The authors report on (1) the absorption of agmatine from the gastrointestinal tract as an important source of this polycation in the organism, (2) its organ distribution, and (3) its putative role in liver regeneration. When rats received 0.5 microCi [(14)C]agmatine contained in 5 grams of standard rat chow after a fasting period of 24 hours, radioactivity was recovered in all organs investigated, in blood, and in urine. In the liver 67% +/- 7% of administered radioactivity was found. After partial (two-thirds) hepatectomy, administration of 250 mg and 500 mg agmatine by gavage for 6 days reduced liver regeneration at day 7 by 20% and 22%, respectively, compared with animals that received no agmatine. Agmatine is absorbed from the gastrointestinal tract, probably by means of a specific transporter. It is likely that agmatine in the chyme of the gut represents an essential source of agmatine in the tissues of the organism. An increase in the availability of gastrointestinal agmatine for absorption impairs liver regeneration and may contribute to the development of liver diseases.

The aims of the present study were: (i) to investigate the accumulation of radioactivity in the stomach wall after luminal exposure of the rat isolated stomach to[(14)C]agmatine and (ii) to determine the distribution of radioactivity in various tissues after oral administration of this radiolabelled polyamine to rats in vivo. In isolated rat stomach, [(14)C]agmatine was accumulated in part by an energy-dependent uptake process that could be inhibited by phentolamine. These findings correspond to properties of the recently identified specific agmatine transporter in human glioma cells, suggesting that in rat stomach [(14)C]agmatine is taken up by such a carrier. In in vivo experiments, rats received 0.5 microCi [(14)C]agmatine adsorbed to 5 g rat standard chow after a fasting period of 24 h. After oral ingestion of [(14)C]agmatine, radioactivity was recovered in all organs investigated as well as in blood and urine. Radioactivity also seemed to be secreted into the pancreaticobiliary fluid, as it was recovered in the luminal content of distal ileum and sigmoid colon. Accumulation of radioactivity in organs and distal gut luminal content was dose-dependently decreased by simultaneous administration of putrescine. In conclusion, the present data are compatible with the view that agmatine can be absorbed in rat at least from the stomach and probably also from the gut by means of an energy-dependent agmatine transport mechanism. Agmatine itself and/or its degradation products, which also have the potential to be pharmacologically active, are unevenly distributed between the organs. Putative secretion of radioactivity into the pancreaticobiliary fluid suggests the potential for an enterohepatic circulation of agmatine. In view of the high intraluminal concentration of agmatine in the stomach and distal gut and the operation of an agmatine transporter, it is rather likely that agmatine in the chyme of the gut represents an important source for agmatine detected in the tissues of the organism.

1 In segments of rat vena cava preincubated with [ 3 H]-noradrenaline and superfused with physiological salt solution, the in¯uence of agmatine on the electrically evoked [ 3 H]-noradrenaline release, the EP 3 prostaglandin receptor-mediated and the a 2D -adrenoceptor-mediated inhibition of evoked [ 3 H]-noradrenaline release was investigated. 2 Agmatine (0.1 ± 10 mM) by itself was without e ect on evoked [ 3 H]-noradrenaline release. In the presence of 10 mM agmatine, the prostaglandin E 2 (PGE 2 )-induced EP 3 -receptor-mediated inhibition of [ 3 H]-noradrenaline release was not modi®ed, whereas the a 2D -adrenoceptor-mediated inhibition of [ 3 H]-noradrenaline release induced by noradrenaline, moxonidine or clonidine was more pronounced than in the absence of agmatine. However, 1 mM agmatine antagonized the moxonidine-induced inhibition of [ 3 H]-noradrenaline release. 3 Agmatine concentration-dependently inhibited the binding of [ 3 H]-clonidine and [ 3 H]-rauwolscine to rat brain cortex membranes (K i values 6 mM and 12 mM, respectively). In addition, 30 and 100 mM agmatine increased the rate of association and decreased the rate of dissociation of [ 3 H]-clonidine resulting in an increased a nity of the radioligand for the a 2D -adrenoceptors. 4 [ 14 C]-agmatine labelled speci®c binding sites on rat brain cortex membranes. In competition experiments. [ 14 C]-agmatine was inhibited from binding to its speci®c recognition sites by unlabelled agmatine, but not by rauwolscine and moxonidine. 5 In conclusion, the present data indicate that agmatine both acts as an antagonist at the ligand recognition site of the a 2D -adrenoceptor and enhances the e ects of a 2 -adrenoceptor agonists probably by binding to an allosteric binding site of the a 2D -adrenoceptor which seems to be labelled by [ 14 C]-agmatine.

Radioligand binding experiments carried out in cell membranes from rat and human stomach revealed the existence of non-adrenoceptor [3H]clonidine and [3H]idazoxan binding sites and of [3H]DTG (1,2-di-(2-tolyl)guanidine) binding sites. In rat stomach, specific binding was inhibited by imidazolines and guanidines and by non-imidazoline sigma-site ligands, respectively, at different rank orders of affinity, suggesting the existence of non-I1/non-I2 [3H]clonidine binding sites, I2-imidazoline binding sites as well as sigma 2-like-sites. These sites are not directly related to a postsynaptic contractile effect on rat gastric smooth muscle or to acid release from isolated gastric glands. Finally, we demonstrated that the gastric pathogen Helicobacter pylori is able to form and to release the endogenous imidazoline receptor ligand agmatine and that considerable amounts of agmatine are present in human gastric juice. The quantities of agmatine were higher in gastric juice from H. pylori-positive than H. pylori-negative patients.

Strips from rat glandular stomach were used to investigate whether the imidazoline derivatives clonidine, tolazoline and BDF 6143 (4-chloro-2(2-imidazolin-2-ylamino)-isoindoline) and the guanidine derivative agmatine are able to release histamine from histamine-storing cells. Histamine was detected and quantified by HPLC. Clonidine, tolazoline and agmatine concentration-dependently induced a release of histamine from the gastric strips, whereas BDF 6143 was ineffective. However, BDF 6143 abolished the release-inducing effect of clonidine and agmatine. It is concluded that imidazoline and guanidine derivatives can induce histamine release from histamine-storing cells in the stomach by a specific mechanism (possibly via imidazoline receptors) which in turn might lead to a relevant increase in gastric acid secretion.