The sequence of glucagon-like peptide-1 (7-36) amide (GLP-1) is completely conserved in all mammalian species studied, implying that it plays a critical physiological role. We have shown that GLP-1 and its specific receptors are present in the hypothalamus. No physiological role for central GLP-1 has been established. We report here that intracerebroventricular (ICV) GLP-1 powerfully inhibits feeding in fasted rats. ICV injection of the specific GLP-1-receptor antagonist, exendin (9-39), blocked the inhibitory effect of GLP-1 on food intake. Exendin (9-39) alone had no influence on fast-induced feeding but more than doubled food intake in satiated rats, and augmented the feeding response to the appetite stimulant, neuropeptide Y. Induction of c-fos is a marker of neuronal activation. Following ICV GLP-1 injection, c-fos appeared exclusively in the paraventricular nucleus of the hypothalamus and central nucleus of the amygdala, and this was inhibited by prior administration of exendin (9-39). Both of these regions of the brain are of primary importance in the regulation of feeding. These findings suggest that central GLP-1 is a new physiological mediator of satiety.

A massive neuronal system was detected by immunocytochemistry and radioimmunoassay with antibodies to neuropeptide Y, the recently isolated peptide of the pancreatic polypeptide family. Immunoreactive cell bodies and fibers were most prevalent in cortical, limbic, and hypothalamic regions. Neuropeptide Y was extracted in concentrations higher than those of any other peptide hitherto discovered in the mammalian brain. Column chromatography of brain extracts and double immunostaining experiments indicate that neuropeptide Y is the endogenous brain peptide responsible for immunostaining of pancreatic polypeptide-like immunoreactivity in the mammalian brain.

Calcitonin gene-related peptide (CGRP) immunoreactivity was found throughout the entire spinal cord of man, marmoset, horse, pig, cat, guinea pig, mouse, rat, and frog. CGRP-immunoreactive fibers were most concentrated in the dorsal horn. In the ventral horn of some species large immunoreactive cells, tentatively characterized as motoneurons, were present. Pretreatment of rats with colchicine enhanced staining of these large cells but did not reveal CGRP-immunoreactive cell bodies in the dorsal horn. In the dorsal root ganglia, CGRP immunoreactivity was observed in most of the small and some of the intermediate sized cells. Substance P immunoreactivity, where present, was co-localized with CGRP to a proportion of the small cells. In the cat the ratio of substance P-immunoreactive to CGRP-immunoreactive ganglion cells was 1:2.7 (p less than 0.001). The concentration of CGRP-immunoreactive material in tissue extracts was determined by radioimmunoassay. In the dorsal horn of the rat spinal cord the levels of peptide were found to range from 225.7 +/- 30.0 pmol/gm of wet weight in the cervical region to 340.6 +/- 74.6 pmol/gm in the sacral spinal cord. In the rat ventral spinal cord, levels of 15.7 +/- 2.7 to 35.1 +/- 10.6 pmol/gm were found. The concentration in dorsal root ganglia of the lumbar region was 225.4 +/- 46.9 pmol/gm. Gel permeation chromatography of this extractable CGRP-like immunoreactivity revealed three distinct immunoreactive peaks, one eluting at the position of synthetic CGRP and the others, of smaller size, eluting later. In cats and rats, rhizotomy induced a marked loss of CGRP-immunoreactive fibers from the dorsal horn of the spinal cord. In the cat, unilateral lumbosacral dorsal rhizotomy resulted in a significant (p less than 0.05) reduction of extractable CGRP from the ipsilateral lumbar dorsal horn (5.6 +/- 1.2 pmol/gm of wet weight) compared to the contralateral side (105.0 +/- 36.0 pmol/gm of wet weight). We conclude that the major origin of CGRP in the dorsal spinal cord is extrinsic, from afferent fibers which are probably derived from cells in the dorsal root ganglia. The selective distribution of CGRP throughout sensory, motor, and autonomic areas of the spinal cord suggests many putative roles for this novel peptide.

Glucagon-like peptide-1 7-36 amide (GLP-1) has been postulated to be the primary hormonal mediator of the enteroinsular axis but evidence has been indirect. The discovery of exendin (9-39), a GLP-1 receptor antagonist, allowed this to be further investigated. The IC5o for GLP-1 receptor binding, using RIN 5AH fl-cell membranes, was found to be 0.36 nmol/l for GLP-1 and 3.44 nmol/l for exendin (9-39). There was no competition by exendin (9-39) at binding sites for glucagon or related peptides. In the anaesthetized fasted rat, insulin release after four doses of GLP-1 (0.1, 0.2, 0.3, and 0.4 nmol/kg) was tested by a 2-min intravenous infusion. Exendin (9-39) (1.5, 3.0, and 4.5 nmol/kg) was administered with GLP-1 0.3 nmol/kg, or saline, and only the highest dose fully inhibited insulin release. Exendin (9-39) at 4.5 nmol/kg had no effect on glucose, arginine, vasoactive intestinal peptide or glucose-dependent insulinotropic peptide stimulated insulin secretion. Postprandial insulin release was studied in conditioned conscious rats after a standard meal. Exendin (9-39) (0.5 nmol/kg) considerably reduced postprandial insulin concentrations, for example by 48% at 15 min (431±21 pmol/ I saline, 224±32 pmol/l exendin, P < 0.001). Thus, GLP-1 appears to play a major role in the entero-insular axis. (J. Clin. Invest. 1995.95:417-421.)

SUMMARY Previous studies have shown that ileal infusion of partially digested triglyceride inhibits jejunal motility. The partial digest used in those studies contained a mixture of glycerol, free fatty acid, mono-, di-, and triglycerides. In Part I of the present study we have separately infused emulsions containing either glycerol 3.1 g (n=6), oleic acid 9-6 g (n=6), triolein 10 g (n=12), or medium chain triglycerides 10 g (n=6) into the ileum and have recorded the effect this has on jejunal motility. Five further subjects received infusions of partial hydrolysates of corn starch 10 g and lactalbumin 7 g. Marked inhibition of jejunal pressure wave activity was seen after all three lipid infusions, per cent activity falling from a control of 37-7 (7.7) to 6-2 (2 1) and 22.4 (8.2)% 30 min after completing the oleic acid and triolein infusions respectively, and from a control value of 39.5 (4.1) to 17.7 (4.7) after MCTs (all p<005). No significant fall occurred after infusion of glycerol, protein or carbohydrate. All three lipid infusions raised plasma concentrations of neurotensin, enteroglucagon and peptide YY equally effectively, although only the rise in peptide YY correlated significantly with the inihibition ofjejunal pressure wave activity (r=0 80, n=6, p