thereafter, wells were washed 8 times with 200 L of 70% ethanol, and Oil Red O was extracted by adding 100 L of dimethyl sulfoxide (Merck) and shaking for 1 minute on a mirotitertek plate shaker, Titramax (Salm en Kip, Breukelen, The Netherlands). The A₅₄₀ was determined using a standard microtiter reader (Bio-Rad, Hercules, CA). A calibration curve of Oil Red O in dimethyl sulfoxide was used as a reference.

Determination of Leptin

Every 4 days, culture medium of aSVC culture was collected. In this culture medium, the amount of leptin was measured with a standard sandwich enzyme-linked immunosorbent assay (40).

Statistical Analysis

Statistical analyses were performed with the SPSS/Mac statistical program (version 8.0 for Macintosh; SPSS, Inc., Chicago, IL). Data in Figures 2 and 3 were analyzed using a standard variance analysis. Differences were considered significant at p < 0.05. The data in Figures 4, and 5 were analyzed with two-tailed Spearman Rank statistics. Significant levels were corrected for multiple testing using the Bonferoni correction and null hypothesis (no association between the variables studied) rejected as p < 0.005.

Figure 2.

Determination of the variability in preadipocyte isolation. Subcutaneous and omental adipose tissues (10 g) from two obese individuals were digested with collagenase for up to 3.5 hours. After 1, 2, and 3.5 hours, SVCs were isolated, and the viability (A) and cell number (B) were determined. Error bars represent SD, n = 3. SVCs were serial diluted and plated in a 96-well plate. (C) After 48 hours, attached cells were fixed, stained, and the number of aSVCs per gram of tissue was determined. Error bars represent SD, n = 8. (D) subcutaneous and omental adipose tissues (0.5, 10, and 20 g) of an obese individual were digested with collagenase for 2 hours. SVCs were isolated and plated in 96-well plates. After 48 hours of settling, the cells were fixed and stained with Giemsa, and the aSVC number was determined. Error bars represent SD, n = 8. (E) Six independent subcutaneous biopsies (10 g each) and two independent omental biopsies (10 g each) were digested with collagenase for 2 hours. SVCS were isolated and plated in a 96-well plate. After 48 hours of settling, cells were fixed and stained with Giemsa, and the aSVCs were determined. Error bars represent SD, n = 8.

Full figure and legend (77K)

Figure 4.

Determination of preadipocyte number and function in 27 obese individuals. Subcutaneous and omental adipose tissue from the abdominal region was obtained during gastric surgery. Ten grams of adipose tissue was digested with collagenase for 2 hours. SVCs were isolated, their viability and number were determined, and they were serial diluted in 96-well plates. After 48 hours of settling, one 96-well plate was used to determine the number of aSVCs. The aSVCs were fixed, stained with Giemsa, and the cell number was determined. The second 96-well plate was used to determine leptin production and fat accumulation. aSVCs were differentiated for 20 days. Every 4 days, culture medium was changed. After 20 days, cell cultures were fixed and stained with Oil Red O. The amount of Oil Red O (oro) was determined, as was the amount of leptin in the culture from day 16 to 20. The following Spearman Rank Correlation Coefficients were determined: the number of SVCs isolated from the subcutaneous and the omental adipose tissue (A), the number of aSVCs isolated from the subcutaneous and the omental adipose tissue (B), the number of aSVCs and age (C), the number of aSVCs and BMI (D), the leptin production for 4 days (day 16 to 20) by the subcutaneous- and the omental-derived aSVCs (E), and the Oil Red O accumulation in 20 days of the subcutaneous- and the omental-derived aSVCs (F). (Figure 4 is continued on the next page.)

Full figure and legend (88K)

Figure 5.

Relationship between preadipocyte number and leptin production. The number of preadipocytes from subcutaneous and omental adipose tissue and the leptin production were determined as described in Figure 4. The Spearman rank correlation coefficients for the associations are displayed in the graph.

Full figure and legend (62K)

Results

Isolation of SVCs: Determination of the Sampling Error

The isolation method of preadipocytes is labor intensive and error prone. Therefore, we determined the sampling error of the preadipocyte isolation procedure. Three variables were considered important: the incubation time with collagenase, the size of the biopsy, and the place of biopsy. In Figure 2, the effect of collagenase incubation time on cell survival and cell recovery is shown. Subcutaneous and omental adipose tissue from two donors was digested with collagenase as described (see "Research Methods and Procedures") for up to 3.5 hours. After 1, 2, and 3.5 hours, the collagenase digest was separated into an adipocyte-containing fraction and a preadipocyte-containing fraction referred to as SVC. After red blood cell lysis, the viability and the number of cells was determined. Cell viability always exceeded 70%, but was dependent on the collagenase incubation time (Figure 2A). Optimal viability of the SVC population was obtained after 2 hours of collagenase digestion. The number of SVCs was determined, as described in "Research Methods and Procedures," and expressed as the number of SVCs per gram of adipose tissue (Figure 2B). The incubation time with collagenase influenced the number of SVCs per gram of adipose tissue. The variation in SVCs per gram of adipose tissue during the total incubation time did not exceed a factor 2. Next, we determined the number of SVC cells able to attach to culture disks. To this end, serial dilutions of SVCs were plated in 96-well plates, and after 48 hours, aSVCs were counted (Figure 2C). The influence of digestion time on the number of aSVCs per gram of adipose tissue is shown. The variation due to collagenase treatment was large. The lowest variation was found at 2 hours of incubation. This was chosen as the optimal incubation time. The size of the biopsy was also found in pilot experiments to be an important variable. Therefore, we determined the variation in different biopsy sizes: 0.5, 1, 10, and 20 grams of adipose tissue (Figure 2D). From these experiments, it became apparent that the estimation of the number of aSVCs per gram of adipose tissue from small biopsies was significantly lower in comparison with large biopsies (p = 0.0047). No significant difference was observed between 10 and 20 grams of tissue (p = 0.46) or between 0.5 and 1 gram of tissue (p = 0.086). The influence of the site of biopsy was also determined (Figure 2E). From six independent subcutaneous and two independent omental biopsies (10 grams of tissue), aSVC number was assessed. Although the number of aSVCs per gram of adipose tissue was dependent on the relative site of the biopsy, variation was found to be small (subcutaneous, 6.2 2.3 10⁴ aSVC/g; omental, 5.7 1.2 10⁵ aSVC/g). Taken together, these data indicate that experimental variation of the isolation procedure is mainly dependent on the size of biopsy and to a lesser extent on the collagenase digestion time. From these data, we concluded that usage of 10-gram biopsies and a digestion time of 2 hours reduced experimental variation considerably.

aSVCs Consist of Preadipocyte Subpopulations

To ascertain whether the aSVCs were indeed preadipocytes, differentiation was induced, and fat accumulation and leptin production were determined. The serum-free protocol of Hauner et al. was taken as a starting point (25). Although adipogenesis occurred, the efficiency of fat accumulation and leptin production was low using this protocol (Figure 3 A and B). To optimize the differentiation protocol, we tested different peroxisome proliferator-activated receptor (PPAR) ligands. One of the PPAR family members, PPAR2, is obligatory for adipogenesis, whereas other PPAR members, such as PPAR and PPAR, have been described as involved in adipogenesis. From the PPAR ligands tested, cPGI₂ (PPAR and PPAR) gave the best adipogenesis, as measured by fat storage and leptin production. The PPAR ligand ciglitazone and the PPAR ligand Wy14624 were ineffective in inducing fat storage and leptin production.

Figure 3.

Determination of the variability in the differentiation protocol. The influence of FCS and PPAR ligands on the amount of fat accumulation (A) and the leptin production (B) was determined. aSVCs were seeded confluent and differentiated for 20 days in the presence of different concentration of FCS and PPAR ligands. On day 20, cells were fixed and stained with Oil Red O (oro), and the amount of Oil Red O present in the cultures and the amount of leptin were measured. Error bars represent SD, n = 8. W, Wy14643; C, ciglitazone. The effect of insulin and cPGI₂ on the leptin production in time (C). aSVCs were seeded confluent and differentiated for 20 days. After 4 days, culture medium was changed, and the amount of leptin was determined. Each graph represents one time-point; on the x axis is the concentration of cPGI₂ (0, 0.1, 1, and 10 M), and on the z axis is the concentration of insulin (0, 5, and 500 nM). Error bars represent SD, n = 3. Effect of cell number on the fat accumulation in time of subcutaneous preadipocytes (D) and omental preadipocytes (E). aSVCs were seeded at different densities (1: 10,000 cells/well; 2: 5000 cells/well; 4: 2500 cells/well; 8: 1250 cells/well; 16: 625 cells/well; 32: 312 cells/well; 64: 156 cells/well). After reaching a confluent stage, cells were differentiated. Cell cultures were fixed and stained with Oil Red O on days 0, 7, 11, and 18. Error bars represent SD, n = 8. Effect of PPAR ligands on fat accumulation of subcutaneous (F) and omental preadipocytes (G). aSVCs were seeded confluent and differentiated in the presence of three PPAR ligands. After 20 days of differentiation, cells were fixed and stained with Oil Red O. Error bars represent SD, n = 8. DMSO, dimethyl sulfoxide. (Figure 3 is continued on the next page.)

Full figure and legend (124K)

No effect of ciglitazone and Wy14624 on the cPGI₂-induced fat storage and leptin production was observed. Culturing of aSVCs under serum-free conditions often resulted in clustering of aSVCs. The addition of low amount of serum (1% FCS) inhibited the clustering completely (data not shown). In Figure 3 A and B, the effect of serum on the PPAR ligand-induced fat accumulation and leptin production is shown. Although 2% FCS gave optimal fat storage, leptin production was highest at 8% FCS.

The kinetics of leptin production clearly showed that adipocytes were formed during differentiation (Figure 3C). Leptin production was not detectable on day 8 of differentiation. On day 20, leptin production had become cPGI₂ and insulin dependent, a characteristic of mature adipocytes.

aSVCs cultured in the presence of serum have been described as losing their ability to become adipocytes (25). This loss of adipogenic ability could be the result of selection of non-preadipocytes, such as fibroblast, mesothelial, or endothelial cells, that are known for contamination of the SVC fraction (33). To test whether fat accumulation in our culture system was dependent on cell density, subcutaneous and omental aSVCs (10,000 cells/well) were serially diluted and allowed to grow to confluence. Next, confluent aSVC cultures were induced to undergo adipogenesis (see "Research Methods and Procedures"), and the kinetics of fat accumulation were studied (Figure 3 D and E). Nonconfluent plated aSVCs had a significantly reduced ability to store fat. Although the kinetics of fat accumulation seemed to differ for subcutaneous and omental aSVC, seeding subcutaneous aSVC at one-half the cell density, resulted in a significantly lower fat accumulation per well on day 18 (p = 0.03). For omental aSVCs, seeding 8 times fewer cells resulted in a significant reduction in fat storage per well on day 18 (p = 0.005). From these data, we inferred that estimation of the number of fat-storing cells could best be performed with 10,000 cells per 0.32 cm² in the presence of 10 M cPGI₂ and 2% FCS.

Differentiation experiments under these conditions with aSVCs from different donors showed a high variability in fat accumulation, indicating that not all aSVCs were preadipocytes or that aSVCs from different donors had different capabilities to store fat. Comparison of the fat accumulation of two new PPAR ligands, the L783983 (PPAR/PPAR) and the L805645 (PPAR), indicated that fat accumulation was strongly dependent on the PPAR ligand present during differentiation (Figure 3 F and G). From four obese individuals, the aSVCs from the subcutaneous and omental adipose depots were obtained, and fat accumulation was determined. Interestingly, fat accumulation during differentiation differed per donor for the three ligands. In the presence of L805645, fat accumulation was present in >90% of the cells (data not shown), indicating that most of the cells were, in fact, preadipocytes, but that the efficiency of fat accumulation was dependent on the signaling pathway triggered.

aSVC Numbers of Subcutaneous and Omental Adipose Tissue Are Related

Having obtained an insight into the procedural consequences of the variation in preadipocyte isolation, we determined the SVC number and the aSVC number in subcutaneous and omental adipose tissue of each individual studied (Figure 4 A and B). Interestingly, large differences (up to 100-fold) in SVC and aSVC number were observed in the subcutaneous and the omental adipose tissue. Plotting the number of SVCs per gram of the subcutaneous adipose tissue against the number of SVCs per gram of the omental adipose tissue of each individual showed a distinct association between SVC number per gram of adipose tissue in both compartments (r = 0.69, p = 0.002, n = 17). Comparable results were obtained for aSVC number per gram of adipose tissue for the two compartments (r = 0.68, p = 0.003, n = 17). Therefore, it seems that part of the SVC consisted of aSVC and that both the number of SVCs and aSVCs were determined independently from the adipose tissue from which they were isolated. The large differences in the number of aSVCs between individuals could be a reflection of the weight or the age of the donors from which the cells were derived. However, no association between BMI (aSVC_sc: r = 0.23, n = 24, p = 0.30; aSVC_om: r = 0.44, n = 19, p = 0.6; where aSVC_sc and aSVC_om represent aSVC(s) from subcutaneous adipose and omental adipose tissues, respectively) or age (aSVC_sc: r = -0.36, n = 24, p = 0.8; aSVC_om: r = -0.25, n = 19, p = 0.31) and the number of aSVCs was observed (Figure 4 C and D).

Tight Correlation of Leptin Production, but Not Fat Accumulation, in Omental and Subcutaneous aSVCs

To determine the commitment to adipogenesis, we measured the leptin secretion of the aSVCs during adipogenesis (Figure 4E). Although the leptin production varied among donors, the amount of leptin produced by the subcutaneous and the omental aSVCs were comparable for each individual (r = 0.832, n = 13, p = 0.001). However, leptin production was not associated with BMI (aSVC_sc: r = -0.19, n = 19, p = 0.45; aSVC_om: r = -0.34, n = 14, p = 0.23) or age (aSVC_sc: r = -0.43, n = 19, p = 0.07' aSVC_om: r = 0.11, n = 14, p = 0.71). In contrast, we observed large differences in the fat accumulation between omental and subcutaneous fat depots and between donors (Figure 4F). Interestingly, for the omental aSVCs, a negative association was found between the ability to store fat and the BMI, indicating that in the omental adipose tissue, the cPGI₂ responsiveness declined (aSVC_sc: r = -0.019, n = 17, p = 0.94; aSVC_om: r = -0.79, n = 14, p = 0.001) with increasing BMI.

Fat Accumulation and Leptin Production Are Uncoupled in cPGI₂-Mediated Differentiation

Leptin production during cPGI₂-induced differentiation showed a clear relationship between the aSVCs from the subcutaneous and omental adipose tissue. However, as discussed above, no such relationship was observed between omental and subcutaneous preadipocytes. Taking a closer look at the fat accumulation and the leptin production during differentiation revealed that leptin production and fat accumulation are uncoupled. Cultures producing high amounts of leptin showed no or little sign of fat accumulation (data not shown). This discrepancy between cPGI₂-induced leptin production and fat accumulation was also observed in the FCS titration experiment (Figure 3). Although leptin production increased with increasing concentration of FCS (Figure 3B), the amount of fat showed a reversed U-shaped curve (Figure 3A). Fat accumulation was maximal at 2% FCS. High levels of FCS (8%) resulted in a decreased storage of fat.

Leptin Production, by Differentiating Subcutaneous and Omental aSVCs, Is Inversely Associated with the Number of Subcutaneous and Omental aSVCs per Gram of Fat Tissue

To test whether or not there was a relation between leptin production and the number of aSVCs per gram of adipose tissue, we determined the Spearman rank correlation coefficient. In obese to severely obese subjects, a negative association between the leptin production by adipocytes and preadipocyte number per gram of adipose tissue was observed (Figure 5; subcutaneous adipose store: r = -0.77, n = 19, p < 0.001; omental adipose store: r = -0.87, n = 14, p < 0.001).

Discussion

The main purpose of this study was the assessment of preadipocyte number and function in the subcutaneous and omental adipose tissue of obese subjects. To this end, the reproducibility of the standard preadipocyte isolation protocol as described by Hauner et al. was determined (25). Important variables in the assessment of preadipocyte number were collagenase digestion time and size of the biopsy. Although viability of the SVCs always exceeded 70%, it was dependent on the duration of the collagenase digestion. Highest viability was observed after a 2-hour digestion, whereas at this time-point the lowest numbers of cells were obtained. One way to explain this observation is that on longer digestion, damaged or dead cells, initially present in the 1-hour digest, are lost and no longer present in the 2- and 3-hour digests.

Our techniques allow only detection of cells that survive the isolation and culture. In principle, this may lead to underestimation of preadipocytes in various tissues. Current technologies, however, do not allow better estimation of preadipocytes. When SVCs were seeded in culture disks to determine the number of aSVC, 10% of the SVCs was recovered. Coating the plates with extracellular proteins, like fibronectin, laminin, and pronectin F (data not shown), did not influence the large discrepancy between SVC and aSVC number. The observation that the number of SVCs and aSVCs for both fat depots are associated (subcutaneous: r = 0.68, N = 17, p = 0.002; omental: r = 0.56, N = 17, p = 0.016) suggests that the aSVCs and nonadherent SVCs are a fixed fraction of the SVC population. However, at the present time, we are not able to exclude the presence of preadipocytes in the nonadherent SVC cell fraction.

Previous studies demonstrated that the velocity of fat storage declines with age (25, 32, 41). In addition, it was shown that preadipocytes from different depots differed in their velocity to store fat (33). In contrast, experiments performed by others showed no differences in the extent of adipogenesis between omental and subcutaneous preadipocytes (35). Also, others observed no effect of age on fat cell differentiation (34). The relative narrow range of age of the subjects in our study makes it difficult to unequivocally establish a relationship between age and fat storage velocity. However, the described contradicting observations could be the result of differences in the differentiation protocol applied. Indeed, this study shows that variations in experimental procedure may result in large variations of preadipocyte function. Preadipocyte function was dependent on PPAR ligand used, amount of serum present, and cell density. It has been reported before that the efficiency of adipogenesis is related to cell density and the presence of serum proteins (25, 42, 43). This observation is of interest because most of the functional studies mentioned are performed with preadipocytes expanded in serum-containing medium for various time spans. To exclude the possibility of loss of phenotype or selection of other cell types present in our SVC fraction, cells were plated at high cell density (confluent). Because the PPAR ligand L805645 committed >90% of the cells to fat accumulation, it can be concluded that the contamination with other cells was low.

Little is known about the mechanism that regulates preadipocyte cell numbers in vivo. The observed association between the number of SVCs and aSVCs in two adipose depots suggest that cell numbers in the two depots are determined at a systemic and/or genetic level. The observation that large differences in SVC and aSVC number are independent of BMI also indicates that preadipocyte numbers are regulated at a systemic and/or genetic level.

From in vitro mouse studies, it has become clear that the adipokine leptin is associated with fat storage strategy. Hypertrophic fat storage has been linked with low leptin production, and hyperplastic fat storage has been linked with high leptin production (44). Hyperplastic fat storage is likely to lead to depletion of preadipocytes, which is in contrast to hypotrophic fat storage. These observations could be explained by assuming that preadipocytes from tissues with hyperplastic fat storage, characterized by low preadipocyte numbers per gram of fat tissue, show a high rate of adipocyte differentiation and, thus, high leptin production. Preadipocytes from tissues with hypertrophic fat storage, characterized by high preadipocyte number per gram of fat tissue, show a low rate of adipocyte differentiation and, thus, low leptin production. Interestingly, in our study population, such a negative association between the leptin production by adipocytes and preadipocyte number per gram of adipose tissue was observed.

If the fat storage strategy, hypertrophic or hyperplasticis associated with leptin production, no association between the amount of fat stored and the amount of the leptin produced is to be expected. Our data showed no relationship between cPGI₂-induced leptin production and fat accumulation for subcutaneous (N = 13, r = 0.134, n = 13, p = 0.65) and omental preadipocytes (r = 0.68, n = 9, p = 0.04). In addition, the cPGI₂-dependent adipogenesis resulted in leptin production independent of the fat storage in two additional experimental settings—the FCS titration experiment (Figure 2B) and the leptin production in time (Figure 3C).

In conclusion, we found that the assessment of SVCs, aSVCs, and preadipocyte number is strongly dependent on the isolation and differentiation protocol used. Using our standardized isolation and differentiation methods, large differences in the number of SVCs, aSVCs, and preadipocytes were still observed. Interestingly, in individuals, the number of SVCs, aSVCs, and preadipocytes in two adipose depots were associated. In obese to severely obese subjects, BMI and age could not explain the large differences in SVCs, aSVCs, and preadipocyte number.

Notes

¹ Nonstandard abbreviations: SVC, stromal vascular cell; DMEM, Dulbecco's modified Eagle's medium; FCS, fetal calf serum; CM, complete medium; aSVC, adherent stromal vascular cell; RT, room temperature; PPAR, peroxisome proliferator-activated receptor; cPGI₂, carbacyclin; aSVC_sc, adherent stromal vascular cell(s) from subcutaneous adipose tissue; aSVC_om, adherent stromal vascular cell(s) from omental adipose tissue.

References

1. Björntorp, P. (1996) New curse of the urbanized world: obesity: a still increasing epidemic. Lakartidningen. 93: 4637–4640.
2. McLellan, F. (2002) Obesity rising to alarming levels around the world. Lancet. 359: 1412 | Article | PubMed | ISI |
3. Calle, E. E., Thun, M. J., Petrelli, J. M., et al (1999) Body-mass index and mortality in a prospective cohort of U.S. adults. N Engl J Med. 341: 1097–1105. | Article | PubMed | ISI | ChemPort |
4. Willett, W. C., Dietz, W. H., Colditz, GA. (1999) Guidelines for healthy weight. N Engl J Med. 341: 427–434. | Article | PubMed | ISI | ChemPort |
5. Elliott, P., Shipley, M. J., Jarrett, RJ. (1988) Obesity and hypertension. Lancet. 1: 995
6. Jackson, IJ. (1999) The mahogany mouse mutation: further links between pigmentation, obesity and the immune system. Trends Genet. 15: 429–431. | Article | PubMed | ISI | ChemPort |
7. Edelman, R. (1981) Obesity: does it modulate infectious disease and immunity? Prog Clin Biol Res. 67: 327–337.
8. Bianchini, F., Kaaks, R., Vainio, H. (2002) Overweight, obesity, and cancer risk. Lancet Oncol. 3: 565–574. | Article | PubMed | ISI |
9. Bouchard, C., Tremblay, A., Despres, J. P., et al (1990) The response to long-term overfeeding in identical twins. N Engl J Med. 322: 1477–1482. | PubMed | ISI | ChemPort |
10. Despres, JP. (2001) Health consequences of visceral obesity. Ann Med. 33: 534–541. | PubMed | ISI | ChemPort |
11. Björntorp, P. (1992) Regional fat distribution: implications for type II diabetes. Int J Obes Relat Metab Disord. 16(Suppl 4): S19–S27. | PubMed |
12. Janjic, D. (1996) Android-type obesity and gynecoid-type obesity. Schweiz Rundsch Med Prax. 85: 1578–1583.
13. Jensen, MD. (1997) Health consequences of fat distribution. Horm Res. 48: 88–92.
14. Frayn, KN. (2000) Visceral fat and insulin resistance: causative or correlative? Br J Nutr. 83(Suppl 1): S71–S77. | PubMed | ISI | ChemPort |
15. Bouchard, C., Perusse, L., Leblanc, C., et al (1988) Inheritance of the amount and distribution of human body fat. Int J Obes. 12: 205–215. | PubMed |
16. Perusse, L., Rice, T., Chagnon, Y. C., et al (2001) A genome-wide scan for abdominal fat assessed by computed tomography in the Quebec Family Study. Diabetes. 50: 614–621. | Article | PubMed | ISI | ChemPort |
17. Brockmann, G. A., Kratzsch, J., Haley, C. S., et al (2000) Single QTL effects, epistasis, and pleiotropy account for two-thirds of the phenotypic F(2) variance of growth and obesity in DU6i DBA/2 mice. Genome Res. 10: 1941–1957. | Article | PubMed | ISI | ChemPort |
18. Chagnon, Y. C., Perusse, L., Weisnagel, S. J., et al (2000) The human obesity gene map: the 1999 update. Obes Res. 8: 89–117. | PubMed | ChemPort |
19. Joyner, J. M., Hutley, L. J., Cameron, DP. (2000) Glucocorticoid receptors in human preadipocytes: regional and gender differences. J Endocrinol. 166: 145–152. | PubMed |
20. Prins, J. B., O'Rahilly, S. (1997) Regulation of adipose cell number in man. Clin Sci. 92: 3–11. | PubMed | ISI | ChemPort |
21. Ailhaud, G., Grimaldi, P., Negrel, R. (1992) Cellular and molecular aspects of adipose tissue development. Annu Rev Nutr. 12: 207–233. | Article | PubMed | ISI | ChemPort |
22. Faust, I. M., Johnson, P. R., Stern, J. S., et al (1978) Diet-induced adipocyte number increase in adult rats: a new model of obesity. Am J Physiol. 235: E279–E286.
23. Niesler, C. U., Siddle, K., Prins, JB. (1998) Human preadipocytes display a depot-specific susceptibility to apoptosis. Diabetes. 47: 1365–1368. | PubMed |
24. Sorisky, A., Magun, R., Gagnon, AM. (2000) Adipose cell apoptosis: death in the energy depot. Int J Obes Relat Metab Disord. 24(Suppl 4): S3–S7. | Article | PubMed | ChemPort |
25. Hauner, H., Entenmann, G., Wabitsch, M., et al (1989) Promoting effect of glucocorticoids on the differentiation of human adipocyte precursor cells cultured in a chemically defined medium. J Clin Invest. 84: 1663–1670. | PubMed | ISI | ChemPort |
26. Deslex, S., Negrel, R., Vannier, C., et al (1987) Differentiation of human adipocyte precursors in a chemically defined serum-free medium. Int J Obes. 11: 19–27.
27. Akazawa, S., Sun, F., Ito, M., et al (2000) Efficacy of troglitazone on body fat distribution in type 2 diabetes. Diabetes Care. 23: 1067–1071.
28. Halvorsen, Y. D., Bond, A., Sen, A., et al (2001) Thiazolidinediones and glucocorticoids synergistically induce differentiation of human adipose tissue stromal cells: biochemical, cellular, and molecular analysis. Metabolism. 50: 407–413. | Article | PubMed | ISI | ChemPort |
29. Stumvoll, M., Haring, HU. (2002) Glitazones: clinical effects and molecular mechanisms. Ann Med. 34: 217–224. | PubMed | ISI | ChemPort |
30. Blickle, JF. (2001) Thiazolidinediones: clinical data and perspectives. Diabetes Metab. 27: 279–285.
31. Hallakou, S., Doare, L., Foufelle, F., et al (1997) Pioglitazone induces in vivo adipocyte differentiation in the obese Zucker fa/fa rat. Diabetes. 46: 1393–1399. | Article | PubMed | ISI | ChemPort |
32. Kirkland, J. L., Dobson, DE. (1997) Preadipocyte function and aging: links between age-related changes in cell dynamics and altered fat tissue function. J Am Geriatr Soc. 45: 959–967.
33. Tchkonia, T., Giorgadze, N., Pirtskhalava, T., et al (2002) Fat depot origin affects adipogenesis in primary cultured and cloned human preadipocytes. Am J Physiol Regul Integr Comp Physiol. 282: R1286–R1296. | PubMed | ISI | ChemPort |
34. Shahparaki, A., Grunder, L., Sorisky, A. (2002) Comparison of human abdominal subcutaneous versus omental preadipocyte differentiation in primary culture. Metabolism. 51: 1211–1215.
35. Hutley, L. J., Herington, A. C., Shurety, W., et al (2001) Human adipose tissue endothelial cells promote preadipocyte proliferation. Am J Physiol Endocrinol Metab. 281: E1037–E1044. | PubMed | ISI | ChemPort |
36. Alles, J. U., Bosslet, K. (1986) Immunohistochemical and immunochemical characterization of a new endothelial cell-specific antigen. J Histochem Cytochem. 34: 209–214.
37. Muller, W. A., Ratti, C. M., McDonnell, S. L., et al (1989) A human endothelial cell-restricted, externally disposed plasmalemmal protein enriched in intercellular junctions. J Exp Med. 170: 399–414. | Article | PubMed | ISI | ChemPort |
38. Berger, J., Leibowitz, M. D., Doebber, T. W., et al (1999) Novel peroxisome proliferator-activated receptor (PPAR) gamma and PPARdelta ligands produce distinct biological effects. J Biol Chem. 274: 6718–6725. | Article | PubMed | ISI | ChemPort |
39. Ramirez-Zacarias, J. L., Castro-Munozledo, F., Kuri-Harcuch, W. (1992) Quantitation of adipose conversion and triglycerides by staining intracytoplasmic lipids with Oil red O. Histochemistry. 97: 493–497. | Article | PubMed | ChemPort |
40. van Dielen, F. M., van't Veer, C., Schols, A. M., et al (2001) Increased leptin concentrations correlate with increased concentrations of inflammatory markers in morbidly obese individuals. Int J Obes Relat Metab Disord. 25: 1759–1766. | Article |
41. Karagiannides, I., Tchkonia, T., Dobson, D. E., et al (2001) Altered expression of C/EBP family members results in decreased adipogenesis with aging. Am J Physiol Regul Integr Comp Physiol. 280: R1772–R1780.
42. Opavsky, R., Haviernik, P., Jurkovicova, D., et al (2001) Molecular characterization of the mouse Tem1/endosialin gene regulated by cell density in vitro and expressed in normal tissues in vivo. J Biol Chem. 276: 38795–38807. | Article | PubMed | ChemPort |
43. Ailhaud, G., Dani, C., Amri, E. Z., et al (1989) Coupling growth arrest and adipocyte differentiation. Environ Health Perspect. 80: 17–23.
44. Kubota, N., Terauchi, Y., Miki, H., et al (1999) PPAR gamma mediates high-fat diet-induced adipocyte hypertrophy and insulin resistance. Mol Cell. 4: 597–609. | Article | PubMed | ISI | ChemPort |

Acknowledgments

This study was supported, in part, by the Dutch Society for Scientific Research to A.H.F.B. (NWO 980-10-012) and by an Assistent Geneeskunde in Opleiding tot Klinisch Onderzoeker stipendium to F.M.H.v.D.