S. R. Myers

Importance of the route of intravenous glucose delivery to hepatic glucose balance in the conscious dog.

To assess the importance of the route of glucose delivery in determining net hepatic glucose balance (NHGB) eight conscious overnight-fasted dogs were given glucose via the portal or a peripheral vein. NHGB was measured using the arteriovenous difference technique during a control and two 90-min glucose infusion periods. The sequence of infusions was randomized. Insulin and glucagon were held at constant basal levels using somatostatin and intraportal insulin and glucagon infusions during the control, portal, and peripheral glucose infusion periods (7 +/- 1, 7 +/- 1, 7 +/- 1 microU/ml; 100 +/- 3, 101 +/- 6, 101 +/- 3 pg/ml, respectively). In the three periods the hepatic blood flow, glucose infusion rate, arterial glucose level, hepatic glucose load, arterial-portal glucose difference and NHGB were 37 +/- 1, 34 +/- 1, 32 +/- 3 ml/kg per min; 0 +/- 0, 4.51 +/- 0.57, 4.23 +/- 0.34 mg/kg per min; 101 +/- 5, 200 +/- 15, 217 +/- 13 mg/dl; 28.5 +/- 3.5, 57.2 +/- 6.7, 54.0 +/- 6.4 mg/kg per min; +2 +/- 1, -22 +/- 3, +4 +/- 1 mg/dl; and 2.22 +/- 0.28, -1.41 +/- 0.31, and 0.08 +/- 0.23 mg/kg per min, respectively. Thus when glucose was delivered via a peripheral vein the liver did not take up glucose but when a similar glucose load was delivered intraportally the liver took up 32% (P less than 0.01) of it. In conclusion portal glucose delivery provides a signal important for the normal hepatic-peripheral distribution of a glucose load.

Role of Brain in Counterregulation of Insulin-Induced Hypoglycemia in Dogs

The role of the brain in directing counterregulation during hypoglycemia induced by insulin infusion was assessed in overnight-fasted conscious dogs. Concomitant brain and peripheral hypoglycemia was induced in one group of dogs (n = 5) by infusing insulin peripherally at a rate of 3.5 mU · kg−1 · min−1. In another group (n = 4), insulin was infused as described above to induce peripheral hypoglycemia, and brain hypoglycemia was minimized by infusing glucose bilaterally into the carotid and vertebral arteries to maintain the brain glucose level at a calculated concentration of 85 mg/dl. Glucose was also infused peripherally as needed so that the peripheral glucose levels in both of the protocols were similar (45 ± 2 mg/dl with and 48 ± 3 mg/dl without brain glucose infusion, both P < .05). The responses (in terms of change of area under the curve) of epinephrine, norepinephrine, cortisol, and pancreatic polypeptide when brain glycemia was controlled during insulin infusion were only 14 ± 6, 39 ± 12, 17 ± 8, and 9 ± 4%, respectively, of those present during insulin infusion without concomitant brain glucose infusion (all P < .05). Of particular interest was the glucagon response that occurred when head hypoglycemia was minimized; the glucagon level was only 21 ± 8% of that present when marked brain hypoglycemia accompanied insulin infusion (P < .05). During hypoglycemia resulting from insulin infusion, endogenous glucose production (EGP), as assessed with [3-3H]glucose, rose from 2.6 ± 0.1 to 4.4 ± 0.5 mg · kg−1 min−1 (P < .05). In contrast, EGP decreased from 2.7 ± 0.2 to 2.0 ± 0.3 mg · kg−1 · min−1 When brain hypoglycemia was minimized. In an additional set of studies, when insulin was infused at 3.5 mU · kg−1 · min−1 and glucose was infused peripherally to maintain both the head and peripheral glucose concentrations at 88 ± 6 mg/dl, EGP decreased from 2.6 ± 0.1 to 1.2 ± 0.2 mg · kg−1 · min−1. These results suggest that under marked hyperinsulinemic conditions the brain is the primary director of glucagon release and that it is responsible for ∼ 75% of the life-sustaining glucose production.

Counterregulation during hypoglycemia is directed by widespread brain regions.

Previous studies have demonstrated the importance of the brain in directing counterregulation during insulin-induced hypoglycemia in dogs. The capability of selective carotid or vertebrobasilar hypoglycemia in triggering counterregulation was assessed in this study using overnight-fasted dogs. Insulin (21 pM · kg−1 · min−1) was infused for 3 h to create peripheral hypoglycemia in the presence of 1) selective carotid hypoglycemia (vertebral glucose infusion, n = 5), 2) selective vertebrobasilar hypoglycemia (carotid glucose infusion, n = 5), 3) the absence of brain hypoglycemia (carotid and vertebral glucose infusion, n = 4), or 4) total brain hypoglycemia (no head glucose infusion, n = 5). Glucose was infused via a leg vein as needed in each group to minimize the differences in peripheral glucose levels (2.6 ± 0.1, 3.0 ± 0.2, 2.7 ± 0.1, and 2.5 ± 0.1 mM, respectively). The humoral responses (cortisol, glucagon, catecholamines, and pancreatic polypeptide) to hypoglycemia were minimally attenuated (< 40%) by selective carotid or vertebrobasilar euglycemia. In addition, the increase in hepatic glucose production, as assessed using [3-3H]glucose, was attenuated by only 41 and 34%, respectively, during selective carotid or vertebrobasilar hypoglycemia. These observations offer support for the hypothesis that more than one center is important in hypoglycemic counterregulation in the dog and that they are located in brain regions supplied by the carotid and vertebrobasilar arteries, because significant counterregulation occurred when hypoglycemia developed in either of these circulations. Counterregulation during hypoglycemia, therefore, is probably directed by widespread brain regions that contain glucose-sensitive neurons such that the sensing sites are redundant.

Intraportal glucose delivery enhances the effects of hepatic glucose load on net hepatic glucose uptake in vivo.

Although the importance of the hepatic glucose load in the regulation of liver glucose uptake has been clearly demonstrated in in vitro systems, the relationship between the hepatic glucose load and hepatic glucose uptake has yet to be defined in vivo. Likewise, the effects of the route of glucose delivery (peripheral or portal) on this relationship have not been explored. The aims of the present study were to determine the relationship between net hepatic glucose uptake (NHGU) and the hepatic glucose load in vivo and to examine the effects of the route of glucose delivery on this relationship. NHGU was evaluated at three different hepatic glucose loads in 42-h fasted, conscious dogs in both the absence (n = 7) and the presence (n = 6) of intraportal glucose delivery. In the absence of intraportal glucose delivery and in the presence of hepatic glucose loads of 50.5 +/- 5.9, 76.5 +/- 10.0, and 93.6 +/- 10.0 mg/kg/min and arterial insulin levels of approximately 33 microU/ml, NHGU was 1.16 +/- 0.37, 2.78 +/- 0.82, and 5.07 +/- 1.20 mg/kg/min, respectively. When a portion of the glucose load was infused into the portal vein and similar arterial insulin levels (approximately 36 microU/ml) and hepatic glucose loads (52.5 +/- 4.5, 70.4 +/- 5.6, and 103.6 +/- 18.4 mg/kg/min) were maintained, NHGU was twice that seen in the absence of portal loading (3.77 +/- 0.40, 4.80 +/- 0.59, and 9.62 +/- 1.43 mg/kg/min, respectively). Thus, net hepatic glucose uptake demonstrated a direct dependence on the hepatic glucose load that did not reach saturation even at elevations in the hepatic glucose load of greater than three times basal. In addition, the presence of intraportal glucose delivery increased net hepatic glucose uptake apparently by lowering the threshold at which the liver switched from net glucose output to net glucose uptake.

Magnitude of Negative Arterial-Portal Glucose Gradient Alters Net Hepatic Glucose Balance in Conscious Dogs

To examine the relationship between the magnitude of the negative arterial-portal glucose gradient and net hepatic glucose uptake, two groups of 42-h fasted, conscious dogs were infused with somatostatin, to suppress endogenous insulin and glucagon secretion, and the hormones were replaced intraportally to create hyperinsulinemia (3- to 4-fold basal) and basal glucagon levels. The hepatic glucose load to the liver was doubled and different negative arterial-portal glucose gradients were established by altering the ratio between portal and peripheral vein glucose infusions. In protocol 1 (n = 6) net hepatic glucose uptake was 42.2 ± 6.7, 35.0 ± 3.9, and 33.3 ± 4.4 μmol · kg−1 · min−1 at arterial-portal plasma glucose gradients of −4.1 ± 0.9, −1.8 ± 0.4, and −0.8 ± 0.1 mM, respectively. In protocol 2 (n = 6) net hepatic glucose uptake was 26.1 ± 2.8 and 12.2 ± 1.7 μmol · kg−1 · min−1 at arterial-portal plasma glucose gradients of −0.9 ± 0.2 and −0.4 ± 0.1 mM, respectively. No changes in the hepatic insulin or glucose loads were evident within a given protocol. Although net hepatic glucose uptake was lower in protocol 2 when compared with protocol 1 (26.1 ± 2.8 vs. 33.3 ± 4.4 μmol · kg−1 · min−1) in the presence of a similar arterial-portal plasma glucose gradient (−0.9 vs. −0.8 mM) the difference could be attributed to the hepatic glucose load being lower in protocol 2 (i.e., hepatic fractional glucose extraction was not significantly different) primarily as a result of lower hepatic blood flow. In conclusion, in the presence of fixed hepatic glucose and insulin loads, the magnitude of the negative arterial-portal glucose gradient can modify net hepatic glucose uptake in vivo.

Interaction Between Insulin and Glucose-Delivery Route in Regulation of Net Hepatic Glucose Uptake in Conscious Dogs

In the presence of fixed basal levels of insulin, the route of intravenous glucose delivery (portal vs. peripheral) determines whether net hepatic glucose uptake (NHGU) occurs. Our aims were to determine if the route of intravenous glucose delivery also plays a role in regulating NHGU in the presence of hyperinsulinemia and to determine if length of fast (18 vs. 36 h) influences regulation of NHGU. Five conscious dogs fasted 18 h were given somatostatin and replacement insulin (245 ± 34 μU · kg−1 · min−1) and glucagon (0.65 ng · kg−1 · min−1) infusions intraportally. After a 40-min control period, the insulin infusion rate was increased fourfold, and glucose was infused for 3 h. Glucose was given either through a peripheral vein or the portal vein for 90 min to double the glucose load reaching the liver. The order of infusions was randomized. NHGU was measured with the arterial – venous difference technique. Insulin and glucagon levels were 12 ± 2, 35 ± 6, and 36 ± 5 μU/ml and 55 ± 12, 61 ± 13, and 59 ± 7 pg/ml during the control, peripheral, and portal infusions, respectively. The glucose infusion rate, the load of glucose reaching the liver, and the arterial-portal plasma glucose gradient were 0, 9.58 ± 2.28, and 10.44 ± 2.94 mg · kg−1 · min−1; 29.4 ± 3.6, 56.8 ± 3.4, and 56.8 ± 2.8 mg · kg−1 · min−1; and 2 ± 1, 5 ± 1, and −51 ± 15 mg/dl during the same periods. The liver switched from net glucose output (2.5 ± 0.4 mg · kg−1 · min−1) to uptake of 1.4 ± 0.7 and 3.5 ± 0.8 mg · kg−1 · min−1 during peripheral and portal glucose delivery, respectively. Despite a similar hormonal milieu and indistinguishable glucose loads, there was significantly (P < 0.01) more NHGU when glucose was infused intraportally. Identical studies in five conscious dogs fasted 36 h gave results similar to those of dogs fasted 18 h. After 36 h of fasting, NHGU was 1.6 ± 0.4 and 4.0 ± 0.4 mg · kg−1 · min−1 during the peripheral and portal glucose infusions, respectively (P < 0.005). In conclusion, the route of intravenous glucose administration plays an important role in regulating NHGU even in the presence of hyperinsulinemia in conscious dogs fasted 18 and 36 h.

Acylation of Human Insulin With Palmitic Acid Extends the Time Action of Human Insulin in Diabetic Dogs

To test whether the binding of insulin to an endogenous serum protein can be used to extend the time action of insulin, human insulin was acylated at the epsilonamino group of Lys(B29) with palmitic acid to promote binding to serum albumin. Size-exclusion chromatography was used to demonstrate specific binding of the resulting analog, [N∈-palmitoyl Lys(B29)] human insulin, to serum albumin in vitro, and the time action and activity of the analog were determined in vivo using overnight-fasted, insulin-withdrawn diabetic dogs. In the diabetic animal model, the duration of action of [N∈-palmitoyl Lys(B29)] human insulin administered intravenously was nearly twice that of unmodified human insulin, and the plasma half-life was nearly sevenfold that of the unmodified protein. Administered subcutaneously, [N∈-palmitoyl Lys(B29)] human insulin had a longer duration of action; a flatter more basal plasma insulin profile; and a lower intersubject variability of response than the intermediate-acting insulin suspension Humulin L (Lilly, Indianapolis, IN). These studies support the concept that modification of insulin to promote binding to an existing serum protein can be used to extend the time action of human insulin. In addition, the time action, pattern, and decreased variability of response to [N∈-palmitoyl Lys(B29)] human insulin support the development and further testing of this soluble insulin analog as a basal insulin to increase the safety of intensive insulin therapy.

Effects of small changes in glucagon on glucose production during a euglycemic, hyperinsulinemic clamp

The aim of this study was to examine the influence of small changes in glucagon on hepatic glucose production during a euglycemic, hyperinsulinemic clamp. During 1.0 mU/kg.min insulin infusion, euglycemia was maintained by glucose infusion and glucagon was infused at various rates so as to cause plasma glucagon levels to increase, decrease, or remain unchanged. Changes in glucagon were found to be positively associated with changes in glucose production and inversely related to the degree of suppression of tracer or arteriovenous difference determined endogenous glucose production. Thus, animals in which the glucagon levels increased, appeared to have decreased hepatic insulin sensitivity, while animals in which glucagon levels decreased, appeared to have increased insulin sensitivity. In conclusion, since glucagon often declines during a euglycemic hyperinsulinemic clamp, and since small changes in glucagon can have marked effects on the suppression of hepatic glucose output even in the presence of high insulin levels, changes in glucagon should be considered when conclusions regarding hepatic insulin sensitivity are being drawn.

Chronic hyperinsulinemia decreases insulin action but not insulin sensitivity.

Hyperinsulinemia and insulin resistance are commonly seen in obese and non-insulin-dependent diabetes mellitus (NIDDM) patients, suggesting a causal link exists between hyperinsulinemia and insulin resistance. In a previous study, we demonstrated that chronic (28 days) intraportal hyperinsulinemia (50% increase in basal insulin levels) resulted in a decrease in insulin action as assessed by a one-step euglycemic hyperinsulinemic clamp. Since only one dose of insulin was used during the clamp, it was not possible to determine if the decrease in insulin action was due to a change in insulin sensitivity and/or maximal insulin responsiveness. In the present study, insulin resistance was induced as before, but insulin action was assessed in overnight fasted conscious dogs using a four-step euglycemic hyperinsulinemic clamp (1, 2, 10, and 15 mU/kg/min). Insulin responsiveness was assessed before the induction of chronic hyperinsulinemia (day 0), and after 28 days of hyperinsulinemia (day 28). Transhepatic glucose balance and whole-body glucose utilization were measured to allow assessment of both the hepatic and peripheral effects of insulin. Chronic hyperinsulinemia increased basal insulin levels from 13 +/- 2 to 21 +/- 4 microU/mL. After 4 weeks of chronic hyperinsulinemia, maximal insulin-stimulated glucose utilization was decreased 23% +/- 4% (P less than .05) and insulin sensitivity (ED50) was not significantly altered. During the four-step clamp, the liver was a major site of glucose utilization. The liver was responsible for 13% of the total glucose disposal rate on day 0 (2.9 mg/kg/min) at the highest insulin infusion rate (15 mU/kg/min; 2,000 microU/mL).(ABSTRACT TRUNCATED AT 250 WORDS)

In the Absence of Counterregulatory Hormones, the Increase in Hepatic Glucose Production During Insulin-Induced Hypoglycemia in the Dog Is Initiated in the Liver Rather Than the Brain

We have previously demonstrated that the liver can release glucose in response to insulin-induced hypoglycemia, despite the absence of glucagon, epinephrine, cortisol, and growth hormone. The aim of this study was to determine whether this is activated by liver or brain hypoglycemia. We assessed the response to insulin-induced hypoglycemia in the absence of counterregulatory hormones in overnight-fasted conscious adrenalectomized dogs that were given somatostatin and intraportal insulin (30 pmol · kg−1 · min−1) for 360 min. Glucose was infused to maintain euglycemia for 3 h and then to allow limited peripheral hypoglycemia for the next 3 h. During peripheral hypoglycemia, five dogs received glucose via both carotid and vertebral arteries to maintain cerebral euglycemia (H-EU group) concurrently with peripheral hypoglycemia, while six dogs received saline in these vessels to allow simultaneous cerebral and peripheral hypoglycemia (H-HY group). Throughout the study, arterial insulin was 1,675 ± 295 and 1,440 ± 310 pmol/l in the H-HY and H-EU groups, respectively. Glucose fell from 6.2 ± 0.3 to 2.1 ± 0.0 mmol/l and from 5.8 ± 0.3 to 1.9 ± 0.1 mmol/l in the last hour in the H-HY and H-EU groups, respectively (P < 0.05 for both). Norepinephrine rose from 1.12 ± 0.35 to 2.44 ± 0.69 nmol/l and from 1.09 ± 0.07 to 1.74 ± 0.16 nmol/l in the last hour in the H-HY and H-EU groups, respectively (P < 0.05 for both; no difference between groups). Glucagon, epinephrine, and cortisol were below the limits of detection. The liver switched from uptake to output of glucose during peripheral hypoglycemia in both the H-HY (–7.1 ± 2.1 to 5.4 ± 3.1 μmol · kg−1 · min−1) and H-EU (–7.9 ± 3.5 to 3.4 ± 1.7 μmol · kg−1 · min−1) groups (P < 0.05 for both; no difference between groups). Alanine levels and net hepatic alanine uptake fell similarly in both groups. There were increases (P < 0.05) in glycerol (12 ± 3 to 258 ± 47 μmol/l) and nonesterified fatty acid (194 ± 10 to 540 ± 80 micromol/l) levels and in total ketone production (0.4 ± 0.1 to 1.1 ± 0.2 μmol · kg−1 · min−1) in the H-HY group, but these parameters did not change in the H-EU group. These data clearly indicate that the lipolytic and hepatic responses to hypoglycemia are driven by differential sensing mechanisms. Thus, during insulin-induced hypoglycemia, when counterregulatory hormones are absent, liver hypoglycemia triggers the increase in hepatic glucose production, whereas cerebral hypoglycemia causes the increases in lipolysis and ketogenesis.