Central administration of glucagon-like peptide-1 activates hypothalamic neuroendocrine neurons in the rat

Within the central nervous system, glucagon-like peptide-1-(7-36) amide (GLP-1) acts as a transmitter, inhibiting feeding and drinking behavior. Hypothalamic neuroendocrine neurons are centrally involved in the regulatory mechanisms controlling these behaviors, and high densities of GLP-1 binding sites are present in the rat hypothalamus. In the present study we have, over a period of 4 h, followed the effect of centrally injected GLP-1 on plasma levels of the neurohypophysial hormones vasopressin and oxytocin. Plasma levels of corticosterone and glucose were also followed across time after central administration of GLP-1. In conscious, freely moving, and unstressed rats, central injection of GLP-1 significantly elevated plasma levels of vasopressin 15 and 30 min after administration (basal, 0.8 +/- 0.2 pg/ml; 15 min, 7.5 +/- 2.0 pg/ml; 30 min, 5.6 +/- 1.1 pg/ml; mean +/- SEM) and elevated corticosterone 15 min after administration (52 +/- 13 vs. 447 +/- 108 ng/ml, basal vs. 15 min; mean +/- SEM). In contrast, plasma oxytocin levels were unaffected by intracerebroventricular (icv) injections of GLP-1 over a period of 4 h after the injection. The animals given a central injection of GLP-1 developed transient hypoglycemia 20 min after the injection, which was fully restored to normal levels at 30 min. Furthermore, we used c-fos immunocytochemistry as an index of stimulated neuronal activity. The distribution and quantity of GLP-1-induced c-fos immunoreactivity were evaluated in a number of hypothalamic neuroendocrine areas, including the magnocellular neurons of the paraventricular (PVN) and supraoptic (SON) nuclei and the parvicellular neurons of the medial parvicellular subregion of the PVN. The number of c-fos-expressing nuclei in those areas was assessed 30, 60, and 90 min after icv administration of GLP-1. Intracerebroventricular injection of GLP-1 induced c-fos expression in the medial parvicellular subregion of the PVN as well as in magnocellular neurons of the PVN and SON. A slight induction of c-fos expression was seen in the arcuate nucleus and the nucleus of the solitary tract, including the area postrema. In contrast, the subfornical organ, which is a rostrally situated circumventricular organ, was free of c-fos-positive cells after central administration of GLP-1. When the GLP-1 antagonist exendin-(9-39) was given before the GLP-1, c-fos expression in these neuroendocrine areas was almost completely abolished, suggesting that the effect of GLP-1 on c-fos expression is mediated via specific receptors. A dual labeling immunocytochemical technique was used to identify the phenotypes of some of the neurons containing c-fos-immunoreactive nuclei. Approximately 80% of the CRH-positive neurons in the hypophysiotropic medial parvicellular part of the PVN coexpressed c-fos 90 min after icv GLP-1 administration. In contrast, very few (approximately 10%) of the vasopressinergic magnocellular neurons of the PVN/SON contained c-fos-positive nuclei, whereas approximately 38% of the magnocellular oxytocinergic neurons expressed c-fos-positive nuclei in response to GLP-1 administration. This study demonstrates that central administration of the anorectic neuropeptide GLP-1 activates the central CRH-containing neurons of the hypothalamo-pituitary-adrenocortical axis as well as oxytocinergic neurons of the hypothalamo-neurohypophysial tract. Therefore, we conclude that GLP-1 activates the hypothalamo-pituitary-adrenocortical axis primarily through stimulation of CRH neurons, and this activation may also be responsible for the inhibition of feeding behavior.