糖代谢与肝性脑病关系的临床观察

目的观察肝硬化患者血糖的改变,探讨糖代谢与肝性脑病（HE）的关系。方法 2008年8月至2010年8月间在嘉定区中心医院消化内科住院的肝硬化患者60人作为肝硬化组,非肝硬化住院患者30人作为对照组。收集一般资料,观察神志,体检有无扑翼样震颤,抽血检测患者血糖,根据患者检查结果,把肝硬化患者分为无扑翼样震颤组和有扑翼样震颤组,对肝硬化组与对照组糖代谢异常发生率进行比较,根据患者血糖情况分为血糖正常组及糖代谢异常组,对两组间扑翼样震颤及HE发生率进行比较。计量资料应用t检验,率的比较应用卡方检验。结果对照组30例,肝硬化组60例中无扑翼样震颤组28例,有扑翼样震颤组32例,对照组糖代谢异常发生率13.33%,肝硬化组糖代谢异常发生率26.67%,肝硬化组患者糖代谢异常发生率高于对照组（χ2=2.058,P〈0.05）。糖代谢正常组扑翼样震颤发生率50.00%,HE发生率34.09%,糖代谢异常组扑翼样震颤发生率62.50%,HE发生率37.50%,两组间差异均无统计学意义（P〉0.05）。结论肝硬化患者存在糖代谢的紊乱,但糖代谢异常不一定易于发生HE,提示存在大脑对糖的利用障碍。     

Role of hepatic artery in haemodynamics in normal and cirrhotic livers in rats

To examine the degree of influence of the hepatic artery on microcirculation in the liver, microscopic observation of blood flow in the hepatic minute blood vessels and the sinusoids and pressure measurements at key points in hepatic vascular pathways in vivo were performed before and after hepatic artery ligation in normal and cirrhotic rats. In normal rats, portal vein pressure (109 mmH2O) fell 10 mmH2O after hepatic artery ligation, but the pressures of the terminal portal venule, the terminal hepatic venule and the inferior vena cava did not change. In cirrhotic rats, portal vein pressure (206 mmH2O) and terminal portal venule pressure (106 mmH2O) fell 23 and 10 mmH2O after hepatic artery ligation respectively: the pressures in the terminal hepatic venule and the inferior vena cava did not change. These results suggests that the pressure transmitted from the hepatic artery was mostly supplied to the intrahepatic portal vein in normal rats and both to the intrahepatic portal vein and to the sinusoids in cirrhotic rats. In both normal and cirrhotic rats, however, the pressure transmitted from the hepatic artery was about 10 per cent of the initial portal vein pressure, and the blood flow in minute vessels and sinusoids did not change after hepatic artery ligation. Accordingly, it is believed that the hepatic artery plays only a small role in the haemodynamics of the liver in both normal and cirrhotic rats, irrespective of the distribution and manner of the hepatic arterial termination.     

Hepatosplanchnic haemodynamics and renal blood flow and function in rats with liver failure

BACKGROUND: Massive liver necrosis, characteristic of acute liver failure, may affect hepatosplanchnic haemodynamics, and contribute to the alterations in renal haemodynamics and function. AIMS: To investigate the relation between hepatosplanchnic haemodynamics, including portal systemic shunting, and renal blood flow and function in rats with acute liver failure. METHODS: Liver failure was induced in male Wistar rats by intraperitoneal injection of 1.1 g/kg of D(+)-galactosamine hydrochloride. The parameters assessed included; systemic, hepatosplanchnic, and renal blood flow (57Co microsphere method); portal-systemic shunting and intrarenal shunting (consecutive intrasplenic, intraportal, or renal arterial injections of 99mTc methylene diphosphonate and 99mTc albumin microspheres); arterial blood pressure and portal pressure; renal function; and liver function (liver function tests and 14C aminopyrine breath test). RESULTS: Progressive liver dysfunction was accompanied by the development of a hyperdynamic circulation, a highly significant decrease in renal blood flow and function, and an increase in intrarenal shunting 36, 42, and 48 hours after administration of D-galactosamine. The alterations in renal blood flow and function were accompanied by significant increases in portal pressure, portal venous inflow, and intrahepatic portal systemic shunting in galactosamine treated rats compared with controls. There was a significant correlation between changes in renal blood flow and changes in portal pressure, intrahepatic portal systemic shunting, and deterioration in liver function (r = 0.8, p < 0.0001). CONCLUSIONS: The results of this study suggest that both increased intrahepatic portal systemic shunting and hepatocyte impairment may contribute to alterations in renal haemodynamics and function.     

Effects of diabetes on cardiac glycogen metabolism in rats

Summary The effects of diabetes on myocardial glycogen metabolism in rats were examined and compared with those of fasting. Male Wistar rats were divided into three groups: controls, streptozotocininduced diabetics, and one-week fasted. Isolated rat hearts were subjected to substrate-free 30-min Langendorff perfusion followed by 60-min working heart perfusion with glucose alone or in combination with insulin or insulin plus -hydroxybutyrate (BHB). Myocardial glycogen contents were determined before or 30 min after Langendorff perfusion, or 60 min after working heart perfusion. Before Langendorff perfusion, tissue glycogen concentrations in control, diabetic, and fasted hearts were 3.3 ± 0.2, 10.0 ± 0.9, and 5.7 ± 0.5 (mg/g wet weight), respectively. In diabetic rats, the myocardial glycogen concentration was markedly decreased after working heart perfusion of any of the substrate combinations, even those with insulin and BHB. In contrast, myocardial glycogen in control or fasted rats was not reduced after the addition of glucose with insulin, and/or glucose with insulin and BHB. These results suggest that degradation of tissue glycogen occurs in isolated perfused hearts from diabetic rats, while a clearly different response is shown by fasted hearts.     

Hepatic microcirculation assessed by positron emission tomography of first-pass ammonia metabolism in porcine liver.

AIMS/BACKGROUND: Intrahepatic branching of the hepatic artery (HA) to liver microcirculatory units, the acini, is more heterogeneous than that of the portal vein (PV). Furthermore, part of HA blood enters the sinusoid partially downstream between the in- and outlets. We examined the effects of these vascular variations on porcine hepatic first-pass ammonia metabolism, which is characterised by high uptake and separate periportal urea and perivenous glutamine formations. METHODS: (13)NH(3) was given via the PV, HA or caval vein, followed by 22 min dynamic liver positron emission tomography (PET) recordings in six pigs. Heterogeneity of liver (13)N-metabolism was quantified by the coefficient of variation of tissue (13)N-radioactivity measured 10 min after tracer infusion. Sinusoidal zonal clearances of (13)NH(3) into (13)N-urea and (13)N-glutamine were calculated by kinetic PET modelling. RESULTS: Liver metabolic heterogeneity was 0.65+/-0.20 (mean+/-SD, n=6) following (13)NH(3)-infusion into HA, 0.34+/-0.17 into PV and 0.10+/-0.02 into the caval vein. Clearance of (13)NH(3) to (13)N-urea was of similar magnitude following (13)NH(3) administration into HA and PV: 0.27+/-0.11 ml/min/g (mean+/-SD) and 0.29+/-0.09 ml/min/g, respectively. Clearances of (13)NH(3) to (13)N-glutamine when (13)NH(3) was given into HA and PV were also similar: 0.47+/-0.18 and 0.50+/-0.13 ml/min/g, respectively. CONCLUSIONS: The present measurements of the hepatic metabolism of (13)NH(3) showed metabolic heterogeneity compatible with variation of the HA supply of the acini. Second, results of PET modelling of the sinusoidal zonation metabolism of (13)NH(3) to (13)N-urea and to (13)N-glutamine did not indicate metabolically important partial downstream arterial entry into the sinusoids.     

The metabolism of recombinant erythropoietin in the isolated perfused rat liver

ABSTRACT— Indirect evidence points to extrarenal organs, presumably the liver, as the site of degradation of erythropoietin (EPO). The metabolism of both fully glycosylated and desialated intrinsically labelled ³⁵S-Cysteine recombinant human erythropoietin (rhEPO) was therefore studied in isolated Wistar rat livers perfused in a recirculating mode for 180 min with a hemoglobin-free medium containing rhEPO. Perfusate and bile levels of rhEPO were measured by RIA. Total ³⁵S-radioactivity in liver, bile and perfusate as well as non-acid precipitable radioactivity in perfusate were determined. In addition, detection of ³⁵S-radioactivity was performed after subcellular fractionation of rat livers perfused with desialo-³⁵S-Cysteine rhEPO. While concentrations of fully glycosylated ³⁵S-Cysteine rhEPO did not exhibit any detectable decrease during perfusion, desialo-³⁵S-Cysteine rhEPO was rapidly cleared from the perfusate. After 60 min of perfusion, only 32% of the initial levels of both immunoreactive rhEPO and total radioactivity remained in the perfusate. Quantitative hepatic accumulation of desialated tracer was demonstrated. Subcellular fractionation showed extensive hepatic degradation of the desialated tracer. Furthermore, during perfusion progressively larger amounts of small molecular weight degradation products of the tracer were found in the perfusate. Bile excretion of both fully glycosylated and desialated tracer was negligible. The significance of hepatic metabolism of desialo-³⁵S-Cysteine rhEPO was supported by reduced removal of desialo-³⁵S-Cysteine rhEPO from plasma in hepatectomized rats. It is hypothesized that continuous in vivo desialation is a crucial rate-limiting step in the degradation of circulating EPO.     

Phlorizin treatment of diabetic rats partially reverses the abnormal expression of genes involved in hepatic glucose metabolism

Summary Liver insulin resistance and glucagon-stimulated hepatic glucose production are characteristics of the diabetic state. To determine the potential role of glucose toxicity in these abnormalities, we examined whether phlorizin treatment of streptozotocin-diabetic rats resulted in altered expression of genes involved in key steps of hepatic glucose metabolism. By inhibiting renal tubular glucose reabsorption, phlorizin infusion to diabetic rats induced normoglycaemia, did not significantly alter low circulating insulinaemia, but caused a marked decrease in hyperglucagonaemia. Glucokinase and L-type pyruvate kinase mRNA levels were reduced respectively by 90% and 70% in fed diabetic rats, in close correlation with changes in enzyme activities. Eighteen days of phlorizin infusion partially restored glucokinase mRNA and activity (40% of control levels), but had no effect on L-type pyruvate kinase mRNA and activity. In contrast to the glycolytic enzymes, mRNA and activity of the gluconeogenic enzyme, phospoenolpyruvate carboxykinase were increased (10- and 2.2-fold, respectively) in fed diabetic rats. Phlorizin administration decreased phospoenolpyruvate carboxykinase mRNA to values not different from those in control rats, while phospoenolpyruvate carboxykinase activity remained 50% higher than that in control rats. The 50% rise in liver glucose transporter (GLUT 2) mRNA and protein, produced by diabetes, was also corrected by phlorizin treatment. In conclusion, we propose that phlorizin treatment of diabetic rats may induce a partial shift of the predominating gluconeogenesis, associated with hepatic glucose overproduction, into glycolysis, by correction of impaired pre-translational regulatory mechanisms. This could be essentially mediated through improved pancreatic alpha-cell function and subsequent lowering of hyperglucagonaemia. These observations suggest that glucagon-stimulated hepatic glucose production may result, in part, from glucose toxicity.     

New insights into sympathetic regulation of glucose and fat metabolism

The autonomic nervous system modulates glucose and fat metabolism through both direct neural effects and hormonal effects. This review presents recent concepts on the sympathetic regulation of glucose and fat metabolism. Focally released norepinephrine from sympathetic nerves is likely to increase glucose uptake in skeletal muscle and adipose tissues independent of insulin but norepinephrine does not contribute so much as epinephrine to hepatic glucose production. Epinephrine increases hepatic glucose production and inhibits insulin secretion and the glucose uptake by tissues that is induced by insulin. Additionally, catecholamines can increase thermogenesis and lipolysis, leading to increased energy expenditure and decreased fat stores. It is likely that β-(β3)-adrenergic receptors mediate these responses. Alterations of central neurotransmission and environmental factors can change the relative contribution of sympathetic outflow to the pancreas, liver, adrenal medulla and adipose tissues, leading to the modulation of glucose and fat metabolism. Recent studies have proposed that leptin, an adipocyte hormone, affects the central nervous system to increase sympathetic outflow independent of feeding. The effects of leptin on glucose and fat metabolism could be in part mediated by the sympathetic nervous system. Studies using mice with a genetic disruption of serotonin 5-HT2 c receptor indicate that central neural mechanisms in the regulation of sympathetic outflow and satiety could be dissociated. Abnormalities of sympathetic effects, including disturbances of leptin and β3-adrenergic receptor signalling, are likely to cause obesity and impaired glucose tolerance in rodents and humans. These findings indicate that dysfunction of the sympathetic nervous system could predispose to obesity and Type II (non-insulin-dependent) diabetes mellitus. [Diabetologia (2000) 43: 533–549]     

Abnormal insulin secretion and glucose metabolism in pancreatic islets from the spontaneously diabetic GK rat

Summary Insulin secretion and islet glucose metabolism were compared in pancreatic islets isolated from GK/Wistar (GK) rats with spontaneous Type 2 (non-insulin-dependent) diabetes mellitus and control Wistar rats. Islet insulin content was 24.5±3.1 U/ng islet DNA in GK rats and 28.8±2.5 U/ng islet DNA in control rats, with a mean (±SEM) islet DNA content of 17.3±1.7 and 26.5±3.4 ng (p < 0.05), respectively. Basal insulin secretion at 3.3 mmol/l glucose was 0.19±0.03 · ng islet DNA^–1· h^–1 in GK rat islets and 0.40±0.07 in control islets. Glucose (16.7 mmol/l) stimulated insulin release in GK rat islets only two-fold while in control islets five-fold. Glucose utilization at 16.7 mmol/l glucose, as measured by the formation of ³H₂O from [5-³ H]glucose, was 2.4 times higher in GK rat islets (3.1±0.7 pmol · ng islet DNA^–1 · h^–1) than in control islets (1.3±0.1 pmol · ng islet DNA^–1 · h^–1; p<0.05). In contrast, glucose oxidation, estimated as the production of ¹⁴CO₂ from [U-¹⁴C]glucose, was similar in both types of islets and corresponded to 15±2 and 30±3 % (p<0.001) of total glucose phosphorylated in GK and control islets, respectively. Glucose cycling, i. e. the rate of dephosphorylation of the total amount of glucose phosphorylated, (determined as production of labelled glucose from islets incubated with ³H₂O) was 16.4±3.4% in GK rat and 6.4±1.0% in control islets, respectively (p<0.01). We conclude that insulin secretion stimulated by glucose is markedly impaired in GK rat islets. Glucose metabolism is also altered in GK rat islets, with diminished ratio between oxidation and utilization of glucose, and increased glucose cycling, suggesting links between impaired glucose-induced insulin release and abnormal glucose metabolism.     

Increase of mexiletine plasma levels due to delayed hepatic metabolism in patients with chronic liver disease

The clinical relevance of changes in pharmacokinetics of oralmexiletine (600 mg daily dose) was studied in 82 patients withventricular arrhythmias and impaired liver, renal or heart function(control group n = 51, patients with liver cirrhosis n = 9,with renal insufficiency n = 14, or heart failure n = 8). Increasedplasma levels of mexiletine were found in patients with chronicliver disease (2.21 ± 0.94 µg/ml, versus, 0.63± 0.22µg/ml of controls, P < 0.01). Plasma levels in patients with renal insufficiency or heartfailure were not significantly different from the controls.The resulting elevated plasma levels in patients with livercirrhosis emphasize the importance of hepatic metabolism inthe elimination of mexiletine. Drug monitoring must be considerednecessary in patients with impaired liver function.     

Comparison of portal and peripheral insulin delivery on carbohydrate metabolism in streptozotocin-diabetic rats

Summary Severely diabetic rats (150 mg streptozotocin/kg) were transplanted with fetal pancreatic islets: (a) under the kidney capsule to model peripheral insulin delivery, and (b) into the splenic pulp to model portal delivery. Long-term normoglycaemia, normal weight gain and normal peripheral insulin levels were achieved in both groups of transplanted animals. In both groups, 24-h fasted blood lactate, pyruvate and alanine were identical to those observed in sham-operated control animals. Blood glucose and plasma insulin responses to 300 mg oral glucose 8 weeks after transplantation were the same as in control animals. Hepatic glycogen concentration was, however, lower in fed rats with islets beneath the kidney capsule compared with control rats (p<0.01), suggesting inadequate hepatic insulinisation in the fed state with peripheral insulin delivery. Muscle glycogen was the same as in controls. Glucose turnover and glucose carbon recycling were not significantly different from results in normal control and splenic pulp islet-transplanted animals. The findings indicate that consistent normoglycaemia, normal glucose flux and normalisation of blood intermediary metabolites can be achieved in the rat with peripheral insulin delivery without associated hyperinsulinaemia.     

Induction of gastric lesions and hypoglycemic response by food deprivation in streptozotocin-diabetic rats

Overnight fasting causes hemorrhagic lesions in the stomach of streptozotocin (STZ)-induced diabetic rats, but the pathogenetic mechanism remains unknown. The present study was performed to investigate the pathogenesis of such lesions developed in STZ-diabetic rats after starvation, mainly in relation to blood glucose changes. A single injection of STZ (70 mg/kg, intraperitoneally) induced hyperglycemic conditions one week after the administration, and high blood glucose levels (BGL: >350 mg%) remained up to three weeks later. The STZ-diabetic rats developed gastric lesions with the marked reduction of BGL after 18 hr of fasting, depending upon the duration of diabetes; the lesion score and BGL reduction in the 3-week-old STZ rats were 32.0±7.8 mm and >250 mg/100 ml, respectively. Acid secretion in the pylorus-ligated rats was not significantly changed in the STZ-induced diabetic conditions for the initial two weeks but slightly decreased at three weeks when compared with normal rats. Fasting of normal rats for 18 hr did not cause either BGL reduction or any lesion in the stomach. In the 3-week-old STZ animals, the severity of gastric lesions increased with the duration of fasting (4–18 hr) and was again closely associated with the degree of BGL reduction. These lesions induced by 18 hr of starvation in 3-week-old STZ rats were significantly inhibited by pretreatment with insulin (4 units/rat/day) for the last one week to maintain BGL within normal ranges or by intravenous infusion of 25% glucose during fasting period. Both of these treatments significantly prevented BGL reduction in response to fasting. These results suggest that gastric lesions induced in STZ-diabetic rats by fasting are insulin-sensitive and may be associated with a profound hypoglycemic response to food deprivation.     

In vitro insulin action on erythrocyte glucose metabolism in normal and diabetic rats

Summary Alloxan-induced diabetes in rats significantly impaired the capacity of the erythrocytes to metabolise glucose in vitro to either lactic acid or CO₂. Both these metabolic activities were initially insensitive to insulin in normal as well as in diabetic animals; but became responsive when these cells were subjected to insulin and glucose starvation for 1 h through incubation in their absence. This action of insulin in starved cells showed concentration dependence and required preincubation with the hormone prior to addition of glucose.     

Exaggeration of acute liver damage by hepatic sympathetic nerves and circulating catecholamines in perfused liver of rats treated with D-galactosamine

Effects of electrical stimulation of the hepatic nerves on acute liver damage were examined using isolated rat liver perfused in situ, 24 hours after intraperitoneal injection with D-galactosamine (800 mg/kg). The leakage of lactate dehydrogenase (LDH) and aspartate aminotransferase (AST) from the liver was used as markers of acute liver damage. In perfused livers after treatment with galactosamine, nerve stimulation (20 V, 20 Hz, 2 ms) increased the leakage of LDH and AST about 3-fold over the basal level accompanied by the decrease in flow rate, whereas with control livers the leakage of LDH and AST into the effluent was almost undetectable throughout the perfusion. The rapid increase in the leakage of LDH and AST was observed during nerve stimulation even under conditions where perfusion flow was maintained constant. Such effects of hepatic nerve stimulation on galactosamine-treated livers were mimicked well by infusion of noradrenaline or phenylephrine, and inhibited by the alpha1-antagonist bunazosin. Artificial reduction of perfusion flow alone did not induce the rapid leakage of LDH and AST into the effluent. On the other hand, low concentration (10 nmol/L) of noradrenaline only minimally decreased the flow rate but apparently augmented liver cell damage. The acute liver damage augmented by noradrenaline was dependent on extracellular Ca²⁺. These results indicate that in the liver, already having been injured slightly, the activation of hepatic sympathetic nerves and circulating catecholamines exaggerates acute liver damage through an action on liver cells, which depends on the influx of extracellular Ca²⁺. (Hepatology 1996 Mar;23(3):524-9)     

The maximal activity of phosphate-dependent glutaminase and glutamine metabolism in the colon and the small intestine of streptozotocin-diabetic rats

Summary The effects of short- and long-term diabetes on the maximal activities of phosphate-dependent glutaminase and glutamine metabolism were studied in the colon and the small intestine of streptozotocin-diabetic rats. The maximal activity of colonic phosphate-dependent glutaminase was decreased [44% in mucosal scrapings (p<0.01); 29% in whole colon (p<0.001)] or unchanged in short- or long-term diabetes respectively. That of the small intestine was increased in both short- (110%) and long-term (200%–500%) diabetes; insulin treatment corrected this increase. Acute insulin-deficiency (using anti-insulin serum) resulted in the increase (18%, p<0.05) of the activity of only intestinal glutaminase. Chemically-induced acidosis and alkalosis decreased (46%, p<0.001) and increased (24%, p<0.001), respectively, the activity of intestinal glutaminase, but had no effect on the colonic enzyme. Changes in glutaminase of the enlarged colon and small intestine were only detectable when activities were measured in whole organ. Arteriovenous-difference measurements showed diminished metabolism of plasma glutamine by the gut which correlated with the duration of the state of diabetes, and was accompanied by enhanced release by skeletal muscle and increased uptake by both kidney and liver. It is concluded that insulin is directly or indirectly involved in the regulation of glutamine metabolism of the gut.     

Possible role of free radicals generated by pseudohypoxia in the regulation of hepatic glucose output. An in vitro model using rat liver microsomal glucose 6-phosphatase

Summary Hepatic glucose output is decreased by hyperglycaemia through an unknown mechanism. We hypothesize that free radicals generated by hyperglycaemic pseudohypoxia might cause glucose output to decrease by inhibiting glucose 6-phosphatase – a key enzyme of gluconeogenesis. To prove this a model experiment was performed on a microsome fraction of rat liver. One of the characteristic features of pseudohypoxia due to hyperglycaemia is an increase in the ratio of NADH/NAD ⁺ , so in the present study the changes in NADH – induced glucose 6-phosphatase activity were investigated as related to the release of inorganic phosphate (Pi) derived from glucose 6-phosphate. After incubation for 50 min, Pi release was significantly reduced by NADH (4.026 ± 0.189 vs 2.696 ± 0.429 μmol · l^–1· mg protein^–1, control vs NADH samples, p < 0.01). The decrease in the activity of glucose 6-phosphatase generated by NADH was prevented by using desferrioxamine, an irreversible ferric chelator, butylated hydroxytoluene and Trolox, two agents which inhibit lipid peroxidation, and reduced glutathione, a non-specific radical scavanger. Superoxide dismutase, catalase and the hydroxyl radical scavenger dimethyl sulphoxide proved to be ineffective. When the above investigations were carried out in the presence of a ferric-EDTA complex the inhibition of glucose 6-phosphatase was found to be inducible by hydrogen peroxide and/or hydroxyl free radicals. These investigations seem to indicate that pseudohypoxia due to hyperglycaemia can inhibit the activity of glucose 6-phosphatase both by lipid peroxidation and by inducing hydrogen peroxide and/or hydroxyl free radicals and thus it may play a part in the glucose-induced decrease of hepatic glucose output. [Diabetologia (1997) 40: 1251–1254] Received: 28 January 1997 and in final revised form: 1 July 1997