Promoting effect of glucocorticoids on the differentiation of human adipocyte precursor cells cultured in a chemically defined medium.

Stromal-vascular cells obtained from adult human subcutaneous adipose tissue were cultured in a chemically defined serum-free medium. In the presence of 0.2 nM triiodothyronine and 0.5 microM insulin, up to 25% of the cells were able to undergo terminal adipose differentiation within 18 d, as assessed by lipid accumulation and the expression of lipoprotein lipase (LPL) and glycerol-3-phosphate dehydrogenase (GPDH) activities. Addition of cortisol resulted in a potent dose-dependent stimulation of the adipose differentiation process. Cortisol could be replaced by dexamethasone and partly by aldosterone, but not by sex steroids. The proportion of differentiated cells was dependent upon the age of the donor; when isolated from young adults, up to 70% of the stromal-vascular cells expressed the adipocyte phenotype as compared with 5-10% when the cells were isolated from the oldest subjects. An inverse relationship was observed between the age of the 27 normal-weight donors and the extent of GPDH expression after maintenance of the cells for 18 d in chemically defined medium supplemented with insulin, triiodothyronine, and cortisol (r = -0.787, P less than 0.001). It is concluded that adult human adipose tissue still contains precursor cells that are able to undergo adipose differentiation in vitro. This improved culture system may offer the opportunity to characterize other adipogenic factors as well as antiadipogenic factors involved in the control of adipose tissue growth.

Prostacyclin as an indicator of preadipocyte transformation: studies in vivo by microdialysis and in vitro.

Nontransformed (Ob1771) and polyoma virus-transformed (Ob17PY) mouse cells from the preadipocyte Ob17 clonal line have been compared in their ability to release prostaglandins in vitro as well as in vivo as assayed by in situ microdialysis. Prostaglandin FE2, prostaglandin-2 alpha and mainly prostacyclin are released in larger amounts (4- to 10-fold) by Ob17PY cells in vitro and Ob17PY-induced tumors in vivo as compared to Ob1771 preadipocytes in vitro and periepididymal adipose tissue in vivo. In contrast to Ob1771 preadipocytes, none of these prostanoids appear to be involved in the control of proliferation or differentiation of Ob17PY cells in serum-free culture medium. However, prostacyclin, the level of which is the most affected by transformation, might be considered as a valuable indicator of fibrosarcoma development.

Fetuin modulates growth and differentiation of Ob17 preadipose cells in serum-free hormone-supplemented medium.

A serum-free hormone-supplemented medium able to support the growth of rodent adipose precursor cells has been used to characterize additional components from serum required for the differentiation of preadipose Ob17 cells into adipose-like cells. Fetuin is shown to behave as a growth-promoting agent for these cells. In addition to growth hormone, triiodothyronine and a low-molecular weight component(s) also purified from serum, fetuin is required for the full expression of the differentiation program. Other serum proteins as well as other mitogenic factors are unable to substitute for fetuin. A possible role of fetuin in the development of adipose tissue is discussed.

Differentiation of ob 17 preadipocytes to adipocytes. Effects of prostaglandin F2alpha and relationship to prostaglandin synthesis.

The adipose conversion of ob 17 preadipose cells can be irreversibly blocked when prostaglandin F2alpha is included post-confluence for a minimum of 24 h in insulin-containing media. Prostaglandins E1 and E2 are inactive. The lack of adipose conversion is accompanied by the maintenance of a fusiform cell shape, by a slow increase in cell number and by a potent rise in de novo prostaglandin synthesis; it is paralleled by the absence of the characteristic phenotypes of adipose conversion. The multiple effects of prostaglandin F2alpha are dose-dependent, with half-maximal concentrations ranging from 10 to 40 nM. The absence of differentiation and the high rate of prostaglandin synthesis in the presence of prostaglandin F2alpha are likely a consequence of a sustained growth, as also observed with other growth-promoting agents (bovine retinal extract and cat serum). Indomethacin, while suppressing endogenous prostaglandin synthesis, is unable to reverse the long-term and multiple effects of prostaglandin F2alpha. Although adipose conversion normally follows a decrease in prostaglandin production (R. Négrel and G. Ailhaud, Biochem. Biophys. Res. Commun. (1981) 98, 768-777), these results indicate that both events can be dissociated.

Binding of lipoproteins and regulation of cholesterol synthesis in cultured mouse adipose cells.

Binding of human lipoproteins to cultured mouse Ob17 preadipose and adipose cells was studied, using labeled VLDL, LDL and apoprotein E-free HDL. In each case, saturation curves were obtained, yielding linear Scatchard plots. The Kd values were found to be respectively 6.4, 31 and 24 micrograms/ml for VLDL, LDL and apoprotein E-free HDL, whereas the maximal numbers of binding sites per cell were 4.2 X 10(4), 1.5 X 10(4) and 2.5 X 10(5). The binding of 125I-LDL was competitively inhibited by LDL greater than VLDL greater than total HDL; human LDL and mouse LDL were equipotent in competition assays. Methylated LDL and apoprotein E-free HDL were not competitors. In contrast, the binding of 125I-apoprotein E-free HDL was competitively inhibited by apoprotein E-free HDL greater than total HDL and the binding of 125I-HDL3 by mouse HDL. Thus, mouse adipose cells possess distinct apoprotein B, E and apoprotein E-free HDL binding sites which can recognize heterologous or homologous lipoproteins. The cell surface receptor of LDL in mouse preadipose cells shows similarities with that described for human fibroblasts, since: (1) the LDL binding initiated the process of internalization and degradation of the apoprotein B and apoprotein E-containing lipoproteins; (2) receptor-mediated uptake of cholesterol LDL led to a parallel but incomplete decrease in the [14C]acetate incorporation into cholesterol and in the activity of HMG-CoA reductase. Growing (undifferentiated) or growth-arrested cells (differentiated or not) showed no significant changes in the Kd values for lipoprotein binding. In contrast, the maximal number of binding sites correlated with the proliferative state of the cells and was independent of cell differentiation. The results are discussed with respect to cholesterol accumulation in adipose cells.

Characterization of high-density lipoprotein binding and cholesterol efflux in cultured mouse adipose cells.

Binding of high-density lipoproteins to cultured mouse Ob1771 adipose cells was studied, using labeled human HDL3, mouse HDL and apolipoprotein AI- or AII-containing liposomes. In each case, saturation curves were obtained, yielding linear Scatchard plots. The Kd values were found to be respectively 18, 42, 30 and 3.4 micrograms/ml, whereas the maximal binding capacities were found to be 160, 100, 90 and 21 ng/mg of cell protein. Apoprotein AI not inserted into liposomes did not bind. The binding of 125I-HDL3 was competitively inhibited by apolipoprotein AI-containing liposomes greater than mouse HDL greater than HDL3. The binding of 125I-labeled apolipoprotein AI- and 125I-labeled apolipoprotein AII-containing liposomes was competitively inhibited by HDL3, apolipoprotein AI- and apolipoprotein AII-containing liposomes. Dimyristoylphosphatidylcholine liposomes containing or not cholesterol did not interfere with the binding of labeled HDL3 or apolipoprotein-containing liposomes. Binding studies on crude membranes of Ob1771 adipose cells revealed the presence of intracellular binding sites for LDL and HDL3. Thus, adipose cells have specific binding sites for apolipoprotein E-free HDL and apolipoprotein AI (or AII) is the ligand for these binding sites. Long-term exposure of adipose cells to LDL cholesterol as a function of LDL concentration led to an accumulation of cellular unesterified cholesterol. This process was saturable and reversible as a function of time and concentration by exposure to HDL3 or apolipoprotein AI-containing liposomes, whereas apolipoprotein AII-containing liposomes did not promote any cholesterol efflux. Since long-term exposure of adipose cells to LDL and HDL3 did not affect the number of apolipoprotein B,E receptors and apolipoprotein E-free binding sites, respectively, it appears that adipose cells do not show efficient cholesterol homeostasis and thus could accumulate or mobilize unesterified cholesterol.

Lipoprotein lipase and monoacylglycerol lipase activities during maturation of ob17 preadipocytes.

Lipoprotein lipase and monoacylglycerol lipase activities of developing ob17 preadipocytes were assayed through the selective inhibition of the former activity by 0.4% Triton X-100. After confluence both enzymes rose tao a peak activity around 13--15 days which was enhanced in insulin-treated cells. Inclusion of differentiation-blocking agents (bromodeoxyuridine, prostaglandin F2 alpha and bovine retinal extract) in the culture medium led in both enzymes to levels similar to those obtained in exponentially growing cells, in contrast to unspecific monoester hydrolase. Marked changes in the relative proportion of released to intracellular lipoprotein lipase activities were observed during adipose conversion. Our studies indicate that lipoprotein lipase and monoacylglycerol lipase are integral part of the program of differentiation of adipose cells.

Autocrine control of adipose cell differentiation by prostacyclin and PGF2 alpha.

The mitogenic-adipogenic effect exerted by arachidonic acid, which leads to terminal differentiation of Ob1771 mouse preadipocytes, has been shown to be (i) blocked by cyclooxygenase inhibitors, (ii) mimicked by a stable analogue of prostacyclin (carbaprostacyclin) and (iii) potentiated by PGF2 alpha. Since these prostanoids are known to be synthesized and secreted by preadipocytes, we have proposed that both prostacyclin as the key mediator and PGF2 alpha as a modulator control the expression of terminal events of adipose conversion by means of an autocrine mechanism (Gaillard, D. et al. and Negrel, R. et al. Biochem. J. (1989) 257, 389-397 and 399-405). In order to test this hypothesis, the release of prostacyclin, characterized under the form of its stable degradation product 6-keto-PGF1 alpha, and that of PGF2 alpha have been studied in the culture medium of Ob1771 cells. A striking increase in the release of 6-keto-PGF1 alpha and to a minor degree of PGF2 alpha was observed when cells were exposed to arachidonic acid as shown by using [3H]arachidonic acid prelabelled cells or by radio-immunoassays. Since antagonists of PGF2 alpha and PGI2 receptors were not available, specific antibodies directed against PGF2 alpha and 6 beta-PGI1, another stable analogue of prostacyclin, were added as neutralizing agents in the culture medium. These antibodies were able to counteract the mitogenic-adipogenic effect of arachidonic acid. Prostacyclin and PGF2 alpha thus appear as autocrine mediators in the process of adipose conversion.

Regulation of cholesterol synthesis and binding of lipoproteins in cultured rat intestinal epithelial cells.

The regulation of cholesterol synthesis has been studied using a rat epithelial intestinal cell line (IRD 98) as a cellular model. As observed in other cell types, mevinolin increases the levels of 3-hydroxy-3-methylglutaryl coenzyme A reductase (EC 1.1.1.34) and concomitantly reduces the incorporation of [14C]acetate into cholesterol. Free cholesterol is able to suppress reductase activity. In contrast, when cells are shifted from standard culture medium to lipoprotein-deficient medium, an increase in hydroxymethylglutaryl-CoA reductase specific activity (2-5-fold) is observed. The possible regulatory roles of the different classes of human lipoproteins were thus compared. The effects of a long-term exposure to LDL and HDL vary according to cell density. In actively growing cells, VLDL and LDL cause a decrease in the level of hydroxymethylglutaryl-CoA reductase, whereas HDL do not have a significant effect. In contrast, in subconfluent preresting cells, HDL provoke large decreases in hydroxymethylglutaryl-CoA reductase activity as compared to VLDL and LDL. While LDL binding is constant, the maximal binding capacity of HDL in subconfluent cells is seven times that of actively growing cells. Altogether, these results suggest an important role for HDL in the regulation of intestinal cholesterol synthesis.

Differentiation of ob 17 preadipocytes to adipocytes. Triggering effects of clofenapate and indomethacin.

Conversion of ob 17 preadipocytes to mature adipose cells is accelerated by addition of clofenapate or of indomethacin, in either the absence or presence of insulin. General stimulation of triacylglycerol-pathway enzymes is observed, as well as dramatic increase in endogenous fatty-acid synthesis. This increase is a function of drug concentration and exposure time. In contrast to indomethacin, the continuous presence of clofenapate after the cells reached confluence was required to observe the effects on adipose conversion. Growth of ob 17 fibroblasts in the presence of 5-bromo-2'-deoxyuridine normally prevents their differentiation to adipose cells. Addition of either clofenapate or indomethacin to these cells at confluence overrides this block. The effects of hypolipidemic drugs such as clofenapate observed on a long-term basis in vitro are consistent with the results of studies on adipose tissue in vivo.

Glucocorticoids and insulin promote plasminogen activator inhibitor 1 production by human adipose tissue.

Plasminogen activator inhibitor 1 (PAI-1) is likely to play a role in vascular disease, primarily in subjects with android obesity. It has been demonstrated that PAI-1 is overexpressed in adipose tissue from obese subjects and that visceral adipose tissue produced more PAI-1 than subcutaneous fat. In the present study, the effect of insulin and glucocorticoids, which are key mediators of adipose tissue metabolism, was examined in relation to PAI-1 synthesis by human adipose tissue explants (HAT), collagenase isolated human adipocytes (IHA), cultured human stromal cells (cSC), and differentiated adipocytes from the murine clonal cell line 3T3-F442A. A significant increase in PAI-1 antigen release (1.5-fold) from HAT was detectable after 16 h of treatment with insulin concentrations of at least 10(-8) mol/l. This was associated with a PAI-1 mRNA increase. Concomitant addition of insulin (10(-8) mol/l) to forskolin (5 x 10(-5) mol/l) reversed the decrease in PAI-1 antigen caused by forskolin alone. No effect on PAI-1 antigen was observed when insulin was incubated with IHA or cSC. 3T3 F442A cells were sensitive to insulin with a four- and twofold increase in PAI-1 antigen and mRNA levels, respectively, after 16 h of stimulation with 10(-8) mol/l. Dexamethasone (DXM) significantly enhanced PAI-1 antigen and mRNA expression by HAT (1.5- and 2.5-fold increase, respectively) at concentrations of at least 10(-8) mol/l. A higher stimulation was observed with IHA (sevenfold increase) and with the differentiated 3T3 F442 cell line. Cortisol was found to be less potent than DXM. No effect was observed when glucocorticoids were incubated with cSC. Coincubation of HAT with insulin (10(-7) mol/l) and DXM (10(-7) mol/l) led to an additive effect on PAI-1 synthesis. These results support the hypothesis that PAI-1 expression in human adipose tissue is controlled by insulin and glucocorticoids and may help to explain the increase in plasma PAI-1 levels observed in patients with android obesity.

Adipose angiotensinogen is involved in adipose tissue growth and blood pressure regulation.

White adipose tissue and liver are important angiotensinogen (AGT) production sites. Until now, plasma AGT was considered to be a reflection of hepatic production. Because plasma AGT concentration has been reported to correlate with blood pressure, and to be associated with body mass index, we investigated whether adipose AGT is released locally and into the blood stream. For this purpose, we have generated transgenic mice either in which adipose AGT is overexpressed or in which AGT expression is restricted to adipose tissue. This was achieved by the use of the aP2 adipocyte-specific promoter driving the expression of rat agt cDNA in both wild-type and hypotensive AGT-deficient mice. Our results show that in both genotypes, targeted expression of AGT in adipose tissue increases fat mass. Mice whose AGT expression is restricted to adipose tissue have AGT circulating in the blood stream, are normotensive, and exhibit restored renal function compared with AGT-deficient mice. Moreover, mice that overexpress adipose AGT have increased levels of circulating AGT, compared with wild-type mice, and are hypertensive. These animal models demonstrate that AGT produced by adipose tissue plays a role in both local adipose tissue development and in the endocrine system, which supports a role of adipose AGT in hypertensive obese patients.

Hormonal regulation of adipose differentiation.

Adipose differentiation is a multistep process with the following sequence: adipoblasts --> preadipocytes --> adipocytes. Adipogenic agents are only involved in the terminal differentiation of preadipocytes to adipocytes by means of circulating hormones (growth hormone, glucocorticoids, or triiodothyronine) and locally produced hormones (prostacyclin). Fatty acids also behave as hormones and act as transcriptional regulators of lipid-related genes. Once differentiated, adipocytes become secretory cells able to synthetize and release an impressive number of peptide and nonpeptide compounds, suggesting a potential link between excess of adipose tissue mass and various physiopathphysiologic consequences.

Differentiation of preadipose cells: paracrine role of prostacyclin upon stimulation of adipose cells by angiotensin-II.

Prostacyclin (PGI2), the major metabolite of arachidonic acid in adipose tissue, has been shown to play a key role in the process of preadipose cell differentiation in vitro. Moreover, angiotensin-II (Ang II) is able to induce the production of PGI2 in suspensions of isolated adipocytes as well as in the interstitial fluid of rat adipose tissue. A possible role of Ang II in the control of the autocrine-paracrine adipogenic effect of PGI2 has been investigated, using cells of the Ob1771 preadipocyte clonal line cultured in serum-free chemically defined medium. Whereas both preadipose and adipose cells were able to produce PGI2 upon exposure to arachidonic acid, only adipose cells were able to do so when challenged with Ang II. In agreement with this observation, the ability of Ang II to induce preadipose cells to differentiate required the simultaneous presence of differentiated cells. Such coculture experiments show that the promoting effect of Ang II on preadipose cell differentiation was strongly reduced by aspirin, antibodies able to neutralize PGI2, and the AT2 receptor antagonist PD123177, but not by the AT1 receptor antagonist losartan. Together, these results support Ang II as being able, by means of binding to a receptor of the AT2 subtype present in adipose cells, to control the adipogenic effect of PGI2 through a paracrine mode of action.

Differential response of preadipocytes and adipocytes to prostacyclin and prostaglandin E2: physiological implications.

The major prostaglandins (PGs) locally produced in adipose tissue both in rodent and man are PGE2 and prostacyclin (PGI2). We have recently described PGI2 as an autocrine promoter and/or amplifier of terminal differentiation of cultured preadipocytes in several species. The effectiveness and specificity of PGI2 as an adipogenic agent are related to its ability to induce in preadipocytes intracellular increases of both cAMP and free calcium. Moreover, PGs of the E series are well known to exert an antilipolytic effect in mature adipocytes. These observations have prompted us to address two questions of physiological interest: 1) Is PGI2 still able to increase cAMP in differentiated adipocytes, behaving thus as a lipolytic agent, and 2) Is PGE2 able to negatively modulate cAMP production in adipose precursor cells, behaving thus as a counteracting effector of PGI2 action? Our results, with respect to cAMP production and/or lipolysis and antilipolysis, demonstrate clearly that in adipose tissue of both rat and man, PGI2 exclusively affects adipose precursor cells whereas PGE2 exclusively affects adipocytes. We propose a model of concerted action for both PGs in the development of adipose tissue mass, PGI2 behaving as an adipogenic-hyperplastic effector and PGE2 as an antilipolytic-hypertrophic effector.

Influence of estrogenic status on the lipolytic activity of parametrial adipose tissue in vivo: an in situ microdialysis study.

Ovarian hormones have been shown to modulate the metabolism of adipose cells obtained from adipose tissue of different animals. The aim of this study was to better understand the short- and long-term influences of estrogens on the in vivo lipolytic response of rat parametrial fat pads, determined by measurement of extracellular glycerol concentrations using in situ microdialysis. Possible direct effects of estrogens on lipolysis were studied by perfusion of a potent estrogenic analogue such as moxestrol. Moxestrol (10(-6) M) failed to increase glycerol concentrations in estrus, diestrus, or 8-day ovariectomized animals. However, the basal glycerol concentrations and the lipolytic responses stimulated by 10(-6) M isoproterenol were decreased in parametrial fat pads of diestrus, compared with estrus, rats. Greater decreases in basal and stimulated glycerol concentrations were observed in rats that had been ovariectomized for 8, 15, or 30 days. In ovariectomized rats, isoproterenol-induced lipolysis was restored to the levels observed in diestrus animals by a daily injection of 17 beta-estradiol for a period of 7 days. These results implicate estrogens as long-term modulators of in vivo basal and stimulated lipolytic responses of rat parametrial fat pad.

Angiotensin II as a trophic factor of white adipose tissue: stimulation of adipose cell formation.

White adipose tissue is known to contain the components of the renin-angiotensin system giving rise to angiotensin II (AngII). In vitro, prostacyclin is synthesized from arachidonic acid through the activity of cyclooxygenases 1 and 2 and is released from AngII-stimulated adipocytes. Prostacyclin, in turn, is able to favor adipocyte formation. Based upon in vivo and ex vivo experiments combined to immunocytochemical staining of glycerol-3-phosphate dehydrogenase (GPDH), an indicator of adipocyte formation, it is reported herein that AngII favors the appearance of GPDH-positive cells. In the presence of a cyclooxygenase inhibitor, this adipogenic effect is abolished, whereas that of (carba)prostacyclin, a stable analog of prostacyclin that bypasses this inhibition, appears unaltered. Taken together, these results are in favor of AngII acting as a trophic factor implicated locally in adipose tissue development. It is proposed that AngII enhances the formation of GPDH-expressing cells from preadipocytes in response to prostacyclin released from adipocytes.

Angiotensinogen-deficient mice exhibit impairment of diet-induced weight gain with alteration in adipose tissue development and increased locomotor activity.

White adipose tissue is known to contain the components of the renin-angiotensin system, which gives rise to angiotensin II from angiotensinogen (AGT). Recent evidence obtained in vitro and ex vivo is in favor of angiotensin II acting as a trophic factor of adipose tissue development. To determine whether AGT plays a role in vivo in this process, comparative studies were performed in AGT-deficient (agt(-/-)) mice and control wild-type mice. The results showed that agt(-/-) mice gain less weight than wild-type mice in response to a chow or high fat diet. Adipose tissue mass from weaning to adulthood appeared altered rather specifically, as both the size and the weight of other organs were almost unchanged. Food intake was similar for both genotypes, suggesting a decreased metabolic efficiency in agt(-/-) mice. Consistent with this hypothesis, cellularity measurement indicated hypotrophy of adipocytes in agt(-/-) mice with a parallel decrease in the fatty acid synthase activity. Moreover, AGT-deficient mice exhibited a significantly increased locomotor activity, whereas metabolic rate and mRNA levels of uncoupling proteins remained similar in both genotypes. Thus, AGT appears to be involved in the regulation of fat mass through a combination of decreased lipogenesis and increased locomotor activity that may be centrally mediated.

Physiology and pathophysiology of the adipose tissue renin-angiotensin system.

The renin-angiotensin system has long been recognized as an important regulator of systemic blood pressure and renal electrolyte homeostasis, and local renin-angiotensin systems have also been implicated in pathological changes of organ structure and function by modulation of gene expression, growth, fibrosis, and inflammatory response. Recently, substantial data have been accumulated in support of the notion that adipose tissue, besides other endocrine functions, also hosts a local renin-angiotensin system. In the first part of this review, we describe the components of the adipose tissue renin-angiotensin system in human and rodent animal models with respect to regulation of angiotensinogen expression and secretion, formation of angiotensin peptides, and the existence of angiotensin II receptors. In the second part, we describe the role of the adipose tissue renin-angiotensin system in the process of adipogenic differentiation and in the regulation of body weight. We also detail the differential regulation of the adipose tissue renin-angiotensin system in obesity and hypertension and thereby also speculate on its possible role in the development of obesity-associated hypertension. Although some findings on the adipose tissue renin-angiotensin system appear to be confusing, its involvement in the physiology and pathophysiology of adipose tissue has been confirmed by several functional studies. Nevertheless, future studies with more carefully described phenotypes are necessary to conclude whether obesity (by stimulation of adipogenic differentiation) and hypertension are associated with changes of renin-angiotensin system activity in adipose tissue. If so, the physiological relevance of this system in animal models and humans may warrant further interest.

Evidence for a novel regulatory pathway activated by (carba)prostacyclin in preadipose and adipose cells.

Prostacyclin, one of the major prostanoids generated in adipose tissue, has been previously described as an autocrine/paracrine adipogenic effector, acting, in preadipose cells, by means of cAMP and free Ca2+ as cell surface receptor-mediated messengers. The present study presents evidence for the first time that its stable analogue, carbaprostacyclin, is unique among prostanoids in regulating the expression of two differentiation-dependent genes in preadipose and adipose cells in a way distinct from that elicited by its cell surface receptor. This regulation is likely mediated by some member(s) of the peroxisome proliferator-activated receptor family and suggests that prostacyclin behaves as an intracrine effector of adipose cell differentiation.