Structural Basis of Agonist-induced Desensitization and Sequestration of the P2Y2 Nucleotide Receptor

Molecular determinants of P2Y2receptor desensitization and sequestration have been investigated. Wild-type P2Y2 receptors and a series of five C-terminal truncation mutants of the receptor were epitope-tagged and stably expressed in 1321N1 cells. These constructs were used to assess the importance of the intracellular C terminus on 1) UTP-stimulated increases in intracellular calcium concentration, 2) homologous desensitization of the receptor, and 3) agonist-induced decreases in cell-surface density (receptor sequestration) of epitope-tagged receptors using fluorescence-activated cell sorting. The potency and efficacy of UTP were similar for the wild-type and all mutant P2Y2 receptors. Truncation of 18 or more amino acids from the C terminus increased by ∼30-fold the concentration of UTP necessary to desensitize the receptor. Both the rate and magnitude of UTP-induced receptor sequestration were decreased with progressively larger truncations of the C terminus. Furthermore, the recovery from sequestration was slower for the most extensively truncated receptor. Complete desensitization was obtained with >50% of the original receptor complement remaining on the cell surface. Protein kinase C activation, which desensitizes the P2Y2 receptor, had no effect on sequestration, consistent with the ideas that desensitization and sequestration are discrete events and that agonist occupancy is required for receptor sequestration.

activation of phospholipase C and increased levels of inositol 1,4,5-trisphosphate and diacylglycerol, leading to elevations in the intracellular free calcium concentration ([Ca 2ϩ ] i ) and the activation of protein kinase C (PKC) (1)(2)(3)(4). The P2Y 2 nucleotide receptor subtype (formerly the P 2U receptor) is distinguished pharmacologically from the other known mammalian P2Y receptor subtypes by the equal potency and efficacy of the naturally occurring agonists ATP and UTP.
Activation of P2Y 2 receptors present in airway epithelium increases Cl Ϫ secretion through Ca 2ϩ -dependent and outwardly rectifying Cl Ϫ channels (5). It has been shown that P2Y 2 receptor activation by UTP in airway epithelia of cystic fibrosis patients can increase Cl Ϫ secretion, thereby effectively bypassing the defective cAMP-dependent Cl Ϫ transport (6). Like other members of the GPCR superfamily, P2Y 2 receptors undergo agonist-induced desensitization (7), but little is known about the mechanisms involved in desensitization of the P2Y 2 receptor. It seems likely that a fuller understanding of desensitization and the signaling pathways that affect the P2Y 2 receptor may lead to improved therapies targeted to this receptor.
Desensitization of GPCRs is a complex process involving phosphorylation of the receptors by multiple protein kinases, including G protein-coupled receptor kinases, cAMP-dependent protein kinase, and PKC. The mechanisms of GPCR desensitization have been studied most extensively with the ␤ 2 -adrenergic receptor (reviewed in Ref. 8). The activation of ␤ 2 -adrenergic receptors by agonist results in G protein-coupled receptor kinase-mediated phosphorylation of serine and threonine residues in the carboxyl-terminal tail (9). The subsequent agonistinduced increase of cyclic AMP can result in the phosphorylation of additional residues within the ␤ 2 -adrenergic receptor by protein kinase A. Phosphorylation of ␤ 2 -adrenergic receptors leads to desensitization and sequestration of the receptors from the cell surface, where they can no longer bind hydrophilic ligands and apparently undergo dephosphorylation prior to recycling to the cell membrane (10 -13). Prolonged exposure to agonist results in down-regulation of the receptor, which is a decrease in receptor number that requires new protein expression to reestablish receptor levels.
We investigated the role of the serine-and threonine-rich C terminus of the P2Y 2 receptor in agonist-induced receptor desensitization and sequestration. The C terminus of the P2Y 2 receptor contains two consensus phosphorylation sites for PKC (another site is present in the third intracellular loop) and possible sites for G protein-coupled receptor kinase-mediated phosphorylation. To determine the role of the C terminus in receptor desensitization and sequestration, truncation mutants of a P2Y 2 receptor cDNA encoding an epitope-tagged protein were constructed by polymerase chain reaction (PCR). The truncated receptor cDNAs were expressed in a clonal human 1321N1 astrocytoma cell line devoid of P2Y nucleotide receptors. UTP-induced desensitization was compared in cells expressing mutant or wild-type P2Y 2 receptors. Cell-surface receptor density and agonist-induced sequestration in 1321N1 cells were assessed by immunofluorescence detection of the epitope-tagged receptors.

EXPERIMENTAL PROCEDURES
Mutagenesis, Epitope Tagging, and Subcloning-Wild-type murine P2Y 2 receptor cDNA was subcloned into the retroviral vector pLXSN at the EcoRI/BamHI sites of the multiple cloning site. The open reading frame of the wild-type P2Y 2 receptor cDNA was modified to incorporate, at the amino terminus of the expressed protein, the hemagglutinin (HA) epitope (YPYDVPDYA) from influenza virus using PCR. The forward and reverse HA primers were 5Ј-gatcgtgaattctgatgtatccatatgatgttccagattatgctgcagcagacctggaaccctgg-3Ј and 5Ј-gatcgtggatcccctgactgaggtgctatagccg-3Ј, respectively. The PCR solution contained primers (0.7 M each), 10 l of 10ϫ Vent polymerase buffer (New England Biolabs Inc., Beverly, MA), 100 ng of template DNA, 2.5 units of Vent (exo Ϫ ) polymerase, 0.25 units of Vent (exo ϩ ) polymerase, and 20 l of dNTP mixture (0.2 mM dATP, dCTP, dTTP, and dGTP) in a final volume of 100 l. The PCR parameters were as follows: 96°C for 1 min, 62°C for 1 min, and 72°C for 2.5 min for 25 cycles. The PCR conditions were identical for creation of the five truncation mutants. In each case, the HA epitope was inserted using the forward HA primer and the following reverse primers: 5Ј-tgctagggatcccgctactatgtccgtcttgagtcgtcact-3Ј (truncation mutant 1), 5Ј-tgctagggatcccgtcatcagtcgtcactactgactgacaaatc-3Ј (truncation mutant 2), 5Ј-tgctagggatcccgctactaacttccagctggtgtggactc-3Ј (truncation mutant 3), 5Ј-tgctagggatcccgtcatcacaaatctttcctcacagttctgttag-3Ј (truncation mutant 4), and 5Ј-gcatcggatcctcatcagttaggcctgtgcaggccc-3Ј (truncation mutant 5). After verification of the PCR amplification by agarose gel electrophoresis, the products were purified using a PCR Wizard kit (Promega, Madison, WI). The purified PCR products and pLXSN DNA were digested overnight with EcoRI and BamHI and ligated together, followed by transformation of competent Escherichia coli and identification of positive clones (14). All mutant DNAs were sequenced on both strands to ensure that mutagenesis had occurred as predicted, using an ABI Prism automated sequencing apparatus (Perkin-Elmer) and fluorescence dideoxynucleotide technology.
Calcium Measurements-Changes in [Ca 2ϩ ] i were detected by dualexcitation spectrofluorometric analysis of P2Y 2 -1321N1 cell suspensions loaded with fura-2 as described previously (16,17). Cells were assayed in 10 mM Hepes-buffered saline (pH 7.4) containing 1 mM CaCl 2 and 1 mM MgCl 2 unless indicated otherwise. Desensitization experiments were performed by incubating P2Y 2 -1321N1 cell suspensions with varying concentrations of UTP for 5 min at 37°C. The cells were pelleted in a microcentrifuge and resuspended in 2 ml of buffer. The EC 50 value of UTP, determined during the concentration-response experiments (see Fig. 3), was used to re-challenge the cells. Concentration-response data were analyzed with the Prism curve fitting program (GraphPAD Software for Science, San Diego, CA).
Sequestration Assays-P2Y 2 -1321N1 cells, grown to 80% confluency in 35-mm 2 dishes, were incubated with UTP (Amersham Pharmacia Biotech) or phorbol 12-myristate 13-acetate (PMA; Sigma) for the times and at the concentrations indicated in the figure legends. Control cells were incubated without UTP (to allow an estimation of the total cellsurface complement of P2Y 2 receptors) for 180 min.
The P2Y 2 -1321N1 cells were washed with 2 ml of ice-cold Hepesbuffered saline (pH 7.4) and incubated at 4°C for 1 h with gentle rocking in 1 ml of Hepes-buffered saline containing 10 g of anti-HA antibody 12CA5 (Boehringer Mannheim) that was affinity-purified as follows. The antibody was added to a HiTrap protein G column (Amersham Pharmacia Biotech) equilibrated with 20 mM phosphate buffer (pH 7.5) and eluted with 1-ml fractions of 100 mM glycine (pH 2.75). The eluted fractions were immediately neutralized with 1 M Tris buffer (pH 8.5), dialyzed against phosphate-buffered saline (PBS), and stored for use at a protein concentration of 1.5 mg/ml as estimated by a modification (18) of the method of Lowry et al. (19).
Cells incubated with antibody 12CA5 were washed with 2 ml of Dulbecco's PBS and incubated in 1 ml of Dulbecco's PBS containing 10 l of Fc-specific fluorescein isothiocyanate-labeled goat anti-mouse antibody (Sigma) in the dark at 4°C with gentle rocking for 1 h. Control cells were incubated with primary or secondary antibody alone to detect nonspecific fluorescence. The P2Y 2 -1321N1 cells were washed with 2 ml of Dulbecco's PBS and detached in 2 ml of Hepes-buffered saline (pH 7.4) containing 2 mM EDTA. The cells were centrifuged and resuspended in 1 ml of 1% (v/v) formaldehyde in PBS and incubated in the dark at 4°C for 10 min.
Cells were centrifuged and resuspended in 0.5 ml of PBS and analyzed on an EPICS 753 flow cytometer (Coulter Corp., Hialeah, FL) with a 5-watt argon laser tuned at 488 nm using 150-milliwatt output. The instrument was optically aligned using DNA-Check fluorospheres (Coulter Corp.) with the coefficients of variation Յ2%. For each analysis, 10,000 cells were collected using a dual-parameter histogram of forward angle light scatter versus log 90°light scatter to gate out debris. Integral green fluorescence emission (525 band path filter) of fluorescein isothiocyanate-labeled cells was displayed as a single-parameter histogram on a three-decade log scale. Data analysis was performed using EASY2 software (Coulter Corp.). Fluorescence intensity was used as a measure of differences in receptor density on the cell surface for the various receptor constructs. The computer cursor was set at a point representing half the peak width at half the peak height, and the channel number of this position was noted as the fluorescence intensity of the histogram. Fluorescence intensity was measured after logarithmic amplification with the use of a three-decade log scale, but illustrated on a linear scale of 253 channels or 85.33 units/decade. Fluorescence intensity (FI) was converted to linear arbitrary units (LAU) using the following formula: LAU ϭ 10 (FI/85.33) . LAU was corrected for fluorescence due to secondary antibody alone. Cells expressing wild-type or mutant P2Y 2 receptors that were not incubated with UTP but were incubated with antibody 12CA5 and secondary antibody were considered to have the maximum number of receptor molecules on the cell surface.
Calculation of Relative Receptor Densities-Cells expressing P2Y 2 receptors were diluted from 1:2 to 1:8 with nontransfected 1321N1 cells, and the amount of fluorescein isothiocyanate fluorescence was measured in a Spex dual-excitation/emission spectrofluorometer using 488 and 525 nm for the excitation and emission wavelengths, respectively. Determinations of total protein/ml of cells were made using a modification (18) of the method of Lowry et al. (19). A standard curve of fluorescence was constructed with HEK-293 cells heterologously expressing N-terminally HA-tagged ␤ 2 -adrenergic receptors (B max ϭ 500 fmol/mg of total cell protein (20)) diluted with nontransfected HEK-293 cells.

RESULTS
Five cDNAs encoding P2Y 2 receptors with different length C-terminal truncations ( Fig. 1) were constructed by PCR and expressed in 1321N1 cells to determine the role of the C terminus in agonist-induced desensitization of receptor-mediated calcium mobilization and receptor sequestration. The most extensively truncated receptor (truncation mutant 5) lacks two of three protein kinase C consensus phosphorylation sites and potential sites of phosphorylation by G protein-coupled receptor kinase present in the intracellular domains of the wild-type P2Y 2 receptor.
Recombinant P2Y 2 Receptor Expression Levels in 1321N1 Cells-In the absence of an established method to detect P2Y 2 receptors on the cell surface, all receptor constructs were Nterminally tagged with an epitope of the HA antigen from influenza virus. The fluorescence intensities of cells expressing the epitope-tagged P2Y 2 receptor constructs were determined by flow cytometry, and the resulting scans were superimposed (Fig. 2). These data indicate similar fluorescence intensities for each receptor, with truncation mutant 3 and wild-type receptors representing the high and low values, respectively, of the range of receptor expression. Thus, the expression of the six receptor constructs was approximately the same, with a maximal 2-fold difference in receptor density (Table I). The number of HA-tagged wild-type P2Y 2 receptors in 1321N1 cells was estimated by comparison with the fluorescence intensity due to HA-tagged ␤ 2 -adrenergic receptors (500 fmol/mg of total protein) in HEK-293 cells (20  same density. The truncated receptors exhibited rapid and transient increases in [Ca 2ϩ ] i in response to UTP, followed by a return to base-line levels within 3 min (data not shown), similar to the wild-type receptor. The desensitization of the calcium response to UTP was determined for each receptor following a 5-min preincubation with the indicated concentration of UTP (Fig. 3). Progressive truncations of the C terminus increased the IC 50 values for UTP-induced desensitization to a maximum of 25-fold with truncation mutant 4 ( Table I).
Truncations of the C Terminus Decrease Receptor Sequestration- Fig. 4 shows that Ͼ80% of the wild-type receptors were sequestered within 10 min. The t1 ⁄2 , the time at which 50% of the cell-surface receptor complement is sequestered, was Ͻ5 min. Progressive truncations of the C terminus of the P2Y 2  Table I). The data presented are expressed as a percentage of the peak calcium response to the EC 50 value of UTP in cells that were not preincubated with the agonist. The peak values were as follows: wild-type, 176 Ϯ 43 nM   Fig. 3 (A-F). Normalized receptor expression levels were calculated by determining the ALI (n ϭ 3) for each receptor construct and comparing each value to the wild-type receptor, which was regarded as 1. Estimated t1 ⁄2 values of sequestration were from Fig. 4 (A-F). Maximal receptor sequestration (the percentage of receptor complement removed from the cell surface) was determined from the data used to construct Fig. 4 (A-F) receptor increased the t1 ⁄2 for sequestration. For the most truncated forms of the receptor (truncation mutants 4 and 5), the t1 ⁄2 values were ϳ25-fold greater, and at least 35% of the receptors remained on the cell surface after 180 min. UTP concentrationeffect curves show that 50% of the wild-type receptors were sequestered after a 10-min incubation with ϳ6 M UTP, whereas sequestration of 50% of the truncation mutant 5 receptors required a 120-min incubation with 1 mM UTP (Fig. 5). Hydrolysis of UTP by 1321N1 cells was determined by HPLC analysis to be negligible, with 96% of the UTP remaining after 3-h incubation (data not shown). The original UTP stock solution contained a similar percentage of UTP, with 4% contamination with UDP. We were thus unable to detect any significant hydrolysis of UTP under these experimental conditions, unlike Lazarowski et al. (21), who demonstrated significant ecto-apyrase activity in 1321N1 cells. It is possible that, during our experiments, too much UTP and too few 1321N1 cells were present to demonstrate significant hydrolysis of nucleotide. P2Y 2 Receptor Sequestration Is Reversible-The reversibility of sequestration of the wild-type P2Y 2 receptor was determined after preincubating cells for 10 min with 1 mM UTP. Following incubation of the cells for various times in the absence of agonist, cell-surface receptor expression was determined. As shown in Fig. 6, recovery was time-dependent and nearly complete (ϳ90%) after 240 min. Half of the receptor complement returned to the cell surface in ϳ60 min. Fig. 7 illustrates the relationship between agonist-induced desensitization and receptor sequestration for the wild-type and truncation mutant 5 receptors. It is apparent for both receptors that 50% of the cell-surface receptor complement was still present when the UTP-induced increase in [Ca 2ϩ ] i was desensitized. Furthermore, the wild-type and truncation mutant 5 P2Y 2 receptors had nearly all of the receptor complement on the cell surface when 50% of the UTP-induced calcium response was desensitized.

UTP-induced Receptor Desensitization and Sequestration Are Distinct Events-
The effects of the protein kinase C activator PMA on desensitization of UTP-stimulated increases in [Ca 2ϩ ] i were examined in cells expressing the wild-type and truncation mutant 5 P2Y 2 receptors. Fig. 8 shows that the wild-type and truncation  However, exposure of cells to 1 M PMA for times ranging from 5 to 180 min failed to induce sequestration of wild-type P2Y 2 receptors (Fig. 9). Furthermore, co-incubation with 1 M PMA and 1 mM UTP produced a pattern of sequestration similar to that obtained with UTP alone. These findings suggest no role for PKC in the sequestration of the P2Y 2 receptor and thus dissociate receptor desensitization from sequestration. DISCUSSION Desensitization of the G protein-coupled P2Y 2 receptor can attenuate responses to nucleotides, which may impair the therapeutic effects of nucleotides in treatment of diseases such as cystic fibrosis. G protein-coupled receptor desensitization can involve several protein kinases, including protein kinase A, PKC, and G protein-coupled receptor kinases (reviewed in Ref. 8). However, the pathway involved in P2Y 2 receptor desensitization is not well understood. P2Y 2 receptor activation is coupled to phospholipase C, leading to the production of diacylglycerol, an activator of PKC. The presence of three consensus sites for phosphorylation by PKC and possible sites for G protein-coupled receptor kinase phosphorylation in the intracellular domain of the P2Y 2 receptor prompted us to investigate the role of the serine/threonine-rich C terminus in agonist-induced P2Y 2 receptor desensitization.
It has been established in this study that the recombinant P2Y 2 receptor expressed in 1321N1 cells desensitizes like the endogenous receptor in bovine aortic endothelial cells (7). Desensitization of the P2Y 2 receptor occurs in response to UTP or activation of PKC with PMA, although in some cells, PKC may not mediate P2Y 2 receptor desensitization (22).
The lack of a reliable radioligand binding assay and antibodies to the endogenous P2Y 2 receptor necessitated the incorporation of the HA epitope of influenza virus at the N terminus of the receptor to monitor sequestration with the 12CA5 monoclonal antibody. Truncation of 30 amino acids from the C terminus of the P2Y 2 receptor (truncation mutant 5) increased the concentration of UTP required for receptor desensitization and decreased the ability of these receptors to sequester as efficiently as the wild-type P2Y 2 receptor. This trend was also exhibited with truncation mutants 3 and 4. However, truncation mutant 2 had a similar UTP concentration response for desensitization, but different sequestration properties compared with the wild-type P2Y 2 receptor, whereas truncation mutant 1 desensitized and sequestered like the wild-type receptor. The cause for the change in sequestration properties that occurred between truncation mutants 1 and 2 requires further investigation, but may involve the loss of one or more essential amino acids from truncation mutant 2 that impact on sequestration, but not desensitization. Clearly, these changes do not affect receptor activation since UTP-induced increases in [Ca 2ϩ ] i for all the recombinant P2Y 2 receptor constructs had similar concentration-response curves. The wild-type P2Y 2 receptor lacks any cysteine residues in its C terminus that are apparently required for palmitoylation of the ␤ 2 -adrenergic receptor (23,24) and the formation of tertiary structure required for agonist-induced receptor desensitization (25).
The wild-type P2Y 2 receptor expressed in 1321N1 cells desensitizes and sequesters in a time-dependent manner similar to other GPCRs, and although the P2Y 2 receptor recovers from desensitization like other GPCRs, its recovery from sequestration is slower (11,26). It may be that the presence of the HA epitope affects the sequestration kinetics of the P2Y 2 receptor, but until reliable radioligands are developed for this receptor, we are unable to address this question. Although de novo protein synthesis may be responsible for recovery of some receptor activity, resensitization is more likely due to receptor recycling given the short incubation period with UTP. How-ever, recovery from receptor sequestration was difficult to evaluate in cells expressing truncation mutant 5 because of the necessity to incubate these cells for 180 min (a period that may involve receptor down-regulation) to get adequate sequestration. Experiments on transcriptional regulation of the endogenous P2Y 2 receptor will be necessary to determine the mechanisms of recovery from receptor down-regulation.
Truncations of the carboxyl-terminal tail of G protein-coupled receptors can have a wide range of effects on receptor desensitization and sequestration. Truncation of the terminal 45 amino acids of the angiotensin II type 1A receptor inhibited receptor internalization with no effect on radioligand binding, downstream signaling, or agonist-induced desensitization (27). Removal of 72 of 86 amino acids from the C terminus of the serotonin type 2 receptor had no effect on the ability of the receptor to stimulate phospholipase C or to be internalized even though three of five PKC consensus phosphorylation sites and potential G protein-coupled receptor kinase phosphorylation sites were removed (28). However, the authors noted the presence in the truncated and wild-type 5HT 2 receptors of the consensus NPXXY sequence, in which the tyrosine residue has been reported to be essential for ␤ 2 -adrenergic receptor sequestration. In contrast, Thomas et al. (27) have reported the presence of an NPXXY sequence in an internalization-resistant mutant of the angiotensin II type 1A receptor. The P2Y 2 receptor has a DPXXY sequence that remains after the most radical truncation (truncation mutant 5), and it will be of interest to see if this tyrosine residue has any role in receptor sequestration.
Another study of the phospholipase C-linked G protein-coupled ␣ 1B -adrenergic receptor indicated the importance of the C terminus in agonist-induced receptor phosphorylation and desensitization (29). A 147-amino acid truncation of the C terminus of the ␣ 1B -adrenergic receptor did not decrease agonistinduced inositol 1,4,5-triphosphate accumulation compared with the wild-type ␣ 1B -adrenergic receptor. However, the truncated receptor was resistant to desensitization and was not phosphorylated in response to activation by its ligand, epinephrine. The truncated ␣ 1B -adrenergic receptor was able to sequester, but more slowly than the wild-type receptor.
Removal of 43 amino acids from the C terminus of the lutropin/choriogonadotropin receptor (30) or 48 amino acids from the C terminus of the ␤ 2 -adrenergic receptor (31) led to a decrease in agonist-induced uncoupling of the receptors from adenylyl cyclase. Compared with the wild-type ␤ 2 -adrenergic receptor, the truncation mutant required longer periods of agonist exposure to desensitize, presumably indicating the importance of ␤-adrenergic receptor kinase phosphorylation sites that were lost upon receptor truncation.
It will be of interest to investigate the specific amino acid residues of the P2Y 2 receptor that are involved in both desensitization and sequestration of this receptor. Fredericks et al. (9) identified G protein-coupled receptor kinase phosphorylation sites in the ␤ 2 -adrenergic receptor and found similarities between these sites and those found in rhodopsin. The most amino-terminal phosphorylation site in both rhodopsin and the ␤ 2 -adrenergic receptor has pairs of acidic residues at its aminoterminal side. Truncation mutant 3 of the P2Y 2 receptor truncates the receptor directly after two acidic residues and directly before a serine residue. It is tempting to speculate the importance of these amino acids in P2Y 2 receptor function since truncation mutant 3 is the least truncated form of the P2Y 2 receptor used in this study that required an increased concentration of UTP to induce desensitization compared with the wild-type receptor.
Very few studies have directly investigated the role of func-tionally important domains of P2Y receptors. Specific amino acids in the transmembrane-spanning domains of the P2Y 2 (32) and P2Y 1 (33) receptors have been shown to affect agonist potency and specificity. The present study has defined the importance of the C terminus of the P2Y 2 receptor in desensitization and sequestration. Considering the widely varying roles of the C terminus in G protein-coupled receptor functions, future studies should better delineate the relevance of this domain to P2Y 2 receptor regulation. Studies are also needed to fully evaluate the role of PKC and other protein kinases in the desensitization and sequestration of the P2Y 2 receptor. A complete understanding of the signaling pathways involved may help minimize receptor desensitization and sequestration to optimize nucleotide therapies, such as those proposed for cystic fibrosis, that are directed at the P2Y 2 receptor.