Receptors coupled to pertussis toxin-sensitive G-proteins traffic to opposite surfaces in Madin-Darby canine kidney cells. A1 adenosine receptors achieve apical and alpha 2A adrenergic receptors achieve basolateral localization.

The α2A adrenergic receptor (α2AAR) previously was shown to be directly delivered to and retained on the lateral subdomain of renal epithelial cells. The present studies demonstrate that, in contrast, wild-type and epitope-tagged canine A1 adenosine receptors (A1AdoR) are apically enriched (65-83%) in Madin-Darby canine kidney (MDCKII) and porcine renal epithelial (LLC-PKI) cells, based on surface biotinylation strategies detecting photoaffinity-labeled A1AdoR. Confocal microscopy corroborated the apical enrichment of the epitope-tagged A1AdoR. Metabolic labeling studies revealed that this steady-state polarization is achieved by direct delivery to both the apical (60-75%) and basolateral surface. Growth of A1AdoR-expressing cells as monolayers was achieved only following Transwell culture in the presence of A1AdoR antagonists, which decreased cell growth, suggesting that A1AdoR elicit MDCKII cell proliferation. The preferential apical but detectable basolateral localization of A1AdoR provides a molecular understanding of published reports that functional responses can be elicited following apical as well as basolateral delivery of adenosine agonists in varying renal preparations. These findings also suggest that receptor chimeras derived from the Gi/Go-protein-coupled α2AAR and A1AdoR will be informative in revealing structural features critical for basolateral versus apical targeting.

It is important to understand the molecular properties that govern polarized expression of epithelial cell proteins, as this polarity is an intrinsic part of the vectorial functioning of these cells. The coordinated cellular functions mediated by endogenous and exogenous ligands depend on the availability of appropriate receptors at the particular surface domains to which the ligand has access. Our laboratory is interested in elucidating the mechanisms and structural regions within G-proteincoupled receptors responsible for polarized expression of these regulatory molecules in renal epithelial cells. We have demonstrated previously that the ␣ 2A AR is predominantly localized (Ͼ85%) to the basolateral surface of Madin-Darby canine kid-ney (MDCKII) 1 cells, a polarized model system for renal epithelia that accurately reflects ␣ 2A AR localization in vivo (1). Immunolocalization studies revealed that the ␣ 2A AR is enriched in the lateral subdomain of the basolateral surface, and metabolic labeling studies indicated that the ␣ 2A AR is directly targeted to the basolateral domain (1).
The present studies characterize the localization and delivery of another seven transmembrane-spanning G-protein-coupled receptor, the A 1 adenosine receptor (A 1 AdoR), in polarized renal epithelial cells. Interestingly, this receptor is enriched on the apical surface in both MDCKII and porcine renal LLC-PKI cells. In the kidney, adenosine, present in both blood and urine (2), acts as a paracrine regulator of renal blood pressure, renin secretion, and renal excretion. The A 1 AdoR, like the ␣ 2A AR, regulates cellular processes by interacting with G-proteins in the G i /G o family. Although some have explored the localization of the adenosine receptors by means of functional studies (3)(4)(5)(6)(7), there is no evidence of the localization of these receptors using direct biochemical and morphological strategies. In an attempt to study the localization of the A 1 AdoR in renal epithelial cells, we have created permanent clonal cell lines of MDCKII and LLC-PKI cells expressing either a wild-type or epitope-tagged A 1 AdoR. The apical versus basolateral distribution of the A 1 AdoR in polarized renal epithelial cells was determined biochemically and morphologically, and the general mechanism by which this predominantly apical localization is achieved was explored.
Male in the laboratory of Carolyn Slayman (Yale University, New Haven, CT). The parental cells and the permanent cell lines were maintained in ␣ minimum essential medium supplemented with 10% heat-inactivated fetal bovine serum (Sigma), 100 units/ml penicillin G, and 100 g/ml streptomycin (␣ minimum essential complete medium) in a 5% CO 2 , 37°C incubator.
For polarity experiments, MDCKII and LLC-PKI cells were seeded at a density of 1 ϫ 10 6 cells/24.5-mm polycarbonate membrane filter (Transwell chambers, 0.4-m pore size, Costar, Cambridge, MA), and cultured for 5-8 days with medium changes every day. Where stated, cells were grown in the presence of 60 M theophylline and 100 M DPSPX. Prior to each experiment, the integrity of the monolayer was assessed by adding [ 3 H]methoxyinulin to the apical medium, and monitoring the leak of [ 3 H]methoxyinulin from the apical compartment to the basolateral compartment by sampling and counting the basolateral medium in a ␤-counter (Packard Tricarb) after a 1-h incubation at 37°C. Chambers with leaks greater than 3%/h were discarded.
Construction of the A 1 and TAG-A 1 Adenosine Receptor-RDC7 (canine thyroid adenosine A 1 receptor cDNA) cloned into the pBluescript SKϩ phagemid at the EcoRI site was obtained from the laboratory of J. E. Dumont and G. Vassart, Universite Libre de Bruxelles, Brussels. A 2.028-kilobase pair KpnI-XbaI fragment containing the A 1 AdoR fulllength coding region, plus 53 bases of 5Ј-noncoding and 997 bases of 3Ј-noncoding sequence, was excised and cloned into the KpnI-XbaI sites of the mammalian expression vector pCMV4. To create the epitopetagged A 1 AdoR, oligonucleotide-directed mutagenesis was utilized to add 13 amino acids (YPYDVPDYALVPR) to the extracellular amino terminus of the canine A 1 AdoR immediately after the initiator methionine (9). The first 9 amino acids constitute a hemagglutinin (HA) epitope (10) recognized by the commercially available monoclonal antibody 12CA5 (Berkley Antibody Co.), while the last 4 amino acids were added as a spacer arm. Positive mutations were verified using dideoxy-DNA sequencing (Sequenase kit, U. S. Biochemical Corp.) utilizing T7 DNA polymerase with [␣-35 S]dATP. Once verified, the insert encoding the epitope tag was subcloned into the pCMV4-A 1 AdoR plasmid (11), and reverified with sequencing. All the resulting pCMV4-A 1 AdoR and pCMV4-TAG-A 1 AdoR constructs were transiently transfected into COSM6 cells, and membranes from these transfectants were assayed for binding of the A 1 AdoR antagonist, [ 3 H]DPCPX (see below). Binding of [ 3 H]DPCPX to the epitope-tagged A 1 AdoR was indistinguishable from binding to the wild-type A 1 AdoR.
Development of Permanent Transformants of LLC-PKI and MDCKII Cells-The pCMV4-A 1 AdoR (20 g) or pCMV4-TAG-A 1 AdoR adenosine receptor (20 g) was co-transfected with 2 g of pRSVneo into 1 ϫ 10 6 MDCKII cells using a modification of a standard Ca(PO 4 ) 2 precipitation method (12), except that 2 days following transfection, 500 g of active G418/ml was added either directly to the original plate of transfected cells or to plates containing different dilutions of the transfected cells. G418-resistant colonies were screened for A 1 AdoR expression by assaying the binding of the A 1 AdoR antagonist [ 3 H]DPCPX (13). Specific A 1 AdoR binding was defined as that binding not competed for by 1 mM theophylline. Untransfected MDCKII cells displayed no detectable specific binding of [ 3 H]DPCPX.
A 1 Adenosine Receptor Binding Assay-MDCKII and LLC-PKI cell membrane particulate preparations were prepared essentially as described (11) with the following modifications. The lysis buffer was 15 mM Tris-HCl, 5 mM EGTA, 5 mM EDTA, pH 8.0 containing 0.5 g/ml leupeptin, 0.7 g/ml pepstatin, 100 g/ml soybean trypsin inhibitor, and 1 mM phenylmethylsulfonyl fluoride. The final pellet was resuspended in 900 l of membrane binding buffer (50 mM Tris-HCl, 10 mM MgCl 2 ⅐6H 2 O, pH 7.4). The assay of [ 3 H]DPCPX binding was performed in 12 ϫ 75-mm polypropylene tubes containing 3 nM [ 3 H]DPCPX (diluted in 1% ethanol in 50 mM Tris-HCl, 10 mM MgCl 2 , pH 7.4) in the absence (total binding) or presence (nonspecific binding) of 1 mM theophylline. Incubations were for 45 min at 25°C. Reactions were terminated by the addition of 3.5-ml ice-cold membrane binding buffer and filtration through Whatman GF/B glass microfiber filters, which were pretreated with 0.3% polyethylenimine in membrane binding buffer at 4°C for 30 min. Upon termination of the binding assay, the filters were dried at 60°C for 20 min and were counted in 5 ml of NEN-963 liquid scintillation fluid.
Steady-state Localization of the A 1 -and TAG-A 1 Adenosine Receptors: Biotin Surface Labeling Strategy-The method for quantitating the apical versus basolateral distribution of the wild-type and epitopetagged A 1 AdoR in polarized MDCKII cells was biotinylation of the apical versus basolateral surface of cells grown in Transwell culture, photoaffinity labeling of the functional A 1 AdoR in harvested membranes, detergent extraction, and isolation of biotinylated receptors via streptavidin-agarose chromatography. The studies were performed essentially as described previously (1). Briefly, MDCKII or LLC-PKI cells grown in Transwell culture were washed three times for 5 min in PBS containing 0.5 mM CaCl 2 and 1 mM MgCl 2 (PBS/CM) and 1 mM theophylline. Following sequential treatment with sodium periodate and biotin-LC-hydrazide, cells were harvested into PBS containing 1 mM EDTA and 0.1 mM phenylmethylsulfonyl fluoride and membranes were prepared as described above. A small sample of the membrane preparation was removed and assayed for [ 3 H]DPCPX binding. To photoaffinity label the A 1 AdoR in these membranes, aliquots containing 25,000 -50,000 cpm of specific [ 3 H]DPCPX binding were incubated with 2 nM A 1 AdoR antagonist, 125 I-azido-BW-A844U (13), for 45 min in membrane binding buffer at room temperature in the dark. Nonspecific photoaffinity labeling was assessed in a separate incubation by inclusion of 1 mM theophylline, added 15 min before 125 I-azido-BW-A844U. Photoincorporation was initiated by irradiation of the membrane preparations with ultraviolet light at 254 nm (Rayonet miniphotochemical reactor) for 3 min. The photoaffinity-labeled membranes, derived from Transwell cultures that had been biotinylated on either the apical or the basolateral side, were collected by centrifugation, resuspended in RIPA buffer (150 mM NaCl, 50 mM Tris, pH 8.0, 5 mM EDTA, 1% Nonidet P-40, 0.5% deoxycholate, 0.1% SDS) and prepared for streptavidin-agarose chromatography as described previously (1). The eluates were then subjected to SDS-PAGE and autoradiography, and the area of the gel corresponding to the A 1 AdoR signal was excised and counted on a Beckman 5000 ␥-counter to quantify the amount of 125 I-azido-BW-A844U-labeled A 1 AdoR on the apical versus basolateral surface.
Metabolic Labeling/Biotinylation Strategy for Determining Surface Delivery of the TAG-A 1 Adenosine Receptor-The amount of newly synthesized A 1 AdoR delivered to the apical versus basolateral surface was quantified by biotinylating one surface or the other of metabolically labeled MDCKII cells in Transwell culture and subsequently isolating radiolabeled receptor on the biotinylated surface through sequential immunoprecipitation and streptavidin-agarose chromatography. These procedures were performed essentially as described (1), except that membranes were prepared in adenosine receptor membrane binding buffer, as described above.
Immunolocalization of A 1 Adenosine Receptor-Primary and secondary immunostaining protocols were optimized using cells plated on coverslips before evaluating cells polarized by culture in Transwells. For immunostaining of MDCKII cells on coverslips, cells were plated at confluence on glass coverslips and cultured for 2-3 days prior to staining. Cells were fixed and stained with a 1:50 dilution of 12CA5 primary antibody as described (1). Upon rinsing the cells after the incubation of the primary antibody, a 1:200 dilution of the secondary Cy3-conjugated donkey anti-mouse IgG was added to the cells in PBS containing 2% bovine serum albumin and incubated for 1 h at room temperature in the dark. The cells then were mounted on glass slides with Aqua-Poly/ Mount (Polysciences Inc., Warrington, PA).
For immunostaining of cells grown in Transwell culture, cultures were grown for 4 -6 days and assessed for [ 3 H]methoxyinulin leak; wells with leaks greater than 3% were discarded. The cells on the polycarbonate filter were fixed and stained via the same protocol as those grown on the coverslips, except that the filter was excised from the Transwell support prior to the first antibody incubation. Samples were visualized by confocal microscopy on a Zeiss Axiovert 135 Micro Systems LSM (Germany). The samples were first visualized in the xy plane, and then in the xz plane; the images were downloaded onto a Silicon Graphics Iris Indigo workstation for analysis using Showcase software.

The Wild-type and the Epitope-tagged A 1 Adenosine Receptor Are Enriched on the Apical Membrane Domain of MDCKII
Cells-To examine the trafficking of the A 1 -and TAG-A 1 AdoR in canine renal epithelial cells, clones of MDCKII cells permanently expressing wild-type and epitope-tagged receptors were made. The cell lines characterized in most detail in this study expressed 17.2 pmol of wild-type A 1 AdoR/mg protein (clonal line A 1 AdoR#10) and 36.9 pmol of TAG-A 1 AdoR/mg of total membrane protein (clonal line TAG-A 1 AdoR#17). Two other independent untagged cell lines of lower densities also were evaluated; the findings with all cell lines evaluated were comparable.
To determine the steady-state localization of the A 1 AdoR, we employed surface biotinylation and subsequent photoaffinity labeling. The A 1 AdoR extracted from membrane preparations and eluted from streptavidin-agarose was evaluated using SDS-PAGE. The A 1 AdoR migrated on SDS-PAGE with an apparent mass of 43-45 kDa; this photolabeled species was not detected when the incubation with 125 I-azido-BW-A844U was carried out in the presence of the A 1 AdoR antagonist, theophylline. The A 1 AdoR was enriched on the apical surface in cell lines expressing both the wild-type (65-68%) and the TAG-A 1 AdoR (75-83%) structures (Fig. 1A). These data indicate that the presence of the HA-epitope did not significantly alter the apical enrichment of the A 1 AdoR at steady state.

A Similar Apical Localization Is Detected for the A 1 Adenosine Receptor in LLC-PKI Cells-Because there are examples
where polarization of a particular cell surface protein differs in different cultured renal epithelial cell backgrounds, 2 we examined the localization of the wild-type A 1 AdoR after transfection into porcine renal epithelial LLC-PKI cells. As seen with MD-CKII cells, the A 1 AdoR was enriched on the apical surface (80 -85%) of LLC-PKI cells (Fig. 2). In the LLC-PKI cells, the A 1 AdoR migrated consistently between 38 and 42 kDa. We presume that the difference in migration of specific A 1 AdoR photolabeling in MDCKII cells (estimated at 43-45 kDa) versus LLC-PKI cells is due to different glycosylation in the two cellular backgrounds, since treatment of the A 1 AdoR in LLC-PKI cells with endoglycosidase F reduces the apparent molecular mass of the photolabeled A 1 AdoR to ϳ36 kDa (data not shown), a value consistent with the mass of the amino acids encoded by the A 1 AdoR cDNA and that previously reported (14,15). The finding that the A 1 AdoR is similarly enriched on the apical surface following expression in MDCKII and LLC-PKI cells is consistent with the interpretation that the polarization observed is a consequence of the A 1 AdoR structure and its interaction with shared renal epithelial trafficking mechanisms. The apical enrichment does not seem to be a function of receptor density as there seems to be little difference between clones 50, 17, and 10, expressing 57, 37, and 17 pmol/mg protein, respectively. Thus, apical enrichment does not seem to be driven by increased receptor density. Because MDCKII cells frequently are used for the study of cell surface protein polarization and because we have studied a basolaterally targeted protein, the ␣ 2A AR, extensively in these cells, the further characterization of A 1 AdoR localization reported here was evaluated in MDCKII cells.
Immunolocalization of A 1 Adenosine Receptor in MDCKII Cells-Because introduction of the HA-epitope did not appear to alter the localization of the A 1 AdoR in MDCKII cells based on the findings from surface biotinylation studies, we performed immunocytochemical studies to independently determine the localization of the A 1 AdoR in MDCKII cells using the introduced epitope as the receptor marker. For these studies, clones of MDCKII cells permanently expressing the hemagglutinin epitope-tagged version of the A 1 AdoR (TAG-A 1 AdoR#17 and TAG-A 1 AdoR#1) were evaluated. (The receptor density of clone 1 was about half that of clone 17 based on binding assays.) Immunocytochemical studies revealed a predominantly apical localization consistent with the biochemical data obtained using photoaffinity-labeled receptor (Fig. 1B). The x-y scan revealed staining over the top surface of the cells, and the z scan reinforced this interpretation by depicting a distinct horizontal band on the top of the cells. Lateral staining, although detectable, was significantly less than apical staining, even though staining of the lateral subdomain reflects the signal emerging from two cell surfaces juxtaposed to each other. No staining was detected in MDCKII cells expressing wild-type A 1 AdoR, indicating that the immunofluorescent signal detected is indeed due to the epitope-tagged amino terminus of the TAG-A 1 AdoR. The specificity of identification of the 2 M. Caplan, personal communication.
FIG. 1. The wild-type and epitope-tagged A 1 AdoR are enriched on the apical membrane of MDCKII cells at steady state. A, the MDCKII clonal cell lines expressing either wild-type A 1 AdoR (A 1 AdoR#10) or epitope-tagged A 1 AdoR (TAG-A 1 AdoR#17) were grown in Transwell culture and treated with biotin hydrazide to selectively label either the apical or the basolateral membrane surface. In the studies shown here, three Transwell cultures were pooled for each surface determination. The A 1 AdoR was identified with photoaffinity labeling using 125 I-azido-BW-A844U as described under "Experimental Procedures"; non-receptor labeling was assessed in the presence of the adenosine receptor antagonist, theophylline. The autoradiogram was exposed for 7 days to Kodak X-Omat film between two Quanta III screens at Ϫ70°C. Gel slices corresponding to the position of the HA epitope with the 12CA5 antibody is further revealed by comparison of localization of the TAG-A 1 AdoR with the basolaterally targeted TAG-␣ 2 AR structures in subsequent figures (cf. Figs. 5 and 6).
Growth of A 1 AdoR-expressing MDCKII Cells as a Polarized Monolayer Requires the Presence of Adenosine Receptor Antagonists-Upon growing the A 1 AdoR-expressing clones in culture, we observed that the growth pattern of the adenosine receptor-transfected MDCKII cells looked uncharacteristic of other MDCKII cell clones, in particular, growth in patches and in multilayers. We noted that when cells were plated and grown in the presence of two adenosine receptor antagonists, DPSPX (a hydrophilic, highly selective A 1 receptor antagonist) and theophylline (a non-subtype selective A 1 /A 2 receptor antagonist), the cells were prevented from forming multilayers in culture. To evaluate more directly the effect of A 1 AdoR antagonists on MDCKII cell growth, subconfluent A 1 AdoR-transfected and untransfected MDCKII cells were grown in the absence and presence of 60 M theophylline, a nonspecific adenosine receptor antagonist, for 2 days, at which point the cells were harvested and counted. Theophylline significantly inhibited the growth rate of the A 1 AdoR-transfected cells, but not untransfected MDCKII cells (Fig. 3A). Furthermore, cell number achieved over time in the A 1 AdoR-transfected cells was greater compared to the untransfected cells when both were plated at the same density and grown in the absence of theophylline (2.77 Ϯ 0.24 A 1 AdoR-transfected cells/dish versus 1.92 Ϯ 0.29 untransfected cells/dish; p ϭ 0.058, where n ϭ 5 for each group). These findings suggest that A 1 AdoR, either via theophylline-sensitive, agonist-independent receptor activation of MDCKII cells or in response to ambient adenosine, a proliferative agent in some cells (16,17), evokes cell growth. To assess whether or not stimulation of growth of MDCKII cells is a characteristic response to other G i /G o -coupled receptors in these cells, we examined the effect of the ␣ 2 AR agonist, UK-14304, on cells in which the ␣ 2A AR had been introduced by transfection (1); the ␣ 2A AR agonist had no significant effect on cell growth (Fig. 3B). These findings indicate that introduction of the A 1 AdoR, but not of the ␣ 2A AR, changes the growth properties of the MDCKII cells, suggesting (albeit not to the exclusion of other explanations) that the A 1 AdoR and the ␣ 2A AR are localized in distinct compartments containing different G i /G o -coupled effector systems. Although the precise mechanisms by which adenosine receptors mediate proliferation of MDCKII cells was not explored in these studies, our studies did establish that growth of Transwell cultures in the presence of adenosine receptor antagonists prevented multilayer cellular accumulation. (This interpretation is corroborated by the immunocytochemical data in Fig. 4B, described below.) Because characterization of receptor polarization by biochemical means requires a non-permeable monolayer, we examined whether the proliferation of cells expressing adenosine receptors grown in the absence of adenosine receptor antagonists might have contributed to the variable (65-85%) apical localization of A 1 AdoR reported in preliminary studies from our laboratory (18). We thus compared the relative apical localization of A 1 AdoR in Transwell cultures grown in the absence and presence of the combined adenosine receptor antagonists, theophylline (60 M) and DPSPX (100 M) using both biotinylation (Fig. 4A) and immunocytochemical (Fig. 4B) strategies. It was apparent from the immunolocalization experiments that in most clonal cell lines, the cells grew in a multilayer in the absence of adenosine receptor antagonists (Fig. 4B). In some cases, as in Fig. 4A, the percent apical polarization estimated from streptavidin-agarose recovery of biotinylated and photolabeled A 1 AdoR increased notably when growth in Transwell culture occurred in the presence of antagonists, and in this example was 56% in the absence and 75% in the presence of antagonists. Since cell growth in multiple layers would confound analysis of receptor localization by biochemical strategies, these two receptor antagonists were included in all subsequent experiments of all cell lines expressing the A 1 AdoR.
Immunolocalization of Epitope-tagged A 1 AdoR, ␣ 2A AR, and Fig. 4) that was suppressed by the inclusion of 60 M theophylline, a non-subtype selective adenosine receptor antagonist, in the culture medium. B, the effect of an exogenous ␣ 2A AR agonist on MDCKII cell growth was examined by looking at the change in cell number for ␣ 2A AR-transfected MDCKII cells grown in the absence or presence 10 M of the ␣ 2A AR agonist, UK-14,304. For both A and B, cells were plated at 5 ϫ 10 5 cells/dish (subconfluent) in 60-mm dishes, at which point drug or buffer was added. Two days later, the cells were harvested and counted using a Coulter counter. ␣ 2C AR in MDCKII Cells-To confirm the apical localization of the epitope-tagged A 1 AdoR, we compared the localization of the A 1 AdoR with endogenous MDCKII markers for the apical and basolateral surfaces, and with the ␣ 2A AR, already characterized in terms of its lateral targeting following expression in MDCKII cells (1). The apical staining observed for TAG-A 1 AdoR#17 (Fig. 5A) was consistent with the immunostaining pattern of an endogenous MDCKII cell protein, gp135 (19), that serves as a marker for apical staining in MDCKII cells (Fig.  5B). In contrast, the epitope-tagged ␣ 2A AR revealed a predominantly lateral staining profile (Fig. 5C) and was consistent with the immunostaining pattern of the endogenous EGF receptor (Fig. 5D), which is localized laterally in MDCKII cells (20). The appearance of areas with no cells in the x-y scan of the A 1 AdoR-stained cells is attributed to the various heights of the cells, such that the apical signal comes into focus at different levels.

FIG. 3. Introduction of A 1 AdoR, but not the ␣ 2A AR, into MDCKII cells causes proliferation of MDCKII cells. A, introduction of A 1 AdoR into MDCKII cells caused an increase in cell growth and multilayer morphology (see
In an attempt to determine if some of the "haze" seen in the z scans of the TAG-A 1 AdoR could be attributable to a genuine signal arising from one or more intracellular compartments, the cells were stained in the presence and absence of the detergent Triton X-100 as a permeabilizing agent. The relative distribution of surface to intracellular fluorescence was not significantly altered when the 12CA5 primary antibody directed against the HA epitope was incubated with fixed, TAG-A 1 AdoR-expressing cells in the presence (Fig. 6A) or absence (Fig. 6B) of Triton X-100. The apical staining is somewhat fainter when the procedure is performed in the absence of Triton X-100, perhaps because the detergent facilitates access of the 12CA5 antibody to the amino-terminal epitope on the A 1 AdoR even on the extracellular surface. To test our ability to detect the HA epitope when expressed intracellularly, we also examined MDCKII cells expressing the ␣ 2C AR receptor subtype, which previously has been demonstrated to be localized on the surface as well as in an intracellular compartment in MDCKII cells, 3 as well as COS and human embryonic kidney cells (21). The ability to readily identify the intracellular compartment containing the TAG-␣ 2C AR structure required incubation of the fixed Transwells with primary antibody in the presence of Triton X-100 (Fig. 6C), as the detection of this intracellular compartment was not clearly visible in the absence of Triton X-100 (Fig. 6D). These control studies indicate that if a substantial fraction of the TAG-A 1 AdoR were enriched in an intracellular compartment, incubation with the 12CA5 primary antibody in the presence of Triton X-100 should have detected it.
The A 1 Adenosine Receptor Is Delivered Directly to the Apical and Basolateral Surfaces of MDCKII Cells-To investigate the mechanism by which apical enrichment is achieved, the delivery of the A 1 AdoR to the apical versus the basolateral surface was determined using metabolic labeling in combination with surface biotinylation (Fig. 7). Newly synthesized, metabolically labeled A 1 AdoR was first detectable after 30 min of pulse labeling. At this time, metabolically labeled receptor was detectable on both the apical and basolateral surfaces, but appeared to be preferentially delivered to the apical surface (63% apical, 37% basolateral). Over all the time points examined (30, 45, 60, 90, and 120 min), the apical to basolateral ratio remained relatively constant, consistent with the interpretation that the amount of 35 S-labeled A 1 AdoR detected at each surface is due to direct delivery to these surfaces (Fig. 7A). We are confident that the 35 S-labeled protein, migrating slightly above 3  the 43-kDa molecular size marker, is the A 1 AdoR, since it comigrates with the photoaffinity-labeled A 1 AdoR and because no radioactive band is visible at this molecular weight when MDCKII cells expressing wild-type A 1 AdoR (A 1 AdoR#10) are metabolically labeled for 60 min and processed exactly as MD-CKII cells expressing TAG-A 1 AdoR (Fig. 7B).
Since the steady-state localization of a receptor is influenced by both the initial delivery to a particular surface and the eventual retention to that specific surface, we examined the half-life of the A 1 AdoR (TA 1 AdoR#17) on both the apical and basolateral surface. The kinetics of the removal of the receptor from the apical plasma membrane were not significantly different from those of the basolateral membrane (Fig. 8). The calculated half-life on the apical surface was about 9 -13 h and on the basolateral surface about 11-12 h. Thus, the apical enrichment observed at steady state cannot be attributed to differential retention of the A 1 AdoR on the apical compared to the basolateral surface, but appears to result from preferential delivery to the apical surface.

DISCUSSION
Endogenous compounds or drugs must first be recognized by the appropriate receptor or, in the case of synthetic compounds, receptor subtype to exert their desired effect. Multiple mechanisms likely contribute to specificity in signal transduction by endogenous agonists: the existence of numerous receptor subtypes that couple to distinct signal transduction pathways; coupling of receptors to distinct effector systems in various tissues; and receptor localization to discrete microdomains on the cell surface, restricting the G-proteins, effectors, and other molecules with which the receptor interacts. The importance of receptor localization in signal transduction is inferred by the number of pathophysiologic states that result from mislocalized receptors (22; 23). For example, one form of retinitis pigmentosa results from intracellular trapping of the G-proteincoupled receptor for light, rhodopsin. It therefore is essential to understand the mechanisms that govern the trafficking of receptors and signal transducing proteins in order to both gain insights into their role in receptor-mediated signal transduction events under physiologic conditions and to probe for potential culprits in pathophysiological states.
We previously have demonstrated that the ␣ 2A AR is localized basolaterally in renal epithelial cells (1). One strategy that could reveal the structural regions of the ␣ 2A AR, and perhaps all G-protein-coupled receptors, necessary for conferring basolateral localization is to make receptor chimeras with other seven-transmembrane-spanning, G-protein-coupled receptors that achieve opposite localization in polarized cells. We chose the A 1 AdoR as a potential candidate since adenosine is available in sufficient concentrations (2) to activate adenosine receptors on the apical surface of renal cells in vivo due to the existence of caveolar 5Ј-nucleotidase in that domain (24). Also, in in vivo studies, Franco et al. (25) found that an A 1 AdoR agonist, when administered lumenally, participated in the tubuloglomerular feedback mechanism. However, it should be noted that responses to basolaterally introduced adenosine receptor agonists also have been demonstrated in renal epithelial cells in vitro, e.g. A 1 AdoR-accelerated phosphatidylinositol turnover and inhibition of forskolin-stimulated cAMP production following basolateral but not apical administration of adenosine (3). Similarly, basolaterally localized A 1 AdoR have been suggested by addition of A 1 analogues to the basolateral compartment (7) or to the basolateral surface of perfused inner medullary collecting duct tubules (5).
Our studies are the first to characterize the localization of the A 1 AdoR based on its structure, using biochemical and morphological detection strategies, rather than relying on receptormediated functions to infer receptor localization. Our studies demonstrate that the A 1 AdoR is preferentially localized to the apical surface of MDCKII and LLC-PKI cells in multiple independent clonal cell lines. The apical enrichment of the A 1 AdoR at steady state is mirrored by the preferential delivery of A 1 AdoR to the apical versus basolateral surface. Since the A 1 AdoR has a similar half-life on the apical and the basolateral surfaces after its delivery, our findings are consistent with the interpretation that the A 1 AdoR localization is achieved by direct delivery to the apical versus basolateral surfaces. It may be that both the apically and basolaterally localized receptor populations elicit activities important for polarized function, albeit by coupling to different effector mechanisms. Our own observations on proliferation of MDCKII cells expressing the A 1 AdoR (Fig. 3) are consistent with this interpretation; both the A 1 AdoR and the ␣ 2A AR elicit responses via G i /G o -coupled GTP-binding proteins, but only introduction of the A 1 AdoR led to an increase in cell number in polarized MDCKII cells. Whether or not the effects of theophylline to slow cell doubling are due to agonist-independent effects of the receptor or result from endogenously produced adenosine acting at the heterologous A 1 AdoR has not been determined. Nevertheless, inclusion of A 1 AdoR antagonists was necessary to obtain a monolayer of cells required for performing polarization studies.
The finding that the A 1 AdoR is delivered to both the apical and basolateral surfaces, but preferentially to the apical surface, distinguishes this receptor from other G-protein-coupled receptors characterized to date. The ␣ 2A AR is delivered exclusively to the basolateral surface (1); in contrast, the ␣ 2B AR FIG. 6. The A 1 AdoR is expressed primarily on the surface of MDCKII cells as revealed by immunolocalization in the absence and presence of Triton X-100. The confocal image of the TAG-A 1 AdoR#17 in MDCKII cells grown on Transwell filters in the presence of A 1 AdoR antagonists and stained in the presence of Triton X-100 shows apical staining, as evidenced by the strong red signal of the Cy3 chromophore present in both the x-y scan and the z scan (A). There is some red "haze" present below the apical signal in the z scan. The confocal image of the TAG-A 1 AdoR#17 stained in the absence of Triton X-100 shows a weaker apical signal, with an attenuated red "haze" (B). The confocal image of the epitope-tagged ␣ 2C AR subtype, known to be expressed on the surface as well as in the interior of MDCKII cells (see Footnote 3), shows a lateral staining pattern with detectable intracellular staining in the presence of Triton X-100 (C) and the absence of detectable intracellular signal in the absence of Triton X-100 (D).
achieves steady-state basolateral localization in MDCKII cells via random delivery and selective retention on the basolateral surface. 3 In studies not characterized as rigorously as those presented here, we observed that the M2 and M3 muscarinic receptors are localized to both the apical and basolateral surfaces of MDCKII cells as revealed by confocal microscopy, but preferentially on the basolateral surface. 4 Thus, of all the Gprotein-coupled receptors studied to date, the A 1 AdoR is the only structure preferentially delivered to the apical surface that appears to be retained at that surface without redistribution to the lateral subdomain. Chimeric structures between A 1 AdoR and the ␣ 2A AR might therefore permit identification of structural regions in G-protein-coupled receptors that confer basolateral versus apical targeting. FIG. 7. The A 1 AdoR is preferentially delivered to the apical membrane domain of MDCKII cells, consistent with its steady-state enrichment on that surface. A, MDCKII cells expressing epitope-tagged A 1 AdoR (TAG-A 1 AdoR#17) grown in Transwell culture in the presence of A 1 AdoR antagonists were metabolically labeled with 1 Ci/l [ 35 S]Met/[ 35 S]Cys protein labeling mix (150 l) for the indicated times and then harvested and processed using sequential immunoprecipitation and streptavidin-agarose chromotography as described under "Experimental Procedures." The autoradiogram shown is for an SDS-PAGE gel exposed for 7 days to Kodak X-Omat film between two Quanta III screens at Ϫ70°C. Gel slices corresponding to the position of the A 1 AdoR signal on autoradiograms were excised and counted in a 10 ml of NEN-963 scintillation fluid in a ␤-scintillation counter. Localization on the apical membrane was not significantly different over the range of time points examined; the findings from multiple individual experiments evaluating varying time points (in percent apical enrichment) are as follows: 30 min (n ϭ 2) mean ϭ 61, range ϭ 59 -63; 45 min (n ϭ 3) mean ϭ 67.3, range ϭ 62-76; 60 min (n ϭ 6) mean ϭ 62.3, range ϭ 56 -66; 90 min (n ϭ 2) mean ϭ 57.5 range ϭ 55-60; 120 min (n ϭ 1) is 62% apical enrichment. B, cells expressing wild-type A 1 AdoR (clone 10) were metabolically labeled for 60 min and then treated as in panel A. The autoradiogram for metabolically labeled wild-type receptor was exposed for 7 days with Kodak X-Omat film at Ϫ70°C. Migration of the photoaffinity-labeled A 1 AdoR was included in the experiment to demonstrate the migration of the A 1 AdoR for comparison with the metabolically labeled, epitope-tagged protein isolated by sequential protein A and streptavidin-agarose chromatography. In all gels where TAG-A 1 AdoR were analyzed, there were visible [ 35 S]Met/Cys-labeled bands migrating at about 55, 67, 80, and 94 kDa; these presumably correspond to receptor aggregates, since they also were seen when the A 1 AdoR was identified by photoaffinity labeling. No [ 35 S]Met/Cys signal was detected at any molecular weight when cells expressing the wild-type receptor were analyzed. ]Cys for a 60-min "pulse" and rapidly rinsed in PBS before the "chase" phase was initiated by addition of complete Dulbecco's modified Eagle's medium supplemented with additional 1 mM methionine, 1 mM cysteine, and A 1 AdoR antagonists. For monitoring of half-life, the time ϭ 0 corresponds to a 6-h chase period, since after this time point no increase in metabolic labeling signal occurs at the surface. A 1 AdoR present on the apical surface of labeled cells at each time point was calculated by surface biotinylation, followed by sequential immunoisolation and streptavidin-agarose chromatography of cell extracts, as described under "Experimental Procedures." Each data point represents the streptavidin-agarose eluate derived from three Transwells per time point per surface evaluated. Gel slices corresponding to the position of the A 1 AdoR signal on autoradiograms were excised and counted in 10 ml of NEN-963 scintillation fluid and counted in a Packard-Tricarb liquid scintillation counter on the ␤-channel. The data also were analyzed by scanning the gel on a phosphorimager, where the results were not significantly different from those obtained from cutting and counting the gel slices. The data (in h) are as follows: the calculated half-life ranged from 9 -13 h (n ϭ 2) on the apical surface and 11-12 h (n ϭ 3) on the basolateral surface.