Studies on the mechanism of insulin action: basic concepts and clinical implications

Hepatic acetyl CoA links adipose tissue inflammation to hepatic insulin resistance and type 2 diabetes

Impaired insulin-mediated suppression of hepatic glucose production (HGP) plays a major role in the pathogenesis of type 2 diabetes (T2D), yet the molecular mechanism by which this occurs remains unknown. Using a novel in vivo metabolomics approach, we show that the major mechanism by which insulin suppresses HGP is through reductions in hepatic acetyl CoA by suppression of lipolysis in white adipose tissue (WAT) leading to reductions in pyruvate carboxylase flux. This mechanism was confirmed in mice and rats with genetic ablation of insulin signaling and mice lacking adipose triglyceride lipase. Insulin's ability to suppress hepatic acetyl CoA, PC activity, and lipolysis was lost in high-fat-fed rats, a phenomenon reversible by IL-6 neutralization and inducible by IL-6 infusion. Taken together, these data identify WAT-derived hepatic acetyl CoA as the main regulator of HGP by insulin and link it to inflammation-induced hepatic insulin resistance associated with obesity and T2D.

Mas receptor deficiency is associated with worsening of lipid profile and severe hepatic steatosis in ApoE-knockout mice

The classical renin-angiotensin system pathway has been recently updated with the identification of additional molecules [such as angiotensin converting enzyme 2, ANG-(1–7), and Mas receptor] that might improve some pathophysiological processes in chronic inflammatory diseases. In the present study, we focused on the potential protective role of Mas receptor activation on mouse lipid profile, liver steatosis, and atherogenesis. Mas/apolipoprotein E (ApoE)-double-knockout (DKO) mice (based on C57BL/6 strain of 20 wk of age) were fed under normal diet and compared with aged-matched Mas and ApoE-single-knockout (KO), as well as wild-type mice. Mas/ApoE double deficiency was associated with increased serum levels of atherogenic fractions of cholesterol, triglycerides, and fasting glucose compared with wild-type or single KO. Serum levels of HDL or leptin in DKO were lower than in other groups. Hepatic lipid content as well as alanine aminotransferase serum levels were increased in DKO compared with wild-type or single-KO animals. Accordingly, the hepatic protein content of mediators related to atherosclerotic inflammation, such as peroxisome proliferator-activated receptor-α and liver X receptor, was altered in an adverse way in DKO compared with ApoE-KO. On the other hand, DKO mice did not display increased atherogenesis and intraplaque inflammation compared with ApoE-KO group. In conclusion, Mas deletion in ApoE-KO mice was associated with development of severe liver steatosis and dyslipidemia without affecting concomitant atherosclerosis. Mas receptor activation might represent promising strategies for future treatments targeting both hepatic and metabolic alterations in chronic conditions clustering these disorders.

The in-vitro effects of insulin and the effects of acute fasting on cardiac beta-adrenoceptor responses in the short-term streptozotocin-diabetic rat

This study demonstrates that in-vitro incubation with insulin results in a reduction of the sensitivity of cardiac tissues from streptozotocin-diabetic rats to isoprenaline. Following incubation with insulin, the sensitivity of left atria and papillary muscles from diabetic animals was not significantly different from those of control animals. Insulin incubation had no effect on the sensitivity of diabetic tissues to forskolin or on ventricular beta-adrenoceptor number. Reduction of blood glucose in the absence of insulin by fasting, did not affect the sensitivity of tissues from diabetic animals to isoprenaline. These results suggest that insulin itself can directly reduce beta-adrenoceptor sensitivity without altering receptor number.

Enhanced hepatic adenylate cyclase activity in rats with portacaval shunt

Male Wistar rats were submitted to a portacaval anastomosis (PCA). Control rats were sham operated and pair fed. After 20 days, PCA led to a decrease in liver weight (-40%) and fasting blood glucose (-35%) and to an increase in fasting glucagonemia (+65%). The in vitro response of adenylate cyclase in hepatic membranes to GTP, Gpp(NH)p, fluoride, and forskolin (in the absence of GTP), and to glucagon (in the presence of GTP) was greater in PCA rats than in controls (by 30-54%) whereas the response to L-isoproterenol (in the presence of GTP) was only slightly increased (by 8%) and that to vasoactive intestinal peptide (in the presence of GTP) was similar in both groups of rats. The binding of [125I]glucagon and [125I]VIP to liver membranes did not differ in both groups of animals. It is concluded that the hepatic adenylate cyclase system from PCA rats responded better to stimuli involving efficiently the guanyl nucleotide stimulatory site Ns. This implies that the fasting hypoglycemia observed in these animals, in spite of the hyperglucagonemia, was due to either the refractoriness of a step distal to adenylate cyclase activation or to limited glucose production by an atrophic liver.

Fasting and refeeding modulate neutral amino acid transport activity in the basolateral membrane of the rat exocrine pancreatic epithelium: fasting-induced insulin insensitivity

The effects of fasting and refeeding on amino acid transport in the perfused rat exocrine pancreas were investigated using a rapid dual tracer dilution technique. Unidirectional amino acid influx (15 s) was quantified (relative to the extracellular tracer D-mannitol) over a wide range of perfusate concentrations in pancreata isolated frm fed and 24 h, 48 h, and 72 h fasted and 72 h fasted and refed (24 h) animals. In fed animals transport of phenylalanine (1-24 mM) and L-serine (1-50 mM) was saturable and weighted non-linear regression analyses of the overall transport indicated an apparent Kt = 10 +/- 3 mM and Vmax = 7.0 +/- 1.0 mumol/min per g (n = 7) for phenylalanine and Kt = 16 +/- 3 mM and Vmax = 20.6 +/- 2.1 mumol/min per g (n = 5) for serine. Fasting animals for 24 h or 48 h did not change the kinetics of either phenylalanine or serine transport. After a 72 h fast the rate of phenylalanine transport (Vmax = 15.9 +/- 2.9 mumol/min per g, n = 5) was enhanced whereas the transport affinity (Kt = 11 +/- 3 mM) remained unaltered. L-Serine transport was essentially unaltered. When 72 h fasted animals were refed for 24 h the Vmax for phenylalanine transport was reduced to values observed in fed animals. In parallel experiments refeeding had no significant effect on serine transport. Perfusion of pancreata isolated from 72 h fasted animals with bovine insulin (1 mU/ml or 1 microU/ml) did not stimulate either phenylalanine or serine transport.(ABSTRACT TRUNCATED AT 250 WORDS)

Ultrastructure of hepatic cholesterol crystals in the hypercholesterolemic - diabetic rat

The cellular morphology of lipid accumulation in the liver was examined in normal rats fed a diet containing cholesterol and cholic acid, and streptozotocin-diabetic rats fed the same diet. The cholesterol-fed non diabetic rats displayed moderate hypercholesterolemia (average cholesterol 317 mg/dl) whereas the cholesterol-fed diabetic rats exhibited severe hypercholesterolemia (cholesterol greater than 1300 mg/dl). Ultrastructural studies were performed on hepatic tissues following in situ fixation and water soluble embedment, which were used to reduce lipid extraction and minimize structural distortions. Although both groups exhibited hepatocyte lipid droplets, the accumulation was markedly accentuated in the diabetic animals. The Kupffer cells of the diabetic animals contained cytosolic lipid crystals that were membrane delimited and showed lattice ordering 3.9 +/- 2.2 nm periodicity. These findings suggest that cholesteryl ester crystals of the cholesteric phase, similar to those found in atherosclerotic lesions, may form in other cellular foci exposed to abnormally high plasma lipid levels.

Hepatic insulin effects independent of changes in Ca++ metabolism

The possibility of Ca++ acting as second messenger for insulin in rat liver was investigated using the net stimulation of 14C-glucose incorporation into glycogen by isolated hepatocytes as an index of insulin action. An insulin effect could be partially sustained in the virtual absence of Ca++ and Mg++ and a maximal insulin effect could be observed in the presence of either Ca++ or Mg++, suggesting that extracellular Ca++ is not required for insulin action. Inhibiting the activity of calmodulin, an intracellular mediator of Ca++ action, with trifluoperazine had little effect on insulin action. The efflux of 45Ca from prelabeled hepatocytes was not altered by the presence of insulin arguing against insulin-induced changes in Ca++ fluxes. Collectively, these results do not support the role of Ca++ as second messenger for insulin action in liver.

Circadian profiles of insulin receptors in insulin-dependent diabetics in usual and poor metabolic control

We have characterized the 24-h changes of insulin receptors on erythrocytes from patients with insulin-dependent diabetes of long duration. These diabetics were studied both in usual and poor metabolic control. Moreover, we have examined daytime changes of insulin receptors on monocytes from newly discovered diabetics. In both erythrocyte and monocyte studies, diabetics were compared to healthy controls. At insulin tracer concentration, insulin receptor binding to erythrocytes from diabetics in usual control and normal volunteers underwent a statistically significant diurnal variation with high binding values in the early morning, low daytime values with a nadir in the late afternoon, and a peak around midnight. Even diabetics in poor metabolic control due to insulin deprivation had preserved a similar 24-h rhythm of erythrocyte insulin receptors. Insulin receptor binding to monocytes at insulin tracer concentration declined significantly during the day both in newly discovered diabetics and in healthy controls. The mechanisms responsible for the acute phase changes of insulin-receptor binding are unknown, but the receptor changes seem related to the fed state. Moreover, analysis of the temporal interrelationship between erythrocyte insulin binding and plasma insulin concentration in diabetics during the 24-h period suggests that in these patients insulin may be one of the factors determining the rapid insulin receptor regulation.

Hepatic effects of glucose and insulin independent of cyclic nucleotide changes

To determine if cGMP might function as a second messenger for insulin, an in situ liver perfusion system was established in which hepatic effects of insulin could be correlated with changes in cyclic nucleotides. Several combinations of insulin (10 mU/ml) and glucose (50 mg/ml) were infused (0.1 ml/min) for 30 min into fasted normal and diabetic rats with removal of a similar volume of blood. Samples of livers were removed at the beginning and end and at various times during the perfusion. In normal animals perfused with buffer alone, hepatic glycogen content fell. When glucose (with or without added insulin) was added to the perfusate, glycogen levels rose. With buffer alone, there was no change in the independent (I) form of glycogen synthase at 10 min but a modest increase at 30 min. With insulin and/or glucose, there as a large increase in the I-form of the enzyme at 10 min and a further rise at 30 min. Neither cGMP nor cAMP changed even though tissue samples were obtained at multiple times throughout the perfusion. Cyclic nucleotides were also measured in liver slices exposed to insulin (1 mU/ml) after 30 min of pre-incubation for stabilization. Although significant increases in cGMP were noted in the tissue exposed to insulin, similar significant rises also occurred in appropriately paired control slices. When glucagon was used in both the in situ perfusion and the paired liver slice systems, the expected rapid and large increases in cAMP levels occurred attesting to the validity of both approaches in evaluating hepatic cyclic nucleotide responses. These results plus the paucity of convincing data in the literature strongly suggest that cGMP can no longer be considered a candidate for the putative second messenger of insulin.

Insulin action

Our understanding of the molecular basis of insulin actin remains incomplete, but important new insights have been achieved recently. All available evidence to date indicates that intracellular signalling by the hormone results from its initial interaction with specific cell surface receptors. Insulin receptors from all tissues studied to date appear to be minimally composed of two Mr 125,000 subunits denoted as alpha and two Mr 90,000 subunits denoted as beta. The beta subunit is extremely sensitive to proteolytic cleavage near the center of its amino acid chain. The four subunits are linked together by disulfide bonds to give a symmetrical configuration with a stoichiometry of (alpha-s-s-beta)-s-s-(alpha-s-s-beta). This structure is remarkably similar to the general subunit composition of immunoglobulin G molecules and provides a structural basis for the postulate that this minimum insulin receptor structure may be divalent for binding hormone. A second area of recent progress involves the successful generation of a soluble factor or factors by insulin that are capable of modulating the activity of insulin-sensitive enzymes such as pyruvate dehydrogenase, glycogen synthase and cyclic adenosine monophosphate (cyclic AMP)-dependent protein kinase in cell-free systems. Indirect evidence indicates that the putative mediator or mediators of insulin action exhibits properties expected of a low molecular weight peptide, including destruction by proteases. The data available are consistent with the hypothesis that insulin-receptor interaction leads to the activation of a membrane protease that catalyzes the release of a peptide mediator or mediators of insulin action.

Regulation of glycogen synthesis in rat-hepatocyte cultures by glucose, insulin and glucocorticoids

The changes in glycogen content and in its rate of synthesis in two-day-old primary cultures of rat hepatocytes were assessed under various conditions. Hepatocytes cultivated in serum-free and hormone-free medium switch from glycogen degradation to glycogen deposition at 10.3 mM glucose. After pretreatment of the cells with glucocorticoids this threshold was reduced, in the absence or presence of insulin, to 5.4 or 1.2 mM glucose, respectively. The rate of glycogen synthesis in the presence of 10 mM glucose was amplified from 5 nmol x h-1 x mg protein-1 to 20 nmol glucose x h-1 x mg protein-1 after pretreatment with triamcinolone. Glucagon pretreatment also significantly increased the subsequent glycogen synthesis rate. Insulin addition accelerated glycogen synthesis about twofold regardless of the pretreatment. The dose-response relationship between insulin concentration and glycogen synthesis rate showed half-maximal effect at 0.62 +/- 0.22 nM (mean +/- S.D.) insulin. Pretreatment of hepatocytes with glucocorticoids, glucagon, insulin or combinations of these hormones did not significantly change the concentration which gives the half-maximal effect.

Investigation of the subcellular distribution of cyclic-AMP phosphodiesterase in rat hepatocytes, using a rapid immunological procedure for the isolation of plasma membrane

The distribution of cyclic-AMP phosphodiesterase was investigated in subcellular fractions prepared from homogenates of rat liver or isolated hepatocytes. When measured at 1 mM or 1 microM substrate concentration, approx. 35% or 50%, respectively, of enzyme activity was particulate. The soluble activity appeared to be predominantly a 'high Km' form, whereas the particulate activity had both 'high Km' and 'low Km' components. The recovery of cyclic-AMP phosphodiesterase was measured using 1 microM substrate concentraiton, in plasma membrane-containing fractions prepared either by centrifugation or by the use of specific immunoadsorbents. The recovery of phosphodiesterase was lower than that of marker enzymes for plasma membrane, and comparable with the recovery of markers for intracellular membranes. It was concluded that regulation of both 'high Km' and 'low Km' phosphodiesterase could potentially make a significant contribution to the control of cyclic AMP concentration, even at microM levels, in the liver. the 'low Km' enzyme, for which activation by hormones has been previously described, appears to be located predominantly in intracellular membranes in hepatocytes. The immunological procedure for membrane isolation allowed the rapid preparation of plasma membranes in high yield. Liver cells were incubated with rabbit anti-(rat erythrocyte) serum and homogenized. The antibody-coated membrane fragments were then extracted onto an immunoadsorbent consisting of sheep anti-(rabbit IgG) immunoglobulin covalently bound to aminocellulose. Plasma membrane was obtained in approx. 40% yield within 50 min of homogenizing cells.

Relationship between body fat mass, carbohydrate tolerance and IRI response during glucose infusion in subjects with early diabetes

We have studied the interrelationship of total body fat mass, carbohydrate tolerance and IRI response in 17 non-obese and obese subjects, who were suspected of having early diabetes. We carried out an i.v. glucose infusion test consisting of a priming injection of 0.33 g/kg followed by constant glucose infusion of 12 mg/kg/min in all persons. Total body fat mass was estimated by the tritium dilution method. There was a positive correlation of body fat mass, fasting glucose concentration and blood glucose concentration at 150 min as well as a strong correlation between body fat mass and BG area 60--120 min as parameters of carbohydrate tolerance in all subjects, i.e. the degree of carbohyrate intolerance was directly related to the quantity of total body fat mass. A similar correlation was found when the non-obese and obese groups were analyzed separately. In neither group did total body fat mass correlate with parameters of IRI response. In obese subjects with pathological carbohydrate tolerance, however, a positive correlation of basal IRI concentration and total body fat mass was found. Furthermore, a close relation between basal IRI level and parameters of carbohydrate tolerance could be demonstrated in obese subjects. The present study failed to demonstrate any correlation of parameters of carbohydrate tolerance and glucose-induced IRI response in either group. Thus, the significant relationship between body fat mass and degree of carbohydrate intolerance indicates that body fat mass plays an important role in the disturbance of blood glucose homeostasis in early diabetes with and without obesity.