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.
Department of Nutrition, University of Tennessee, Room 229, Jessie Harris Building, 1215 West Cumberland Avenue, Knoxville, TN 37996. E-mail: firstname.lastname@example.org
Objective: 1, 25-Dihydroxyvitamin D3 dose dependently increases intracellular calcium in human adipocytes. We have demonstrated that suppression of circulating 1, 25-dihydroxyvitamin D3 levels by increasing dietary calcium reduces adipocyte intracellular calcium and reduces adiposity in both humans and rodents, with preferential loss of trunk fat. Autocrine production of cortisol by adipocytes of mice overexpressing 11β-hydroxysteroid dehydrogenase type 1 (11β-HSD 1) in adipose tissue increases visceral adiposity, whereas knockout of 11β-HSD 1 appears to attenuate truncal obesity. Accordingly, our objective was to investigate the role of 1, 25-dihydroxyvitamin D3 in the modulation of adipocyte glucocorticoid metabolism.
Research Methods and Procedures: We examined the effect of 1, 25-dihydroxyvitamin D3 or angiotensin II on cortisol production and expression using real-time reverse transcriptase-polymerase chain reaction of 11β-HSD 1, angiotensin II receptor type 1 (AT1), and AT2 receptor in human adipocytes.
Results: Adipocytes produced negligible cortisol in the absence of substrate (cortisone). In the presence of cortisone (1 to 10 nM), there was significant cortisol production, which was dose dependently augmented (2- to 6-fold, p < 0.001) by 1, 25-dihydroxyvitamin D3 (0.1 to 10 nM). 1, 25-Dihydroxyvitamin D3 dose dependently increased 11β-HSD 1 expression up to 2-fold (p < 0.01) in both the presence and absence of cortisone. In contrast, 1, 25-dihydroxyvitamin D3 dose dependently decreased adipocyte AT1 expression (by 30% to 50%, p < 0.001) in both the presence and absence of cortisone, suggesting compensatory down-regulation of AT1.
Discussion: We conclude that 1, 25-dihydroxyvitamin D3 directly regulates adipocyte 11β-HSD 1 expression and, consequently, local cortisol levels and that this may contribute to the preferential loss of visceral adiposity by high-calcium diets.
Calcium signaling is an important modulator of adipocyte metabolism. Previous data from our laboratory demonstrate that increasing intracellular calcium ([Ca2+]i)1 in human adipocytes promotes energy storage and accumulation of fat mass by stimulating de novo lipogenesis and inhibiting lipolysis (1,2,3), whereas treatment of obesity-prone mice with a calcium channel antagonist (nifedipine) significantly reduces adipose tissue mass (4). We initially investigated the effects of high- vs. low-calcium diets on body weight and adiposity in obesity-prone transgenic mice (5,6). Low-calcium diets accelerated body weight and fat gain in animals fed an obesity-promoting diet and impeded fat loss during energy restriction (5,6). In contrast, high-calcium diets suppressed fat deposition and fat loss during modest energy restriction (5). More recently, we have reported that increasing dietary calcium accelerates weight and fat loss in energy restricted obese adults, with preferential loss of trunk fat (7). To further explore the mechanism involved in these effects, we investigated the potential for calcitrophic hormones to modulate adipocyte Ca2+ and lipid metabolism. 1, 25-Dihydroxyvitamin D3 elicited marked increases in [Ca2+]i in human adipocytes; this is a rapid, membrane-initiated effect and appears to be mediated by the 1, 25-dihydroxyvitamin D3 membrane-associated rapid response to steroid (1, 25D3-MARRS) binding protein (8,9,10). The 1, 25-dihydroxyvitamin D3-mediated increase in adipocyte [Ca2+]i enhanced lipogenesis and reduced lipolysis (8). Thus, increasing [Ca2+]i exerts both lipogenic and antilipolytic actions on adipocyte lipid metabolism, which promotes triglyceride accumulation and expansion of adipose tissue mass.
1, 25-Dihydroxyvitamin D3 promotes translocation of the 1, 25D3-MARRS protein to the nucleus, suggesting that 1, 25D3-MARRS relays signals between the plasma membrane and nuclear vitamin D receptor (nVDR) (11). Work from this laboratory has demonstrated that 1, 25-dihydroxyvitamin D3 acts through the nVDR to inhibit expression of uncoupling protein 2, whereas suppression of 1, 25-dihydroxyvitamin D3 levels by feeding high-calcium diets to mice results in increased adipose tissue uncoupling protein 2 expression and increased thermogenesis (12). These data suggest the existence of a metabolic link between the membrane-initiated events mediated through 1, 25D3-MARRS after exposure to 1, 25-dihydroxyvitamin D3 and genomic events involving the classical nVDR that may subsequently influence energy partitioning and adipose tissue mass. Accordingly, we suggest that such an interaction may implicate suppression of 1, 25-dihydroxyvitamin D3 levels as a viable target for interventions to regulate adiposity.
Local adipose tissue glucocorticoid levels and intracellular glucocorticoid availability are controlled by the activity of 11β-hydroxysteroid dehydrogenase type 1 (11β-HSD 1), which functions almost exclusively as a high-affinity reductase in vivo, generating active cortisol from inactive cortisone (13). Cortisol stimulates the activity of 11β-HSD 1 in several cell lines, including adipocytes, and the induction of 11β-HSD 1 expression during preadipocyte differentiation is synergistically enhanced by cortisol and insulin (14,15). Although plasma cortisol concentrations are not consistently elevated in obese subjects (16,17), dysregulation of adipose tissue glucocorticoid metabolism may be involved in regulation of adipose tissue mass and site of deposition, as suggested by the specific increase of adipose tissue 11β-HSD 1 activity in obese Zucker rats (18). In addition, overexpression of 11β-HSD 1, specifically in white adipose tissue of transgenic mice, recapitulates features of the metabolic syndrome, including central obesity, hypertension, dyslipidemia, and insulin resistance (19,20), whereas homozygous 11β-HSD 1 knockout mice are protected from developing these clinical abnormalities associated with the metabolic syndrome (21). Obesity increases the expression of 11β-HSD 1 in human subcutaneous and visceral adipose tissue, and both 11β-HSD1 transcription and activity are higher in omental versus subcutaneous adipocytes from obese subjects, suggesting that the pattern of central fat deposition associated with chronic glucocorticoid excess, as in Cushing's syndrome, may be due, in part, to the greater capacity for regeneration of active glucocorticoids in the omental depot (22,23). Despite these findings, little else is known about the regulation of 11β-HSD 1 expression in human adipose tissue.
Although not yet studied in adipocytes, angiotensin II (AT II) stimulates cortisol release by two distinct calcium-dependent angiotensin receptor type 1 (AT1)-mediated mechanisms in bovine adrenal zona fasciculate cells (24). The stimulatory effect of AT II on cortisol production and the appearance of hypertension in subjects with Cushing's syndrome suggest an interaction between the renin-AT system and glucocorticoid metabolism. This interaction is consistent with the appearance of hypertension, which is alleviated by an AT1 receptor blocker in mice overexpressing 11β-HSD 1 in adipose tissue (20,23,25).
Given our previous data demonstrating greater loss of trunk fat in obese humans consuming energy-restricted high-calcium diets, we designed studies to investigate the role of 1, 25-dihydroxyvitamin D3-mediated increases in [Ca2+]i in the regulation of adipocyte glucocorticoid metabolism in vitro.
Research Methods and Procedures
These studies utilized human adipocytes differentiated in thiazolidinedione-containing medium from human adipose tissue stromal vascular cells obtained from Zen-Bio (Research Triangle, NC) (26). Cells are received from Zen-Bio 25 days postdifferentiation into mature adipocytes and were used in these experiments 5 days after receipt. Before each experiment, cells were incubated overnight in nominally serum-free medium [Dulbecco's modified Eagle's medium/Ham's F-10, 1:1 (vol/vol) containing 0.2% fetal bovine serum, 15 mM HEPES, 33 μM biotin, 17 μM Pantothenate, 100 U/mL penicillin, and 100 μg/mL streptomycin].
Cells were subjected to treatments of various combinations of 1, 25-dihydroxyvitamin D3 (0.1 to 100 nM), cortisone (0.1 to 100 nM), AT II (0.1 to 100 nM), saralasin (0.1 to 100 nM, nonspecific AT II receptor antagonist), losartan (0.1 to 100 nM, AT1-specific antagonist), PD123177 (0.1 to 100 nM, AT2-specific antagonist), lumisterol (0.1 to 100 nM, specific 1, 25D3-MARRS agonist), and/or 1β, 25-dihydroxyvitamin D (0.1 nM, specific 1, 25D3-MARRS antagonist).
Determination of Adipocyte Secretory Function
To determine optimum exposure times, aliquots of media were taken at 30 minutes, 1 hour, 90 minutes, and 24 hours after exposure to treatments and used for determination of cortisol (Assay Designs, Ann Arbor MI) and AT II (Phoenix Pharmaceuticals, Belmont, CA) release into the medium. Assays were performed immediately after the incubation period, without freezing the samples. Data were normalized for cellular DNA content measured by the CyQuant cell proliferation assay kit (Molecular Probes, Eugene, OR) according to the manufacturer's instructions using a microplate fluorometer (Packard Instrument Co., Inc., Downers Grove, IL).
Total RNA was extracted using the Totally RNA kit (Ambion, Inc., Austin, TX) according to the manufacturer's instructions. RNA was resuspended in nuclease-free water before quantification using dual-wavelength (260/280 nm) spectrophotometry.
Expression of 11β-HSD 1 (forward primer, 5′- GGAATATTCAGTGTCCAGGGTCAA-3′; reverse primer, 5′-TGATCTCCAGGGCACATTCCT-3′; probe, 5′-FAM-CTTGGCCTCATAGACACAGAAACAGCCA-BHQ-3′), the AT1 receptor (forward primer, 5′-ACCCAATGAAGTCCCGCCT-3′; reverse primer, 5′-AGCAGCCAAATGATGATGCAG-3′; probe, 5′-FAM-CGACGCACAATGCTTGTAGCCAAAGTCA-BHQ-3′), and the AT2 receptor (forward primer, 5′-CCAAGTCCTGAAGATGGCAGC-3′; reverse primer, 5′-GGAAGGTCAGAACATGGAAGGG-3′; probe, 5′-FAM-TTGTTCTGGCCTTCATCATTTGCTGGC-BHQ-3′) was quantitated by real-time reverse transcriptase-polymerase chain reaction (PCR) using the Smart Cycler Real Time PCR System (Cephid, Sunnyvale, CA) with a TaqMan 1000 Reaction Core Reagent Kit (Applied Biosystems, Foster City, CA).
To establish a standard curve for each gene, 1 μL of total RNA (concentration, 1 ng/μl) from each sample was pooled and, subsequently, serial-diluted in the range of 1.5625 to 25 ng for each gene of interest. The PCR reaction mixture contained 10× PCR buffer (500 mM KCl and 100 mM Tris-HCL, pH 8.3), 25 mM MgCl2, 75 mM of each deoxynucleotide triphosphate (dATP, dGTP, dCTP, and dTTP), 10 μM forward primer, 10 μM reverse primer, murine leukemia virus reverse transcriptase (50 U/μl), RNase inhibitor, AmpliTaq DNA polymerase, and 10 μM probe. The polymerase chain reaction was performed according to the instructions for the Smart Cycler System (Cephid) and TaqMan Real Time PCR Core Kit (Applied Biosystems) using 50 to 200 ng of total RNA per sample. Quantitation for each sample was normalized using corresponding 18s expression (forward primer, 5′-AGTCCCTGCCCTTTGTACACA-3′; reverse primer, 5′-GATCCGAGGGCCTCACTAAAC-3′; probe, 5′-FAM-CGCCCGTCGCTACTACCGATTGG-BHQ-3′).
Data were analyzed by ANOVA to compare overall group means with subsequent analysis using least significant difference test to identify significantly different group means (SPSS Statistics Package) (27). Variables that did not follow a normal distribution were transformed according to accepted statistical practices (log base 10, square, square root, inverse, and inverse square root) to achieve normal distribution before ANOVA. When normality could not be achieved, significantly different group means were identified by Kruskall-Wallis nonparametric test (28).
Cortisol Release by Human Adipocytes
No cortisol release was detected in the absence of exogenously added cortisone. Time (0 to 24 hours) and dose response (0.1 to 100 nM) studies indicated significant cortisol release plateaus 90 minutes after exposure to 10 nM cortisone (data not shown). To determine the role of [Ca2+]i in modulating cortisol release by human adipocytes in the presence of 10 nM cortisone, [Ca2+]i levels were increased by exposure to 1, 25-dihydroxyvitamin D3 (0.1 to 100 nM), depolarization with 100 nM KCl, or treatment with the Ca2+ ionophore BAYK8644 (0.2 to 0.5 μM). Figure 1A demonstrates that all calcium agonists studied increased cortisol release by human adipocytes (3- to 6-fold, p < 0.001) within 90 minutes. The rapid stimulation of cortisol release by 1, 25-dihydroxyvitamin D3 suggested that this effect was mediated by the 1, 25D3-MARRS protein. To confirm this, we determined cortisol release after addition of either lumisterol (specific 1, 25D3-MARRS agonist) or 1β-25(OH)2-D (an antagonist of the rapid effects of 1, 25-dihydroxyvitamin D3) to the treatment medium containing 10 nM 1, 25-dihydroxyvitamin D3 and 10 nM cortisone (Figure 1B). The addition of 5 nM lumisterol increased cortisol release >3-fold compared with 10 nM 1, 25-dihydroxyvitamin D3 alone. As was the case for other calcium agonists, cortisol release plateaued within 90 minutes. Furthermore, we detected a dose-responsive inhibition of cortisol release in the presence of 1β-25(OH)2-D, which was maximally effective at a concentration of 5 nM.
AT II also dose responsively increased cortisol release within 90 minutes (Figure 2). This effect was mediated by the AT1 receptor, as indicated by the complete inhibition of cortisol release by 100 nM losartan (AT1 receptor antagonist), whereas 100 nM PD123177 (AT2 receptor antagonist) exerted no effect (Figure 2).
AT II Production by Human Adipocytes
AT II release was significantly suppressed in adipocytes exposed to 10 nM 1, 25-dihydroxyvitamin D3, 10 nM cortisone, or both hormones simultaneously compared with control (10 nM AT II) cells (Figure 3) at all intervals tested (90 minutes, 4 hours, and 24 hours).
We examined the effects of 24-hour exposure to 1, 25-dihydroxyvitamin D3 on the expression of 11β-HSD I in both the presence and absence of 10 nM cortisone. Although we did not detect cortisol release in the absence of exogenous cortisone with any of our treatments, expression of 11β-HSD I was increased by 50% in cells exposed to 1 and 10 nM 1, 25-dihydroxyvitamin D3 vs. untreated cells or cells exposed to 0.1 nM 1, 25-dihydroxyvitamin D3 (Figure 4A). Furthermore, the 1, 25-dihydroxyvitamin D3-mediated increase in 11β-HSD I expression was augmented by the addition of 10 nM cortisone (Figure 4B).
Although cortisol release increased after exposure to AT II, expression of 11β-HSD I was not significantly affected by 24-hour exposure to AT II (0.1 to 100 nM) when cortisone was included in the medium (Figure 5). However, in the absence of added cortisone, AT1 receptor expression was attenuated by exposure to 1, 25-dihydroxyvitamin D3 for 24 hours (Figure 5). Similarly, cortisone (10 nM, 24 hours) resulted in a 46% reduction in AT1 receptor expression in the absence of 1, 25-dihydroxyvitamin D3, and simultaneous exposure to both hormones attenuated AT1 receptor expression (Figure 6). We did not detect expression of the AT2 receptor with any treatment.
Data from these studies indicate that 1, 25-dihydroxyvitamin D3 in human adipocytes stimulates expression of 11β-HSD 1 and increases the conversion of cortisone to cortisol, which is subsequently released by adipocytes. The rapid nongenomic action of 1, 25-dihydroxyvitamin D3 to stimulate cortisol release in adipocytes is mediated, in part, through the 1, 25D3-MARRS because this response is mimicked by lumisterol but is prevented by 1β-25(OH)2-D, which antagonizes the rapid, membrane-associated signaling events resulting from exposure to 1, 25-dihydroxyvitamin D3. The net effect of 1, 25-dihydroxyvitamin D3-mediated increases in adipocyte [Ca2+]i is an increase in the amount of cortisol available to bind to activate glucocorticoid receptors. Accordingly, we suggest that strategies designed to suppress circulating 1, 25-dihydroxyvitamin D3 levels, such as increasing dietary calcium, may suppress adipocyte cortisol levels and mediate the preferential loss of visceral fat in obese humans.
The 1, 25D3-MARRS protein, also known as 58-kDa glucose-regulated protein and ERP57, belongs to a superfamily of multifunctional glucose-regulated and redox-sensitive proteins, which includes scaffolding proteins, thyroid hormone and estrogen-binding proteins, and factors involved in glycoprotein synthesis and immune response (9,29,30). Structural analysis of the mature 1, 25D3-MARRS protein predicts multiple sites for posttranslational modification by several signal transduction cascades associated with the rapid effects of 1, 25-dihydroxyvitamin D3, including protein kinase A and protein kinase C, both of which can modify the activity of calcium channels and, consequently, plasma membrane ion permeability (9). Although the membrane-initiated actions of 1, 25-dihydroxyvitamin D3 influence cellular secretory activity within minutes, these events seem to be inextricably linked to a functional nVDR. Zanello et al. (31) found that the normal fusion of individual secretory granules to the plasma membrane of osteoblasts exposed to 1, 25-dihydroxyvitamin D3 does not occur in osteoblasts from VDR knockout mice. Furthermore, the 1, 25-dihydroxyvitamin D3-mediated activation of L-type Ca2+ channels, and subsequent calcium influx, is impaired in osteoblasts from mice lacking a functional nuclear VDR (31). 1, 25D3-MARRS also influences the activation of the protein kinase C and protein kinase A signaling cascades that are capable of immediately influencing cellular functions but are also capable of affecting gene transcription mediated through a variety of second messengers (9,10).
Our studies in human adipocytes suggest that an increase in [Ca2+]i after exposure to 1, 25-dihydroxyvitamin D3 may enhance both the activity and expression of 11β-HSD 1. We are currently investigating the role of the 1, 25D3-MARRS protein in regulation of 11β-HSD 1 by 1, 25-dihydroxyvitamin D3. Studies from other investigators suggest tissue-specific regulation of 1, 25D3-MARRS protein by 1, 25-dihydroxyvitamin D3 and/or dietary calcium (32,33,34). Specific binding of [3H]1, 25-dihydroxyvitamin D3 is higher in basolateral membranes prepared from intestines of vitamin d-replete chicks than binding to plasma membranes from kidney or brain (32). In addition, 1, 25-dihydroxyvitamin D3 status influences the extent of specific binding and membrane receptor levels in the intestine and kidney but not in brain (33). Specifically, 1, 25D3-MARRS protein was upregulated in intestinal enterocytes from chicks fed 1, 25-dihydroxyvitamin D3-deficient diets, suggesting that the protein is involved in countering the effects of hormone deficiency by increasing absorption of available dietary calcium (32). In contrast, 1, 25D3-MARRS may be reduced in intestinal enterocytes from chicks fed a diet deficient in 1, 25-dihydroxyvitamin D3, phosphate, and calcium, as suggested by a reduction in specific ligand binding in the plasma membrane and attenuated tissue responsiveness after pharmacological dosing of 1, 25-dihydroxyvitamin D3 (34). We are particularly interested in determining the relationships between 1, 25-dihydroxyvitamin D3-mediated increases in adipocyte [Ca2+]i and 1, 25D3-MARRS in the preferential deposition of adipose tissue in the abdominal region and the potential for this interaction to amplify glucocorticoid action specifically in abdominal adipose tissue. In vitro, cortisol promotes adipocyte triglyceride accumulation by reducing basal and catecholamine-stimulated lipolysis while also enhancing the activity of lipoprotein lipase (35). Similarly, previous work from this laboratory demonstrates that increases in adipocyte [Ca2+]i are both antilipolytic and lipogenic (3). Based on our clinical observations demonstrating preferential loss of central adiposity in obese subjects consuming high-calcium, energy-restricted diets compared with lower calcium diets with the same level of energy restriction, we suggest that suppression of 1, 25-dihydroxyvitamin D3 levels by increasing dietary calcium may reduce local cortisol levels by reducing expression of 11β-HSD 1 and, thereby, attenuate local glucocorticoid action specifically in the truncal region (5).
AT II was originally identified as an important regulator of blood pressure, sodium balance, and extracellular fluid volume (36,37). The discovery of active renin-angiotensin systems in several tissues, including adipose tissue, has raised the possibility that locally produced AT II may play a prominent role in the pathophysiology of obesity and provides a potential pathway connecting obesity and the development of hypertension (38,39). The classical actions of AT II, such as vasoconstriction, sodium and water retention, cellular proliferation, and endothelial dysfunction, are mediated through the AT1 receptor (40). In addition, AT II has been shown to regulate adipogenesis and fat pad mass by increasing the activities of two key lipogenic enzymes, glycerol-3-phosphate and fatty acid synthase, and inducing differentiation of preadipocytes mediated through the AT1 receptor (41,42). These findings, along with others demonstrating that plasma angiotensinogen, plasma renin activity, and plasma angiotensin converting enzyme activity are elevated in obese humans, suggest that expanded adipose tissue stores may contribute to enhanced systemic renin-AT activity and, thereby, provide a physiological link between obesity and hypertension (43,44). However, this conclusion is complicated by studies demonstrating that infusion of high doses of AT II to rats markedly reduces body weight (45). Our data demonstrate that 1, 25-dihydroxyvitamin D3 is a negative regulator of the adipose tissue renin-AT system, reducing both AT II release and AT1 receptor expression. The effect of 1, 25-dihydroxyvitamin D3 to suppress AT II release was independent of the presence of either cortisone or AT II but was enhanced when these agents were included into the culture medium. Similarly, cortisone and 1, 25-dihydroxyvitamin D3 down-regulated AT1 expression, with a greater suppression observed when cells were exposed to both agents simultaneously. This may inhibit the mechanism by which mature adipocytes induce differentiation of preadipocytes and instead promote the accumulation of larger adipocytes.
Thus, the actions of 1, 25-dihydroxyvitamin D3 at the plasma membrane and nuclear level may effectively enhance local cortisol availability and action in adipose tissue and may suppress pathways designed to limit expansion of adipose tissue stores. These data demonstrate that reciprocal regulation of two pathways is likely involved. First, the enhanced expression of 11β-HSD 1 increases cortisol availability and, as a consequence, glucocorticoid signaling in adipose tissue. Secondly, a potential mechanism designed to limit further expansion of adipose tissue stores, specifically the expression of the AT1 receptor, is down-regulated. Consequently, the greater density of the components of these systems in the visceral depot would favor the development of central adiposity by stimulating factors biased toward adipose tissue deposition while constraining pathways that would counter lipogenesis. In contrast, strategies designed to lower 1, 25-dihydroxyvitamin D3 levels would be expected to reduce adipocyte [Ca2+]i and exert an antiobesity effect in vivo.
This study was supported by grants from the National Dairy Council.
Nonstandard abbreviations: [Ca2+]i, intracellular calcium; 1, 25D3-MARRS, 1, 25-dihydroxyvitamin D3 membrane-associated rapid response to steroid; nVDR, nuclear vitamin D receptor; 11β-HSD 1, 11β-hydroxysteroid dehydrogenase type 1; AT II, angiotensin II; AT1, angiotensin receptor type 1; PCR, polymerase chain reaction.