Dose-kinetics of pancreatic glucagon responses to arginine and glucose in subjects with normal and impaired pancreatic B cell function

Hyperglucagonaemia in diabetes: altered amino acid metabolism triggers mTORC1 activation, which drives glucagon production

AIM/HYPOTHESIS

Hyperglycaemia is associated with alpha cell dysfunction, leading to dysregulated glucagon secretion in type 1 and type 2 diabetes; however, the mechanisms involved are still elusive. The nutrient sensor mammalian target of rapamycin complex 1 (mTORC1) plays a major role in the maintenance of alpha cell mass and function. We studied the regulation of alpha cell mTORC1 by nutrients and its role in the development of hyperglucagonaemia in diabetes.

METHODS

Alpha cell mTORC1 activity was assessed by immunostaining for phosphorylation of its downstream target, the ribosomal protein S6, and glucagon, followed by confocal microscopy on pancreatic sections and flow cytometry on dispersed human and mouse islets and the alpha cell line, αTC1-6. Metabolomics and metabolic flux were studied by 13C glucose labelling in 2.8 or 16.7 mmol/l glucose followed by LC-MS analysis. To study the role of mTORC1 in mediating hyperglucagonaemia in diabetes, we generated an inducible alpha cell-specific Rptor knockout in the Akita mouse model of diabetes and tested the effects on glucose tolerance by IPGTT and on glucagon secretion.

RESULTS

mTORC1 activity was increased in alpha cells from diabetic Akita mice in parallel to the development of hyperglycaemia and hyperglucagonaemia (two- to eightfold increase). Acute exposure of mouse and human islets to amino acids stimulated alpha cell mTORC1 (3.5-fold increase), whereas high glucose concentrations inhibited mTORC1 (1.4-fold decrease). The mTORC1 response to glucose was abolished in human and mouse diabetic alpha cells following prolonged islet exposure to high glucose levels, resulting in sustained activation of mTORC1, along with increased glucagon secretion. Metabolomics and metabolic flux analysis showed that exposure to high glucose levels enhanced glycolysis, glucose oxidation and the synthesis of glucose-derived amino acids. In addition, chronic exposure to high glucose levels increased the expression of Slc7a2 and Slc38a4, which encode amino acid transporters, as well as the levels of branched-chain amino acids and methionine cycle metabolites (~1.3-fold increase for both). Finally, conditional Rptor knockout in alpha cells from adult diabetic mice inhibited mTORC1, thereby inhibiting glucagon secretion (~sixfold decrease) and improving diabetes, despite persistent insulin deficiency.

CONCLUSIONS/INTERPRETATION

Alpha cell exposure to hyperglycaemia enhances amino acid synthesis and transport, resulting in sustained activation of mTORC1, thereby increasing glucagon secretion. mTORC1 therefore plays a major role in mediating alpha cell dysfunction in diabetes.

DATA AVAILABILITY

All sequencing data are available from the Gene Expression Omnibus (GEO) repository (accession no. GSE154126; https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE154126 ).

Impaired counter regulation of hypoglycemia in a group of insulin-dependent diabetics with recurrent episodes of severe hypoglycemia

The counterregulatory response to insulin-induced hypoglycemia was investigated in 22 insulin-dependent diabetics (IDD) with recurrent hypoglycemia and in 6 healthy volunteers. Hypoglycemia was induced by a constant rate infusion of insulin (2.4 U/h) up to four hours. Conventional insulin therapy was changed to an i.v. infusion of regular insulin 24 hours prior to the experiment. The presence of diabetic autonomic neuropathy was evaluated by respiratory sinus arrhythmia and Valsalva maneuver. In healthy subjects, blood glucose was decreased to 2.5 mmol, here reaching steady state level and giving rise to marked glucagon and growth hormone (GH) responses. The majority of IDD (group A) reached a slightly lower steady state glucose level and exhibited similar glucagon and GH responses while the epinephrine response was augmented. Six IDD (group B) showed a continuous decrease in blood glucose to 1.2 +/- 0.1 mmol/l at which level the infusion of insulin was discontinued due to neuroglucopenic symptoms. These subjects had no glucagon and epinephrine responses while their GH and cortisol responses were normal. A comparison of the diabetic groups revealed a longer duration of diabetes and a more impaired autonomic nervous function in group B while glycosylated hemoglobin was similar. It is concluded that most IDD have normal hormonal responses (epinephrine, glucagon, GH, cortisol) and normal counterregulartory capacity to hypoglycemia induced by a prolonged infusion of a moderate dose of insulin. Some patients with long-term diabetes and impaired capacity to counteract hypoglycemia exhibit deficient glucagon and epinephrine responses to hypoglycemia.

Insulin responses to glucose in healthy males are associated with adult height but not with birth weight

OBJECTIVE

To investigate the relations between height, birth weight and insulin secretion.

SUBJECTS AND DESIGN

Subjects were selected from a register of all male healthy volunteers who had previously participated in insulin secretion studies. All men in whom a 1-h glucose infusion test had been performed on two or more occasions were selected (n = 88). Subjects were divided into two equally sized groups according to 0-10-min insulin responses.

MAIN OUTCOME MEASURES

Insulin responses were measured by standardized glucose infusion tests. Heights and weights were measured on these occasions. Birth weights were obtained from questionnaires and validated from obstetric records.

RESULTS

The average height for the 50% of subjects with the lowest insulin response was 3.5 cm less [95% confidence interval (CI): 1.1 to 5.9] than in those with the highest response, P < 0.005. These differences were also significant when expressed relative to heights of national cohorts with the same years of birth [1.7% less in those with lowest response (95% CI: 0.3 to 3.0), P < 0.02 for difference]. However, birth weight (known to 69% of subjects) was not associated with insulin response (3706 +/- 126 g in the 50% with lower insulin response, 3590 +/- 136 g in those with higher insulin response).

CONCLUSIONS

An early insulin response to glucose associates with postnatal growth. This suggests that physiological variations in postnatal insulin secretion can influence growth and height in healthy subjects. Furthermore, in the present study group, a low birth weight is not an important determinant of postnatal insulin secretion.

Dexamethasone treatment fails to increase arginine-induced insulin release in healthy subjects with low insulin response

We have compared insulin responses to L-arginine before and during dexamethasone treatment in healthy subjects, previously classified as subjects with either high or low insulin response according to a standardized glucose infusion test. Arginine stimulation was administered as a 150 mg/kg bolus followed by 10 mg.kg-1.min-1 to six subjects with high insulin response and to seven subjects with low insulin response. Before dexamethasone treatment the incremental insulin level during 0-10 min of arginine was higher in subjects with high (36.5 +/- 6.8 microU/ml) than in subjects with low response (14.5 +/- 2.3 microU/ml), p less than 0.01 for difference. Dexamethasone treatment (6 mg/day for 60 h) markedly enhanced the insulin response to arginine in subjects with high response (+99% 0-30 min) but failed to affect the subjects with low response (+4% 0-30 min). The C-peptide response to arginine exhibited similar differences between groups. Decreased responsiveness to arginine in subjects with low insulin response, especially during dexamethasone treatment, suggests a Beta-cell capacity defect although a decreased potentiating-sensing effect of glucose cannot be completely ruled out.

Plasma amino acids and the insulin/glucagon ratio as an explanation for the dietary protein modulation of atherosclerosis

The amino acid composition of the diet influences the postprandial levels of plasma amino acids along with the hormones insulin and glucagon in humans fed single test meals identical in composition except for protein source. Soy protein (hypocholesterolemic), vs. casein (hypercholesterolemic), contains a higher amount of arginine and glycine and induces an increase in postprandial arginine and glycine. Soy protein induces a low postprandial insulin/glucagon ratio in both hypercholesterolemic and normocholesterolemic subjects. Casein induces a high postprandial insulin/glucagon ratio among hypercholesterolemic subjects. Amino acids such as arginine and glycine are associated with a decrease, while lysine and branched-chain amino acids are associated with increased serum cholesterol levels. Our data are consistent with the hypothesis that the control of cholesterol by insulin and glucagon is regulated by dietary and plasma amino acids. From this hypothesis the insulin/glucagon ratio is proposed as an early metabolic index of the effect of dietary proteins on serum cholesterol levels, a risk factor and a common mechanism through which dietary and lifestyle factors influence cardiovascular disease.

Plasma amino acids and the insulin/glucagon ratio as an explanation for the dietary protein modulation of atherosclerosis

The amino acid composition of the diet influences the postprandial levels of plasma amino acids along with the hormones insulin and glucagon in humans fed single test meals identical in composition except for protein source. Soy protein (hypocholesterolemic), versus casein (hypercholesterolemic), contains a higher amount of arginine and glycine and induces an increase in postprandial arginine and glycine. Soy protein induces a low postprandial insulin/glucagon ratio in both hypercholesterolemic and normocholesterolemic subjects. Casein induces a high postprandial insulin/glucagon ration among hypercholesterolemic subjects. Amino acids such as arginine and glycine are associated with a decrease, while lysine and branched-chain amino acids are associated with increased serum cholesterol levels. Our data are consistent with the hypothesis that the control of cholesterol by insulin and glucagon is regulated by dietary and plasma amino acids. From this hypothesis the insulin/glucagon ratio is proposed as an early metabolic index of the effect of dietary proteins on serum cholesterol levels, a risk factor and a common mechanism through which dietary and lifestyle factors influence cardiovascular disease.

Metabolic tolerance to arginine: implications for the safe use of arginine salt-aztreonam combination in the neonatal period

Two similar cohorts of low birth weight infants whose size was appropriate for gestational age randomly received either aztreonam-arginine plus ampicillin (n = 15) or gentamicin plus ampicillin (n = 15) for empiric treatment of neonatal sepsis. The regimens were infused together with glucose at greater than 5 mg/kg per minute, and immediate (4 hours) and cumulative (3 days) effects were assessed. Serum arginine and insulin values rose immediately after administration of aztreonam (containing 0.15 mmol of arginine per kilogram), but there were no changes in the gentamicin-treated cohort; no differences occurred in either cohort in serum concentrations of glucose, ammonia, potassium, creatinine, and bilirubin. After 3 days of antibiotic therapy (n = 13), the baseline serum arginine concentration was almost twice as high in the aztreonam group and showed a similar further rise and fall during the 4 hours after infusion; arginine urinary fractional excretion (normalized to creatinine clearance) decreased in the gentamicin group. The indirect bilirubin concentration rose more (p less than 0.001) in the aztreonam-treated infants (5.1 to 11.5 mg/dl (87 to 196 mumol/L] than in the gentamicin-treated infants (5.5 to 8.1 mg/dl (94 to 138 mumol/L)). Thus a modest differential bilirubin response and modestly elevated baseline serum arginine level occurred after the 3-day low-arginine doses of this study; serum ammonia and glucose concentrations were not affected. Aztreonam-arginine in neonates was well tolerated metabolically, and we believe that it can be used safely in conjunction with attention to glucose and bilirubin metabolism.

Dose-kinetics of pancreatic alpha- and beta-cell responses to a protein meal in normal subjects

A protein meal is well known to induce a prompt secretion of insulin and glucagon. However, the data regarding the dose-response relationship between the protein meal and the insulin and glucagon responses are sparse. This study assessed the effects of ingestion of protein meals of varying amounts on plasma glucose [S], insulin [I], and glucagon [G] concentrations in eight normal subjects. Protein meals were administered after an overnight fast in a randomized sequence at intervals of 10 days in four different quantities: 250 mg/kg body weight (BW) (A), 500 mg/kg BW (B), 1 g/kg BW (C), and 2 g/kg BW (D). Mean S levels were not significantly altered following A, B, or C, although significant decreases in S responses were noted after C and D as reflected by absolute changes (delta) and/or the cumulative responses (CR) and the areas under the curve (sigma). Mean I increased promptly to peak concentration by 30 minutes, although in individual subjects the peak was achieved either at 30 or 60 minutes following all protein meals. The increase was progressively greater and the return was delayed with increasing quantities resulting in progressive elevations in delta I and percent increase from basal concentration (%), as well as CRI and sigma I. G increased following all protein meals as well. The mean peak G concentrations were achieved by 90 minutes, although in individual subjects the peak G was reached at 90 or 120 minutes, a significant delay in comparison to the peak I levels. G returned to base line only following ingestion of A during the study period.(ABSTRACT TRUNCATED AT 250 WORDS)

Response of hormones modulating plasma cholesterol to dietary casein or soy protein in minipigs

To elucidate the mechanism mediating the effect of dietary casein or soy protein on serum cholesterol concentrations we followed the endocrine response to the intake of these dietary proteins. The hormones analyzed were those known to modulate serum cholesterol concentration. A 7-wk crossover nutrition study was performed with six adult Göttingen minipigs consuming semisynthetic diets based on either 20 wt% casein or soy isolate. At d 42 and 49, concentrations of six hormones were determined in 22 blood samples taken over the whole day. There were no significant differences in insulin, glucagon, the insulin/glucagon ratio, hydrocortisone or triiodothyronine among dietary groups. In the late postprandial phase (5 h after the meal and later) there were significantly higher growth hormone concentrations in soy-fed animals. At all times of the day, total and free thyroxine concentrations were higher after soy feeding than after casein feeding. On average, total and free thyroxine concentrations were 34 and 26% higher with soy protein feeding than with casein feeding. Our data agree with other reports of protein-dependent changes of thyroid hormones and may explain why different dietary proteins have different effects on serum cholesterol levels in sensitive species.

Retrograde perfusion as a model for testing the relative effects of glucose versus insulin on the A cell

In order to determine whether the A cell may be directly suppressed by glucose in the absence of insulin, canine pancreata were perfused in vitro, both antegrade, through the arterial system and retrograde, through the venous system. Studies of the islet microvasculature have suggested that blood flows from the B cell core to the mantle; thus, the A cell may be tonically inhibited by intra-islet insulin. Retrograde perfusion may then be expected to prevent insulin from reaching the A cell, releasing it from inhibition. Retrograde perfusion with 88 mg/dl glucose markedly increased both insulin and glucagon secretion relative to antegrade levels. In a series of experiments, glucose concentrations were changed from 88 to 200 mg/dl. An antegrade glucose change resulted in increased insulin (134+/-21%; P less than 0.0025) and decreased glucagon (-26+/-9%, P less than 0.025) secretion. A retrograde glucose increase resulted in increased secretion of both insulin (91+/-15%; P less than 0.0005) and glucagon (23+/-9%; P less than 0.0125). To confirm that retrograde perfusion deprived the A cell of endogenous core derived, vascularly delivered insulin, possibly resulting in increased insulin sensitivity, 0.3 mU/ml exogenous porcine insulin was infused. Antegrade, 0.3 mU/ml insulin, had no effect on glucagon secretion (P less than 0.250), while retrograde infusion of 0.3 mU/ml insulin significantly inhibited glucagon secretion (-31 + 8%; P less than 0.0005). The results of our study support the concept that the direction of blood flow and of flow-dependent intra-islet hormone interactions are from the islet B cell core to the mantle. It was further concluded that the normal A cell may not be suppressed by glucose in the absence of insulin.

Time-dependent inhibition of insulin release: suppression of the arginine effect by hyperglycaemia

Brief stimulation of the pancreas with arginine causes a refractory state which reduces the insulin response to subsequent stimulations (time-dependent inhibition). In control subjects, a pair of arginine injections (75 mg/kg) at a 30-min interval resulted in 20% reduction of peak and integrated insulin responses to the second injection. In Type 2 (non-insulin-dependent) diabetic patients and in obese subjects, the inhibitory effect of repeated arginine stimuli was abolished. Healthy subjects were made acutely hyperglycaemic (9.3 +/- 0.3 mmol/l) by the glucose clamp technique. This induced a three- to fivefold greater insulin response to arginine. Compared to the response of diabetic subjects with similar hyperglycaemia, the control subjects secreted four to nine times more insulin. When the arginine stimulation was repeated 30 min later, no inhibition was observed, the second insulin response being instead augmented 1.5- to 1.8-fold. We conclude that (1) the insulin response to arginine is markedly reduced in Type 2 diabetes; (2) arginine-induced time-dependent inhibition of insulin release is abolished in patients with minimal to moderate hyperglycaemia; (3) this is probably due to the acute synergistic action of glucose and arginine on the B cell; (4) time-dependent inhibition of insulin release may be a protective mechanism against insulin oversecretion following repetitive stimulation of the pancreas; its abolition in hyperglycaemic states may be a compensatory mechanism, allowing substantial increases in insulin output.