Contribution of galanin to stress-induced impairment of insulin secretion in swimming mice

Energy homeostasis targets chromosomal reconfiguration of the human GH1 locus

Levels of pituitary growth hormone (GH), a metabolic homeostatic factor with strong lipolytic activity, are decreased in obese individuals. GH declines prior to the onset of weight gain in response to excess caloric intake and hyperinsulinemia; however, the mechanism by which GH is reduced is not clear. We used transgenic mice expressing the human GH (hGH) gene, GH1, to assess the effect of high caloric intake on expression as well as the local chromosome structure of the intact GH1 locus. Animals exposed to 3 days of high caloric intake exhibited hyperinsulinemia without hyperglycemia and a decrease in both hGH synthesis and secretion, but no difference in endogenous production of murine GH. Efficient GH1 expression requires a long-range intrachromosomal interaction between remote enhancer sequences and the proximal promoter region through "looping" of intervening chromatin. High caloric intake disrupted this interaction and decreased both histone H3/H4 hyperacetylation and RNA polymerase II occupancy at the GH1 promoter. Incorporation of physical activity muted the effects of excess caloric intake on insulin levels, GH1 promoter hyperacetylation, chromosomal architecture, and expression. These results indicate that energy homeostasis alters postnatal hGH synthesis through dynamic changes in the 3-dimensional chromatin structure of the GH1 locus, including structures required for cell type specificity during development.

Loss-of-function mutation of the galanin gene is associated with perturbed islet function in mice

The neuropeptide galanin is expressed in sympathetic nerve terminals that surround islet cells and inhibits insulin secretion. To explore its role for islet function, we studied mice with a loss-of-function mutation in the galanin gene [galanin knockout (KO) mice]. Intravenous 2-deoxy-glucose, which activates both the sympathetic and parasympathetic branches of the autonomic nervous system, caused an initial (1-5 min) inhibition of insulin secretion that was impaired in galanin KO mice (P = 0.027), followed by a subsequent stimulation of insulin secretion that was augmented in galanin KO mice (P < 0.01). Similar effects were seen after chemical sympathectomy by 6-hydroxydopamine. In contrast, galanin KO mice had a reduced insulin response to glucose, both in vivo (P < 0.001) and in isolated islets (P < 0.001), and to arginine, both in vivo (P = 0.012) and in vitro (P = 0.018). During an iv glucose tolerance test, galanin KO mice had impaired glucose disposal (P = 0.005) due to a reduced insulin response (P < 0.001) and a reduced insulin-independent glucose elimination (glucose effectiveness; P = 0.040). Insulin sensitivity, as judged by a euglycemic, hyperinsulinemic clamp technique, was slightly increased in galanin KO mice (P = 0.032). We conclude that 1) galanin may contribute to sympathetic influences inhibiting insulin secretion in mice, and 2) galanin KO mice have a reduced glucose-induced insulin secretion.

Pancreatic innervation is not essential for exercise-induced changes in glucagon and insulin or glucose kinetics

The purpose of this study was to determine the role of pancreatic innervation in mediating exercise-induced changes in pancreatic hormone secretion and glucose kinetics. Dogs underwent surgery >16 days before an experiment, at which time flow probes were implanted on the portal vein and the hepatic artery, and Silastic catheters were inserted in the carotid artery, portal vein, and hepatic vein for sampling. In one group of dogs (DP) all nerves and plexuses to the pancreas were sectioned during surgery. A second group of dogs underwent sham denervation (SHAM). Pancreatic tissue norepinephrine was reduced by >98% in DP dogs. Each study consisted of basal (-30 to 0 min) and moderate exercise (0 to 150 min, 100 m/min, 12% grade) periods. Isotope ([3-(3)H]glucose) dilution and arteriovenous differences were used to assess hepatic function. Arterial and portal vein glucagon and insulin concentrations and the rate of net extrahepatic splanchnic glucagon release (NESGR) were similar in DP and SHAM during the basal period. Arterial and portal vein glucagon and NESGR increased similarly in DP and SHAM during exercise. Arterial and portal vein insulin were similar during exercise. Arterial glucose, tracer-determined endogenous glucose production, and net hepatic glucose output were similar in DP and SHAM during the basal and exercise periods. These results demonstrate that pancreatic nerves are not essential to pancreatic hormone secretion or glucose homeostasis during rest or moderate exercise.

Stress and colchicine do not induce the release of galanin from the external zone of the median eminence

The aim of the present study was to verify the hypothesis that stress exposure modifies the content and release of galanin in the hypothalamic paraventricular nucleus and the median eminence. Colchicine and immobilization served as stress stimuli, and the changes in galanin immunoreactivity were compared with those in corticotropin-releasing hormone and vasopressin. In control animals, a limited number of galanin perikarya were identified in the paraventricular nucleus. The high dose (75 microg) of colchicine enhanced galanin in both parvicellular and magnocellular subdivisions, as analysed 72 h later. In the median eminence, galanin accumulated only in the external zone. High-dose colchicine did not affect galanin, while corticotropin-releasing hormone and vasopressin were depleted from the median eminence. Immobilization (120 min) neither alone nor in combination with colchicine influenced galanin immunoreactivity in the external zone. The low dose of colchicine induced an unexpected accumulation of galanin in the internal zone of the median eminence, which was further increased by subsequent immobilization. In the external zone, low-dose colchicine induced a complete disappearance of vasopressin, substantial depletion of corticotropin-releasing hormone and no changes in galanin immunoreactivity. The present studies demonstrate that galanin in the external zone of the median eminence is not influenced by colchicine or by immobilization stress.

Plasma galanin immunoreactivity in the rat after swimming

The purpose of our experiment was to study plasma immunoreactive galanin in the rat after swimming. Four groups of rats were used. At rest, one group was studied after an intravenous injection of D-glucose while another group received a corresponding saline injection. The two remaining groups, treated respectively with glucose and saline, were investigated after a 30-min swimming session. After 30 min of rest in their respective cages, or after swimming, the animals were anesthetized for immediate blood sampling. The main observation was that plasma galanin was higher after swimming than after rest only in glucose-treated rats, in the vena cava (11.82 +/- 2.90 vs. 5.05 +/- 1.65 pM) and the portal vein (15.75 +/- 3.74 vs. 6.58 +/- 1.75 pM). Both saline- and glucose-treated groups had a significant increase of plasma norepinephrine from rest to swimming in the vena cava. A decrease of plasma insulin was observed in the portal vein of exercised glucose-treated rats, while plasma glucose was higher in the portal vein of this group. In conclusion, after swimming, plasma galanin level was increased only in glucose-treated rats. A significant correlation was also observed between plasma galanin and glucose (r = 0.69, p < 0.01), suggesting that the presence of glucose is necessary to induce galanin liberation during exercise.

Possible evidence for endogenous production of a novel galanin-like peptide

Galanin mRNA and peptide are not detectable in normal islets. We studied the effect of galanin antagonists on insulin secretion in the rat beta cell line, RIN5AH, and in perifused rat islets. In RIN cell membranes galanin and its antagonists showed high affinity for 125I-galanin binding sites [Kd: (galanin) 0.03+/-0.01; Ki for galanin antagonists: (C7) 0.12+/- 0.02, (M35) 0.21+/-0.04, and (M40) 0.22+/-0.03 nM, mean+/- SEM, n = 4]. Galanin (1 microM) inhibited glucose-induced insulin release in islets (control 21.2+/-1.5 vs. galanin 4.5+/-0.2 fmol/islet per min, P < 0.001, n = 6) and RIN5AH cells (control 0.26+/-0.01 vs. galanin 0.15+/-0.02 pmol/10(6) cells per h, P < 0.001, n = 9). In RIN5AH cells, all antagonists blocked the inhibitory effects of galanin and stimulated insulin release in the absence of galanin. C7 and M40 (1 microM) alone significantly stimulated glucose-induced insulin secretion. Purified porcine galanin antibody (GAb) enhanced glucose-induced insulin release from islets (control 100+/- 16.3% vs. GAb 806.1+/-10.4%, P < 0.001, n = 6), and RIN5AH cells (control 100+/-9.6% vs. GAb 149+/-6.8%, P < 0. 01, n = 6). Western blotting of dexamethasone-treated islet extracts using GAb showed a specific band of similar molecular weight to porcine galanin not detected using a rat specific galanin antibody. One possible explanation for these results is the presence of an endogenous galanin-like peptide.

Immunologically induced sympathectomy of preganglionic nerves by antibodies against acetylcholinesterase: increased levels of peptides and their messenger RNAs in rat adrenal chromaffin cells

Systemic administration of murine monoclonal acetylcholinesterase antibodies to rats has been shown to cause selective degeneration of sympathetic preganglionic neurons. In the present study rats were subjected to a single i.v. injection of these acetylcholinesterase antibodies, or to normal IgG or saline for control. Exophthalmos, piloerection and eyelid-drooping (ptosis) were observed within 1 h after administration of the antibodies. Rats were killed at different time-points after antibody administration, and the adrenal glands were analysed by means of indirect immunohistochemistry and in situ hybridization histochemistry. As soon as 3 h after the antibody treatment, a marked increase in the number of chromaffin cells expressing mRNA encoding, respectively, enkephalin, calcitonin gene-related peptide, galanin, neurotensin and substance P was seen. At 12 h the peptide mRNA levels were still elevated and there was a concomitant increase in the number of peptide-immunoreactive cells. All peptide levels remained high for at least 48 h; however, 77 days after the antibody treatment only enkephalin-immunoreactive cells could be encountered. A disappearance of acetylcholinesterase- and enkephalin-immunoreactive cells could be encountered. A disappearance of acetylcholinesterase- and enkephalin-positive fibers was already seen 3 h after the antibody treatment, and after 24 h no fibers were encountered. In contrast, up until 48 h there was no apparent change in the number or intensity of immunofluorescent fibers expressing calcitonin gene-related peptide, galanin, neurotensin or substance P. However, 77 days after the antibody treatment the number of calcitonin gene-related peptide- and substance P-immunoreactive fibers was increased as compared to controls. In addition, reappearance of acetylcholinesterase- and enkephalin-immunoreactive fibers was seen 77 days after antibody administration, although their number was still low as compared to controls. Double-labeling immunohistochemistry revealed that the chromaffin cells expressing peptides after the antibody treatment preferentially were adrenaline storing cells (noradrenaline-negative). The majority of these cells expressed only one peptide. Both surgical transection of the splanchnic nerve as well as treatment with acetylcholine receptor antagonists mimicked the effects seen after the acetylcholinesterase-antibody treatment, although changes were less pronounced. The present results show that interruption of splanchnic transmission induces fast, marked, and selective increases in peptide expression in rat adrenal chromaffin cells.

N-terminally elongated fragments of galanin(1-16) inhibit insulin secretion from isolated mouse islets

The neuropeptide galanin inhibits insulin secretion and has been suggested to be an adrenergic co-transmitter in the endocrine pancreas. Recently, N-terminally elongated forms of galanin have been identified in both porcine brain and adrenals. Whether these elongated peptides show galanin-like biological effects is not known. We therefore synthesized two N-terminally elongated fragments of galanin(1-16), which contains the active site of galanin. The synthesized peptides were galanin(-9-16) and galanin(-7-16), which correspond to amino acids 24-61 and 26-61 in the galanin precursor molecule. Both these peptides were found to potently inhibit glucose-(11.1 mM)-stimulated insulin secretion from isolated mouse islets of Langerhans in all concentrations studied (1-1000 nM) (P < 0.0001). The potency of the peptides was not different from that of synthetic rat galanin. Thus, at 100 nM, insulin secretion was inhibited by galanin(-7-16) by 83 +/- 7% and by galanin(-9-16) by 71 +/- 17% and by rat galanin by 93 +/- 4% (not statistically different). Furthermore, the galanin receptor antagonist, M35 (10 nM), prevented the inhibitory action of the two N-terminally galanin fragments. This study thus shows that N-terminally elongated galanin-fragments as entire galanin inhibits insulin and thus indicates that the effect of galanin on insulin secretion is not dependent on a free amino-terminus.

Galanin message-associated polypeptide (GMAP) does not affect insulin secretion from isolated islets

Galanin message-associated polypeptide (GMAP) is processed from preprogalanin. Recently, GMAP-like immunoreactivity was demonstrated in insulin cells in the endocrine pancreas. We therefore examined whether synthetic rat GMAP, like galanin, inhibits glucose-stimulated insulin secretion from isolated rat and mouse islets. We found, however, that over a wide dose range (0.1 nM to 1 microM) GMAP did not affect insulin secretion stimulated by 8.3 or 11.1 mM glucose during a 60-min incubation of single rat or mouse islets. In contrast, rat galanin, as expected, completely abolished glucose-stimulated insulin secretion at 100 nM. Thus, in contrast to galanin, GMAP does not affect insulin secretion in isolated rodent islets.

Tissue-specific reduction of galanin content in the pancreas in alloxan diabetes in the mouse

Galanin inhibits insulin secretion and has been proposed to function as a sympathetic neurotransmitter in the endocrine pancreas in some species, for example in the dog. In this study, pancreatic and adrenal gland galanin content were measured following experimental diabetes induced by alloxan in mice. Three days after administration of alloxan (70 mg kg-1, i.p.) in normal mice, pancreatic content of galanin-like immunoreactivity (GLIR) was reduced to 65 +/- 11% of that in untreated controls (P < 0.01), whereas adrenal gland GLIR was unchanged. Similarly, 8 days after alloxan administration, pancreatic GLIR was reduced (P < 0.002), whereas adrenal gland GLIR was unaffected. Pancreatic GLIR also inversely correlated with plasma glucose levels (r = -0.5055, P < 0.005). To distinguish between the direct effects of alloxan vs. indirect metabolic effects induced by the drug, alloxan-diabetic mice were treated with insulin twice daily, which normalized the plasma glucose levels (7.6 +/- 0.3 mmol l-1). Pancreatic GLIR was then not significantly different from controls. Thus pancreatic but not adrenal gland GLIR content is reduced in alloxan-induced diabetes in mice. The data support a role for galanin as a pancreatic sympathetic neurotransmitter which may participate in the metabolic alterations seen in alloxan diabetes in mice.

Blockade of galanin-induced inhibition of insulin secretion from isolated mouse islets by the non-methionine containing antagonist M35

The neuropeptide galanin occurs in pancreatic adrenergic nerves and has been suggested to be the adrenergic mediator of the stress-induced inhibition of insulin release. To study its physiological function, we recently synthesized a galanin-like galanin receptor antagonist, galantide. However, this antagonist contains a methionine moiety, and is therefore easily oxidized. We have now synthesized another galanin antagonist which does not contain methionine. This peptide, M35, is a chimeric 21 amino acid peptide in which galanin-(1-13) is coupled to bradykinin-(2-9). M35 (10 microM to 1 pM) had no effect by itself on glucose (11.1 mM)-stimulated insulin secretion in isolated mouse islets, but potently counteracted the inhibitory action of galanin (100 nM). The lowest effective dose of M35 was 10 nM. M35 did not counteract the inhibitory action of clonidine (1 microM) or somatostatin (1 microM) on insulin secretion. Furthermore, M35 displaced 125I-monoiodo-[Tyr26]galanin from membranes of insulin producing RINm5F cells. The displacement curve fitted to a two-site model in which 60% of label bound with a K1 of 0.1 +/- 0.01 nM and 40% with a K2 of 3 +/- 0.5 nM. In conclusion, M35 is a specific, non-methionine-containing galanin receptor antagonist on insulin-producing cells.

Plasma neuropeptide Y (NPY) and galanin before and during exercise in type 1 diabetic patients with autonomic dysfunction

Plasma neuropeptide Y (NPY), plasma galanin and plasma catecholamines were determined before and during an ergometer exercise test in 11 type 1 diabetic patients (age 19-36 years, mean 30; duration of diabetes 2-18 years, mean 9) with autonomic dysfunction and in 13 age-matched healthy controls (age 24-36 years, mean 29). Before exercise, plasma NPY (100 +/- 6 pmol/l vs 144 +/- 7 pmol/l; P less than 0.001) and plasma galanin (54 +/- 3 pmol/l vs 77 +/- 5 pmol/l; P less than 0.005) were significantly lower in patients than in controls. During exercise, plasma NPY, plasma adrenaline, and plasma noradrenaline increased in patients and controls while galanin only increased in patients. Since there was a direct correlation between plasma NPY before exercise and the increment (delta 80%) in noradrenaline during exercise (r = 0.54; P less than 0.01), it is suggested that plasma NPY determined in the basal situation may be a useful marker of sympathetic nerve failure in diabetic patients.