166

Expression of Receptors for Corticotropin Releasing Factor in Guinea Pig Enteric Nervous System Sumei Liu, Xiang Gao, Na Gao, Xiy 9 Wang, Hong-Zhen Hu, Scott W. Arlin, Guo-Du Wang, Xiyu Fang, Yun Xia, Jackie D. Wood

Acoustic or cold-restraint stress increases colonic transit and passage of fecal pellets in rodents, intracerebroventricular injection of corticotropin releasing factor (CRF) mimics the effects of stress in the colon while reducing gastric acid secretion, stimulating gastric bicarbon- ate secretion and preventing the gastric mucosal lesions that result from cold-restraim stress. The aim of the present study was to test a hypothesis that CRF acts in the enteric nervous system as well as the brain. We combined intraceflular electrophysiologic recording and intraneuronal injection of the tracer biocytin, with indirect immunofluorescence and RT- PCR to test the hypothesis. Application of CRF or urocortin evoked a slowly-activating depolarizing response associated with elevated excitability in 60% (24/40) of ileal or colonic myenteric neurons. Histologic~analysis of biocytin-filled neurons revealed that both Dogiel morphologic Type I neurons with S-type electrophysiologic behavior and Dogiel morphologic Type I1 neurons with AH-type electrophysiologic behavior were stimulated by CRF or urocortin. Depolarizing responses to CRF or urocortin were concentration-dependent with an EC~0 of 19.3 _+ 4.7nM for ,CRF and 20.3 + 3.4nM for urocortin. The CRF,/CRF2 receptor antagonist, astressin, suppressed the CRF-evoked responses. Similar EC~ for CRF and urocortin suggest that the CRFI receptor subtype mediates the excitatory action of CRF. Enteric neurotransmission (i.e., fast EPSPs, slow EPSPs, IPSPs) was unaffected by CRF or urocortin. Immunoreactivity for the CRF~ receptor was distributed widely surrounding ganglion cell bodies in the myenteric plexus of the stomach and small and large intestine and in the submucosal plexus of the small and large intestine. Double-labeling of the CRF 1 receptor with anti-Hu antibody, which marks all ganglion cells, suggested that 41% of myenteric and 36% of submucosal neurons expressed CRFI receptors. CRF 1 receptor immu- noreactivity was co-local/zed with ganglion ceils that expressed calbindin, choline acetyhrans- krase, substance P or VIP. RT-PCR revealed expression of CRF~ receptor mRNA in both myenteric and submucosal plexuses. The results support the hypothesis that neuronal actions of CRF in both the brain and enteric nervous system may underlie the effects of CRF on gut motility and secretion. (Supported by NIH RO1 DK57075 and RO1 DK37238)

167

Activation Of Protease-Activated Receptor-1 (PAR-l) Inhibits Neurally Evoked Chloride Secretion in the Mouse Colon Michelle C. Buresi, Nathalie Vergnolle, Patricia Andrade-Gordon, Morley D. Hollenberg, John L. Wallace, Wallace K. Macnaughion

PARs are G-protein coupled receptors that are activated by serine proteases. We studied the effect of PAR-1 activation in the regulation of chloride secretion in mouse colon. Methods: 1. Expression of PAR-1 in mouse colon was assessed by RT-PCR and immunohistochemistry. 2 To study the role of PAR-1 activation in chloride secretion, mouse colon was mounted in Ussing chambers. Changes in short circuit current (is(:) were measured in tissues exposed to either thrombin, saline, the PAR-1-activating peptide TFLLR-NH2, or the inactive reverse peptide, prior to electrical field stimulation (EFS), carbachol (CCh), and forskolin (FSK) treatment. 3. To determine the role of PAR-1 activation in secretory function in vivo, colonic loops were prepared in fasted, anesthetized mice. PAR-1 agonist or vehicle (0.1 ml) was introduced intraluminally, recovered after 20 minutes, and changes in volume determined gravimetrically. Colonic loop experiments were repeated in PAR-l-deficient and wildtype nuce 7 days after induction of colitis with TNBS. Results: 1. PAR-1 mRNA was expressed in full thickness and mucosal specimens of mouse colon. PAR- 1 immunoreactivity was found on epithelial cells and on submucosal ganglia. 2. In in vitro Ussmg chamber studies of isolated colon, PAR-1 activation by thrombin or the activating peptide sigmficantly reduced secretory responses to EFS. However, secretory responses to CCh or FSK were unaffected in these studies. 3. In in vivo experiments in wildtype mice, instillation of thrombin or TFLLR-NH2 to the colonic lumen resulted in significantly reduced recovery of fluid compared 10 respective controls. Furthermore, in TNBS colitis studies conducted in vivo, colitic PAR- 1 deficient mice had significantly greater secretory function than oolitic wildtype mice. Conclusion: PAR-1 is expressed on submucosal ganglia and activation leads to decreased secretory responses to neural stimulation. PAR-1 activation appears to contribute to the deficient secretory function observed in models of colitis and thus has important implications for inflammatory conditions of the gut where thromhin levels are elevated.

168

FAK Phosphorylat ion at Y397 mediates GRP's Morphogenic Properties of Enhancing Cell Attachment and Promoting Tissue Remodeling Sarah Glover, Melissa Delaney, Kristin Keller, Roger Tran-Son-tay, Richard V. Benya

Background: Differentiation in cancer describes the degree to which tumor cells histologically resemble the non-malignant tissues whence they originated. GRP and its receptor (GRPR) are mntpbogens that contribute to normal intestinal development; and, when aberrantly re- expressed in colon cancer promote the assumption of a better-differentiated appearance (reviewed in Peptides 2001; 22: 689). Since better-differentiated cells metastasize less fre- quently, this suggested that GRP/GRPR should promote cell adherence to the extracelhilar matrix (ECM) and motility in the context of tissue remodeling. Although the mechanism regulating these actions is unknown, our study of colon cancer in GRPR / mice suggested a role for focal adhesion kinase (FAK). Hence the aim of this study was to elucidate the role of GRP-induced FAK activation on regulating cell attachment and motility Methods: 293 cells are a non-malignant epithehal cell line that natively express GRP and GRPR. To dLssect out the role of FAK, 293 cells were modified to inducibly express the dominant negative enzyme FRNK under control of a Tet-On (or dox-sensinve) promoter. To measure cell attachment we designed and manufactured a rheoscope, a cone-plate viscometer coupled to a microscope that directly measures the shear stress required to detach cells from the ECM. To evaluate tissue remodeling confluent cells were disrupted and their behavior

assessed by time-lapse photography. Results: FAK was pbosphorylated at Y397 in control 293 cells, and this was completely inhibited by culturing them in the presence of GRP antagonist or dox. Exposure to a shear of 0.25 dynes/cm 2, similar to what exists in a small venule, had minLmal impact on the attachment of control 293 cells. In contrast, both antagonist or dox induced >50% of cells to detach when exposed to the same shear force. Removing segments of confluent ceils caused those remaining to rapidly enlarge and migrate into the gap without causing any increase in cell proliferation (see Video #1 at www.uic.edu/ com/dom/gastro/labvideos). Ceils migrated only until their lameUipodia detected others, at which point motion was arrested or reversed. In contrast, addition of GRP antagonist or dox equally and completely inhibited cell motility without affecting lamellipodia formation (see Video #2). Conclusions: GRP-induced pbosphorylation of FAK at Y397 increases cell- ECM attachment and promotes cell motility without affecting cell proliferation.

169

Relative Contributions of RGS4 and RGS Domain of GRK2 to Inhibition of PLC- Activity by PKA: Direct Evidence from Site Directed Mutagenesis of RGS4 and

GRK2 Huiping Zhou, Karuam S. Murthy

We have previously shown both PKA and PKG phospborylate RGS4 (regulator of G protein signaling) and enhance its ability to inactivate GTP-bound Gaq and thus, inhibit PLC-t3 activity and IP3 generation, in addition, PKA, but not PKG, phosphorylates GRK2 (contain RGS-like domain), which contributes to inhibition of PLC-13. Aim: To determine the relative contributions of GRK2 and RGS4 in PKA-mediated inhibition of PLC-I~ activity. Methods: Cultured rabbit gastric muscle cells were separately transfected with an RGS4 mutant that does not bind Ga q (L159F); an RGS4 mutant that is devoid of GTPase-activating properties, i.e., GAP activity (N888), and a GRK2 mutant deficient in PKA phosphorylation site ($685A). The cells were treated with acetylcholine (ACh) in the presence or absence of isoproterenol (1 ~,M) or Na nitroprusside (1 IxM). At these concentrations, the two agents selectively activate PKA and PKG, respectively. PLC-13 activity was measured as total inositol phosphates in cells labeled with [3H]inositol, and GRK2 phospborylation was determined in [32]P-labeled cells. Association of Getq with GRK2 was determined by co-immunoprecipitation using GRK2 antibody. Results: ACh (0.1 ~M) in the presence of methoctramine (m2 receptor antagonist) activated Gaq and stimulated PLC-~ activity (256 + 43% above basal level). Isoproterenol (PKA) and Na nitroprusside (PKG) inhibited ACh-stimulated PLC-I~ activity in cells express- ing wild type RGS4 or GRK2. Inhibition via PKG was completely blocked m cells expressing RGS4 mutants that lack GAP activity or ability to bind Gaq, but not in cells expressing phosphorylation-dcficient GRK2 mutant. In contrast, inhibition of PLC-13 activity via PKA was partly blocked in cells expressing either RGS4 mutants (36 +- 4% inhibition) or phospllor- ylation-deficient GRK2 mutant (58+7% inhibition). Inhibition via PKA was completely blocked in ceils co-expressing phosphorylation-deficient GRK2 with either RGS4 mutant. PKA but not PKG stimulated GRK2 phosphorylation and increased GRK2 association with Gaq: both events were suppressed in cells expressing pbosphorylation-deficient GRK2. Conclusion: Pbospborylation of RGS4 and GRK2 by PKA stimulates the GAP activities of both proteins, thereby accelerating inactivation of Gcq and inhibition of PLC-13 activity. GRK2 phosphorylation contributes .-40% and RGS phosphorylation -60% to inhibition of PLC-13 activity by PKA

170

Modeling GI Electrical Activity Martin Buist, Leo Cheng, William Richards, Alan Bradshaw, Andrew Pullan

INTRODUCTION. Our aim is to construct a model of the human GI system and to use this model to investigate the electrical control activity of the stomach and small bowel. METHODS. Photographic images from the visible human data set were used to construct initial descriptions of the GI system components. Over 60000 data points were taken from these images and used to create models of the skin surface, esophagus, stomach, small and large bowel. The modeling framework allows cellular electrical models to be embedded into tissue and organ models, thus allowing the electrical control activity (ECA) to be simulated on a realistic geometry. Equations governing current flow in active tissues are solved for the transmembrane and extracellular potential fields. The resulting ECA creates electric and magnetic fields throughout the torso allowing electrogastrograms and magnetogastrograma (MGGs) to be calculated. RESULTS. An anatomically realistic model of the basic components of the GI system has been developed. For all organs considered, the average spatial error between the model and the data was less than l m m This model was used to simulate the normal electrical and magnetic fields generated from GI electrical activity. The spatial distributions of the magnetic field components were in qualitative agreement with those measured using a Superconducting Quantum Interference Device. The effects of mechanical gastric uncoupling were also investigated. Simulations were perfomed on a tissue slab with a model of the ICC network embedded between the longitudinal and circular muscle layers. The electrical activity was calculated at 12800 points with an average spacing of 0.125mm. Mechanical uncoupling resulted in bradygastria (f= 2.2cpm) distal to the intervention while the proximal section was unchanged (f=3.0cpm). CONCLUSIONS. We have developed the first anatomically based model of the human GI electrical activity that allows us to investigate a variety of fundamental mechanisms associated with the ECA. Initial simulations have produced qualitatively normal MGGs as well as reproducing bradygastria that is seen experimentally distal to the site of mechanical uncoupling.

A-23 AGA Abstracts