The subcellular localization of G(s) alpha, G(i) alpha 1&2, G(i) alpha 3, and G beta was studied in primary-cultured undifferentiated and differentiated, lipid replete, adipose cells. The results show a distinct distribution for each of these G-proteins and differences between differentiated and undifferentiated cells. All the G-proteins examined had a cytoplasmic localization; only G(i) alpha 1 and 2 showed a significant colocalization with the plasma membrane and this only in differentiated cells. Most studies using cells in culture have reported an intracellular localization for C-proteins, whereas in tissue sections the localization has been reported to be largely with the plasma membrane, with some intracellular localization. The results suggest that the cell-cell interactions or the specific geometry imposed by culture conditions favor the intracellular compared to peripheral localization of C-proteins. Alternately the posttranslational modifications necessary for C-protein insertion in the plasma membrane may be deficient in cultured cells. (C) 1997 Wiley-Liss, Inc.

The effects of glucose, K+, and cholecystokinin octapeptide (CCK-8) on intracellular free Ca2+ concentration ([Ca2+]i) and membrane-associated protein kinase C (PKC) activity were examined in cultured islet cells from neonatal rats. Raising the glucose concentration from 2.8 to 22.2 mM or external K+ (from 5 to 45 mM), or adding CCK-8 (200 nM) all triggered a [Ca2+]i surge that peaked between 3 and 10 min afterward, depending on the stimulus, and then declined, either to a suprabasal plateau (glucose and K+) or to basal levels (CCK-8). These same manipulations triggered a burst of membrane-associated PKC activity that peaked between 5 and 10 min and then variously declined. Incubation in Ca2+-free medium abolished both the effects of glucose and K+ on [Ca2+]i and the stimulation of membrane-associated PKC activity. The K+-triggered stimulation of PKC activity was also inhibited by pretreating the cells with the general Ca2+ entry blocker lanthanum (1 mM). However, incubation in Ca2+-free medium did not affect the CCK-8-induced release Ca2+ from internal stores, although it abolished the burst of membrane-associated PKC activity, which showed the importance of Ca2+ influx as opposed to internal release for PKC activation. Thus, glucose, the principal stimulator of insulin secretion, rapidly stimulates Ca2+ influx into islet cells from neonatal rats, and it is probably this influx that stimulates membrane-associated PKC activity.

The subcellular localization of the heterotrimeric C-proteins in hepatocytes in situ was compared to that in hepatocytes in primary culture. The ability of various ligands to activate adenylyl cyclase (AC) in membrane preparations was also investigated. In hepatocytes in situ the C proteins were mainly localized at the plasma membrane while in hepatocytes in culture they were predominantly cytoplasmic. The localization of the C-proteins in hepatocytes in situ correlates with their role in signal transduction. In homogenates prepared from the cultured cells, ligands which stimulate AC via G(s) alpha were without effect, which was consistent with the localization of G(s) alpha in the cytoplasmic and nuclear compartments. The ''relocalization'' of the G proteins to the cytoplasm when cells are cultured suggests that transmembrane signalling may be regulated by cell differentiation and cell-cell and cell-extracellular matrix interactions. (C) 1996 Wiley-Liss, Inc.

The effects of glucose, K+, and cholecystokinin octapeptide (CCK-8) on intracellular free Ca2+ concentration ([Ca2+]i) and membrane-associated protein kinase C (PKC) activity were examined in cultured islet cells from neonatal rats. Raising the glucose concentration from 2.8 to 22.2 mM or external K+ (from 5 to 45 mM), or adding CCK-8 (200 nM) all triggered a [Ca2+]i surge that peaked between 3 and 10 min afterward, depending on the stimulus, and then declined, either to a suprabasal plateau (glucose and K+) or to basal levels (CCK-8). These same manipulations triggered a burst of membrane-associated PKC activity that peaked between 5 and 10 min and then variously declined. Incubation in Ca2+-free medium abolished both the effects of glucose and K+ on [Ca2+]i and the stimulation of membrane-associated PKC activity. The K+-triggered stimulation of PKC activity was also inhibited by pretreating the cells with the general Ca2+ entry blocker lanthanum (1 mM). However, incubation in Ca2+-free medium did not affect the CCK-8-induced release Ca2+ from internal stores, although it abolished the burst of membrane-associated PKC activity, which showed the importance of Ca2+ influx as opposed to internal release for PKC activation. Thus, glucose, the principal stimulator of insulin secretion, rapidly stimulates Ca2+ influx into islet cells from neonatal rats, and it is probably this influx that stimulates membrane-associated PKC activity.

The ob/ob mouse white epididymal adipose tissue is endowed with very low lipolytic activity, due to abnormally low adenylyl cyclase activation in response to beta-adrenergic agents. The abundance of the two principal G-proteins that are responsible for the transduction of adenylyl cyclase is also decreased in several tissues of the ob/ob mouse, compared to levels in the lean mouse. By contrast, beta-adrenergic receptor levels appear normal in adipose tissue (Am J Physiol 1992; 263: C121-C129) and are elevated in liver (Am J Physiol 1994; 265: C1664-C1672), suggesting that the diminished abundance of G-proteins was responsible for the low lipolytic activity. We reassessed the relative importance of beta-adrenergic receptors and G-proteins in view of the discovery of the beta 2-adrenergic receptor. The major beta-AR isoform in mouse white adipose tissue is the beta 3-AR and its levels is severely decreased in the obese mouse. This indicates that the lipolytic defect in the ob/ob mouse is due to lack of beta 3-receptor function. Furthermore the extremely high sensitivity of this receptor to the ambient concentrations of GTP, explains the lack of response of adenylyl cyclase activity to the inhibitory effect of GTP in adipose tissue of the ob/ob mouse.

The metabolic response of adipose tissue to stimuli leading to lipid mobilization is important in determining the direction of metabolism and the degree to which adipose tissue can store lipids and release fatty acids in times of need. The lipolytic machinery is controlled by the activity of hormone-sensitive lipase, which in turn is controlled by the cellular levels of cAMP. The production of cAMP is abnormal in the adipose tissue of some animal models of obesity. In the ob/ob mouse, the defective cAMP production has been associated with deficient levels of some of the isoforms of the guanine nucleotide transducing G-proteins and also with the low expression and functionality of the beta 3-adrenergic receptor (beta 3-AR). The recent discovery of the ob gene product leptin calls into question the role of the ob gene in the regulation of the cAMP cascade in adipose tissue. The importance of the beta 3-AR and leptin in regulating human adipose tissue metabolism remains to be clarified.

Norepinephrine-induced thermogenesis was substantial in adipocytes from brown adipose tissue (BAT) of cold-acclimated guinea pigs but absent in adipocytes from BAT of warm-acclimated guinea pigs. There was no thermogenic response to any beta 3-adrenergic agonist (CL-316,243, ZD-7114, BRL-28410, CGP-12177). The receptor was characterized as a beta 1-adrenoceptor. Adrenergic agonists stimulated adenylate cyclase in membranes from BAT of both warm- and cold-acclimated guinea pigs also via a beta 1-adrenoceptor; beta 3-adrenergic agonists had no effect. Glucose transport by brown adipocytes from warm-acclimated guinea pigs was not stimulated by either norepinephrine or insulin. Cold acclimation induced the appearance of stimulation of glucose transport by norepinephrine in association with the appearance of a large capacity for thermogenesis, but there was little improvement in response to insulin. GLUT4 was present in membranes from BAT of both warm- and cold-acclimated guinea pigs. Insulin is known to have an antilipolytic effect on both BAT and white adipose tissue of guinea pigs. Thus there is a selective lack of insulin-regulated glucose transport that is not improved by cold acclimation. Guinea pigs may have a mutated component of the translocation mechanism for GLUT4. beta 3-Adrenoceptors appear to be absent in brown adipocytes of adult guinea pigs, as in white adipocytes of guinea pigs, yet are known to be present in the gut. Tissue-specific expression of beta 3-adrenergic receptors in guinea pigs may differ from that in rats, in which receptors are expressed in the adipose tissues and gut.

Lipolysis and adenylyl cyclase (AC) activation in response to β‐adrenergic agents are abnormally low in white epididymal adipose tissue (WAT) of the ob/ob mouse. The abundance of G‐proteins (G s α and G i α) linked to AC is also abnormally low. By contrast, β‐adrenergic receptor (β‐AR) levels were previously found to be normal in WAT and elevated in liver. The relative importance of various forms of the β‐AR in mouse WAT was reassessed in view of the discovery of the β 3 ‐AR. The results show that (1) the β 3 ‐AR is mainly responsible for AC activation in lean‐mouse WAT; (2) the β 3 ‐AR is only partly responsible for AC activation in obese mouse WAT; and (3) GTP modulates β 3 —‐but not β 1 —‐or β 2 ‐AR activation of AC in a biphasic manner. Therefore, the β 3 ‐AR appears responsible for the well‐known bimodal effect of GTP on β‐adrenergic receptor‐mediated AC activity in WAT.

The effects of extracellular divalent cations on the intracellular Ca2+ concentration ([Ca2+]i) in neonatal rat islet cells were investigated to determine whether these cells, like several others, have signal-generating surface cation sensors. Raising the external Ca2+ concentration by 1 mM increments triggered either sustained increases in [Ca2+]i or large sharp [Ca2+]i spikes followed by return to a suprabasal level. The external Ca(2+)-triggered [Ca2+]i responses were abolished by treating the cells with the inhibitor of inositol phospholipid hydrolysis, neomycin (1.5 mM), but not by another phospholipase C inhibitor, U-73,122 (2.5 microM), or the voltage-sensitive Ca2+ channel blockers nifedipine (20 microM) and methoxyverapamil (D600; 50 microM). [Ca2+]i responses were also triggered by barium (Ba2+; 1 mM) and cobalt (Co2+; 1 mM). The Ba2+ responses were also inhibited by neomycin and unaffected by nifedipine or D600 and the Co2+ response required external Ca2+. Therefore, neonatal rat pancreatic islet cells may display divalent cation receptors/sensors on their surfaces. Activation of these putative receptors, which are coupled to neomycin-sensitive, voltage-independent, dihydropyridine-insensitive channels, by Ca2+, Ba2+ or Co2+ would trigger [Ca2+]i responses by opening these channels to admit external Ca2+ into the cell. The physiological function(s) of such cell-surface divalent cation receptors/sensors and the [Ca2+]i surges they generate in pancreatic islet cells is not known.

Stromal vascular cells from epididymal fat pads of lean and obese mice were cultured in a medium (α‐MEM) containing fetal bovine serum (FBS) and cell replication followed for 11 days. In both types of cells, confluence occurred at 4–5 days, after which virtual growth arrest occurred in lean‐mouse cells while replication continued, albeit at a slower rate in obese‐mouse cells. Little or no lipid accumulation or glycerol‐3‐phosphate dehydrogenase (GPDH) activity was observed under these conditions. When a differentiation mixture consisting of insulin, corticosterone and isobutylmethylxanthine was added to the serum‐containing α‐MEM, a proportion of the lean‐mouse cells accumulated triglycerides and GPDH activity increased significantly, indicating differentiation. By contrast, little or no differentiation occurred in obese‐mouse cells. When cells grown in serum‐containing α‐MEM were transferred to a serum‐free defined medium at confluence, extensive differentiation and maturation occurred in lean‐mouse cells but not in obese‐mouse cells. Similar experiments were conducted in cells isolated from the retroperitoneal fat pad. Although the growth pattern was similar to that of epididymal preadipocytes, the retroperitoneal lean‐ and obese‐mouse cells differentiated more readily than epididymal cells, as shown by the GPDH specific activity. These data suggest that cells from obese mice are resistant to differentiation under conditions that support extensive differentiation in lean‐mouse cells.

The modulation of intracellular free calcium concentration ([Ca2+]i) by cAMP was compared in pancreatic beta-cells of lean (+/+) and obese (ob/ob) mice. Neither forskolin nor 8-bromo-cAMP significantly affected basal [Ca2+]i in unstimulated lean and obese mouse beta-cells. In obese, but not in lean mouse beta-cells, adding forskolin or 8-bromo-cAMP during the glucose-induced [Ca2+]i response triggered external Ca(2+)-dependent [Ca2+]i oscillations with a duration of 5-11 s and a frequency of 2.3-4.8 min-1. The induction of oscillations by cAMP required both a stimulatory glucose concentration and membrane depolarization. (Sp)-cAMPS, did induce oscillations in lean mouse beta-cells. However, these oscillations were different from those seen in obese mouse beta-cells and required higher concentrations of (Sp)-cAMPS. The inducibility of fast oscillations in obese mouse beta-cells indicates hypersensitivity of these cells to cAMP, and suggests an abnormal behavior of K+ and/or Ca2+ channels.

The ob and db genes produce similar hormonal anomalies in mice. Although the expression of the syndromes diverges with age, at 8-12 wk both ob/ob and db/db mice are hyperglycemic and hyperinsulinemic and show evidence of hypercorticoidism. Nevertheless, membranes isolated from livers of ob/ob and db/db mice behave differently in terms of adenylyl cyclase activity and beta-adrenergic receptor function. There are three times as many beta 2-adrenergic receptor binding sites and a threefold increase in the response to catecholamines in ob/ob mouse liver membranes than in comparable preparations from normal controls or db/db mice. By contrast, the two main G proteins of liver membranes (Gs alpha and Gi alpha 2) are less abundant in the mutants, ob/ob and db/db, than in their respective lean controls. Adrenalectomy normalizes the exaggerated response to beta-adrenergic agonists and the number of beta-adrenergic binding sites in the ob/ob mouse. This shows that the enhanced beta-adrenergic receptor response is linked to hypercorticoidism. Cellular maturation and differentiation (D. C. Watkins, J. K. Northrup, and C. C. Malbon, J. Biol. Chem. 262: 10651-10657, 1987) and diseases such as obesity and diabetes (cf. N. McFarlane-Anderson, J. Bailly, and N. Bégin-Heick, Biochem. J. 282: 15-23, 1992) have been associated with modifications in the complement of G proteins detected in cells. However, the relationship among levels, types, and intracellular localization of G proteins in tissues and their influence on the transduction of the message to an effector system, such as adenylyl cyclase, are not yet well understood.

Glucose-induced insulin release, glucose-induced rises in intracellular free Ca2+ concentration ([Ca2+]i), and voltage-dependent Ca2+ channel activity were assessed in monolayer cultures of beta-cells from 3-5-day-old rats. The glucose-stimulated insulin secretory responses and [Ca2+]i rises were like those in adult rat beta-cells rather than fetal rat beta-cells. Voltage-dependent Ca2+ channel antagonists decreased glucose-induced insulin secretion, aborted the [Ca2+]i rise and, like deprivation of extracellular Ca2+, prevented the glucose-induced rise in [Ca2+]i when added before the glucose challenge. The presence of nifedipine-sensitive, voltage-dependent Ca2+ channels was demonstrated directly by measuring Ca2+ currents using the whole-cell configuration of the patch-clamp technique and indirectly by measuring [Ca2+]i after membrane depolarization by 45 mM K+ or 200 microM tolbutamide. Thus, in cultured beta-cells of 3-5-day-old rats the coupling of glucose stimulation to Ca2+ influx is essentially mature, in contrast to what has been reported for fetal or very early neonatal cells.

HGFu and Ob17 are cell lines derived from adipose tissue of lean (+/?) and ob/ob mice, respectively. Neither adenylyl cyclase activity nor G protein abundance and subcellular distribution have been assessed previously in these cells. Cyclase activity was low and resistant to catecholamine stimulation in both cell lines. However, the enzyme could be stimulated to high levels by forskolin and Mn 2+ . G s α (largely the long isoform), G i α2, and Gβ were the major G protein subunits identified. The levels of G protein mRNA expression were similar in both cell lines and, unlike actin expression, did not change as a result of differentiation. Immunoblotting and ADP‐ribosylation of the G peptides corroborated these results. Assessment of the subcellular localization of the subunits by indirect epifluorescence and scanning confocal microscopy showed that each of the subunits had a characteristic subcellular pattern. G s α showed vesicular cytoplasmic and nuclear staining; G i α2 colocalized with actin stress fibers and disruption of these structures altered the distribution of G i α2; β subunits showed some colocalization with the stress fibers as well as a cytoplasmic vesicular and nuclear pattern. As a result of differentiation, there was reorganization of the actin, together with the G i α2 and β fibrous patterns. Both cell lines showed similar modifications. The induction of differentiation in these cells is therefore not associated with changes in adenylyl cyclase activity nor of the abundance of G‐protein subunits, although reorganization of some of these subunits does accompany actin reorganization.

The HGFu and Ob17 cell lines, derived from adipose tissue of lean (+/?) and ob/ob mice, respectively, express several G-protein peptides. Investigation of the expression and subcellular localization of the G(i) alpha 3 subunit showed that this peptide is associated with the Golgi apparatus. These findings indicate a role for this subunit in vesicular traffic and are in agreement with the view of the adipocyte as a secretory cell.

The HGFu and Ob17 cell lines, derived from adipose tissue of lean (+/?) and ob/ob mice, respectively, express several G-protein peptides. Investigation of the expression and subcellular localization of the Giα3 subunit showed that this peptide is associated with the Golgi apparatus. These findings indicate a role for this subunit in vesicular traffic and are in agreement with the view of the adipocyte as a secretory cell.Key words: preadipocytes, differentiation, immunofluorescence, confocal microscopy, obesity, Golgi system.

Glucose-induced shifts in intracellular free Ca2+ concentration ([Ca2+]i) were quantitatively and temporally the same in ob/ob and +/+ beta-cells. In both, epinephrine promptly and protractedly inhibited the glucose-induced [Ca2+]i surge via a pertussis toxin-sensitive alpha 2-adrenergic mechanism that was reversible by potassium depolarization. When added before glucose, epinephrine blocked completely in the ob/ob beta-cells, but in the +/+ beta-cells it produced a delayed, reduced, and transient intracellular Ca2+ (Ca2+i) surge. Neither the ATP-sensitive K+ channel blocker tolbutamide nor the large-conductance Ca(2+)-activated K+ channel (Kmaxi) blocker charybdotoxin reversed the effect of epinephrine. Tetraethylammonium (TEA), a blocker of both the Kmaxi and the delayed-rectifier K+ channel, and forskolin attenuated the effect of epinephrine in +/+ but not in the ob/ob beta-cells. The data show that 1) alpha 2-adrenoreceptor activation decreases the glucose-stimulated Ca2+i surge in +/+ beta-cells primarily by activating a tolbutamide- and charybdotoxin-insensitive, TEA- and forskolin-sensitive K+ channel; 2) the hypersecretion of insulin in ob/ob beta-cells is not due to enhanced glucose-induced Ca2+ influx; and 3) the ob/ob beta-cells are aberrant with regard to alpha 2-adrenergic modulation.

G-protein levels were assessed in liver and brain membranes of lean and obese mice. ADP-ribosylation and immunodetection studies revealed a decrease in the abundance of Gs and Gi alpha-subunits in the liver membranes of obese mice compared with lean mice. In contrast, in brain membranes, the abundance of these proteins was not significantly different between lean and obese mice. Studies at the mRNA level in both liver and brain revealed no difference in gene expression between lean and obese mice. Protein and mRNA studies both showed that Gs, Gi alpha 1, Gi alpha 2, Go alpha and G beta subunits are present in brain membranes, and Gi alpha 3 is barely detectable. In liver, Ga alpha, Gi alpha 2 and G beta subunits are the major constituents, whereas Gi alpha 1, Gi alpha 3 and Go alpha are barely detectable. It is possible that the differences observed at the protein level are due to different rates of translation of the mRNA. Different rates of release of the alpha-subunits from the membrane and/or different rates of degradation would also explain these results.