Obesity Is Accompanied by Disturbances in Peripheral Glucocorticoid Metabolism and Changes in FA Recycling

Authors


(kotryna.simonyte@medicin.umu.se)

Abstract

The glucocorticoid activating enzyme 11β-hydroxysteroid dehydrogenase type 1 (11βHSD1) is of major interest in obesity-related morbidity. Alterations in tissue-specific cortisol levels may influence lipogenetic and gluco/glyceroneogenetic pathways in fat and liver. We analyzed the expression and activity of 11βHSD1 as well as the expression of phosphoenolpyruvate carboxykinase (PEPCK), sterol regulatory element binding protein (SREBP), and fatty acid synthase (FAS) in adipose and liver and investigated putative associations between 11βHSD1 and energy metabolism genes. A total of 33 obese women (mean BMI 44.6) undergoing gastric bypass surgery were enrolled. Subcutaneous adipose tissue (SAT), omental fat (omental adipose tissue (OmAT)), and liver biopsies were collected during the surgery. 11βHSD1 gene expression was higher in SAT vs. OmAT (P = 0.013), whereas the activity was higher in OmAT (P = 0.009). The SAT 11βHSD1 correlated with waist circumference (P = 0.045) and was an independent predictor for the OmAT area in a linear regression model. Energy metabolism genes had AT depot–specific expression; higher leptin and SREBP in SAT than OmAT, but higher PEPCK in OmAT than SAT. The expression of 11βHSD1 correlated with PEPCK in both AT depots (P = 0.05 for SAT and P = 0.0001 for OmAT). Hepatic 11βHSD1 activity correlated negatively with abdominal adipose area (P = 0.002) and expression positively with PEPCK (P = 0.003). In human obesity, glucocorticoid regeneration in the SAT is associated with central fat accumulation indicating that the importance of this specific fat depot is underestimated. Central fat accumulation is negatively associated with hepatic 11βHSD1 activity. A disturbance in peripheral glucocorticoid metabolism is associated with changes in genes involved in fatty acid (FA) recycling in adipose tissue (AT).

Introduction

Tissue-specific glucocorticoid metabolism has come into focus as having a possible role in the development of obesity and its related diseases. 11β-Hydroxysteroid dehydrogenase type 1 (11βHSD1) catalyzes the interconversion of biologically inactive cortisone to active cortisol, thereby regulating its access to glucocorticoid receptors in target tissues (1,2). Most, but not all (3,4), studies have shown that 11βHSD1 expression, enzyme activity, and protein levels in subcutaneous adipose tissue (SAT) are positively related to BMI in both men and women (5,6,7,8,9,10,11,12). A direct measurement of 11βHSD1 activity or protein levels has not been performed in the human liver, but a negative association between BMI and the hepatic cortisone conversion rate has been shown, indicating a downregulation of hepatic 11βHSD1 in obesity (4,5).

Transgenic mice selectively overexpressing 11βHSD1 in adipose tissue (AT) develop central obesity, hypertension, insulin resistance, and dyslipidemia. In addition, they have disproportionately elevated serum levels of leptin, free fatty acids (FAs), and triglycerides and the increased expression of the lipoprotein lipase gene, whereas adiponectin mRNA levels are markedly decreased (13). Similar to AP2-11βHSD1, mice overexpressing hepatic 11βHSD1 exhibit insulin resistance, elevated blood pressure, and a fatty liver, but unaltered visceral fat accumulation (14). In contrast to mice overexpressing 11βHSD1, 11βHSD1 knockout mice exhibit reduced visceral fat accumulation and a protective glycemic and lipid profile driven by the increased hepatic expression of enzymes involved in fat catabolism and their transcription factors (15).

Differences in glucocorticoid metabolism also exist between different AT depots, as was first shown in obese Zucker rats; 11βHSD1 enzyme activity was found to be higher in omental AT (OmAT) compared to SAT (16). Higher 11βHSD1 activity in visceral AT (VAT) compared to the superficial/deep SAT has also been shown in obese South African women (17). Data regarding the expression of the 11βHSD1 gene in different human fat depots are heterogeneous; some studies report no depot-specific differences on 11βHSD1 gene expression (17,18), whereas others have found it to be higher in VAT (4).

Disturbances in glucocorticoid metabolism influence lipogenesis, lipolysis, and gluconeogenesis, possibly contributing to the development of alterations in the metabolic syndrome. However, data from humans for the molecular mechanisms and possible tissue-specific differences are scarce. During the fasting state, glucocorticoids induce the expression of hepatic phosphoenolpyruvate carboxykinase C (PEPCK-C), a key enzyme in gluconeogenesis, but attenuates transcription of the gene in AT. This reduces AT glyceroneogenic activity and increases the rate of free FA release (19,20). Obesity has recently been shown to be positively associated with PEPCK gene expression in SAT (21). The effect of glucocorticoids on lipogenic genes, such as sterol regulatory element binding protein (SREBP) and FA synthase (FAS), is not clear. Interestingly, mice overexpressing hepatic 11βHSD1 have increased levels of FAS in the liver (14), and in obese humans decreased SREBP-1 expression has been found in AT (22), in addition to AT depot–specific expression of higher mRNA levels in SAT compared to OmAT (23). Notably, the pharmacological inhibition of 11βHSD1 in rodents induces AT depot–specific changes in the expression of the FAS and PEPCK genes with increased levels in mesenteric but decreased in epididymal fat (24).

We hypothesized that obesity-related disturbances in the peripheral glucocorticoid metabolism is associated with changes in the expression of genes involved in FA recycling and lipogenesis in AT and the liver. The gene expression of the 11βHSD1 and its enzyme activity, as well as the expressions of the PEPCK, SREBP, and FAS genes in SAT, OmAT, and the liver were analyzed. Putative associations between 11βHSD1 and key genes involved in energy metabolism were investigated.

Methods and Procedures

Subjects

A total of 31 (30 white and 1 black) obese women accepted for gastric bypass surgery were consecutively enrolled in the study. Exclusion criteria were untreated endocrine, liver, or kidney diseases, malignancy, pregnancy, and/or a body weight ≥160 kg. Two women were excluded due to uncontrolled hypertension or malignancy. Three women were postmenopausal and nine were on gestagen treatment in the form of an intrauterine contraceptive device (5), subcutaneous implant (1), or oral supplement. One woman used combined contraceptive pills and one was on hormone replacement therapy. Three participants had been diagnosed with type 2 diabetes mellitus and were treated by a nonpharmacological regimen, oral hypoglycemic agents, or insulin; none of the women had type 1 diabetes. The study was approved by the local ethics committee and all subjects gave written informed consent.

Clinical protocol

Anthropometrical measurements, computed tomography (CT) scans, and dual-energy X-ray absorptiometry were performed on the same morning.

For the anthropometrical measurements, height and waist circumference were measured to the nearest 0.5 cm and weight to the nearest 0.1 kg. Blood pressure (BP) was measured in the supine position with a triple sphygmomanometer cuff (tricuff 0410; Speidel-Keller, Berlin, Germany). The body fat percentage was estimated using dual-energy X-ray absorptiometry, fat distribution was evaluated by abdominal CT at the level of lumbar vertebra 4 (L4) and thigh CT, and the liver attenuation was estimated from CT scans.

Approximately 1 g of SAT and OmAT and 0.5 g of liver were collected during the surgery after an overnight fast. The tissue was divided into smaller pieces, snap frozen in liquid nitrogen, and stored at −80 °C until further analysis.

At the time of the health examination (during the first visit), venous blood was drawn for a standard serum analysis. Later, venous blood was drawn at the time of the operation immediately after the induction of anesthesia but before the surgical procedure. All samples were aliquoted and stored at −20 °C until analyzed.

Standard laboratory immunoassays were used to analyze serum insulin, glucose, free thyroxin (fT4), thyroxin-stimulating hormone, aspartate aminotransferase, alanine aminotransferase, and C-peptide. The homeostasis model assessment for insulin resistance was calculated using HOMA calculator version 2.2 (http:www.dtu.ox.ac.ukhoma).

RNA extraction

Total RNA was extracted according to the manufacturer's instructions from a maximum of 100 mg AT or liver using the RNeasy lipid tissue mini kit (Qiagen Nordic, West Sussex, UK). The RNA concentrations were measured on a ND-1000 spectrophotometer (NanoDrop Technologies, Wilmington, DE). The RNA integrity was evaluated by 1% agarose gel electrophoresis and visualized with ethidium bromide under UV light.

Real-time reverse transcription–PCR

Two micrograms of RNA was reverse transcribed using TaqMan reverse transcription reagents (Roche Molecular Systems, Branchburg, NJ). Relative quantification real-time PCR was carried out using an ABI Prism 7000 Sequence Detection System (Applied Biosystems, Foster City, CA) according to the manufacturer's instructions using Universal PCR Master Mix 2× (Roche Molecular Systems, Branchburg, NJ) and TaqMan Gene expression assays for target genes 11βHSD1 (assay no. Hs00194153_m1), leptin (assay no. Hs00605917_m1), adiponectin (assay no. Hs00605917_m1), SREBP (assay no. Hs01088691_m1), FAS (assay no. Hs00188012_m1), PEPCK (assay no. Hs00356436_m1), and the endogenous control cyclophilin A (peptidylprolyl isomerase A (PPIA)) (assay no. Hs99999904_m1) (Applied Biosystems, Foster City, CA). All reactions were performed in triplicate and a nontemplate control and nonpolymerase control were included in every plate. Data were normalized against PPIA, which had the lowest coefficient of variation out of three endogenous controls that were tested (PPIA, LRP10, and RPLP0, data not shown) (25).

Enzyme activity assay

The activity of 11βHSD1 was measured in the dehydrogenase direction with an excess of cofactor. In 750 and 300 µl of Krebs-Ringer buffer (118 mmol/l NaCl, 3.8 mmol/l KCl, 1.19 mmol/l KH2PO4, 2.54 mmol/l CaCl2, 1.19 mmol/l MgSO4·7H2O, 2.5 mmol/l NaHCO3, pH 7.4), 250 mg of AT and 50 mg of liver were homogenized, respectively, with 0.1 mmol/l dithiothreitol and centrifuged at 4 °C. Protein concentrations were determined using the Bradford technique (Bio-Rad protein assay; Bio-Rad Laboratories, Hercules, CA). Duplicate samples of 0.987 mg/ml protein from AT were incubated at 37 °C with 10 mmol/l nicotinamide adenine dinucleotide phosphate and 50 nmol/l [1,2,6,7-3H]-cortisol for 30 h. Samples were withdrawn at 20, 24, and 30 h and frozen at −80 °C. For liver samples the same procedures were performed for 0.0987 mg/ml protein and samples were withdrawn at 1, 2, and 3 h. Subsequently, glucocorticoids were extracted with dichloromethane and the organic phase evaporated. The extracts were dissolved in ethanol and separated by thin layer chromatography (thin layer chromatography aluminum sheets, 20 × 20 cm2, Silica gel 60 F254; Merck, Darmstadt, Germany; mobile phase, chloroform and ethanol (92:8)). Radiolabeled glucocorticoids were detected by the exposure of the thin layer chromatography sheet to a tritium storage phosphor screen and subsequently scanned on a Typhoon 9400 scanner (GE Healthcare Europe, Freiburg, Germany). The activity was expressed as the percent conversion of cortisol to cortisone.

Statistical analysis

Data are presented as mean ± s.d. or mean ± s.e.m., as indicated in the text. For comparisons between AT depots the nonparametric Wilcoxon test for paired samples was used. Spearman's ρ-test was used to estimate bivariate correlations between different variables. P values <0.05 were considered significant. Correlation P values were adjusted according to Bonferroni, as indicated in the text. Multiple linear regression analysis was used to estimate independent predictors of variables. SPSS 14.0 for Windows was used for statistical analysis (SPSS, Chicago, IL).

Results

Subject characteristics

Anthropometric and biochemical characteristics are summarized in Table 1. All patients had a BMI >35 kg/m2 and a waist circumference >100 cm. The AT distribution measured by abdominal CT scan revealed a 3.7-fold larger area of SAT depot compared to OmAT (P = 0.0001). Normal values of serum thyroid hormones (fT4, thyroxin-stimulating hormone) excluded thyroid-related metabolic abnormalities. Fasting blood glucose, insulin, and serum liver transaminases (aspartate aminotransferase, alanine aminotransferase) levels were within normal ranges.

Table 1.  Subject characteristics (n = 31)
inline image

AT depot–specific differences in 11βHSD1 expression/activity and expression of energy metabolism genes

AT depot–specific differences in the expression of the 11βHSD1 gene and genes related to energy metabolism are given in Table 2. The expression of 11βHSD1 was significantly higher in SAT compared to OmAT, but enzyme activity was significantly higher in OmAT (Figure 1). The expression of 11βHSD1 and its activity positively correlated in OmAT (r = 0.47, P = 0.04) but not in SAT (P = 0.21). The leptin mRNA levels in SAT were twofold higher compared to OmAT (P = 0.0001), whereas the expression of the adiponectin gene did not differ between depots. The SREBP mRNA levels were higher in SAT compared to OmAT (P = 0.03), whereas no differences were seen for FAS. The PEPCK mRNA levels were higher in OmAT compared to SAT (P = 0.001).

Table 2.  Adipose tissue depot–specific gene expression of 11βHSD1 and genes involved in energy metabolism
inline image
Figure 1.

11β-Hydroxysteroid dehydrogenase type 1 (11βHSD1) enzyme activity in adipose tissue depots (n = 20). Data presented as mean ± s.e.m. P = 0.009, P = 0.004, P = 0.01, for 18, 24, 30 h, respectively. OmAT, omental adipose tissue; SAT, subcutaneous adipose tissue.

Associations between AT 11βHSD1 to anthropometrical measurements, and energy metabolism genes

Correlations between 11βHSD1 gene expression and activity in AT depots and anthropometrics are presented in Table 3. There was a positive association between 11βHSD1 mRNA levels in SAT and systolic BP, and between 11βHSD1 enzyme activity in SAT and waist circumference and AT areas, i.e., both L4 total and L4 OmAT. After adjustment for multiple comparisons according to Bonferroni, the association between 11βHSD1 activity and L4 OmAT area remained significant (P = 0.03). SAT 11βHSD1 enzyme activity was also an independent predictor for the L4 OmAT area (β = 0.56, P = 0.039), when waist and age were included in the model (β = 0.14, P = 0.562 and β = 0.41, P = 0.098, for waist and age, respectively).

Table 3.  Correlations between 11βHSD1 levels in adipose tissue and anthropometric variables
inline image

No associations were found between OmAT 11βHSD1 expression/activity and parameters of the metabolic syndrome, including s-insulin, plasma glucose, and BMI. We also found that 11βHSD1 gene expression correlated positively to PEPCK gene expression in both AT depots (Figure 2). The OmAT 11βHSD1 mRNA levels were associated with leptin expression (r = 0.45, P = 0.02).

Figure 2.

Associations between 11β-hydroxysteroid dehydrogenase type 1 (11βHSD1) and phosphoenolpyruvate carboxykinase (PEPCK) in adipose tissue depots (n = 24 for SAT and n = 28 for OmAT, respectively). Open circles represent SAT; filled circles represent OmAT; where r = 0.41, P = 0.05 for SAT and r = 0.67, P = 0.0001 for OmAT. AU, arbitrary unit; OmAT, omental adipose tissue; SAT, subcutaneous adipose tissue.

Hepatic 11βHSD1

Hepatic 11βHSD1 gene expression and enzyme activity correlated positively (r = 0.50, P = 0.04) (Figure 3). There were no associations between the gene expression and anthropometrical measurements or insulin/glucose levels. However, 11βHSD1 activity correlated negatively with abdominal and OmAT areas (r = −0.70, P = 0.002 and r = −0.60, P = 0.01, respectively). Liver 11βHSD1 gene expression correlated strongly with PEPCK gene expression (r = 0.58, P = 0.003).

Figure 3.

Correlation between hepatic 11β-hydroxysteroid dehydrogenase type 1 (11βHSD1) gene expression and enzyme activity (n = 17). r = 0.49, P = 0.043. AU, arbitrary unit.

Discussion

Increased 11βHSD1 gene expression and activity in AT and the liver result in higher local cortisol levels, which has been suggested to be a key mediator of the development of obesity and obesity-related diseases, such as insulin resistance, type 2 diabetes mellitus, and cardiovascular diseases (1,2). There is an ongoing debate as to whether metabolic changes in SAT or VAT are the most harmful. Studies regarding AT depot–specific differences in glucocorticoid metabolism are few and partly conflicting. Our data strengthens earlier findings that 11βHSD1 activity is increased in OmAT compared to SAT in the context of human obesity (17). We also found a correlation between 11βHSD1 mRNA and activity in OmAT, but not in SAT as was previously reported (6,7,17). The disagreement between enzyme activity and gene expression may be related to AT depot–specific, post-transcriptional, or post-translational regulation possibly involving mRNA degradation in SAT or the presence of putative activators or inhibitors in either of depots (5,8,26).

It has been suggested, that AT depot–specific differences in cortisol inactivation, through differences in 11βHSD2 gene expression/activity, may contribute to local availability of cortisol (26). The gene expression of 11βHSD2 was thus reported to be higher in OmAT suggesting a higher glucocorticoid turnover in this depot vs. SAT. However, the 11βHSD2 expression was only found in the stromal cell fraction, whereas 11βHSD1 expression was exclusive for fat cells (26). Notably, 11βHSD2 expression levels in AT are 30-fold lower vs. 11βHSD1 and the physiological importance in humans of 11βHSD2 in fat thus remains to be studied further (27).

Despite a higher enzyme activity of 11βHSD1 in OmAT, the subcutaneous depot appears to be more predictive for obesity and metabolic parameters. Importantly, SAT 11βHSD1 correlated with measures of central fat accumulation and the multiple linear regression analysis showed that SAT 11βHSD1 activity is an independent predictor for central fat accumulation, emphasizing the important role of 11βHSD1 enzyme in SAT depot. We could not verify previously reported associations between SAT 11βHSD1 and BMI, which is a less precise and more general measurement of obesity. The 11βHSD1 in SAT also correlated with BP, which is in line with previous findings in humans and rodents (13,17).

In addition to BMI and BP, insulin resistance has also been reported to be associated with fat accumulation in the SAT of adults and prepubertal children (5,6,17,28). We did not find any association between 11βHSD1 and insulin levels or homeostasis model assessment for insulin resistance in our study, which may be dependent on the fact that the participants were all obese with a limited BMI range.

To further evaluate local AT differences and a putative role of glucocorticoids on adipogenesis, gluco/glyceroneogenesis, and lipogenesis, we estimated the expression of the leptin, adiponectin, PEPCK, SREBP, and FAS genes. Increased adiposity is associated with altered leptin and adiponectin levels, and glucocorticoids have been proposed as potential upregulators of leptin (29). Only the production of leptin in SAT correlates with circulating leptin levels (30), and we found a twofold higher leptin expression in SAT compared to OmAT, which highlights the putative importance of leptin production in SAT as a strong marker of the risk for the development of stroke and cardiovascular disease (31,32). The expression of adiponectin did not differ between the AT depots, though there was a negative association between OmAT adiponectin expression and waist circumference, as well as insulin levels (data not shown). This finding supports the hypothesis that increased fat accumulation is accompanied by reduced adiponectin levels (13,33). This reduction may contribute to the progression of obesity and its comorbidities by modulating the lipolysis and FA oxidation pathways (34).

The SREBP is a key transcription factor involved in adipogenesis, insulin sensitivity, and FA homeostasis; it is negatively associated with obesity (22,35). In obese subjects, SREBP-1c has been shown to be higher expressed in SAT compared to OmAT (23), whereas the opposite has been shown for FAS mRNA levels (36). We found that both genes were higher expressed in SAT, suggesting that this depot is more involved in insulin and/or polyunsaturated FA–regulated de novo FA synthesis and the increased release of free FAs into the blood (23). In vitro and rodent studies have shown that glucocorticoids increase the expression of FAS (14,37), but we could not identify any associations between 11βHSD1 and SREBP/FAS, suggesting that increased local cortisol levels do not have a direct effect on the expression of genes involved in lipogenesis.

The PEPCK protein is a key enzyme involved in the regulation of FA release via the triglyceride–FA cycle within adipocytes. Glucocorticoids are positive regulators of PEPCK gene transcription in the liver and negative regulators in AT (19,20,38). The overexpression of PEPCK-C in AT increases FA re-esterification, leading to obesity (39,40), and a positive association between PEPCK expression in SAT and BMI was recently reported (21). On the other hand, we found that PEPCK is higher expressed in the OmAT depot, which could be due to a parallel increase in local cortisol levels. Interestingly, we found a positive correlation between 11βHSD1 and PEPCK in both AT depots, suggesting that, in the context of obesity, there is a possible dysregulation of PEPCK gene transcription by glucocorticoids in AT, leading to an increased rate of FA re-esterification in fat.

In this study, we had a unique opportunity to evaluate hepatic 11βHSD1 and key genes involved in energy metabolism in human samples. Because we lack a control group, it is not possible to estimate whether there is an up- or downregulation of different genes in the liver of obese vs. subjects of average weight. The methods that we used to estimate gene expression/activity (semiquantitative PCR, thin layer chromatography) do not allow us to make direct comparisons regarding 11βHSD1 between AT and liver. However, in a recent study by Stimson et al. the role of SAT, VAT, and liver in cortisol regeneration by 11βHSD1 was estimated using deuterium labeled cortisol. This showed that splanchnic cortisol production was entirely due to liver 11βHSD1 activity and that VAT 11βHSD1 activity was insufficient to increase portal vein cortisol concentration, and hence influence intrahepatic glucocorticoid receptor signaling (41). However, our finding of a negative association between hepatic 11βHSD1 activity and central fat accumulation supports the hypothesis that 11βHSD1 activity is decreased in obesity (6). As there were no associations between 11βHSD1 and SREBP/FAS gene expression, we suggest that in the liver as well as in fat, glucocorticoids are not directly involved in the regulation of the transcription of genes involved in lipogenesis. Importantly, the physiologically relevant association of 11βHSD1 and PEPCK is maintained in our cohort.

We suggest that, in human obesity, (i) the capacity for glucocorticoid regeneration in SAT, rather than OmAT, is associated with central fat accumulation; (ii) hepatic 11βHSD1 is negatively associated with central fat accumulation; and (iii) extreme obesity is not only accompanied by disturbances in peripheral glucocorticoid metabolism, but also makes major changes in FA recycling pathways in AT.

Disclosure

T.O. has received consulting fees from Wyeth Pharmaceuticals within the previous 12 months.

Ancillary