April 2000 3224 CHOLINERGIC NERVE DYSFUNCTION IN A MODEL OF MU- RINE INTESTINAL INFLAMMATION IS MEDIATED BY MAC· ROPHAGE-COLONY STIMULULATING FACTOR (M-CSF)- DE- PENDENT MACROPHAGES . Francesca Galeazzi, Paola Lovato. Patricia A. Blennerhassett, Bruce A. Vallance, Stephen M. Collins, Univ of Padova, Padova, Italy; McMaster Univ, Hamilton, ON. Canada. Macrophage-colon y stimulating factor (M-CSF) is a key cytokine in the development and differentiation of the monocyte-macrophage lineage . We have previously shown that macrophages playa pivotal role in mediating functional impairment of cholinergic enteric nerves during intestinal in- flammation. However. the macrophage subset responsible for these changes, as well as the mediators involved in the recruitment of these cells during intestinal inflammation are unknown. Thus, the aim of this study was to evaluate the role of M-CSF dependent macrophages on cholinergic enteric nerve function and to determine the involvement of M-CSF in the macrophage proliferation during inflammation. by using nematode-infec - tion in the M-CSF deficient (op/op) mice. M-CSF-deficient op/op mice and their normallillermates (opt?) were infected with 375 T.spiralis larvae. The presence of F4/8D+ macrophages and ACh release from longitudinl mus- cle myenteric plexus preparations were evaluated at day 6 post-infection (PI). Peripheral blood monocytes were also evaluated . In parallel experi- ments, op/op mice were reconstituted with human recombinant M-CSF administered daily during the infection by intraperitoneal injections . In opt? mice, infection caused a 500 % increase in circulating monocytes. There was an infiltrate of F4/8D macrophage s in the muscularis externa and myenteric plexus, and an increased number of lamina propria macro- phages. In contrast, in op/op mice the number of circulating monocytes remained unchanged after infection, and no F4/8D macrophages were evident in the muscularis externa. As expected , infection caused a 7D% suppression of ACh release in normal and opt? mice. In contrast, cholin- ergic nerve function remained normal after infection in op/op mice. How- ever, treatment of op/op mice with M-CSF returned numbers of circulat ing monocytes to normal and restored the presence of macrophages in the muscularis externa and myenteric plexus. In addition, this was accompa- nied by a suppression in ACh release (68%), similar to that seen in infected M-CSF expressing mice. We conclude that M-CSF is critical for the recruitment and proliferation of macrophages in the inflamed intestine , and that the subset of M-CSF dependent macrophages is directly responsible for the functional suppression of cholinergic nerves during intestinal in- flammation. Supported by MRC Canada. 3225 5-HT J RECEPTORS ON SPINAL PRIMARY AFFERENTS PRO- JECTING TO THE COLON OF THE RAT - A PERIPHERAL TAR- GET FOR ALOSETRON. Gareth A. Hicks, Marcus Schindler , Philip A. Bland-Ward , Sarah Lummis , Patrick Pa Humphrey, Glaxo Institute of Applied Pharmacology, Cam- bridge, United Kingdom; Dept of Biochemistry, Cambridge, United King- dom. The 5-HT3 receptor antagonist, alosetron , relieves the pain associated with irritable bowel syndrome (Mangel & Northcutt, 1999 Aliment. Pharmacol. Ther. 13 Suppl. 2, 77-82) and inhibits spinal cord c-fos expression, and the vasodepressor reflex, in response to noxious colorectal distension in rats (Scott et al., 1997 Gastroenterol . 112: A822). Agonists at 5-HT 3 receptors depolarize dorsal root ganglion (DRG) neurones (Todorovic & Anderson , 199D Brain Res. 511,71-79), and excite intestinal afferents (Hillsley et al ., 1999 J. Physiol. 15(2),551-61 ). We have combined retrograde neuronal labelling and immunohistochemical techniques to determine whether the spinal primary afferent fibres, projecting specifically to the gut, express 5-HT3 receptors. Injections (2-5 /-tl ) of the fluorescent neuronal tracer "fast blue" (5% in saline) were made into the wall of the descending colon in adult male isoflurane-anaesthetised Wistar rats (120-18Dg), following lap- arotomy. 10-14 days later, animals were killed and Tl2-S2 DRG were dissected and fixed overnight in 4% PFA. Sections (12/-tm) were cut and mounted for immunohistochemistry and observation of fluorescent label. Cell size was determined with image analysis software (UTHSCSA Im- agetool v2.D). Data is presented as mean := SD. The majority of fast blue-labelled cells were found in 4 specific DRG: L1, L2, L6 and S1. In other ganglia, occasional weak labelling. or no labelling, was observed. Most labelled cells had cross-sectional areas between 3OD-900/-tm 2 (818:=277/-tm 2 ), but a small proportion of larger cells were also labelled (up to 1500/-tm 2 ). 5-HT 3 receptor immunohistochemistry was performed in S I DRG sections and this receptor was localised to cells with cross- sectional areas ranging from 25D-175D/-tm 2 (816:=357/-tm 2 ). In these SI DRG sections, 29% of cells labelled with fast blue were positive for 5-HT3 receptors. The mean cross sectional area of this dual-labelled neuronal population was 773±218/-tm 2 • Many fast blue-labelled cells contained CGRP, but little or no co-localisation was found with isolectinB4. The above data provide anatomical evidence for the existence of 5-HT 3 recep- tors on cell bodies of a sub-population of spinal primary afferent neurones projecting to the descending colon . 5-HT released in the gut wall could activate 5-HT 3 receptors on the peripheral terminals of these neurones , which represent a potential site of action for the visceral analgesic effect of alosetron. AGAA631 3226 CONTROL OF PERISTALSIS IN THE GUINEA-PIG ISOLATED INTESTINE BY CYCLOOXYGENASE ISOFORMS COX-! AND COX-2. Peter Holzer, Anaid Shahbazian, Rufina Schuligoi, Bernhard A. Peskar, Univ of Graz, Graz, Austria. Background : Prostanoid formation is catalyzed by cycloox ygenase (COX) which occurs in two isoforms: COX-l and COX-2. Since the COX·II COX-2 nonselective inhibitor indomethacin stimulates circular muscle activity in the guinea-pig small intestine, we set out to analyze whether endogenous COX-l and/or COX-2 products modulate intestinal peristalsis. The peristaltic motor action of exogenous prostanoids was also examined . Method s: The experiments were carried out with isolated fluid-perfused segments of the guinea-pig small intestine . Peristalsis was elicited by a rise of the intraluminal pressure and recorded via the pressure changes associ- ated with peristaltic contractions . Drugs were added to the organ bath, i.e., to the serosal surface of the segments. The release of prostaglandin 1 2 from intestinal segments was measured by radioimmunoassay of 6-keto prosta- glandin Fl a' Results: Indomethacin (D. I•3 I-tM) stimulated peristalsis as deduced from a concentration-related decrease in the peristaltic pressure threshold (PPT) at which peristaltic contractions were triggered. This action of indomethacin was mimicked by the COX-l selective inhibitor SC-56D (D.l - 1 I-tM) and the COX-2 selective inhibitor NS-398 (D.I - 1 I-tM). Peristaltic motor stimulation caused by indomethacin, SC-56D and NS-398 (each at 1 /-tM) was associated with suppression of the release of 6-keto prostaglandin F la from the intestinal segments. Prostanoids modi- fied peristalsis in a manner opposite to that of COX inhibitors. Prostaglan- dins E, and as well as the thromboxane A 2 agonist U-46,619 (0.01 • I I-tM) were particularly effective in depressing peristalsis as reflected by a concentration dependent increase in PPT and a reduction of the amplitude and efficiency of the peristaltic waves. Conclusions : These results show that products of both COX-I and COX-2 control peristaltic motor activity in the guinea-pig isolated small intestine, their effect being a tonic depres- sion of propulsive motility. It would seem that prostaglandins of the E series as well as thrornboxanes are responsible for this action. Our findings imply that pathological alterations in the expression of COX will lead to profound disturbances of peristaltic motor activity. This study was sup- ported by FWF grant P1l834-MED and P13512-MED. 3227 ADENOSINE Al RECEPTOR AGONIST AUGMENTS MECHANI· CALLY·EVOKED RELEASE OF 5·HYDROXYTRYPTAMINE FROM HUMAN BON CELLS. Minsoo Kim, Najma H. Javed, Helen Raybould, Helen J. Cooke, The Ohio State Univ, Columbus, OH; Ball State Univ, Muncie, IN; Ucla/Cure , Los Angeles, CA. 5-Hydroxytryptamine (5-HT) released from enterochromaffin cells by a mechanical stimulus is essential in activating secretory and peristaltic reflexes necessary for lubrication and propulsion of intestinal luminal contents. In order to identify intracellular signaling mechanisms for regu- lation of 5-HT release, BON cells with pro,perties similar to enterochro- maffin cells were seeded at a density of IxlO cells per well and grown for 48 hrs. Rotational shaking with a platform shaker, which generate s a combination of shear stress and hydrostatic pressure changes, was used to release 5-HT into the buffer solution. Aliquots of buffer were frozen for subsequent analysis of 5-HT by enzyme immunoassay. Increasing the speed of shaking from 50 to 100 rpm caused a rpm-depend ent increase in 5-HT release without any evidence of cell injury and detachment deter- mined by trypan blue exclusion, protein concentration and cell number. Shaking at 80 rpm maximally increased 5-HT release to 212% (p< 0.05) of basal levels of 0.8:=0.08 pmollwel1l20 min. The adenosine AI receptor agonist, (100 nM CCPA), enhanced 5-HT release caused by shaking to 295% of basal levels (p< 0.05). In the absence of shaking, the AI receptor agonist had no effect on basal 5-HT release. While the Al receptor agonist significantly augmented 5-HT re- lease in response to shaking, it inhibited by 43% 5-HT release caused by isoproterenol 's activation of {3-adrenergic receptors , which is transduced by Gs and generation of cyclic AMP. The synergistic interaction between shaking and CCPA on 5-HT release was abolished by the AI receptor antagonist. 8-cyclo-pent yl-l .3-dimethylxanthine, and not by the A 2 recep- tor antagonist, 3.7-dimethyl-propargylxanthine. 5-HT release stimulated by shaking was nearly abolished by GDP-f3-S in BON cells permeabilized with streptolysin 0 (20 U/ml) for 5 min compared to 5-HT release with streptolysin 0 alone. These findings suggest that mechanical stimulation of the enterochromaffin-like BON cells releases 5-HT by activating a G protein-coupled signaling pathway that differs from the isoproterenol- stimulated, Gs protein-coupled pathway. These results imply that when adenosine is elevated during hypoxia, intestinal anaphylaxis or inflamma- tion, synergistic cross-talk occurs between the signaling pathway activated by shaking and adenosine signaling through Gi/Go in 5-HT-producing enteroendocrine cells. Supported by NIH Grants: DK37240 (HJC) and DK41004 (HR).