Regulation of electrolyte transport with IL-1β in rabbit distal colon

Interletrkin-1β levels are elevated in inflammatory bowel disease. In this study the mechanism by which interleukin-1β affects electrolyte transport in the rabbit distal colon, was investigated. Interleukin-1β caused a delayed increase in short-circuit current (Isc) which was attributed to protein synthesis since the effect was inhibited by cycloheximide. The interleukin-1β induced increase in Isc was not affected by amiloride treatment but was completely inhibited by bumetanide or in chloride-free buffer and by indomethacin. Prostaglandin E2 levels increased in tissue treated with interleukin-1β, but this increase was reversed by cycloheximide. These data suggest that interleukin-1β causes its effect via a yet to be identified second messenger, by increasing chloride secretion through a prostaglandin E2 mediated mechanism.


Introduction
IL-1 is a cytokine of around 17 kDa which is responsible for mediating a variety of processes in host defence, inflammation, and response to injury. It is produced by diverse cell types following infection or injury. On the basis of several overlapping activities and similar patterns of production, IL-1 may be grouped with tumour necrosis factor-(TNF0t, and IL-6 as factors mediating common effects. 2,3 It is produced by a variety of cells such as fibroblasts and T and B lymphocytes and consists of two distinct but related molecules, IL-10t and IL-I[ which are encoded by separate genes; in most human tissues IL-1[3 mRNA predominates over IL-10t mRNA. Both forms appear to function by binding to the same membrane-associated receptor. IL-1 appears to be a primary molecule in inflammatory reactions through its induction of other inflammatory metabolites, e.g. it induces production of prostaglandin E (POE2) and phospholipase a (PEA2).
In the intestine, increased IL-1 production has been reported in the mucosa of patients with inflammatory bowel disease and in intestinal tissue of animal models of colitis TM suggesting that IL-1 may be involved in the pathophysiology of inflammatory bowel disease. IL-1 has been shown to stimulate el: and HCO3secretion in the chicken intestine 12 and inhibit Na and C1absorption in rabbit ileum, 1 however, the role and mechanism of action of IL-1 in affecting secretory function in the normal colon has not been fully elucidated.
More than one second messenger seems to be involved in mediating the effects of IL-1. Some of these mediators include nitric oxide (NO), cAMP and phospholipases C and A2 . 14 '15 In this study, the actions ( 1995 Rapid Communications of Oxford ktd of IL-113 on electrolyte transport in normal rabbit distal colon and the second messenger(s) that might be involved in its mechanism of action are characterized. It was also determined whether TNF0t has similar effects to IL-1 on ion transport in the rabbit colon since, although TNFz and IL-1 are biochemically distinct cytokines, they have many related and overlapping biological functions, and they play important and similar roles in many immune responses. 2 Transport studies: Male New Zealand albino rabbits weighing 2-3 kg were maintained on a standard rabbit chow diet with free access to water. The animals were killed by an overdose of pentobarbital sodium (i.v.). The distal colon was removed and epithelial sheets devoid of serosa and muscularis propria were prepared for transport studies. Colonic mucosa was mounted as a fiat sheet between two Lucite modified Ussing chambers having an aperture of 1.13 cm2, and oxygenated and maintained at 37C. Short-circuit current (Isc) which is equivalent to the electrical sum of all ion transport processes occurring simultaneously was determined. An automatic voltage clamp (W.P.I., Sarasota, FL), corrected for fluid resistance between the potential difference sensing bridges, provided continuous short-circuiting of the tissue.
Unless specified, the bathing solution consisted of Krebs-HCO (KBS) composed of (in mM): KC1 4.8, CaCl2 2.5, NaC1 118.1, NaH2PO 1.2, MgSO 1.2, NaHCO3 25, glucose 11, pH 7.4 after gassing with 95% 02/5% CO2. Human recombinant IL-I[ was used in this study. It has previously been shown that rabbit and human IL-1 cross-react with an antibody to human IL-116 and that rabbit and human IL-1 have approximately 65% homology. 17 Amiloride and bumetanide were dissolved in 100% ethanol and indomethacin dissolved in 10 mM sodium carbonate. All reagents were added to the serosal bathing solution. Ion substitution studies were performed by either replacing Clions by gluconate ions or replacing HCO= ions by HEPES solution.
Measurement ofPGEe levels in tissue treated with IL-lfJ: Stripped distal colon mucosal sheets were washed with KBS, placed on ice and cut into small but equal pieces. For all assays, pieces of tissue were transferred to polypropylene vials containing 5 ml of oxygenated KBS at 37C in a slow shaking water bath. The tissue was continuously oxygenated during the experiment. Aliquots were transferred from the supernatant to Eppendorf tubes at specific times and centrifuged at 4C, for 2 min at 7 000 g in the Eppendorf centrifuge. Samples were kept at-70C for later determination of PGE levels using radioimmunoassay (RIA). Commercially available RIA kits for measurement of PGE2 were used (Advanced Magnetics Inc., MA). The sample was precipitated with cold acetone, centrifuged and the precipitate discarded. Petroleum ether was then added to the supernatant and the aqueous phase separated and acidified to pH 3-4 and then extracted further with ethyl acetate. The sample was lyophilized and reconstituted in a known amount of assay buffer. In brief, the assay involved incubating the sample with the corresponding antiserum overnight at 4C and then centrifuging for 15 min at 4C at 1 000 x g. The supernatant was then transferred to scintillation vials, scintillant was added and the vial counted in a liquid scintillation counter. Statistical analyses: Statistical analyses were performed with Student's t-test for paired and unpaired data; half-maximal and maximal effects of IL-I[ on transport were determined by the method of Woolf-Hanes 18 by plotting the concentration/change in Isc vs. concentration. In these calculations, the halfmaximal concentration and the maximal effect for each experiment were determined, and the results are presented as the means of these + S.E. Unless specified all results are reported as mean + S.E. 62 Mediators of Inflammation. Vol 4. 1995

Results
Effects ofIL-l on Isc in rabbit distal colon: IL-1 (0.1 to 10 ng/ml) was added to the serosal bathing solution and Isc monitored for 60 min. IL-I[ caused a significant, concentration dependent increase in Isc ( btM), a neuronal Na channel inhibitor which prevents neural input to the epithelial cells. IL-13 (1 ng/ ml) was added to the serosal side of the colon in the Ussing chamber after pretreating the tissue with either of these inhibitors and the peak change in Isc was measured. Indomethacin decreased the Isc by 7.3 + 2.1 A/cm 2. It completely abolished the effects of IL-1[3 on Isc (% Alsc caused by IL-lJ3 in the presence of indomethacin was-10.2 + 13.2 ILtA/cm2, Fig. 2) suggesting that IL-113 mediates most of its effects through the production of prostaglandins.
Atropine alone had no effect on Isc and had no effect on IL-l-induced increases in Isc. TTX, which alone caused a decrease in Isc of 8.2 + 0.7 a/cm2, inhibited the, effect of IL-1[3 by 56 + 17% (Fig. 2) indicating that while there is a large portion of the effect being mediated through enteric nerves, there is a possibility of a direct effect exerted on epithelial cells.
The nature of the ion involved in IL-1 effect: To determine the nature of the ion(s) whose transport is affected by IL-1 treatment, treatment of the tissue with amiloride, bumetanide or ion substitution studies were performed. Chloride ion was substituted for g|uconate to determine if it is involved in the action of the cytokine. Amiloride, at 1 btM, an inhibitor of the apical Na channel, had no effect on the IL-l[3 induced increase in Isc (14.3 + 1.2 La/cm in control tissue vs. 14.3 + 0.9 A/cm in tissue treated with amiloride, n 3, p > 0.1, NS), suggesting that Na ion movement is not affected, while C1replacement completely reversed the increase in Isc (Fig. 3).
Bumetanide (10 laM), an inhibitor of the Na+/K+/2Cl transporter present on the basolateral membranes of the crypt cells, had no effect on Isc alone, also caused complete inhibition of the effect of IL-lJ3 on Isc (16 + i lLtA/cm in control tissues vs. 0 + 2 tA/cm in bumetanide treated tissue, n 5, p < 0.005; and -9 + 3 btA/cm in chloride-free buffer, n 5, p < 0.001). From these data we conclude that IL-1[3 causes a significant increase in anion secretion, mainly CI-, which causes the increase in Isc observed in the Ussing chamber.
The effect ofIL-l on the production ofPGE2" Treatment of stripped tissue with IL-1]3 (1 ng/ml) caused a significant increase in the production of PGE (Fig.  4) compared with a time control (5 + I ng/mg protein in control tissue vs. 12 + 2 ng/mg protein in treated tissue, n 6, p < 0.05). The time course for PGE production followed a similar course to the effect observed on Isc in Fig. lb. The levels of LTB were undetectable in control and in IL-l-treated tissues.
Effects oflL-l in the presence of cycloheximide and actinomycin D: To test if protein synthesis is the reason for the delayed effect observed for the action of IL-I, the tissue was treated with cycloheximide (100 l.tg/ml), a protein synthesis inhibitor, or with actinomycin D (10 g/ml), a transcription inhibitor. The effect of TNFc on Isc in rabbit distal colon: We found no evidence in the rabbit colon for a similar action of TNF to that observed by IL-1. TNF at different concentrations (0.1-100 ng/ml) added to the serosal bathing solution, had no effect on Ic (TNF at 100 ng/ml, caused a change in Ic -2.3 + 1.2 [tA/cm2, n 3, not significantly different to zero) in rabbit distal colon (data not shown).

Discussion
Cytokines such as IL-1, possess a broad spectrum of biological activity. They are now being recognized as essential mediators of both normal and pathological immune responses such as temperature regulation, bone and cartilage remodelling, and regulation of extracellular matrix products. However, increased levels of these cytokines can.lead to predominance of proinflammatory effects that characterize disease states such as sepsis and inflammatory bowel disease, and to the generation of inflammatory mediators such as leukotriene B or PGEs. 19 In this study it is reported that IL-1[ stimulates ion secretion through a mechanism which involves the production of PGE 2. study is consistent with the in vivo experiments reported by Cominelli et al. 2 In the latter report, it was shown that in the isolated rabbit distal colon a 10 h infusion of IL-1 progressively increases production of PGE2, 6-keto PGFla and thromobxane B 2. In the present study it was found that the effect of ILl on Isc and on PGE production were not only concentration-and time-dependent but required protein synthesis. IL-113 caused an increase in Isc mainly by increasing Clsecretion significantly with no effect on Na ion movement through the apical Na channel. In pancreatic islet cells, IL-1 was shown to activate the Na//H exchanger and cause a rapid increase in cytosolic Na concentration. 21 Since there is no evidence for the presence of the Na//H exchanger in rabbit distal colon, the only possible explanation for the increased Isc is an increase in C1secretion.
Most of the effects of IL-113 on Isc and PGE2 production were found to be through protein synthesis. Induction of protein synthesis in response to IL-1 treatment has been reported in a variety of cells. [22][23][24][25] In these studies, alterations in phospholipid metabolism were observed. These were primarily due to increased PLA2 activity which occurred in response to IL-1 and which was attributed to induced synthesis of a phospholipase A activating protein. 26 In human umbilical vein endothelial cells, IL-1 has recently been shown to induce cyclooxygenase-2 expression through a post-transcriptional regulation. 27 In insulinsecreting cells IL-1 modulates the expression of specific proteins, not yet identified, but might represent heat shock protein, superoxide dismutase, or other proteins involved in the functional response of the islet to oxidative stress and free radical formation. 28 Although the cell surface receptor for IL-1 has been cloned, 29 it is still largely unknown how it transmits information to the inside of the cell after IL-1 binding. The intracellular signals induced by IL-1 appear to be complex because IL-1 seems to use multiple signal transduction pathways. 14,15 In the beta cell, for example, nitric oxide (NO) has been implicated as the effector molecule responsible for some of the effects of IL-1 on cell function. NG-monomethyl-L-arginine (NM.MA) and NAME, both of which are competitive inhibitors of nitric oxide synthase, prevent the inhibitory effects of IL-I[ on glucose-stimulated insulin secretion by isolated islets. 3 In our study, however, using a NOS inhibitor, we found no evidence for NO being a mediator for IL-1 action.
More than one second messenger seems to be involved in mediating the effects of IL-1. In different cell types, such as islets, pituitary, mesangial cells, chondrocytes and fibroblasts, different mediators have been described in association with cytokine action. Some of these mediators included Ca2/, inositol phosphates, cAMP, protein kinase A, PLC and PLA as well as alteration in gene transcrip-tion. 24,[31][32][33][34][35][36] In the intestine, the second messenger(s) involved in mediating the effect(s) of IL-1 is yet to be determined.
The human intestinal mucosa has been shown to be a rich source of IL-1 under normal conditions and that production of both forms, IL-lot and IL-l[3, is dramatically increased during pathological conditions such as inflammation. IL-1 has been found to originate from mononuclear cells in the lamina propria but not from epithelial cells in normal or inflamed mucosa. Whether the receptor for IL-1 is present on epithelial cells, or not, is not yet known, although some evidence has emerged that IL-113 and IL-6 receptors have been identified on human epithelial cells suggesting that these cells can be regulated directly by proinflammatory cytokines. 37 In conclusion, this study demonstrates that IL-1 has profound effects on the secretory function of the normal distal colon. Since IL-1 is significantly increased in inflammatory bowel disease, an understandin'g of its mechanism of action on electrolyte transport in the colon will lead to greater insights into the pathophysiology of inflammatory bowel disease and possibly new therapeutic modalities.