INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

ADENOSINE IS AN IMPORTANT mediator of various physiological responses including muscle tone, neuronal firing, immune function, and secretion of various hormones and cytokines (24, 27, 33). In addition to its role in the regulation of these physiological processes, adenosine is released during inflammatory conditions and acts as a paracrine factor with diverse effects on a variety of organ systems including cardiovascular, nervous, urogenital, respiratory, and digestive systems (31, 33, 37). Adenosine release acts as an autocoid by interacting with the adenosine receptor belonging to the family of seven transmembrane G protein-coupled group of cell surface receptors (27, 31). Biochemical and molecular cloning studies have demonstrated the existence of four adenosine receptor subtypes designated A1, A2a, A2b, and A3 (12, 18, 24, 27, 31). In the intestine, where neutrophil transmigration into the lumen is characteristic of the active phase of many intestinal disorders including inflammatory bowel disease, adenosine has been shown to mediate electrogenic chloride secretion, the event underlying secretory diarrhea (2, 3, 19, 23). With the use of molecular, pharmacological, and biochemical approaches we characterized the intestinal adenosine receptor as the A2b subtype in both T84 cells, a model intestinal epithelial cell line, and intact human intestinal epithelia (36). Furthermore, the A2b receptor appears to be the only adenosine receptor present in T84 cells. In these cells, A2b is functionally coupled to Gs, and the stimulation of apical or basolateral surface with adenosine results in increased cAMP in a polarized manner (36).

Prolonged exposure of G protein-coupled receptors to an agonist results in a decrease in receptor responsiveness, a process termed desensitization. This agonist-induced desensitization or refractoriness is a universal feature of G protein-coupled receptors (5, 13). Several studies have indicated that, for many receptors, including the adenosine receptors (7, 8, 16, 25, 26, 28-30), desensitization can be divided into two temporally and mechanistically distinct phases: 1) short-term exposure to agonist resulting in uncoupling of the receptor from G proteins and mediated by receptor phosphorylation, and 2) long-term agonist exposure resulting in receptor downregulation and mediated by internalization of the receptor and/or decreased receptor synthesis. Although much is known about the phenomenon of desensitization, it is not known if, in polarized cells, desensitization events can be influenced by the domain on which the receptor resides. Because polarized epithelial cells such as T84 cells are able to seal the apical from the basolateral domains, potential cross-domain receptor desensitization can be studied using these cells. The T84 monolayers used for these studies have high electrical resistance (1,200-1,500  · cm²), as is typical for this cell line (10). Such severe restriction on passive permeation of small hydrophilic solutes permits sidedness of responses to be clearly separated (9). We have made use of the apical and basolateral expression of the A2b receptor in T84 cells to examine whether biological events determining receptor desensitization might be dictated by the membrane domain on which the receptor is expressed. Here, we report the characteristics of adenosine receptor desensitization and the effect of polarity on the desensitization. We demonstrate that a single receptor subtype, here modeled by the A2b receptor, may differentially desensitize based on the membrane domain on which it is expressed. Furthermore, agonist exposure on one domain can result in desensitization of receptors on the opposite domain. In the intestine, potential effects of desensitization of apical receptors on basolateral receptors or vice versa might have therapeutic implications.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Reagents. All tissue culture supplies were obtained from Life Technologies (Grand Island, NY). Adenosine and 5'-(N-ethylcarboxamido)adenosine (NECA) were obtained from Research Biochemicals International (Natick, MA). IBMX, vasoactive intestinal peptide (VIP), and nocodazole were obtained from Sigma (St. Louis, MO). Forskolin and carbachol (an acetylcholine receptor agonist) were from Calbiochem (La Jolla, CA).

Cell culture. T84 cells were grown and maintained in culture as previously described (19) in a 1:1 mixture of DMEM and Ham's F-12 medium supplemented with penicillin (40 mg/l), ampicillin (8 mg/l), streptomycin (90 mg/l), 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 cm² in area (inserts). This permits differentiation of cells and development of tight junctions with high electrical resistance (1,200-1,500  · cm²). Inserts with rat tail collagen-coated polycarbonate membrane filter (5-µm pore size; Costar, Cambridge, MA) rested in wells containing media until steady-state resistance was achieved, as previously described (23). This permits apical and basolateral membranes to be separately interfaced with apical and basolateral buffer, a configuration identical to that previously developed for various microassays (9).

Short-circuit current measurements. Inserts were rinsed and placed in Hanks' balanced salt solution (HBSS) in a new 24-well plate containing 0.5 ml of HBSS in the serosal surface and 0.2 ml of HBSS in the apical compartment. To determine currents, transepithelial potentials, and resistance, a set up consisting of a commercial voltage clamp (Bioengineering Department, University of Iowa) was interfaced with an equilibrated pair of calomel electrodes submerged in saturated KCl and a pair of Ag-AgCl electrodes submerged in HBSS (20). For electrical determinations, agar bridges were used to interface one calomel and one Ag-AgCl electrode on each side of the monolayer incubated at 37°C, and a pulse of 25 µA of current was passed across the monolayer. With the use of Ohm's law (V = I × R), the transepithelial resistance and the short-circuit current (I_sc) was calculated.

Statistical analysis. Data are expressed as mean ± SD. Student's t-test or ANOVA with Student-Newman-Keuls post hoc test were used to compare mean values as appropriate. P values < 0.05 were considered to represent significant differences.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Desensitization of the A2b receptor by NECA. An I_sc response representing electrogenic chloride secretion was elicited from T84 cells by adenosine or its analogs applied to either apical or basolateral membrane (2, 3, 36). Either apical or basolateral adenosine (100 µM) rapidly stimulated I_sc response from T84 cells. Because the natural ligand of the A2b receptor, adenosine, is degraded with time by ectoadenosine deaminase, NECA, a nonmetabolizable analog of adenosine, was used for desensitization studies. After exposure to NECA, cells were washed, and desensitization was assessed by functional response as measured by I_sc 10 min after the addition of adenosine [we have shown earlier that apical and basolateral stimulation of T84 cells with adenosine results in maximal I_sc response at 10 min (36)].

View larger version (15K): [in this window] [in a new window]   Fig. 1.   Desensitization of apical or basolateral A2b receptor. T84 cells were treated with 5'-(N-ethylcarboxamido)adenosine [NECA; 10 µM, apical (Ap) or basolateral (Bs)] for a period of 12 h. Cells were washed with Hanks' balanced salt solution (HBSS), equilibrated at 37° for 20 min, and stimulated with adenosine (100 µM Ap or 100 µM Bs). Short-circuit current (Isc) was measured 10 min poststimulation. Data are represented as percentage of maximal response of Isc-induced apical or basolateral adenosine stimulation. In µA/cm2: unstimulated, 9.0 ± 2.8; Ap adenosine (Ado), 70.0 ± 4.0; Bs Ado, 82.0 ± 9.8; Ap NECA + Ap Ado, 28.0 ± 3.0; Bs NECA + Bs Ado, 39.0 ± 6.0. Data represent the responses observed in 4 separate experiments plotted as mean ± SD, n = 6 per treatment group. # Significantly different from cells treated with adenosine P < 0.001 by ANOVA; NS, not significantly different from Ap NECA + Ap Ado; ** significantly different from Ap NECA + Bs Ado, P < 0.001.

Time course of desensitization. We next studied the time course of desensitization of apical and basolateral adenosine receptors. Figure 2A shows that apical NECA inhibited subsequent I_sc responses to apical adenosine. This inhibition began within 20 min and plateaued at 2-3 h after NECA treatment. A similar time course of desensitization was observed with basolateral NECA pretreatment and subsequent stimulation with basolateral adenosine (i.e., maximum I_sc inhibition of 60% ± 3% compared with basolateral adenosine treatment alone was seen 2-3 h after exposure to basolateral NECA). Figure 2B shows that the inhibition of I_sc response to apical adenosine by pretreatment with basolateral NECA began around 3 h, and >90% inhibition of apical adenosine response occurred between 6 and 12 h. Desensitization of the apical receptor by exposure to NECA did not inhibit subsequent I_sc response to basolateral adenosine at any time.

View larger version (24K): [in this window] [in a new window]   Fig. 2.   Time course of adenosine receptor desensitization. T84 cells were exposed to apical or basolateral NECA (10 µM) at the various times indicated. Cells were subsequently washed with HBSS, equilibrated at 37° for 20 min, and stimulated with adenosine (100 µM apical and 100 µM basolateral). Isc before NECA washout varied depending on the time of exposure to NECA. 20 min: Ap NECA, 10 ± 2; Bs NECA, 11 ± 3; 1 h: Ap NECA, 7.1 ± 1; Bs NECA, 5.4 ± 0.9; 2 h: Ap NECA, 5.8 ± 2; Bs NECA, 3 ± 0.5; 3-12 h: values ranged from 4.8 to 5.2 ± 2 for both apical and basolateral NECA. Isc was measured 10 min after stimulation with adenosine. Experiment was repeated at least 4 times and data from 1 representative experiment are plotted as mean ± SD, n = 3 per treatment group. A: cells were desensitized apically or basolaterally and stimulated with apical or basolateral adenosine, respectively (i.e., stimulation added to the same domain as desensitization). B: cells were desensitized apically or basolaterally and stimulated with basolateral or apical adenosine, respectively (i.e., stimulation added to opposite domain as desensitization). Significantly different from cells stimulated with adenosine alone, * P < 0.05, ** P < 0.001 by ANOVA.

Recovery from desensitization. To study the effect of polarity on the pattern of resensitization, T84 monolayers were desensitized with NECA for 12 h, washed, and then stimulated with adenosine at various time intervals. As seen in Fig. 3A, monolayers pretreated with apical NECA were able to elicit maximal I_sc response to apical adenosine stimulation (compared with control monolayers treated with apical adenosine alone) within 6 h of NECA washout. Similar time course of recovery was seen in cells pretreated with basolateral NECA and then stimulated with basolateral adenosine. In contrast, as shown in Fig. 3B, the desensitization of apical adenosine receptors by basolateral NECA showed only a partial recovery. Apical adenosine stimulation resulted in only 60% ± 2% maximal response compared with apical adenosine alone even 6 h after NECA washout. Full recovery to apical adenosine response occurred around 12 h after NECA washout.

View larger version (21K): [in this window] [in a new window]   Fig. 3.   Time course of recovery from adenosine receptor desensitization. T84 cells were exposed to apical or basolateral NECA (10 µM) for 12 h. Cells were subsequently washed with HBSS and stimulated with adenosine (100 µM apical and 100 µM basolateral) at the times indicated. Isc was measured 10 min after stimulation with adenosine. Experiment was repeated at least 3 times, and data were plotted as mean ± SD, n = 3 per treatment group. A: cells desensitized apically or basolaterally were stimulated with apical and basolateral adenosine, respectively. B: cells desensitized apically or basolaterally were stimulated with basolateral and apical adenosine, respectively. * Significantly different from cells stimulated with adenosine alone, P < 0.001 by ANOVA.

Heterologous desensitization. Because heterologous desensitization has been reported for other G protein-coupled receptors such as chemoattractant receptors of neutrophils (1) and adenosine receptors in pulmonary epithelial cells (14), we next sought to determine whether the desensitization of adenosine receptors affected signaling of other receptors associated with chloride secretion. For this purpose, we used VIP and carbachol, which elicit I_sc via cAMP- and calcium-dependent signaling pathways, respectively (6, 11). Monolayers were desensitized with apical or basolateral NECA for 12 h, washed, and then stimulated with VIP (2 nM, basolateral) or carbachol (10 µM, basolateral). Figure 4 shows that VIP-induced I_sc was not affected by apical or basolateral NECA pretreatment. Similarly, I_sc induced by carbachol was not affected by apical or basolateral NECA pretreatment. In addition, I_sc induced by the direct activation of Gs protein by cholera toxin (20 nM) (17) was not affected by apical or basolateral NECA treatment under conditions where the I_sc induced by adenosine is inhibited (Fig. 4). The addition of forskolin, a direct activator of adenylate cyclase, elicited the usual substantial increase in I_sc in the desensitized cells, suggesting that no form of desensitization of A2b receptors interferes with the adenylate cyclase-dependent signaling pathway (in µA/cm²): Fsk alone, 62.2 ± 3.9; Ap NECA + Fsk, 57.8 ± 5.0; Bs NECA + Fsk, 78.7 ± 3.1; unstimulated, 4.1 ± 1.0. These results suggest that desensitization observed here is specific to the adenosine receptor, and heterologous desensitization as reported for other adenosine receptors does not occur.

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Effect of adenosine receptor desensitization on other receptors. T84 cells were exposed to apical and basolateral NECA (10 µM) for 12 h. Cells were washed and stimulated with vasoactive intestinal peptide (VIP; 2 nM basolateral), carbachol (10 µM basolateral), or cholera toxin (C.T.; 20 nM). Isc was measured 10 min after stimulation with VIP, 3 min after carbachol, and 1 h after cholera toxin. Unstimulated cells had an Isc of 1.9 ± 0.2 µA/cm2. Isc induced by VIP, carbachol, and cholera toxin were significantly different from unstimulated cells, P < 0.001. Experiment was repeated at least 3 times, and data from 1 representative experiment were plotted as mean ± SD, n = 6 per treatment group.

Effect of low temperature or nocodazole on adenosine-induced Isc. One possible mechanism of desensitization of apical receptors as a late consequence of basolateral receptor desensitization (and not vice versa) could relate to a possible trafficking step in which the A2b receptor is first targeted basolaterally upon synthesis. The receptor could then be internalized and transported to the apical membrane domain via vesicular transport/microtubules. Such trafficking has been reported for the A1 receptor synthesis and trafficking in renal epithelia (34). To test the hypothesis that a similar event occurs for the intestinal A2b receptor, cells were incubated at 15-20°C for 12 h to inhibit vesicular movement (17), and responses of cells to apical or basolateral adenosine were subsequently tested. Under these conditions, a selective inhibition of the apical response to adenosine was observed while the basolateral response was preserved. As shown in Fig. 5, incubating cells at 15-20°C inhibited apical response to adenosine by 54%, whereas the basolateral response remained 96% compared with control response (in µA/cm²): apical adenosine, 76.4 ± 7.2; low temperature + apical adenosine, 41.2 ± 2.9; basolateral adenosine, 72.1 ± 4.0; low temperature/basolateral adenosine, 67.8 ± 4.2. Similarly, pretreatment of cells with nocodazole (33 µM equivalent to 10 µg/ml, 12 h), which disrupts microtubules by preventing repolymerization of tubulin heterodimers (15), also inhibited subsequent response to apical adenosine stimulation by 48% (in µA/cm²): nocodazole + apical adenosine, 38.3 ± 4.4; apical adenosine, 75.9 ± 6.8, whereas the response to basolateral adenosine was not affected [93% of control response (in µA/cm²): basolateral adenosine, 71.9 ± 3.9; nocodazole + basolateral adenosine, 66.6 ± 6.3]. Figure 6 shows the dose-response curve to apical (Fig. 6A) and basolateral adenosine (Fig. 6B) after nocodozole treatment of cells. As seen in the figure, the dose response to apical adenosine and not basolateral adenosine is blunted in cells exposed to nocodozole. These observations are consistent with the notion that apical A2b receptors traffic to the apical domain from the basolateral domain by a process that is dependent upon microtubule-mediated vesicular traffic.

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Effect of inhibition of vesicular movement/microtubule on Isc response to adenosine. T84 cells were incubated at low temperature (15-20°C), nocodazole (33 µM), or vehicle for 12 h. Cells were washed with HBSS, equilibrated at 37° for 10 min, and stimulated with adenosine (100 µM apical and 100 µM basolateral), and Isc was measured 10 min after stimulation with adenosine. Experiment was repeated 3 times, and data from 1 representative experiment were plotted as mean ± SD, n = 6 per treatment group.

View larger version (19K): [in this window] [in a new window]   Fig. 6.   Effect of nocodazole on adenosine dose response. T84 cells were treated with nocodazole (33 µM) for 12 h. Cells were washed and stimulated with various doses of apical adenosine (A) or basolateral adenosine (B). Isc was measured 10 min after stimulation with adenosine. Experiment was repeated at least 2 times. Data were plotted as mean ± SD, n = 4 per treatment group. Significantly different from cells treated with vehicle + Ap adenosine treatment, * P < 0.05; ** P < 0.01 by ANOVA followed by Student-Newman-Keuls post hoc test.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Our study provides evidence that, in polarized epithelial cells, agonist-induced desensitization of the natively expressed A2b receptor on one domain is not only able to desensitize the receptor on the opposite membrane domain but also induces a differential desensitization pattern, i.e., the basolateral receptor desensitization affects the apical receptor-induced I_sc but not vice versa. In addition, this study underlines the selectivity of A2b receptor-mediated desensitization because neither the cAMP-mediated Cl secretion induced by agonists such as VIP, cholera toxin, and adenylate cyclase activator, forskolin nor the calcium-mediated Cl secretion induced by carbachol are affected by A2b receptor desensitization.

Attenuation of agonist-induced signaling is classically observed after chronic stimulation of G protein-coupled receptors. Two patterns of desensitization have been described: 1) a rapid event that occurs after brief exposure to agonists (seconds to minutes) resulting in the loss of agonist-induced functional response with no change in receptor number or affinity for the ligand, and 2) a slower event following long-term exposure to agonist (hours to days) that results in the downregulation of the receptor caused by sequestration, internalization, or decreased synthesis of the receptor (5, 13, 31, 32). Our results show that the A2b receptor is subject to both of these patterns of desensitization albeit in a polarized manner. The desensitization of the A2b receptor in the same membrane domain is rapid and is totally reversible within 6 h of agonist washout. Moreover, such desensitization reaches a plateau of 40-60% maximal response within 3 h of exposure to NECA, and no further desensitization occurs despite the presence of agonist. The rapidity and reversible nature of desensitization of the A2b receptor in the same membrane domain is suggestive of uncoupling of the A2b receptor from G protein after receptor phosphorylation either by specific receptor kinase (GRKs) or nonspecific phosphorylation of the receptor (protein kinase A or C). Indeed, rapid termination of signaling by G protein receptors is typically initiated by such phosphorylation events catalyzed by either second messenger-activated kinases or G protein receptor coupled kinases (GRKs), which in turn promotes high-affinity binding of arrestins (13). Interestingly, recent studies on short-term desensitization of A2 receptor in nonpolarized NG-108-15 (21) and HEK-293 (22) cells have shown that GRK-2 and arrestin-2 are both involved in the rapid desensitization process (21, 22).

In contrast to the desensitization of the A2b receptor in the same membrane domain, the down modulation of I_sc response to apical A2b receptor stimulation after desensitization of basolateral A2b receptor is slow, complete (100% inhibition of apical I_sc response to adenosine), and only partially reversible at 6 h after NECA washout. Complete recovery did not occur until 12 h after NECA washout. On the other hand, NECA pretreatment of the apical A2b receptor does not affect subsequent I_sc response to basolateral adenosine stimulation at any time. Interestingly, Barrett et al. (2, 3) observed that, after 1 wk of culture with apical and basolateral NECA, the apical response to adenosine was completely abolished, whereas a significant component of the basolateral response to adenosine persisted. However, the polarity of desensitization of the apical receptor by the basolateral receptor was not observed, possibly because these investigators did not assess the compartmentalization on the desensitization process. The polarized effect of the basolateral A2b receptor on the apical receptor may reflect distinct receptor proteins, differences in receptor density, differences in linkages to postreceptor signaling mechanisms, or compartmentalization of intracellular target proteins. Our observation cannot be explained by the presence of different types of adenosine receptor in the apical and basolateral membrane domain because, using molecular and pharmacological approaches, we have previously shown that A2b is the only known adenosine receptor subtype and is present in both the apical and basolateral membrane domains of T84 cells (36). Moreover, we and others have shown that both apical and basolateral A2b receptors have similar K_m for activation by various agonists (3, 36) and, therefore, it is unlikely that the ligand-receptor interaction on both surfaces is different (36). The cAMP response of the apical and basolateral receptors in response to agonist differs markedly (36). Whereas the basolateral A2b receptor results in a severalfold increase in cAMP, the apical A2b receptor results in a small but significant increase in cAMP for comparable I_sc response to the same dose of agonist. However, our data show that cAMP levels alone do not influence desensitization because dibutyryl cAMP does not induce desensitization of apical or basolateral adenosine receptors. By Northern blot analysis, the levels of A2b receptor mRNA are not altered in monolayers treated with NECA under conditions that produced maximal desensitization for I_sc (preliminary data). These data do not exclude the possibility that receptor density may be affected by posttranscriptional or posttranslational mechanisms.

A possible explanation of the polarized effect of the basolateral A2b receptor on the apical I_sc response is that the A2b receptor is first targeted to the basolateral domain upon synthesis and then moves to the apical membrane via vesicular transport. If so, desensitization of the basolateral A2b receptor resulting in downregulation of the basolateral receptor would affect the subsequent expression of the receptor on the apical domain. Consistent with this hypothesis is our observation that the inhibition of vesicular movement or microtubules, which affect intracellular trafficking, selectively inhibit I_sc induced by apical adenosine while not affecting I_sc induced by basolateral adenosine exposure. These data suggest that the A2b receptor may first be targeted basolaterally and thereafter transported to the apical domain, as has been documented for a subset of other apical membrane proteins including the adenosine A1 receptor (4, 34). Although the polarized effect of low temperature and nocodazole on adenosine-induced I_sc is consistent with our hypothesis, this result could also be explained by the effect of these treatments on chloride secretion that is independent of adenosine receptor trafficking. We are in the process of developing antibodies to the A2b receptor to explore these possibilities.

In conclusion, we have demonstrated that a single receptor subtype, here modeled by the A2b receptor, may differentially desensitize based on the membrane domain on which it is expressed, and agonist exposure on one domain can result in desensitization of receptors on the opposite domain. This occurs in a polarized fashion, i.e., desensitization of the basolateral A2b receptor causes desensitization of the apical A2b receptor and not vice versa.