articles
nature publishing group
adipocyte biology
Subcutaneous Abdominal Adipose Tissue Subcompartments: Potential Role in Rosiglitazone Effects Gillian E. Walker1, Paolo Marzullo2, Barbara Verti2, Gabriele Guzzaloni2, Sabrina Maestrini1, Francesco Zurleni3, Antonio Liuzzi2 and Anna Maria Di Blasio1 Abdominal visceral tissue (VAT) and subcutaneous adipose tissue (SAT), comprised of superficial-SAT (sSAT) and deep-SAT (dSAT), are metabolically distinct. The antidiabetic agents thiazolidinediones (TZDs), in addition to their insulin-sensitizing effects, redistribute SAT suggesting that TZD action involves adipose tissue depot–specific regulation. We investigated the expression of proteins key to adipocyte metabolism on differentiated first passage (P1) preadipocytes treated with rosiglitazone, to establish a role for the diverse depots of abdominal adipose tissue in the insulin-sensitizing effects of TZDs. Adipocytes and preadipocytes were isolated from sSAT, dSAT, and VAT samples obtained from eight normal subjects. Preadipocytes (P1) left untreated (U) or treated with a classic differentiation cocktail (DI) including rosiglitazone (DIR) for 9 days were evaluated for strata-specific differences in differentiation including peroxisome proliferator–activated receptor-γ (PPAR-γ) and lipoprotein lipase (LPL) expression, insulin sensitivity via adiponectin and glucose transport-4 (GLUT4), glucocorticoid metabolism with 11β-hydroxysteroid dehydrogenase type-1 (11βHSD1), and alterations in the adipokine leptin. While depot-specific differences were absent with the classic differentiation cocktail, with rosiglitazone sSAT had the most potent response followed by dSAT, whereas VAT was resistant to differentiation. With rosiglitazone, universal strata effects were observed for PPAR-γ, LPL, and leptin, with VAT in all cases expressing significantly lower basal expression levels. Clear dSAT-specific changes were observed with decreased intracellular GLUT4. Specific sSAT alterations included decreased 11βHSD1 whereas secreted adiponectin was potently upregulated in sSAT with respect to dSAT and VAT. Overall, the subcompartments of SAT, sSAT, and dSAT, appear to participate in the metabolic changes that arise with rosiglitazone administration. Obesity (2008) 16, 1983–1991. doi:10.1038/oby.2008.326
Introduction
In the 1950s, Vague et al. were the first to suggest that regulation of the endocrine and metabolic functions of abdominal adipose tissue were controlled, in part, by the anatomical distribution of fat with android obesity associated to diabetes and atherosclerosis (1). It is now accepted that the anatomical distribution of abdominal adipose tissue is dependent on adipose tissue depots (2). These include visceral adipose tissue (VAT) centrally located and enclosed by the peritoneum (2) and the subcutaneous adipose tissue (SAT) located directly below the skin. Both VAT and SAT show subject-to-subject variations with respect to distribution and volume, being dependent on age, gender, nutritional intake, and the autonomic regulation of energy homeostasis (2). However, between the two, VAT accumulation is central to android obesity and is an independent risk factor
for obesity-related metabolic and cardiovascular disorders (3), in particular insulin resistance and dyslipidemia (3). An evaluation of the association of SAT accumulation with obesity-associated complications has generated many contradictory studies. SAT has been conventionally regarded as a homogenous depot; however, it is anatomically divided by the scarpas’s fascia to form two subcompartments, the superficial-SAT (sSAT) directly below the skin and the deepSAT (dSAT) that comes into contact with the preperitoneal adipose tissue (4). Each is histologically distinct (5), with conventional imaging procedures illustrating subject-to-subject variations in distribution, particularly in association with obesity and insulin resistance (4,6,7). In a previous investigation, cellular studies of biopsies from the two SAT subcompartments in lean subjects revealed a diverse pattern of expression
1 Laboratory of Molecular Biology, I.R.C.C.S. Istituto Auxologico Italiano, Piancavallo (VB), Italy; 2Division of Internal Medicine, I.R.C.C.S. Istituto Auxologico Italiano, Piancavallo (VB), Italy; 3Ospedale di Circolo di Busto Arsizio, Busto Arsizio, Italy. Correspondence: Gillian E. Walker (
[email protected])
Received 19 June 2007; accepted 7 March 2008; published online 19 June 2008. doi:10.1038/oby.2008.326 obesity | VOLUME 16 NUMBER 9 | SEPTEMBER 2008
1983
articles Adipocyte Biology of key proteins that have been associated with obesity and its complications, particularly glucose transport-4 (GLUT4), 11β-hydroxysteroid dehydrogenase type-1 (11βHSD1), resistin, and leptin (8). This raised the possibility that the SAT subcompartments may have distinct, yet unknown, metabolic activities, which may explain differences in the pattern of fat deposition and the controversial findings of differential lipolysis that has been observed between the sSAT and dSAT strata (7,9,10). Thiazolidinediones (TZDs) are a family of antidiabetic pharmacological agents including rosiglitazone, pioglitazone, and troglitazone that have been proven to ameliorate insulin sensitivity in insulin-resistant subjects (11). Insulin resistance is a common complication of obesity, yet paradoxically TZDs are potent inducers of preadipocyte differentiation in both human (12) and murine (13) cell lines in vitro. The association of TZDs with preadipocyte differentiation can be explained by the fact that they are high-affinity ligands for peroxisome proliferator–activated receptors (PPARs), in particular PPAR-γ (14). The PPARs are nuclear hormone receptors abundantly expressed in adipose tissue and central to the regulation of preadipocyte differentiation through transcriptional control of adipocyte-specific genes (15). The insulin-sensitizing effects of TZDs have been shown to correlate to their ability to activate PPAR-γ (16). Interestingly, a number of in vivo human studies evaluating the effect of TZDs as a treatment for diabetes observed that although there was an improvement in glycemic control there was also a modest weight gain in the study subjects (17,18). This weight gain was reflected in a relative increase of abdominal SAT over VAT, suggesting that TZDs may elicit depot-specific effects promoting the redistribution of adipose tissue (17,18). Previous studies evaluating the effects of rosiglitazone on SAT and VAT preadipocytes documented a depot-specific responsiveness, with SAT preadipocyte differentiation enhanced by the inclusion of the TZD, whereas VAT preadipocytes were resistant (12,19). In the studies described, the effects of TZDs were not tested on preadipocytes isolated from dSAT, a subcompartment of SAT that appears to have independent metabolic functions in a normal physiological state (8) and is correlated to insulin resistance and lipolysis (6,10). As dSAT is correlated to insulin resistance (6) and reflects physiologically a protein profile intermediate to both sSAT and VAT in lean subjects (8), we sought to explore the effect of TZDs, specifically rosiglitazone, on preadipocyte differentiation from sSAT, dSAT, and VAT and evaluate the expression of key proteins associated with differentiation and obesity-associated complications. Methods and PROCEDURES Patients and sampling Adipose tissue samples from the three anterior abdominal adipose depots sSAT, dSAT, and VAT were obtained, as previously described (8), from eight healthy normal subjects at the peri-umbilical level during abdominal surgery unrelated to inflammatory or neoplastic abdominal disorders. Of the subjects, five were postmenopausal women and three were men with a collective BMI 23.3 ± 0.65 kg/m2 1984
and age 55.3 ± 4.9 years. At the time of surgery, fasting blood samples were collected for chemical analysis. All chemical measurements were performed using commercially available kits, as described previously (8). The study protocol was approved by the individual Institution Ethics Committees, with the aim and the design of the study explained to each subject, who in turn gave his/her informed consent. Cell isolation and differentiation Both adipocytes and preadipocytes were isolated as previously described (8). Adipocytes were stored directly at −80 °C, whereas first passage (P1) preadipocytes were left to reach confluency, which took between 4 and 6 days from the point of isolation. Differentiation was initiated on P1 preadipocytes that had clear cellular contact with an overnight incubation in serum-free medium (1:1 DMEM:Ham’s-F12) containing 1× concentration of insulintransferrin-selenium (Gibco BRL, Rockville, MD). Preadipocytes were then either untreated (U; serum-free medium + 1× insulintransferrin-selenium) or treated for 3 days with one of two differentiation cocktails, dexamethasone with 3-isobuty-1-methylxanthine (DI; 1:1 DMEM:Ham’s-F12, 1× insulin-transferrin-selenium, 1 µmol dexamethasone (Sigma Chemical, St Louis, MO) 100 µmol 3-isobuty-1-methylxanthine (Sigma)) or DI with 1 µmol rosiglitazone (DIR; rosiglitazone (GlaxoSmithKline, West Sussex, UK)). At 3 days the conditioned mediums (CMs) were collected and substituted with the DI and DIR cocktails minus 3-isobuty-1-methylxanthine. The CM collection was repeated at 6 days and on completion at 9 days when the cells were harvested for total RNA and protein as previously described (8). Assessment of differentiation Differentiation was assessed by cell morphology and triglyceride (TG) content using Oil Red O staining. At 9 days, cells were dehydrated with 60% isopropenol and stained in 1% Oil Red O (Sigma) in 60% isopropenol. Cells were washed in 60% isopropenol and left to dry for image capture or they were washed with water and treated with 100% isopropenol to release the TGs which were measured spectrophotometrically at A500 normalized to an empty well. Results are given as optical density A500. Semiquantative reverse transcriptase–PCR Semiquantitative reverse transcriptase–PCR was performed with as previously described (8) with hypoxanthine phosphoribosyl transferase serving as the internal control for the basal expression of PPAR-γ (forward: 5ʹ-AGACAACAGACAAATCACCAT, reverse: 5ʹ-CTTCACAGCAAACTCAAACTT) and lipoprotein lipase (LPL) (forward: 5ʹ-GAGATTTCTCTGTATGGCACC, reverse: 5ʹ-CTGCAAA TGAGACACTTTCTC) in adipocytes and preadipocytes and glyceraldehyde-3-phosphate dehydrogenase (forward: 5ʹ-AGCCTC AAGATCATCAGCAATG, reverse: 5ʹ-ATGGACTGTGGTCATGA GTCCTT) serving as the internal control for differentiation studies. All samples were analyzed in duplicate and quantified by densitometry using QuantityOne software (Bio-Rad, Hercules, CA) normalized to the expression of hypoxanthine phosphoribosyl transferase or glyceraldehyde-3-phosphate dehydrogenase and a standard intra-experimental control. Results are presented as arbitrary units relative to the expression of hypoxanthine phosphoribosyl transferase or glyceraldehyde3-phosphate dehydrogenase. Protein evaluation The CM collected at 3, 6, and 9 days of differentiation were centrifuged at 3,000 r.p.m. for 10 min at room temperature and stored at −20 °C for the evaluation of secreted proteins by western immuno blot. Protein evaluations were made on cell lysates and CM as previously described (8). VOLUME 16 NUMBER 9 | SEPTEMBER 2008 | www.obesityjournal.org
articles Adipocyte Biology Statistical evaluation Data are expressed as mean ± s.e.m. Differences and trends within each strata were evaluated using a repeated measures ANOVA with between strata differences evaluated using a one-way ANOVA, each with a Bonferroni post hoc test. Statistical significance was set at P
