In this study we investigated the effects of insulin-induced hypoglycaemia on tyrosine hydroxylase (TH) protein and TH phosphorylation in the adrenal gland, C1 cell group, locus coeruleus (LC) and midbrain dopaminergic cell groups that are thought to play a role in response to hypoglycaemia and compared the effects of different concentrations of insulin in rats. Insulin (1 and 10 U/kg) treatment caused similar reductions in blood glucose concentration (from 7.5-9 to 2-3 mmol/L); however, plasma adrenaline concentration was increased 20-30 fold in response to 10 U/kg insulin and only 14 fold following 1 U/kg. Time course studies (at 10 U/kg insulin) revealed that in the adrenal gland, Ser31 phosphorylation was increased between 30 and 90 min (4-5 fold), implying that TH was activated to increase catecholamine synthesis in adrenal medulla to replenish the stores. In the brain, Ser19 phosphorylation was limited to certain dopaminergic groups in the midbrain, while Ser31 phosphorylation was increased in most catecholaminergic regions at 60 min (1.3-2 fold), suggesting that Ser31 phosphorylation may be an important mechanism to maintain catecholamine synthesis in the brain. Comparing the effects of 1 and 10 U/kg insulin revealed that Ser31 phosphorylation was increased to similar extent in the adrenal gland and C1 cell group in response to both doses whereas Ser31 and Ser19 phosphorylation were only increased in response to 1 U/kg insulin in LC and in response to 10 U/kg insulin in most midbrain regions. Thus, the adrenal gland and some catecholaminergic brain regions become activated in response to insulin administration and brain catecholamines may be important for initiation of physiological defences against insulin-induced hypoglycaemia.

No abstract

Recurrent insulin-induced hypoglycaemia is a major limitation to insulin treatment in diabetes patients leading to a condition called hypoglycaemia-associated autonomic failure (HAAF). HAAF is characterised by reduced sympathoadrenal response to subsequent hypoglycaemia thereby predisposing the patients to severe hypoglycaemia that can lead to coma or even death. Despite several attempts being made, the mechanism of HAAF is yet to be clearly established. In order for the mechanism of HAAF to be elucidated, establishing a human/animal model of the phenomenon is the foremost requirement. Several research groups have attempted to reproduce the phenomenon in diabetic and nondiabetic humans and rodents and reported variable results. The success of the phenomenon is marked by a significant reduction in plasma adrenaline response to subsequent hypoglycaemic episode relative to that of the antecedent hypoglycaemic episode. A number of factors such as the insulin dosage, route of administration, fasting conditions, blood sampling methods and analyses, depth, duration, and number of antecedent hypoglycaemic episodes can impact the successful reproduction of the phenomenon and thus have to be carefully considered while developing the protocol. In this review, we have outlined the protocols followed by different research groups to reproduce the phenomenon in diabetic and nondiabetic humans and rodents including our own observations in rats and discussed the factors that have to be given careful consideration in reproducing the phenomenon successfully.

No abstract

Low blood glucose activates brainstem adrenergic and cholinergic neurons, driving adrenaline secretion from the adrenal medulla and glucagon release from the pancreas. Despite their roles in maintaining glucose homeostasis, the distributions of insulin-responsive adrenergic and cholinergic neurons in the medulla are unknown. We fasted rats overnight and gave them insulin (10 U/kg i.p.) or saline after 2 weeks of handling. Blood samples were collected before injection and before perfusion at 90 min. We immunoperoxidase-stained transverse sections of perfused medulla to show Fos plus either phenylethanolamine N-methyltransferase (PNMT) or choline acetyltransferase (ChAT). Insulin injection lowered blood glucose from 4.9 ± 0.3 mmol/L to 1.7 ± 0.2 mmol/L (mean ± SEM; n = 6); saline injection had no effect. In insulin-treated rats, many PNMT-immunoreactive C1 neurons had Fos-immunoreactive nuclei, with the proportion of activated neurons being highest in the caudal part of the C1 column. In the rostral ventrolateral medulla, 33.3% ± 1.4% (n = 8) of C1 neurons were Fos-positive. Insulin also induced Fos in 47.2% ± 2.0% (n = 5) of dorsal medullary C3 neurons and in some C2 neurons. In the dorsal motor nucleus of the vagus (DMV), insulin evoked Fos in many ChAT-positive neurons. Activated neurons were concentrated in the medial and middle regions of the DMV beneath and just rostral to the area postrema. In control rats, very few C1, C2, or C3 neurons and no DMV neurons were Fos-positive. The high numbers of PNMT-immunoreactive and ChAT-immunoreactive neurons that express Fos after insulin treatment reinforce the importance of these neurons in the central response to a decrease in glucose bioavailability.