Interferon-gamma down-regulates adenosine 2b receptor-mediated signaling and short circuit current in the intestinal epithelia by inhibiting the expression of adenylate cyclase.

Adenosine is an endogenous signaling molecule that is highly up-regulated in inflammatory states. Adenosine acts through the A2b receptor, a G protein-coupled receptor that couples positively to Galpha(s) and activates adenylate cyclase. This leads to cAMP-mediated electrogenic chloride secretion in intestinal epithelia. To better understand the regulation of the A2b receptor in intestinal epithelia, we studied the effects of interferon-gamma (IFN-gamma), a potent immunomodulatory cytokine, in the T84 cell line. Pretreatment of cells with 500 units/ml IFN-gamma for 12 h inhibited an adenosine-induced short circuit current (Isc) without affecting the transepithelial resistance. Under these conditions, IFN-gamma did not inhibit the protein expression or membrane recruitment of the A2b receptor, shown to be essential for its function. Interestingly, IFN-gamma inhibited cAMP levels as well as its downstream signaling pathway as shown by the inhibition of adenosine-induced phosphorylation of cAMP response element-binding protein and protein kinase A activity. Similar studies with forskolin, a direct activator of adenylate cyclase, also demonstrated inhibition of cAMP and its downstream response by IFN-gamma. However, IFN-gamma did not affect secretory responses to the calcium-dependent secretagogue carbachol or cAMP analog 8-bromo-cAMP, indicating that normal secretory responses to adequate second messengers in IFN-gamma-treated cells are achievable. Moreover, IFN-gamma inhibited the expression of adenylate cyclase isoforms 5 and 7. In conclusion, we demonstrate that IFN-gamma down-regulates adenosine-mediated signaling possibly through the direct inhibition of adenylate cyclase expression. We propose that IFN-gamma may acutely affect global cAMP-mediated responses in the intestinal epithelia, thereby decreasing secretory responses, which may consequently aggravate inflammatory processes.

Adenosine is an endogenous signaling molecule that is highly up-regulated in inflammatory states. Adenosine acts through the A2b receptor, a G protein-coupled receptor that couples positively to G␣ s and activates adenylate cyclase. This leads to cAMP-mediated electrogenic chloride secretion in intestinal epithelia. To better understand the regulation of the A2b receptor in intestinal epithelia, we studied the effects of interferon-␥ (IFN-␥), a potent immunomodulatory cytokine, in the T84 cell line. Pretreatment of cells with 500 units/ml IFN-␥ for 12 h inhibited an adenosine-induced short circuit current (I sc ) without affecting the transepithelial resistance. Under these conditions, IFN-␥ did not inhibit the protein expression or membrane recruitment of the A2b receptor, shown to be essential for its function. Interestingly, IFN-␥ inhibited cAMP levels as well as its downstream signaling pathway as shown by the inhibition of adenosine-induced phosphorylation of cAMP response element-binding protein and protein kinase A activity. Similar studies with forskolin, a direct activator of adenylate cyclase, also demonstrated inhibition of cAMP and its downstream response by IFN-␥. However, IFN-␥ did not affect secretory responses to the calciumdependent secretagogue carbachol or cAMP analog 8-bromo-cAMP, indicating that normal secretory responses to adequate second messengers in IFN-␥-treated cells are achievable. Moreover, IFN-␥ inhibited the expression of adenylate cyclase isoforms 5 and 7. In conclusion, we demonstrate that IFN-␥ down-regulates adenosine-mediated signaling possibly through the direct inhibition of adenylate cyclase expression. We propose that IFN-␥ may acutely affect global cAMP-mediated responses in the intestinal epithelia, thereby decreasing secretory responses, which may consequently aggravate inflammatory processes.
Adenosine is an important modulator of physiological as well as inflammatory responses in humans. Adenosine is generated during active inflammation, and levels increase both in the intestinal lumen and in tissue during inflammation to as high as 500 -600 nM (1). The intestinal adenosine 2b receptor (A2bR), 1 one of the four adenosine receptor subtypes (A1, A2a, A2b, and A3), mediates the biological effects of adenosine (2,3). Interestingly, the A2bR is the predominant adenosine receptor expressed in the caecum and colon in both the model colonic cell line T84 and in intact human colonic mucosa (3,4). Indeed, in the model colonic epithelia T84 cells, the A2bR is the only adenosine receptor expressed (3,5). Depending on the organ, the A2bR activates pro-or anti-inflammatory pathways. For example, in the joints and cardiovascular system, the A2bR is anti-inflammatory, whereas in the lung the A2bR is a potent proinflammatory mediator, and the A2bR antagonists are evolving as potential drugs for reactive airway disease (2). In the intestine, apical or basolateral stimulation of the A2bR results in cAMP-dependent electrogenic chloride secretion through the activation of chloride channels (3,6). Active chloride secretion by intestinal crypt enterocytes is known to be the central pathophysiological disturbance in acute and chronic diarrheal illnesses (7). In addition to chloride secretion, adenosine also induces apically directed interleukin-6 secretion and fibronectin secretion (8,9). Thus adenosine, acting through the A2bR, modulates intestinal secretion and inflammatory response in a cAMP-dependent manner.
In the intestinal epithelial cells, the A2bR couples positively to G␣ s and activates adenylate cyclase. Apical or basolateral stimulation of the A2bR induces an increase in intracellular cAMP and downstream cAMP signaling including phosphorylation and activation of the transcription factor, CREB and the activation of PKA (3,10). The former is involved in interleukin-6 secretion in the intestine in response to adenosine (11), and PKA is involved in the chloride secretory pathway activated by adenosine (11,12). Unlike mast cells, where the A2bR also couples to G q and increases intracellular calcium (13), cAMP is the only signaling pathway mediated by the A2bR in the intestinal epithelial cells (3). Studies in our laboratory have demonstrated that the A2bR is recruited to the membrane upon agonist stimulation and exists as a multiprotein complex at the membrane with the PDZ domain-containing protein, NHERF-2, cytoskeletal anchoring protein, ezrin, and PKA (11). Although the expression and biological effect of adenosine in the intestine have been characterized, the regulation of the A2b is not known.
The major goal of this study is to characterize the regulation of the A2bR by interferon-␥ (IFN-␥). IFN-␥ is an immunoregulatory cytokine produced by T helper 1 cells. It is highly upregulated during chronic inflammatory diseases such as inflammatory bowel disease as well as during acute viral or bacterial enteritis and is thought to play a central role in the pathogenesis of inflammation and diarrhea associated with these diseases (14 -16). During inflammation, IFN-␥ directly affects the enterocytes, including the Cl Ϫ -secreting crypt cells, and regulates enterocyte functions including barrier regulation and ion secretion (17)(18)(19). The effect of chronic exposure to IFN-␥ (Ͼ24 h) on enterocyte function has been studied extensively. Prior studies using vasoactive intestinal peptide (VIP), cholera toxin (cAMP-mediated I sc ), and carbachol (calciummediated I sc ) demonstrated significant decrease in transepithelial resistance (TER) and chloride secretion after prolonged treatment with IFN-␥ (Ͼ24 -48 h) without altering the morphology of cells. This inhibition in secretory response has been attributed to decreased synthesis/expression of cystic fibrosis conductance regulator (CFTR) (20), Na ϩ K ϩ -ATPase, and/or Na ϩ K ϩ -2Cl Ϫ cotransporter required for anion secretion (21). Although the chronic effects of IFN-␥ have been well characterized, the acute effect of this cytokine on secretory response in intestinal epithelia is not known. In this study we investigated the effect of acute IFN-␥ exposure on the expression, signaling, and secretory function of the A2bR.
Cell Culture-T84 cells were grown and maintained in culture as described previously (22) in a 1:1 mixture of Dulbecco's modified Eagle's medium and F-12 medium supplemented with 40 mg/liter penicillin, 90 mg/liter streptomycin, and 5% newborn calf serum. Confluent stock monolayers were subcultured by trypsinization. Experiments were done on cells plated for 7-8 days on permeable supports of 0.33-cm 2 , 4.5-cm 2 inserts or 1-cm 2 snap well filters (Costar, Cambridge, MA).
Inserts (0.4-m pore size, Costar) rested in wells containing medium until steady-state resistance was achieved, as described previously. This permits apical and basolateral membranes to be interfaced separately with apical and basolateral buffer, a configuration identical to that developed previously for various microassays. The T84 cells had a high electrical resistance (900 -1,200 ⍀/cm 2 n ϭ 50 monolayers). All experiments were performed on T84 cells between passages 69 and 76.
Ussing Chamber Electrophysiology Studies-T84 cells were plated on snap well filters (12-mm diameter, 0.4-m pore size, Costar), surface area of 1 cm 2 , and were grown to confluence (ϳ8 days) until steady-state resistance was achieved. The filter rings were detached and mounted in an Ussing chamber and were incubated with Hanks' balanced salt solution at 37°C and bubbled continuously with 95% O 2 and 5% CO 2 . The fluid volume on each side of the filter was 5 ml. Voltage-sensing electrodes consisting of Ag/AgCl pellets and current-passing electrodes of silver wire were connected by agar bridges containing 3 M KCl and interfaced via head-stage amplifiers to a microcomputer-controlled voltage/current clamp VCC-MC6, respectively (Physiologic Instruments, San Diego). Voltage-sensing electrodes were matched to within 1-mV asymmetry and corrected by offset-removal circuit. The current between the two compartments (values reported are referenced to the apical side) were monitored and recorded at 20-s intervals, whereas the voltage was clamped to 0. The voltage was measured with blank filters first in buffer to be used for the experiments. The values obtained, generally less than 1 mV in magnitude, represent the difference in junction potentials between the two voltage-sensing bridges summed with any potential that might exist across the filter membrane. These values were subtracted from all subsequent measurements with filters containing attached cell monolayers. The total resistance between apical and basal compartments was determined throughout the experiment from the current evoked by a 5-A bipolar voltage pulse. Before Ussing chamber experiments the cell monolayers were pretreated basolaterally with or without IFN-␥ with various doses (1-1,000 units/ml) and various time intervals starting from 1 to 48 h. After a sustained base-line I sc cells were stimulated with 10 Ϫ2 mM apical or basolateral adenosine and 10 M FSK. The increase in I sc was then determined.
Preparation of Plasma Membrane-The plasma membrane fraction was prepared from T84 cells plated in 4.5-cm 2 inserts as described previously (23). Monolayers were washed in phosphate-buffered saline scraped with a rubber policeman and homogenized with a glass/Teflon homogenizer in ice-cold buffer containing 250 mM sucrose, 10 mM Tris, pH 7.5. The cell suspension was centrifuged at 700 ϫ g for 10 min at 4°C. The supernatant was centrifuged at 17,000 ϫ g for 45 min at 4°C (23). The pellet enriched in plasma membrane was recovered in lysis buffer containing protease inhibitors. Protein quantitation was done using the Lowry method (Bio-Rad).
SDS-PAGE and Western Blot-Cells were lysed with phosphatebuffered saline containing 1% Triton X-100 and 1% Nonidet P-40 (v/v), protease inhibitor mixture (Roche Applied Science), EDTA, SDS, sodium orthovanadate, and sodium fluoride. SDS-PAGE was performed according to the Laemmli procedure using acrylamide gel. Proteins were electrotransferred to nitrocellulose membranes and probed with primary antibody. Membranes were then incubated with corresponding peroxidase-linked secondary antibody diluted 1:2,000, washed, and subsequently incubated with ECL reagents (Amersham Biosciences) before exposure to high performance chemiluminescence films (Amersham Biosciences). For molecular mass determination, polyacrylamide gels were calibrated using standard proteins (Bio-Rad) with markers within the range 10 -250 kDa.
Quantitation of Western Blot-The band intensity of Western blot was quantitated using a gel documentation system (Alpha Innotech Co., San Leandro, CA).
cAMP Measurement-T84 cells were treated with or without basolateral IFN-␥ (1-1,000 units/ml), and cells were stimulated with 10 Ϫ2 mM apical and basolateral adenosine and 10 M FSK. cAMP measurements were done in whole cell lysates using a competitive cAMP immunoassay kit (Applied Biosystems, Bedlford, MA). Luminescence was read with luminoscan Ascent Thermo Labsystems (Needham Heights, MA).
PKA Assay-For the determination of PKA activity in IFN-␥-treated cells, T84 cells were treated with or without basolateral IFN-␥ (500 units/ml) for 12h and were stimulated with apical, basolateral adenosine or FSK. PKA activity was determined using a protein kinase assay kit from Calbiochem. An equal amount of cell lysate was added to the PKA reaction mixture containing ATP solution, biotinylated Leu-Arg-Arg-Ala-Ser-Leu-Gly (Kemptide), cAMP solution, PKA reaction buffer, and 0.2 Ci/l [␥-32 P]ATP. Both enzyme and substrate controls were included. Reactions were terminated, and Kemptide was precipitated TABLE I AC isoform-specific primers with avidin. Samples were washed repeatedly, and the radioactivity was counted using a liquid scintillation counter (LKB Wallac 1219 Rackbeta). Aliquots from each sample were taken for protein determination using the Bio-Rad protein assay. PKA specific activity was measured as pmol of phosphate incorporated/min/g of protein, and the relative increase in PKA activity was calculated in relation to the untreated control.
RT-PCR-Total RNA was extracted from monolayers of T84 cells by the TRIzol extraction method (TRIzol Reagent, Molecular Research Center, Cincinnati, OH). The RNA was then used to amplify fragments of the cDNA of AC-1-9 by RT-PCR employing the Qiagen One-step RT-PCR kit. The primers were designed on the basis of the AC-1-9 nucleotides sequences available in the GenBank data base (Table I). A positive control was performed by using primers specific for glyceraldehyde-3-phosphate dehydrogenase (sense, gccaaggtcatccatgacaac; antisense, gtccaccaccctgttgctgta; product size 494 bp). One-step RT-PCR was performed with the following program. A reverse transcription reaction was initiated at 50°C for 30 min. PCR activation at 94°C for 15 min was followed by 40 cycles, each consisted of 94°C for 30 s, 60°C for 30 s, 72°C for 1 min, and final extension time was set at 72°C for 10 min.
Statistical Analysis-The data are presented as the mean Ϯ S.E. Statistical analysis was performed using Graphpad Instat 3 software (www.graphpad.com). Groups were compared using parametric tests (paired Student's t test or one-way analysis of variance with post-test following statistical standards). p values Ͻ0.05 were considered statistically significant. A/cm 2 and TER 900 Ϯ 250 ⍀/cm 2 , and cells treated with IFN-␥ I sc 2.9 Ϯ 0.5 A/cm 2 and TER 1,000 Ϯ 200 ⍀/cm 2 ). These results demonstrate that the barrier function as determined by TER measurements was not affected by IFN-␥ treatment. To investigate the effect of IFN-␥ on the regulation of adenosineinduced I sc , monolayers were stimulated with 100 M apical (AA) or basolateral (BA) adenosine or basolateral 100 M (Bs) carbachol after a sustained baseline current. As shown in Fig.  1B, IFN-␥ treatment significantly decreased both apical and basolateral adenosine induced I sc (IFN-␥ ϩ AA ϭ 4 Ϯ 1, IFN-␥ ϩ BA ϭ 3 Ϯ 1 A/cm 2 ) compared with adenosine alone (AA ϭ 13 Ϯ 4, BA ϭ 19 Ϯ 4 A/cm 2 ). On the other hand, IFN-␥ had no effect on carbachol-mediated I sc (carbachol ϭ 59 Ϯ 7, IFN-␥ ϩ carbachol ϭ 57 Ϯ 5 A/cm 2 ). We next investigated the dose response of IFN-␥-induced inhibition of adenosine-mediated I sc response. T84 cells plated on snap well filters were pretreated with IFN-␥ (1, 5, 10, 50, 100, and 500 units) for 12 h. Snap wells were then mounted on an Ussing chamber, stimulated with 100 M adenosine for 5 min, then I sc and TER were measured. The inhibition of adenosine-induced I sc (19 Ϯ 4 A/cm 2 ) began at 10 units (40% Ϯ 5) and was maximal at 500 units (90% Ϯ 5) with an EC 50  Earlier studies in our laboratory have demonstrated that the A2bR is recruited to the membrane upon agonist stimulation (11,24). To explore the effect of IFN-␥ on the recruitment of the receptor to the membrane, we performed Western blot analysis of the A2bR on plasma membrane fractions. The cells were stimulated with apical or basolateral adenosine after pretreatment with or without IFN-␥ for 12 h. Plasma membrane was isolated as described under "Materials and Methods." As shown in Fig. 2B, adenosine stimulation resulted in recruitment of the receptor to the plasma membrane within 5 min (center lane), and IFN-␥ pretreatment did not alter the adenosine-induced recruitment of the receptor to the membrane (right lane). The bar chart shows densitometric quantification of A2bR recruitment to the plasma membrane. ␤-Actin as loading control and Na ϩ K ϩ -ATPase as membrane marker are shown.

IFN-␥ Inhibits Adenosine-induced I sc in T84 cells-To
IFN-␥ Inhibits A2b Receptor Signaling-the A2bR couples positively to G␣ s and activates adenylate cyclase. We have shown previously that apical or basolateral stimulation of the A2bR induces an increase in intracellular cAMP (3). We next studied the effect of IFN-␥ on adenosine-induced cAMP. T84 cells were pretreated with or without 500 units/ml IFN-␥ for 12 h, and cAMP levels stimulated by 100 M apical adenosine and basolateral adenosine were quantitated using a luminometric assay as described under "Materials and Methods." As expected, both apical and basolateral adenosine increased cAMP levels, which were maximum at 5 min after stimulation (AA ϭ 0.24 Ϯ 0.07, BA ϭ 5.1 Ϯ 0.1) (pmol/10 6 cells). As shown in Fig. 3, pretreatment of cells with IFN-␥ resulted in the inhibition of both apical and basolateral adenosine-stimulated cAMP levels by ϳ60 and 85%, respectively (IFN-␥ ϩAA ϭ 0.1 Ϯ 0.02, IFN-␥ ϩ BA ϭ 0.32 Ϯ 0.01 pmol/10 6 cells, respectively). IFN-␥ alone did not affect basal cAMP compared with unstimulated cells (IFN-␥ 0.07 Ϯ 0.04 and unstimulated ϭ 0.1 Ϯ 0.04 pmol/10 6 cells, respectively). To study the effect of IFN-␥ on PKA activity, T84 monolayers were pretreated with 500 units/ml IFN-␥ for 12 h before the addition of 100 M apical or basolateral adenosine for 5 and 30 min. The PKA activity was measured as described under "Materials and Methods." Both the apical and basolateral adenosine induced PKA activity, which was maximal at 5 min and returned to base line at 30 min. As shown in Fig. 4B, IFN These data suggest that IFN-␥ may inhibit A2bR signaling by affecting the activity of adenylate cyclase or by inducing the degradation of cAMP via the activation of phosphodiesterases.

IFN-␥ Inhibits cAMP-mediated Downstream Signaling-Phosphorylation of CREB and PKA Activation-Activation
To evaluate the involvement of phosphodiesterase, we studied the effect of various phosphodiesterase inhibitors 8-methoxymethyl-isobutylmethylxanthine, trequinsin, and rolipram on reversing the IFN-␥-induced inhibition of adenosine-mediated I sc . Interestingly, phosphodiesterase activity was unchanged, and phosphodiesterase inhibitors did not reverse IFN-␥-mediated inhibition of I sc (data not shown). These data led us to hypothesize that IFN-␥ directly inhibited adenylate cyclase activity and/or its expression.

IFN-␥ Inhibits FSK (a Direct Activator of Adenylate Cyclase)induced Short Circuit Current in a Dose-and Time-dependent
Manner-We investigated the effect of IFN-␥ on FSK-induced cAMP and I sc . T84 cells were pretreated with different doses of IFN-␥ (1, 5, 10, 50, 100, 500 units/ml, respectively) before mounting the snap wells onto the Ussing chamber. The cells were then stimulated with 10 M FSK, and I sc and TER were measured after sustained base-line I sc . A dose-dependent inhibition of FSK-induced I sc was observed compared with untreated cells (Fig. 5A). The inhibition of FSK-induced I sc (45 Ϯ 5 A/cm 2 ) was seen beginning at 10 units/ml (ϳ40% inhibition) and was maximal at 500 units/ml (80% inhibition). We investigated further whether IFN-␥ inhibited FSK-induced I sc in a time-dependent manner. As seen in the Fig. 5B, the inhibition of FSK-induced I sc by IFN-␥ started at 4 h (62%) and was maximal at 12 h (80%). To verify further that adenylate cyclase activity and not the cAMP-dependent transporters involved in adenosineinduced I sc was affected by IFN-␥, we studied the effect of 8-Br-cAMP, a cAMP analog, on the I sc response. T84 cells were pretreated with IFN-␥ (500 units/ml) for 12 h and then stimulated with 2 mM 8-Br-cAMP. As seen in Fig. 7, 8-Br-cAMP-induced I sc was not inhibited by IFN-␥ pretreatment, suggesting that the I sc response to adequate cAMP was not inhibited by IFN-␥.

IFN-␥ Inhibits FSK-induced cAMP, CREB, and PKA Activity-We
IFN-␥ Inhibits Expression of Adenylate Cyclase-Because IFN-␥ inhibited the activity of adenylate cyclase, we studied the IFN-␥ effect on expression of adenylate cyclase. AC exists in 9 isoforms (AC-1 through 9) of which 1, 3, and 8 are exclusively neuronal. We first characterized the expression of AC in T84 cells using RT-PCR and Western blot of plasma membrane with recently commercialized AC isoform-specific antibodies. Total RNA was isolated from T84 cells and reverse transcription and PCR amplification using isoform specific primers were done as described under "Materials and Methods." AC-5 with a prod- uct size of 116 bp, AC-6, 246 bp, AC-7, 227 bp, and AC-9, 240 bp, were detected by RT-PCR (Fig. 8A). Western blot showed expression of AC-5, 7, and 9 ( Fig. 8B) but AC-6 and other isoforms could not be detected. We next determined the effect of IFN-␥ on the expression of AC isoforms. T84 cells were pretreated with 500 units/ml IFN-␥ for 12 h, and total RNA was subjected to RT-PCR. IFN-␥ significantly inhibited the expression of AC-5 RNA while not affecting AC-6, AC-7, or AC-9 (Fig. 8A). Western blot showed that IFN-␥ decreased the expression of AC-5 and AC-7 but did not affect AC-9. The foregoing data collectively suggest that IFN-␥ directly inhibits the expression and activity of AC-5 and AC-7, thus inhibiting its downstream signaling pathway. DISCUSSION In this study, we addressed the regulation of the A2bR by IFN-␥, the most critical inflammatory cytokine that is highly up-regulated in acute and chronic colitis in human and is known to play an important role in chloride secretion and barrier function in the intestine (18,19). We demonstrate that IFN-␥ down-regulates A2bR signaling and function without affecting its expression or membrane recruitment. IFN-␥ sequentially inhibited downstream signaling of cAMP such as phosphorylation of the transcription factor CREB as well as PKA activity, which has been shown to be involved in chloride secretion. It is known that cAMP activates PKA by dissociating its regulatory subunit from the catalytic subunit (25). The free catalytic subunit thereupon initiates a series of enzymatic reactions leading to a phosphorylation cascade, activating multiple proteins including CFTR (26,27). Our data suggest that IFN-␥ significantly inhibited PKA activity essential for adenosine-induced chloride secretion primarily through CFTR. Using FSK, a direct activator of adenylate cyclase, we demonstrated that IFN-␥ inhibited the activity of adenylate cyclase in a time-and dose-dependent manner similar to adenosine. Further, phosphodiesterase activity was unaffected, and phosphodiesterase inhibition did not reverse the effects of IFN-␥ on I sc , suggesting that the inhibition of I sc induced by IFN-␥ pretreatment is associated with decreased synthesis of cAMP rather than increased degradation of cAMP. We demonstrate that, indeed, IFN-␥ directly inhibited the expression and activity of adenylate cyclase.
In T84 cells, we found by RT-PCR that AC-5, 6, 7, and 9 are the most abundantly expressed isoforms. Western blot using recently commercialized AC-specific antibodies showed expression of AC-5, 7, and 9. IFN-␥ pretreatment down-regulated the expression of AC-5 at both the RNA and protein level and AC-7 at the protein level. Interestingly, AC-9, the only isoform that does not respond to FSK, was not inhibited by IFN-␥. Our results are in line with Freeman and MacNaughton (47), who demonstrated that inducible nitric-oxide synthase-derived nitric oxide inhibits cAMP-dependent chloride secretion through inhibition of AC-5 and/or AC-6 in the intestinal epithelia. Recent reports have shown that AC-5 and 6 could be differentiated functionally from other isoforms based on the fact that their activities can be inhibited by increases in intracellular calcium or the activation of G␣ i such as by the muscarinic agonist, carbachol (48 -55). Our preliminary data show that carbachol pretreatment inhibits FSK-induced cAMP synthesis as well as I sc . Interestingly, AC-5 but not AC-7 has been localized to membrane microdomains (55). Along with the data that the A2bR and its signaling complex including AC are compartmentalized to apical membrane microdomains (11,56), we speculate that AC-5 inhibition by IFN-␥ may be directly related to the down-regulation of FSK/adenosine-mediated cAMP synthesis and I sc . Further studies need to be done to elucidate the AC isoform associated with the A2bR in intestinal epithelia and the mechanism by which IFN-␥ inhibits AC expression.
Chloride secretion in intestinal epithelial cells results from the activation of ion transporters and channels located at apical and basolateral surfaces. Basolateral transporters, which include Na ϩ K ϩ -ATPase, Na ϩ K ϩ -2Cl Ϫ cotransporter, and K ϩ channels, act coordinately to elevate intracellular Cl Ϫ concentration to levels above its electrochemical equilibrium potential, so that enterocytes are primed to secrete Cl Ϫ through apical anion channels. Agonists that increase intracellular Ca 2ϩ (e.g. carbachol) and cAMP (e.g. adenosine, VIP, cholera toxin, prostaglandin E 2 ) regulate the activities of these transporters and channels and thus regulate ion secretion (57)(58)(59)(60). Apical or basolateral adenosine, like VIP or cholera toxin, has been demonstrated to activate cAMP-dependent electrogenic chloride secretion through the activation of apical chloride channels (3,6). Several studies have addressed the effect of IFN-␥ on secretory response and TER in T84 cells (7,15,18,31). These studies have used IFN-␥ at higher doses (1,000 units/ml) for longer periods of time (Ն24 h). Under such conditions, IFN-␥ significantly decreased TER and chloride secretion in response to VIP and cholera toxin (cAMP-mediated I sc ) and carbachol (calcium-mediated I sc ) without altering the morphology of cells. This decrease in secretary response has been attributed to decreased synthesis/expression of CFTR (20) or Na ϩ K ϩ -ATPase and/or Na ϩ K ϩ -2Cl Ϫ cotransporter required for anion secretion (21). Our data show that acute exposure to IFN-␥ (12 h or less) in lower doses (10 -500 units/ml) significantly decreased cAMP-mediated I sc without affecting calciummediated I sc or TER. Further, our data demonstrate intact I sc response to exogenous cAMP (8-Br-cAMP), suggesting that normal secretory responses to adequate second messengers are possible in IFN-␥-treated cells. To our knowledge, the effect of IFN-␥ (acute or chronic) on AC has not been investigated. Our results on the AC inhibition by acute IFN-␥ exposure may have implications for decrease fluid secretion not only to adenosine but other cAMP-dependent secretogogues such as VIP, cholera toxin, and prostaglandin E 2 as they induce chloride secretion using the same signaling pathway as adenosine. The AC inhibition by IFN-␥ at later time points (measured up to 72 h) suggests that the AC inhibition may contribute to the overall decreased fluid secretion reported during chronic IFN-␥ exposure. Because chloride secretion plays an important role in regulating water transport across epithelia in various organs (61), dysregulation may result in alterations in chloride secretion across epithelia, which can result in significant pathology, such as acute diarrheal illnesses, cystic fibrosis, secretory diarrhea, and inflammatory bowel disease (62). The ability of crypt epithelial cells to respond to secretagogues with chloride and hence water secretion is an important component of epithelial barrier that protects the intestine by preventing translocation of bacteria, bacterial products, and antigens to lamina propria (47,63). Hence, it is possible that the early inhibitory effect of IFN-␥ on AC may further aggravate the proinflammatory process in acute or chronic colitides. Our finding may thus be relevant to understanding pathophysiology of diarrhea and may have implications for anti IFN-␥ therapeutic strategies in the acute clinical setting.