Identification of omentin mRNA in human epicardial adipose tissue: comparison to omentin in subcutaneous, internal mammary artery periadventitial and visceral abdominal depots.

OBJECTIVE: The purpose of this study was to determine the relative distribution of omentin and visfatin mRNA in human epicardial, peri-internal mammary, upper thoracic, upper abdominal and leg vein subcutaneous adipose tissue as well as the distribution of omentin in the nonfat cells and adipocytes of human omental adipose tissue. BACKGROUND: Omentin is found in human omentum but not subcutaneous fat. Omentin and visfatin are considered markers of visceral abdominal fat. RESEARCH DESIGN AND METHODS: The mRNA content of omentin and visfatin was measured by qRT-PCR analysis of fat samples removed from humans undergoing cardiac or bariatric surgery. RESULTS: Omentin mRNA in internal mammary fat was 3.5%, that in the upper thoracic subcutaneous fat was 4.7% while that in the other subcutaneous fat depots was less than 1% of omentin in epicardial fat. The distribution of visfatin mRNA did not vary between the five depots. Omentin mRNA was preferentially expressed in the nonfat cells of omental adipose tissue since the omentin mRNA content of isolated adipocytes was 9% of that in nonfat cells, and similar results were seen for visfatin. The amount of omentin mRNA in differentiated adipocytes was 0.3% and that of visfatin 4% of that in nonfat cells. The amount of omentin mRNA in preadipocytes was virtually undetectable while that of visfatin was 3% of that in freshly isolated nonfat cells from omental adipose tissue. CONCLUSION: Omentin mRNA is predominantly found in epicardial and omental human fat whereas visfatin mRNA is found to the same extent in epicardial, subcutaneous and omental fat.

TNFalpha release by the nonfat cells of human adipose tissue.

OBJECTIVE: The primary aim was to investigate the relative importance of the adipocytes vs the nonfat cells present in human adipose tissue with respect to release of immunoreactive tumor necrosis factor-alpha (TNFalpha). The second aim was to examine the correlation between body mass index (BMI) and the subsequent release of adiponectin and TNFalpha by explants of human subcutaneous and visceral adipose tissue incubated in primary culture for 48 h. RESULTS: We found that the maximal release of TNFalpha was seen during the first 4 h of a 48-h incubation by explants of human adipose tissue in primary culture. Over 95% of the TNFalpha released to the medium by human adipose tissue explants over a 4-h incubation came from the nonfat cells present in the adipose tissue. The release of TNFalpha by the nonfat cells released during collagenase digestion was slightly higher than that by the cells present in the adipose tissue matrix after collagenase digestion. TNFalpha release by the combined matrix and isolated nonfat cells was greater than that by explants of tissue indicating some upregulation induced by collagenase digestion. Immunoreactive TNFalpha disappeared from the medium with a half-time of approximately 10 h. There was a positive correlation coefficient of 0.79 between TNFalpha release by tissue explants and the BMI of the fat donors as well as a correlation of 0.52 between BMI and release by adipocytes. TNFalpha release negatively correlated [-0.60] with adiponectin release by adipose tissue. The release of TNFalpha was far less than that of adiponectin or IL-6, and less than that of plasminogen activator inhibitor-1, hepatocyte growth factor, or leptin over a 4-h incubation of human adipose tissue explants. TNFalpha release over 4 h was enhanced by lipopolysaccharide and inhibited by a cyclooxygenase-2 inhibitor. CONCLUSION: The release of TNFalpha by adipose tissue of obese humans is primarily due to the nonfat cells present in adipose tissue. TNFalpha is a short-lived adipokine whose release by human adipose tissue in primary culture correlates with the BMI of the fat donors.

Comparison of PGE2, prostacyclin and leptin release by human adipocytes versus explants of adipose tissue in primary culture.

The present studies were designed to investigate the sites of PGE(2), prostacyclin and leptin formation in human adipose tissue. Most of the PGE(2) and prostacyclin formation by adipose tissue explants from obese humans after 48 h in primary culture was due to blood vessels and other tissues not digested by collagenase. However, there was appreciable PGE(2) formation by adipocytes over a 48 h incubation and leptin formation was only seen in adipocytes. An increase in COX-2 immunoreactive protein was also seen after incubation of isolated human adipocytes for 48 h. The release of PGE(2) by adipocytes incubated for 48 h was about 4% that by intact adipose tissue explants while the release of prostacyclin was about 1.5% that by tissue. However, in a different experimental design where PGE(2) formation was measured over 2 h in the presence of 20 microM arachidonic acid the formation of PGE(2) by adipocytes after 48 h prior incubation in primary culture was 38% of that by tissue explants. Dexamethasone enhanced leptin release by adipocytes while inhibiting PGE(2) release and COX-2 up-regulation. The mechanisms involved in up-regulation of COX-2 activity during primary culture of adipocytes and the inhibition of this by dexamethasone do not appear to involve p38 MAPK or p42-44 MAPK. Interleukin I(beta) further enhanced PGE(2) formation by adipocytes but did not affect leptin formation. In conclusion, these data indicate that leptin release is exclusively a function of adipocytes while prostanoids are made by both adipocytes and the other cells present in human adipose tissue

Regulation of human beta(1)-adrenergic receptors and their mRNA in neuroepithelioma SK-N-MC cells: effects of agonist, forskolin, and protein kinase A.

We determined the effect of long-term exposure to beta-agonists on beta(1)-adrenergic receptors (beta(1)-AR) in human neuroepithelioma SK-N-MC cells because earlier studies have indicated that beta(1)-AR in this cell line are resistant to agonist-induced down-regulation. Exposing SK-N-MC cells to isoproterenol for 24 hr reduced the density of beta(1)-AR by 72%, whereas forskolin, an activator of all the isoforms of adenylyl cyclase, failed to affect the density of beta(1)-AR. Measurement of beta(1)-AR mRNA levels by the ribonuclease protection assay revealed that isoproterenol-induced down-regulation of beta(1)-AR was associated with a sharp decline in beta(1)-AR mRNA, while forskolin also failed to affect this parameter. The differences between the effects of isoproterenol and forskolin on beta(1)-AR were unrelated to cyclic AMP levels, since both agents increased cyclic AMP equally. Next, we determined the role of cyclic AMP-dependent protein kinase A (PKA) in this phenomenon. Inhibition of PKA by its specific inhibitor, H-89 [N-[2-((p-bromocinnamyl)amino)ethyl]-5-isoquinolinesulfonamide, 2HCl], markedly reduced the magnitude of the isoproterenol-mediated down-regulation of the beta(1)-AR and its mRNA. Transient expression of the catalytic subunit of PKA in SK-N-MC cells down-regulated beta(1)-AR independently of isoproterenol. Therefore, PKA is central to the effect of beta-agonists in down-regulating beta(1)-AR, and its spatial compartmentalization and access to the receptor appear to be essential components of its action.

Obesity is induced in mice heterozygous for cyclooxygenase-2.

In mice heterozygous for the cyclooxygenase-2 gene (COX-2+/-) the body weight was enhanced by 33% as compared to homozygous COX-2-/- mice. The weights of the gonadal fat pads in COX-2+/- mice were enhanced by 3.5 to 4.7 fold as compared to COX-2-/- mice and by 1.5 to 3.5 fold as compared to wild-type controls+/+ Serum leptin levels and leptin release by cultured adipose tissue of COX-2+/- mice were both elevated as compared to either control or COX-2-/- animals. The basal release of PGE2 or 6 keto PGF1alpha per fat pad over a 24 h incubation of adipose tissue was reduced by 80% and 95% respectively in tissue from COX-2-/- mice. NS-398, a specific COX-2 inhibitor, inhibited leptin release by 27% in adipose tissue from control mice, 31% in tissue from COX-1-/- mice and by 23% in tissue from COX-2+/- mice while having no effect on leptin release by adipose tissue from COX-2-/- mice. These data indicate that heterozygous COX-2 mice develop obesity which is not secondary to a defect in leptin release by adipose tissue.

Stimulation of leptin release by arachidonic acid and prostaglandin E(2) in adipose tissue from obese humans.

The purpose of this study was to examine the effect of arachidonic acid and its metabolites on leptin formation by explants of human adipose tissue over a 48-hour incubation in primary culture. We found that arachidonic acid or prostaglandin E(2) (PGE(2)) stimulated leptin release by explants of subcutaneous adipose tissue from obese humans. The stimulatory effect of arachidonic acid on leptin formation was blocked by NS-398, a cyclooxygenase-2 (COX-2) inhibitor. There was appreciable release of PGE(2) to the medium over 48 hours, and this was inhibited by 99% in the presence of 200 nmol/L dexamethasone or 5 micromol/L NS-398. The increase in PGE(2) release correlated with induction of COX-2 activity during the 48-hour incubation. The increase in COX-2 activity was blocked by 200nmol/L dexamethasone. The level of leptin mRNA at 48 hours was reduced by 28% if PGE(2) was added in the absence of dexamethasone, while in the presence of dexamethasone, the amount of leptin mRNA was enhanced by 156%. These data suggest that when upregulation of COX-2 is blocked by dexamethasone, exogenous PGE(2) enhances both leptin release and leptin mRNA accumulation by explants of human adipose tissue in primary culture.

Differential regulation of carnitine palmitoyltransferase-I gene isoforms (CPT-I alpha and CPT-I beta) in the rat heart.

Carnitine palmitoyltransferase-I (CPT-I) is a major control point for fatty acid oxidation. Two kinetically different isoforms, CPT-I alpha and CPT-I beta, have been identified. Cardiac ventricular myocytes are the only cells known to express both CPT-I isoforms. In this study, we characterized the differential regulation of CPT-I alpha and CPT-I beta expression in the heart. Expression of the CPT-I alpha gene was very high in the fetal heart and declined following birth. CPT-I beta was also highly expressed in fetal myocytes and remained so throughout development. CPT-I alpha mRNA abundance was increased in both the liver and heart of diabetic or fasted rats, but CPT-I beta mRNA levels were not altered in these states. A high fat diet elevated expression of the CPT-I alpha gene in the liver but not in the heart. The fat content of the diet did not affect the expression of CPT-I beta. Cultures of neonatal rat cardiac myocytes were transfected with luciferase reporter genes driven by CPT-I alpha or CPT-I beta promoters. Two regions of the CPT-I alpha promoter, including an upstream region (-1300/-960) and a region in the proximal promoter (-193/-52) contributed equally to basal expression in cardiac myocytes. Basal transcription of CPT-I alpha was dependent on Sp1 sites and a CCAAT box in the proximal promoter. Our data indicate that the CPT-I beta gene is expressed in a tissue specific manner, but that it is not subject to the same developmental or hormonal controls imposed on CPT-I alpha. In addition some aspects of CPT-I alpha expression are confined to the liver. The data presented here thus suggest that two types of differential regulation of CPT-I genes exist: (a) differential control of CPT-I alpha and CPT-I beta gene expression in the heart and (b) differential regulation of CPT-I alpha expression in the heart and liver.

Regulation of leptin release by troglitazone in human adipose tissue.

In pieces of human subcutaneous adipose tissue incubated in primary culture for 48 hours, the release of leptin was stimulated by 50% in the presence of 3.3 micromol/L troglitazone. Insulin (0.1 nmol/L) and dexamethasone (200 nmol/L) stimulated leptin release by 30% and 300%, respectively. Troglitazone in combination with either insulin or dexamethasone had no effect on leptin release. Instead, troglitazone inhibited leptin release in the presence of both dexamethasone and insulin. The stimulatory effect of troglitazone on leptin release was also mimicked by 1 micromol/L 15-deoxy-delta(12-14)prostaglandin J2 (dPGJ2). However, if the concentration of dPGJ2 was increased to 10 micromol/L in the presence of dexamethasone, there was a decrease in leptin release, as well as of lactate formation and lipolysis. These data indicate that both stimulatory and inhibitory effects of troglitazone and dPGJ2 can be seen on leptin release by human adipose tissue.

Regulation of leptin release and lipolysis by PGE2 in rat adipose tissue.

The role of eicosanoids formed by adipose tissue from rats was examined in the presence of the specific cyclooxygenase-2 inhibitor NS-398. This agent totally blocked the release of prostaglandin E2 (PGE2) by rat adipose tissue over a 24-h incubation in primary culture. The final concentration of PGE2 after 24 h was 12 nM, and half-maximal inhibition of PGE2 formation required 35 nM NS-398. While inhibition of PGE2 formation by NS-398 had no effect on basal leptin release or lipolysis, it enhanced the lipolytic action of 10 nM isoproterenol by 36%. The in vivo administration of PGE2 doubled serum leptin. PGE2 also directly stimulated leptin release by rat adipose tissue incubated in the presence of 25 nM dexamethasone, which inhibited endogenous PGE2 formation by 94%. The inhibition of lipolysis as well as the stimulation of leptin release by PGE2 were mimicked by N6-cyclopentyladenosine (CPA). These data indicate that exogenous PGE2 can stimulate leptin release by adipose tissue when the basal formation of PGE2 is blocked by dexamethasone. However, while the endogenous formation of PGE2 does not appear to regulate basal lipolysis or leptin release, it may play a role in the activation of lipolysis by catecholamines.

Eicosanoids as endogenous regulators of leptin release and lipolysis by mouse adipose tissue in primary culture.

Prostaglandin E(2) (PGE(2)) stimulated leptin release over a 24-h incubation of mouse adipose tissue in primary culture. The maximal stimulation of leptin release was seen with 100 nm PGE(2). The role of endogenous eicosanoids in the regulation of lipolysis and leptin formation was examined in the presence of NS-398, a selective cyclooxygenase-2 inhibitor. NS-398 at a concentration of 5 microm enhanced lipolysis by 30% and lowered leptin release by 24%. This concentration of NS-398 almost completely inhibited PGE(2) formation. An inhibition of basal lipolysis by PGE(2) or N(6)-cyclopentyladenosine (CPA) was seen in the presence but not in the absence of NS-398. CPA, whose receptor, like that of PGE(2) inhibits cyclic AMP accumulation in adipose tissue, also enhanced leptin release. These data indicate that PGE2 can stimulate leptin release and suggest that endogenous eicosanoids affect both lipolysis and leptin formation by mouse adipose tissue.

Synergism between insulin and low concentrations of isoproterenol in the stimulation of leptin release by cultured human adipose tissue.

The release of leptin by pieces of human adipose tissue incubated in primary culture for 24 or 48 hours in the presence of dexamethasone was reduced by isoproterenol. An inhibition of leptin release was observed at 24 hours in the presence of isoproterenol and was mediated by beta1-adrenergic receptors, since it was blocked by the specific beta1-adrenoceptor antagonist CGP-20712A. The inhibitory effect of 33 nmol/L isoproterenol on leptin release was reversed in the presence of 0.1 nmol/L insulin to a 2-fold stimulation of leptin release. These data suggest that the primary mechanism by which insulin stimulates leptin release is to blunt the inhibitory effects of beta1-adrenergic receptor agonists, and low concentrations of catecholamines actually enhance the stimulation of leptin release by insulin.

Regulation of lipolysis and leptin biosynthesis in rodent adipose tissue by growth hormone.

The present study examined the effects of growth hormone (GH) on lipolysis and leptin release by cultured adipose tissue from rats and mice incubated for 24 hours in primary culture. A stimulation of leptin release by GH in rat adipose tissue was found in the presence of 25 nmol/L dexamethasone, and this was accompanied by a 28% increase in leptin mRNA content. GH stimulated lipolysis in rat adipose tissue in the presence of 0.1 nmol/L CL 316,243. In contrast, basal lipolysis in mouse adipose tissue was stimulated by GH, but this was not accompanied by an increase in leptin release. However, in the presence of insulin plus triiodothyronine (T3), the stimulation of lipolysis by GH was abolished and GH increased leptin release. These results indicate that GH can stimulate leptin release by both mouse and rat adipose tissue in the absence of a stimulation of lipolysis. In contrast, under conditions in which lipolysis is stimulated by GH, there is no effect on leptin release.

Stimulation of lipolysis but not of leptin release by growth hormone is abolished in adipose tissue from Stat5a and b knockout mice.

The present studies examined the effects of growth hormone (GH) on lipolysis and leptin release by adipose tissue from mice incubated for 24 h in primary culture. In adipose tissue from control mice GH enhanced lipolysis without affecting leptin release. The lipolytic action of GH was unaffected in adipose tissue from Stat5b-/- male mice but leptin release was enhanced by GH in fat from Stat5b-/- mice. In adipose tissue from Stat5ab-/- female mice no significant lipolytic action of GH was seen but leptin release was enhanced by GH. An insulin-like effect of GH on glucose conversion to lactate was also seen in mice deficient in Stat5ab-/-. These results suggest that the lipolytic action of GH involves the Stat5 proteins while the insulin-like effects of GH on glucose metabolism and leptin release involve different mechanisms.

Hormonal regulation of 18S RNA, leptin mRNA, and leptin release in adipocytes from hypothyroid rats.

The present studies were designed to examine the regulation of leptin release in primary cultures of adipocytes from fed hypothyroid rats incubated with hormones for 24 hours. Leptin release was increased in the presence of dexamethasone, while the decrease in leptin mRNA content over a 24-hour incubation was reduced by dexamethasone. Dexamethasone did not affect the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA or 18S RNA content of adipocytes. Insulin increased leptin release by adipocytes in both the absence and presence of dexamethasone. Although insulin also prevented the loss of leptin mRNA, this effect was less than that observed for GAPDH mRNA or 18S RNA content. In isolated adipocytes, the loss of almost half the 18S RNA content over a 24-hour incubation was prevented in the presence of insulin but not oxytocin or epidermal growth factor (EGF). The specific beta3 catecholamine agonist CI 316,243 inhibited the effects of dexamethasone on leptin release and leptin mRNA accumulation, as did EGF, without affecting 18S RNA content. Oxytocin inhibited the increase in leptin release due to dexamethasone without affecting leptin mRNA levels. These data indicate that although dexamethasone and insulin are positive regulators of leptin release, only dexamethasone specifically prevented the loss of leptin mRNA in cultured rat adipocytes. In contrast, insulin, but not dexamethasone, prevented the marked loss in 18S RNA observed over a 24-hour incubation of rat adipocytes.

Stimulation of leptin release by actinomycin D in rat adipocytes.

A greater understanding of the factors causing the enhanced release of leptin by adipocytes in obesity is needed. Experiments were designed to determine the effects of actinomycin D on leptin release by isolated rat adipocytes during primary culture for 24 hr. In adipocytes from fed hypothyroid rats, the initial rate of leptin release over the first 6 hr was not maintained over the next 18 hr. The decline in leptin release by adipocytes in primary culture between 6 and 24 hr was reduced markedly by either dexamethasone or actinomycin D. Both actinomycin D and dexamethasone also reduced the loss of leptin mRNA seen over the 24-hr incubation. Maximal effects on leptin release and leptin mRNA accumulation required only 0.1 microM of actinomycin D, a concentration that had no significant effect on the 18S RNA content of adipocytes at the end of a 24-hr incubation. In contrast to the reduced loss of leptin mRNA seen at 24 hr, the loss of glyceraldehyde-3-phosphate dehydrogenase messenger ribonucleic acid (GAPDH mRNA) was enhanced in the presence of 0.1 microM of actinomycin D. The effects of dexamethasone could be differentiated from those of actinomycin D by the finding that cycloheximide blocked the reduced loss of leptin mRNA due to dexamethasone while having no effect on that due to actinomycin D. These results point to a unique regulation of leptin release and leptin mRNA levels by actinomycin D.

Effect of tri-iodothyronine on leptin release and leptin mRNA accumulation in rat adipose tissue.

Leptin, the product of the obese gene, is produced by white adipocytes. The release of leptin, as well as leptin mRNA content, was enhanced in adipocytes isolated from hypothyroid rats. The administration of tri-iodothyronine (T3) 8 h before death inhibited leptin release by adipocytes incubated for 6 or 24 h. Direct addition of T3 to pieces of adipose tissue enhanced the loss of leptin mRNA seen over 24 h in the presence of dexamethasone plus the beta3-adrenergic agonist Cl 316,243. In contrast, if pieces of adipose tissue were incubated with dexamethasone plus insulin, enhanced the T3 accumulation of leptin mRNA. These results indicate that T3 enhances net adipocyte leptin mRNA accumulation in a condition that approximates the fed state (presence of insulin) but inhibits leptin mRNA accumulation in a condition that approximates the fasted state (absence of insulin).

Identification of a retinoic acid response domain involved in the activation of the beta 1-adrenergic receptor gene by retinoic acid in F9 teratocarcinoma cells.

The density of beta 1-adrenergic receptors (beta 1-AR) is up-regulated upon differentiation of embryonic F9 teratocarcinoma cells by retinoic acid (RA) to the primitive endodermal phenotype. To identify the domains involved in RA-mediated activation of beta 1-AR gene transcription, three kb of 5'-flanking sequence of the beta 1-AR gene were ligated to a luciferase reporter gene and transiently transfected into F9 cells that were pre-exposed to 100 nM RA for 2 days. By generating deletions in the beta 1-AR promoter, a region between -125 and -100 was found to mediate a 3-fold induction in cells exposed to RA for an additional 2 days. Through site-directed mutagenesis of this region, it was determined that the RA responsive element (RARE) was organized as a direct repeat separated by 5 nucleotides in which the 5'-most AGGTCG half-site was between nucleotides -106 and -101 and the 3'-most AGGTCA half-site was between nucleotides -117 and -112. The RA receptor alpha (RAR alpha) isoform bound to the oligomer representing the sequences between -125 and -100 as a heterodimer complex with the retinoid X receptor alpha (RXR alpha). In a separate study, it was determined that the nucleotides between -125 and -100 are involved in thyroid hormone-mediated activation of the beta 1-AR gene in ventricular myocytes. Therefore, transcriptional activation of the beta 1-AR gene by thyroid hormone or RA involves a single binding site in the promoter.