HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 49, 46891-46899, December 6, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/49/46891    most recent M205274200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Koshimizu, T.-a. Articles by Tsujimoto, G. Search for Related Content PubMed PubMed Citation Articles by Koshimizu, T.-a. Articles by Tsujimoto, G. Social Bookmarking                      What's this?

Heteromultimerization Modulates P2X Receptor Functions through Participating Extracellular and C-terminal Subdomains^*

Taka-aki Koshimizu^§, Susumu Ueno^¶, Akito Tanoue, Nobuyuki Yanagihara^¶, Stanko S. Stojilkovic, and Gozoh Tsujimoto^**

From the  Department of Molecular and Cell Pharmacology,  Endocrinology and Reproduction Research Branch, NICHD, National Institutes of Health, Bethesda, Maryland 20892 and the ^¶ Department of Pharmacology, University of Occupational and Environmental Health, Japan School of Medicine, Tokyo 154-8567, Japan

Received for publication, May 29, 2002, and in revised form, September 16, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

P2X purinergic receptors (P2XRs) differ among themselves with respect to their ligand preferences and channel kinetics during activation, desensitization, and recovery. However, the contributions of distinct receptor subdomains to the subtype-specific behavior have been incompletely characterized. Here we show that homomeric receptors having the extracellular domain of the P2X₃ subunit in the P2X_2a-based backbone (P2X_2a/X₃ex) mimicked two intrinsic functions of P2X₃R, sensitivity to -methylene ATP and ecto-ATPase-dependent recovery from endogenous desensitization; these two functions were localized to the N- and C-terminal halves of the P2X₃ extracellular loop, respectively. The chimeric P2X_2aR/X₃ex receptors also desensitized with accelerated rates compared with native P2X_2aR, and the introduction of P2X₂ C-terminal splicing into the chimeric subunit (P2X_2b/X₃ex) further increased the rate of desensitization. Physical and functional heteromerization of native P2X_2a and P2X_2b subunits was also demonstrated. In heteromeric receptors, the ectodomain of P2X₃ was a structural determinant for ligand selectivity and recovery from desensitization, and the C terminus of P2X₂ was an important factor for the desensitization rate. Furthermore, [-³²P]8-azido ATP, a photoreactive agonist, was effectively cross-linked to P2X₃ subunit in homomeric receptors but not in heteromeric P2X₂ + P2X₃Rs. These results indicate that heteromeric receptors formed by distinct P2XR subunits develop new functions resulting from integrative effects of the participating extracellular and C-terminal subdomains.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

ATP and other purine nucleotides have widespread and potent extracellular actions on excitable and non-excitable membranes. Synaptic and hormonal messenger functions of extracellular purine nucleotides are mediated by two types of cell-surface P2 purinergic receptors (1). The P2X receptors (P2XRs)¹ are ligand-gated channels selectively permeable to cations, and the P2Y receptors are members of G protein-coupled heptahelical receptor superfamily (2). In addition to the fast excitatory synaptic signaling, P2XRs participate in control of slower biological processes, including smooth and cardiac muscle contraction, exocytosis, and blood cell functions (3-6). These cellular processes depend on calcium ions as a critical intracellular messenger. Activation of P2XRs leads to an increase in intracellular free calcium concentration ([Ca²⁺]_i) indirectly, through depolarization of plasma membrane and activation of voltage-dependent Ca²⁺ influx, in addition to Ca²⁺ entry through the pores of P2XRs (7). The cation-conducting pore of P2XRs is formed through multimerization of at least three subunits (8). Versatile expression patterns of seven P2X subunits identified so far (9) and their combinations through heteromeric multimerization in a single cell account for channel-specific calcium signaling patterns. Further diversity is produced by alternative splicing of primary transcript for some subunits, including P2X₂ (10-12). Finally, the active duration of ATP, a common agonist for P2XR, is under the control of tissue-specific ecto-ATPase activity, and modification of ATP analogs by this enzyme at the triphosphate moiety can severely influence the agonistic potencies (9).

The P2XR subunits are composed of two putative transmembrane domains critical for ion permeability, a large extracellular loop, and with N and C termini in cytoplasmic face (9). The identification of cDNAs for P2XR subunits in rat vas deferens and PC12 cells was followed by the discovery of additional subunit members and homologs in divergent mammalian species (2, 13). Their primary amino acid sequences exhibit a strong degree of conservation, especially within the extracellular regions. Ten cysteines in the extracellular region are well conserved, and sulfhydryl bonds are proposed to connect these cysteines (2). Recently, residues critical for ATP binding were localized in the extracellular loop near the first transmembrane domain of rat P2X₂Rs and at the well conserved lysine residue of P2X₁R (14, 15). These data confirmed that ATP interacts with the extracellular part of P2XR to induce conformational changes needed for receptor activation.

Recombinant P2XRs are divided into two groups according to their relative potency for ATP and its analog -methylene-ATP (-meATP). P2X₁R and P2X₃R exhibit high sensitivity for both ligands and rapidly desensitize, whereas ATP is more potent than -meATP in other receptors, including P2X₂R, P2X₄R, and P2X₇R, and generates slow or non-desensitizing Ca²⁺ signals (2, 16). The heteromeric assembly of two subunits, P2X₂ and P2X₃, in sensory neurons results in a slow desensitizing channel that is sensitive to -meATP (17, 18). Heteromultimerization of P2X₄/P2X₆ subunits and P2X₁/P2X₅ subunits is also functionally distinguished (19, 20). Furthermore, the intracellular subunit regions have modulatory effects on channel behavior, especially in controlling the rate of receptor desensitization. For example, a C-terminal deletion by alternative splicing and substitution of threonine at the N terminus individually accelerate desensitization of P2X₂Rs (11, 12, 21-23).

However, the contributions of other receptor subdomains to the generation of subtype-specific ligand selectivity and activation/desensitization properties have been incompletely characterized. Also, the effects of mutual interactions among different receptor subdomains on the overall channel activity have not been clarified. Here we studied a functional consequence of altering the C-terminal and extracellular structures of P2XRs, in terms of agonist sensitivity, time course of receptor desensitization, and recovery from desensitization. To this end, we used extracellular chimeric mutants between P2X₂ and P2X₃ subunits with full-length or spliced P2X₂ C terminus. The results of these investigations revealed an additive effect of intracellular and extracellular parts of P2XR subunits in regulating desensitization of homomeric and heteromeric receptors.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Construction of Chimeric Subunits-- Domain swapping between P2X₂ and P2X₃ subunits was conducted using the PCR-based overlap extension method as described by Horton et al. (24). To swap the extracellular domains, SacI and EcoRI sites were introduced into the coding sequences of P2X₂ and P2X₃ subunits. The PCR primers with silent nucleotide substitutions (underlined) are listed as follows: X2EcoL, 5'-AACATCGATTCGAATTCCATAGGCTTTGAT-3'; X3SacU, 5'-ATTGAGAGCTCAGTAGTTACAAAGGTG-3'; X3SacL, 5'-GCTCTCAATGGCGGTGTCCCTCACTTG-3'; X3EcoU, 5'-GGAATTCGCTTTGATGTGCTGGTA-3'; and X3EcoL, 5'-GCGAATTCCAAAAGCCTTCAGGAGTGT-3'. The intrinsic SacI site located in the coding sequence of extracellular domain of P2X_2a and P2X_2b subunits was used to facilitate the swapping of the extracellular sequence flanked by EcoRI and SacI sites. By using PCR conditions described previously (22), the open reading frames for P2X_2a, P2X_2b, and P2X₃ receptors were isolated and subcloned into HincII and SmaI sites of a pBluescript vector (Stratagene, La Jolla, CA) with the SacI site in the multiple cloning sites removed. The extracellular domain of the P2X₃ subunit (Val⁶⁰-Phe³⁰¹) flanked by EcoRI and SacI sites was excised and used to replace the corresponding region (Ile⁶⁶-Tyr³¹⁰) of P2X_2a and P2X_2b subunits. The chimeric constructs of P2X_2a and P2X_2b subunits with a P2X₃ subunit extracellular domain were designated as P2X_2a/X₃ex and P2X_2b/X₃ex, respectively. By using a similar strategy, the extracellular domain of P2X₂ subunit (Ile⁶⁶-Tyr³¹⁰) was also obtained and used to substitute the corresponding sequence (Val⁶⁰-Phe³⁰¹) of P2X₃ subunit. The resulting construct encoding a P2X₃ subunit with a P2X₂ subunit extracellular domain was designated as P2X₃/X₂ex (Fig. 1).

A series of extracellular chimeric mutants were generated using the restriction site-independent method as described previously (25). Briefly, the coding sequences for the extracellular regions of P2X_2a and P2X₃ subunits were subcloned into a pBluescript vector at SacI/EcoRI and KpnI sites, respectively, in a head-to-tail configuration. Two µg of the construct carrying two extracellular sequences was linearized by digesting it with ClaI and XhoI. The linearized plasmid was then directly used for transformation into competent DH5 Escherichia coli strain (Invitrogen). During transformation and recovery, processes for chimera formation involve partial exonuclease digestion of linearized plasmid and base pairing between exposed ends of two inserts, followed by bacterial repair to a single sequence and ligation to recircularize the chimera plasmid (25). Ampicillin-resistant bacterial colonies containing circular plasmids were screened for a series of chimeric constructs by restriction enzyme digestion and PCR. In these mutant extracellular domains, chimeric junctions were in regions of high sequence conservation between P2X_2a and P2X₃ subunits.

The coding sequences of these chimeric constructs were subsequently confirmed with DNA sequencing using a fluorescence-based sequencing kit (Amersham Biosciences). For functional expression of chimeric receptors, the coding sequences were excised by XhaI and KpnI digestion and subcloned into the eukaryotic expression vector pcDNA 3.1 (Invitrogen). In parallel studies, the coding sequences of the P2X₃ receptor and the newly constructed chimeric receptors were also subcloned into the bicistronic vector pIRES2-EGFP (Clontech, Palo Alto, CA). These pIRES2 constructs contain the internal ribosome entry site of viral origin for constitutive expression of GFP. The fluorescence signal of GFP was used to identify transfected GT1 cells for single cell [Ca²⁺]_i measurement.

Cell Culturing and Functional Expression Studies-- GT1 cells were cultured in Dulbecco's modified Eagle's medium and Ham's F-12 medium (1:1), supplemented with 10% fetal calf serum, 100 µg/ml ampicillin, and 100 µg/ml streptomycin. Transient transfection was performed as described previously (22) with slight modifications. Briefly, 1 million GT1 cells were plated on coverslips coated with poly-L-lysine and allowed to attach for 24 h. On the day of transfection, 3 µg of the plasmid DNA was mixed with 15 µl of Plus Reagent^TM and 3 µl of LipofectAMINE^TM (Invitrogen) in 3 ml of serum-free Opti-MEM medium, incubated for 15 min at room temperature, and then applied to cells. After 3 h of incubation, the medium was replaced with fresh culture medium. Single cell Ca²⁺ recordings were performed 24-48 h after transfection. In co-transfection experiments, equal amounts of two expression constructs were used. GT1 cells transfected with pcDNA 3.1 vector without insert did not show any detectable [Ca²⁺]_i response upon 100 µM ATP application.

Measurements of Ca²⁺ Ion Concentration-- Single cell [Ca²⁺]_i measurements were performed as described previously (22). Briefly, cells were incubated at 37 °C for 60 min with 2 µM fura-2 AM in phenol red- and ATP-free medium 199 with Hanks' salt solution, subsequently washed with assay buffer containing 137 mM NaCl, 5 mM KCl, 1.2 mM CaCl₂, 1 mM MgCl₂, 10 mM HEPES, pH 7.4, and 10 mM glucose, and kept for at least half an hour in this medium prior to measurements. Apyrase (grade I) was purchased from Sigma and used at 10 µg/ml throughout the incubation process. Coverslips with cells were mounted on the stage of an Axiovert 135 microscope (Carl Zeiss, Oberkochen, Germany) attached to the Attofluor Digital Fluorescence Microscopy System (Atto Instruments, Rockville, MD). [Ca²⁺]_i responses were examined under a 40× oil immersion objective during the exposure to alternating 340 and 380 nm light beams, and the intensity of light emission at 520 nm was measured. The ratio of light intensities, F₃₄₀/F₃₈₀, which reflects changes in [Ca²⁺]_i, was simultaneously followed in several single cells. Cells expressing fluorescence protein were optically detected by an emission signal at 520 nm when excited by 488 nm ultraviolet light and were not detectable by 340 or 380 nm excitations. In co-transfection experiments with pIRES vectors, about 75% of fluorescent protein-positive cells responded to agonist stimulation and were considered to be co-transfected. Lower co-transfection efficiency below this level was excluded from further analysis.

Expression in Xenopus laevis Oocytes and Electrophysiological Recordings-- Oocytes at stage V and VI were isolated from adult X. laevis as described previously (26) and placed in modified Barth's saline (MBS) consisting of 88 mM NaCl, 1 mM KCl, 10 mM HEPES, 0.82 mM MgSO₄, 2.4 mM NaHCO₃, 0.91 mM CaCl₂, and 0.33 mM Ca(NO₃)₂ at pH 7.5. The oocyte nuclei were directly injected with 1.5 ng of expression constructs for P2X_2a/X₃ex or P2X₃/X₂ex in 30 nl of injection buffer (88 mM NaCl, 1 mM KCl, 15 mM HEPES, pH 7.0). Injected oocytes were maintained for 3 days at 18 °C in sterilized incubation medium containing MBS plus 10 µg/ml streptomycin, 10 units/ml penicillin, 50 µg/ml gentamicin, and 2 mM sodium pyruvate. For electrophysiological recording, oocytes were placed in a rectangular chamber of 100 µl volume and perfused with Ba²⁺-Ringer's solution (115 mM NaCl, 2.5 mM KCl, 1.8 mM BaCl₂, and 10 mM HEPES, pH 7.4) at a rate of 2 ml/min. The oocytes were then impaled with two glass electrodes (0.5-10 megohms) filled with 3 M KCl and voltage-clamped at 70 mV using OC-725C Oocyte Clamp Amplifier (Warner Instruments, Inc., Hamden, CT). Currents were digitally recorded with PowerLab/200 and Chart software (ADInstruments, Grand Junction, CO). ATP was first dissolved in distilled water and then diluted in Ba²⁺-Ringer's solution immediately before use and was applied for 30 s. All measurements were performed at ambient temperature.

Immunological Detection of Epitope-tagged P2XRs-- Hemagglutinin (HA) epitope, YPYDVPDYA, was added to the C-terminal end of P2X₂, P2X₃, and chimeric subunits, and FLAG (FL) epitope, DYKDDDDK, was added to the N terminus of P2X_2a and P2X_2b. The 5'-primer sites for HA tagging corresponded to the nucleotide sequence 803-824 of P2X_2a (27) and 525-544 of the P2X₃ (17). The FL epitope was inserted between the initiative methionine residue and the second amino acid of the P2X₂ by PCR. The 5'-primer sequence contained 6 bases of XhoI site, optimized translational sequence (28), 3 bases for methionine, 24 bases encoding the 8-amino acid FL-peptide sequence, and 21 bases encoding 7 amino acids next to initiator methionine. The 3'-primer for FL tagging has nucleotide sequence 541-562 of P2X_2a. PCR were performed using P2X₂SE and P2X₃SE as templates, and amplified products were subcloned into pBluescript II vector for sequencing. Correctly tagged fragments were transferred to expression constructs using EcoRI/KpnI and XhoI/NarI for C-terminal and N-terminal substitutions, respectively.

Crude membranes were prepared from GT1 cells 24-48 h after transfection as follows. Cultured cells in 60-mm dishes were washed once with ice-cold PBS and collected in TE buffer (50 mM Tris-HCl, pH 7.4, containing 5 mM EDTA, 10 µg/ml aprotinin, 10 µg/ml leupeptin, and 0.1 mM phenylmethanesulfonyl fluoride). Cells were homogenized on ice, and membrane fractions were collected by centrifugation at 10,000 × g for 10 min at 4 °C. The pellets were then lysed in the TE buffer containing 1% Triton X-100 on ice for 30 min and centrifuged at 30,000 × g for 30 min. The supernatant and antibody for immunoprecipitation were incubated for 1 h at 4 °C and additionally for 2 h with protein G. The immunocomplexes were washed four times with the lysis buffer, denatured, and subjected to gel electrophoresis. After separation, protein was electrically transferred to polyvinylidene fluoride membranes. The blots were incubated with 3% bovine serum albumin in TBS (10 mM Tris-HCl, pH 7.5, and 150 mM NaCl) and then with monoclonal anti-HA antibody (Babco, Richmond, CA) at a dilution of 1:3500 or anti-FL M2 antibody (Eastman Kodak Co.) at 1:2000 dilution. For secondary antibody, peroxidase-conjugated anti-mouse antibody was diluted 1:5000, and signals were visualized with enhanced chemiluminescence ECL (Amersham Biosciences). Protein concentration in the membrane protein samples was determined using the Pierce BCA protein assay (Pierce).

For biotinylation of cell-surface protein, cells were treated with 0.5 mg/ml N-hydroxysulfosuccinimide-LC-biotin (Pierce) in assay buffer for 30 min at ambient temperature. The reaction was terminated by washing cells once with assay buffer containing 50 mM ammonium chloride, followed by two washings with assay buffer only. Cells were then lysed with TE buffer containing 1% Triton X-100 and subsequently immunoprecipitated with antibody against epitope. Biotinylated protein on the membrane was detected by peroxidase-conjugated streptavidin (Amersham Biosciences) at 1:5000 dilutions and visualized.

Photoaffinity Labeling of P2XR-- Cells grown in 60-mm dishes were transfected with expression constructs for FL-tagged P2X_2a and HA-tagged P2X₃ as described above. Twenty four hours after transfection, cells were washed with ice-cold phosphate buffer and incubated in assay buffer containing 1 µM [-³²P]8-azido-ATP (370 GBq/mmol, ICN, Costa Mesa, CA) on ice for 10 min. UV irradiation was performed at 254 nm wavelength with 4 milliwatts/cm² for 3 min. After washing with phosphate buffer to remove unlabeled ligand, cells were collected in 0.5 ml of lysis buffer and subjected for immunoprecipitation with anti-FL antibody. Equal amounts of protein samples (20 µg) were denatured, loaded onto 7.5% SDS-polyacrylamide gel, and visualized by autoradiography using x-ray film (Eastman Kodak).

Calculations-- Where appropriate, the results were expressed as means ± S.E. The time course of [Ca²⁺]_i signaling was fitted to a single exponential function, and dose-response curves obtained by [Ca²⁺]_i measurements were fitted to three-parameter logistic function using a non-linear curve-fitting program (Igor, WaveMetrics, Lake Oswego, OR). Significant differences were determined by either Student's t test or one-way analysis of variance followed by Scheffe's test, if applicable, and p < 0.05 was considered as significantly different.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Effect of Extracellular Chimeric Mutations on Agonist Specificity-- To analyze the importance of extracellular P2XR domains in calcium signaling, three native subunits, P2X_2a, P2X_2b, and P2X₃, and two chimeric subunits, P2X_2a/X₃ex and P2X_2b/X₃ex (Fig. 1), were individually expressed in GT1 cells, and agonist-induced [Ca²⁺]_i responses were monitored in single cells. As shown in Fig. 2, A and B, 10 µM -meATP evoked only a small rise in [Ca²⁺]_i in P2X_2aR- and P2X_2bR-expressing cells, whereas the subsequent stimulation with 100 µM ATP resulted in a rapid and large increase in [Ca²⁺]_i. The declining rates in [Ca²⁺]_i toward steady levels were faster in P2X_2bR-expressing cells than in cells expressing P2X_2aR. The differences in the macroscopic channel kinetics judged from these single cell [Ca²⁺]_i recordings correlate well with the structural differences of the putative C-terminal regions and are in accord with previously published data (21) using [Ca²⁺]_i and current measurements.

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Schematic representation of extracellular chimeric P2XR subunits. The putative extracellular regions Val60 to Phe300 of P2X3 and Ile66 to Tyr310 of P2X2 were mutually exchanged to make extracellular chimeric subunits termed P2X2a/X3ex, P2X2b/X3ex, and P2X3/X2ex. The difference between P2X2a and P2X2b is in the C-terminal region, where P2X2b lacks a stretch of 69 amino acids achieved by an alternative splicing reaction. In P2X2a/X3V60-R180, the N-terminal half of the P2X3 extracellular region was transferred to the equivalent part of P2X2a subunit.

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Sequential stimulation of P2X2aR, P2X2bR, and P2X3R by -meATP and ATP. Transfected cells were optically identified by the GFP-derived fluorescence signals, and P2X2aR (A), P2X2bR (B), and P2X3R (C) were stimulated by the sequential application of 10 µM -meATP and 100 µM ATP. In this and the following figures, each trace represents the mean value from five to eight single cell recordings, and experiments were repeated in three to six preparations. Dye-loading solutions with (filled circles) or without (open circles) 10 µg/ml apyrase were indicated at the end of the recordings.

In contrast to P2X₂R-expressing cells, ATP and -meATP were unable to elevate the [Ca²⁺]_i in P2X₃R-expressing cells when recordings were done under identical experimental conditions. Because neurons frequently secrete ATP, which in turn could desensitize P2XRs, we evaluated this hypothesis by preincubating GT1 cells with 10 µg/ml apyrase, an ecto-ATPase (22). After the incubation with apyrase, stimulation of P2X₃Rs with 10 µM -meATP resulted in a rapid and transient increase in [Ca²⁺]_i (Fig. 2C) (n = 118). To fully regain agonist responsiveness in P2X₃R-expressing cells, incubation in ecto-ATPase containing buffer was necessary for at least 20 min. On the other hand, this pretreatment had no apparent effect on P2X_2aR and P2X_2bR desensitization rates (Fig. 2, A and B).

The intrinsic characteristics of P2X₃Rs, high sensitivity to -meATP and rapid desensitization, were assessed in two extracellular chimeric receptors, P2X_2a/X₃ex and P2X_2b/X₃ex, in which the extracellular region of P2X₃ subunit was transferred to the analogous parts of P2X_2a and P2X_2b subunits (Fig. 1). When expressed as homomeric channels, both chimeric receptors responded to application of 10 µM -meATP with a rapid increase in [Ca²⁺]_i, but only after preincubating cells with apyrase (Fig. 3, filled circles versus open circles). In GFP-positive cells transfected with P2X_2a/X₃ex or P2X_2b/X₃ex, 92 (n = 65) and 86% (n = 76) of the cells also responded to ATP, respectively, whereas no apparent increase in [Ca²⁺]_i was observed in both cell types without apyrase treatment (82 cells with P2X_2a/X₃ex and 68 cells with P2X_2b/X₃ex). Therefore, we considered that these two P2X₃R-specific characters are largely dependent on the structure of the ectodomain and could be transferred to the extracellular mutants when expressed as homomeric channels.

View larger version (21K): [in this window] [in a new window]   Fig. 3.   Apyrase and -meATP sensitivity of chimeric P2XR. Cells expressing P2X2a/X3ex and P2X2b/X3ex subunits were stimulated with 10 µM -meATP after 1 h of preincubation with apyrase (filled circles) or solvent (open circles).

The desensitization rate of [Ca²⁺]_i signals induced by ATP was significantly accelerated in chimeric receptors compared with native P2X_2aRs and P2X_2bRs (Fig. 4A). The calculated time constants of signal desensitization from these measurements were as follows (in 10³/s): 4 ± 0.17 versus 12 ± 1.3 for P2X_2aRs (n = 41) and P2X_2a/X₃ex (n = 40), respectively, and 20 ± 2.0 versus 23 ± 7.1 for P2X_2bRs (n = 37) and P2X_2b/X₃ex (n = 47), respectively (Fig. 4B). Although the desensitization rates for both mutant receptors significantly increased by the substitution of extracellular region in P2X₂ subunits, this modification alone was not sufficient to mimic the rapid desensitization rate of homomeric P2X₃Rs. This suggests that the other subunit domains, including the transmembrane domains and the C terminus, may also participate in desensitization, as already suggested (21, 29, 30). Notably, the modulatory effects of C-terminal splicing and extracellular chimeric mutation on the rate of receptor desensitization were additive; both C-terminal splicing and transfer of extracellular domain increased the desensitization rate seen in P2X_2b/X₃ex-expressing cells.

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Desensitization patterns of homomeric channels during continuous stimulation with 100 µM ATP. A, the declining phase of [Ca2+]i signals was fitted to a single exponential decay curve. B, the time constants (in 103/s) of signal decay were calculated from three to five experiments performed with at least 15 cells in each measurement. *, p < 0.05, when wild-type and extracellular mutant were compared.

Effects of Extracellular and C-terminal Domains on Heteromeric Channel Functions-- In general, functional P2X channels are formed by homomeric and heteromeric multimerization, depending on subunits expressed in single cells. In further experiments, we used receptors in heteromultimeric configurations to investigate the interactive effects of extracellular and intracellular domains on agonist potency and receptor desensitization. Co-expression of wild-type P2X_2a and chimeric P2X_2a/X₃ex subunits in single cells increased sensitivity to -meATP compared with the cells expressing homomeric P2X_2aRs (Figs. 2A and 5A). Furthermore, in contrast to homomeric P2X_2a/X₃exRs, 10 µM -meATP-induced [Ca²⁺]_i response by heteromeric P2X_2a/X₃ex + P2X_2a receptors was not dependent on preincubation of cells with apyrase. Because [Ca²⁺]_i responses induced by 10 µM -meATP were small in cells expressing homomeric P2X_2a and not detectable in cells expressing homomeric P2X₃R when incubated without apyrase, it was reasonable to conclude that a newly developed feature of -meATP-sensitive but apyrase-insensitive [Ca²⁺]_i response in co-transfected cells was largely due to activation of heteromeric channels.

We also compared the pattern of -meATP-induced [Ca²⁺]_i signaling in co-transfected cells with different sets of wild-type and chimeric subunits (Fig. 5A). Heteromeric P2X_2a + P2X_2a/X₃ex receptors showed slow desensitization, whereas the channels formed by P2X_2b + P2X_2b/X₃ex subunits desensitized in a remarkably strong manner (Fig. 5B). Physical associations between chimeric and naturally occurring P2X₂ subunits were confirmed using a co-immunoprecipitation method (Fig. 5C). In this and following experiments, the immunological detection of channel molecule was done using the epitope-tagged subunits at either the N or C terminus, a procedure that had no significant effect on agonist-induced [Ca²⁺]_i (Table I) and current (31) responses. The calculated molecular weight of P2X subunit monomers from Western blot was consistently larger than predicted from the polypeptide sequence. For example, P2X_2a and P2X_2b subunit resulted in 70- and 59-kDa bands (Fig. 5, C-E), whereas the expected mass was 53 and 45 kDa, respectively. Thus, P2XR subunits expressed in GT1 cells were considered to undergo extensive glycosylation during their post-translational process.

View larger version (46K): [in this window] [in a new window]   Fig. 5.   Co-expression of P2X2 and chimeric subunits. A, equal amounts of the expression constructs for wild-type and extracellular chimeric subunits were used for each transfection. Cells expressing P2X2a (upper) and P2X2b (lower) subunits together with the corresponding chimeric mutants were stimulated with 10 µM -meATP (open triangles) and subsequently with 100 µM ATP (arrows). In this experiment, cells were not treated with apyrase. B, calculated time constants for desensitization of heteromeric channels. Decay curve of 10 µM -meATP-induced [Ca2+]i signals was fitted to single exponential function, and time constants were calculated. Significant differences of desensitization rates between cells co-expressing P2X2a + P2X2a/X3ex and each one of the other transformants were indicated as * (p < 0.05) and ** (p < 0.01). C, heteromeric assembly of P2X2 and chimeric subunits. Cells co-transfected with FL-tagged P2X2 subunits and HA-tagged mutant subunits were co-immunoprecipitated with anti-FL antibody, and blots were probed with anti-HA (upper panel) and anti-FL (lower panel) antibodies. D, heteromeric assembly of P2X2a and P2X2b subunits. FL-tagged P2X2a was co-expressed with HA-tagged P2X2a or P2X2b subunits. Cell lysates was subjected to immunoprecipitation with anti-FL antibody, and blot was probed with anti-HA (upper panel) and anti-FL (lower panel) antibodies. E, homo- and heteromeric assembly of P2X2 subunits located in the plasma membrane. GT1 cells expressing tagged-P2X subunits were treated with plasma membrane-impermeable biotinylation reagents, immunoprecipitated with anti-FL antibody, separated on SDS-PAGE, blotted onto polyvinylidene difluoride membrane, and probed with peroxidase-conjugated streptavidin. Kd, molecular mass markers.

Both mutations, at the extracellular regions and C-terminal domain, did not perturb association between wild-type and chimeric P2X₂ subunits (Fig. 5C), as suggested previously (32). The actual association between P2X_2a and P2X_2b subunits was also confirmed (Fig. 5D) and possibly occurred at the plasma membrane, because the subunits consisting of cell-surface receptors were modified with the plasma membrane-impermeable biotinylation reagents and were co-immunoprecipitated (Fig. 5E). These results suggest that in cells expressing P2X_2a and P2X_2b subunits, pituitary somatotrophs for example (10), their heteromultimeric association could generate a functional channel that desensitizes faster than the homomeric P2X_2aRs.

Localization of -meATP and Apyrase Sensitivities within the Ectodomain of P2X₃ Subunit-- To identify the regions in the extracellular domain of P2X₃ subunit responsible for -meATP and apyrase sensitivities, a series of extracellular chimeras between P2X₃ and P2X_2a subunits were prepared. The following amino acid regions of P2X_2a were replaced to the corresponding extracellular parts of P2X₃: 66-82, 66-101, 66-192, 66-273, 66-296, 98-310, 159-310, 192-310, 274-310, and 295-310. When expressed in GT1 cells individually or together with P2X_2a, however, all these mutant receptors were not functional except one, termed P2X_2a/X₃V⁶⁰-R¹⁸⁰, in which the Ile⁶⁶ to His¹⁹² sequence of P2X_2a was replaced with Val⁶⁰ to Arg¹⁸⁰ sequence of P2X₃ (Fig. 1). Cells expressing this particular mutant responded to 10 µM -meATP with a rapid increase in [Ca²⁺]_i and slow desensitization, similar to those observed in P2X_2aR-expressing cells stimulated by ATP (Fig. 6). Furthermore, agonist-induced [Ca²⁺]_i response was not dependent on preincubation with apyrase (Fig. 6). Therefore, the two specific features of P2X₃Rs, -meATP sensitivity and apyrase-dependent recovery from desensitization, are separately localized to their responsible extracellular regions; the N-terminal half of the extracellular loop accounts for high potency of -meATP and the C-terminal half for recovery from desensitization.

View larger version (15K): [in this window] [in a new window]   Fig. 6.   Agonist-specific pattern of [Ca2+]i response in P2X2a/X3V60-R180-expressing cells. The apyrase pretreatment was not required for -meATP-induced [Ca2+]i response in cells expressing homomeric P2X2a/X3V60-R180 channels. The sequential stimulations using 10 µM -meATP and 100 µM ATP were performed to access relative potency of two agonists in indicated dosages.

Restoration of Ca²⁺ Signaling Function by Heteromultimerization of Extracellular Mutants-- The effect of mutual substitution at the extracellular region was further examined using P2X₃/X₂ex subunits, in which extracellular loop of P2X₂ subunit was transferred to the equivalent region of P2X₃ (Fig. 1). When expressed in X. laevis oocytes, mutant P2X₃/X₂ex subunits formed a rapidly desensitizing channel (Fig. 7A). However, the amplitude of ATP-induced inward current was less than half compared with that of P2X_2a/X₃ex. The functional expression of channels was always less successful in P2X₃/X₂ex compared with P2X_2a/X₃ex. In GT1 cells transfected with P2X₃/X₂ex subunit, [Ca²⁺]_i response was not detectable in GFP-positive cells (n = 91), although the amount of subunit protein expressed was comparable with that of P2X_2a/X₃ex. Furthermore, the fluorescent signal from HA-tagged subunit was detected from plasma membrane in immunohistochemical studies, indicating cell-surface localization of the subunit (not shown).

View larger version (21K): [in this window] [in a new window]   Fig. 7.   Characterization of agonist-induced current and [Ca2+]i signals in cells co-expressing extracellular mutants. A, Xenopus oocytes expressing P2X2a/X3ex (left trace) and P2X3/X2ex subunit (right trace) demonstrate inward current upon the application of 100 µM ATP for 30 s. B, HA-tagged P2X2a/X3ex and P2X2b/X3ex subunits were co-expressed with FL-tagged P2X3/X2ex and immunoprecipitated with an anti-FL antibody and probed with an anti-HA antibody. Kd, molecular mass markers. C, the P2X3/X2ex subunit was co-transfected into GT1 cells together with P2X2a/X3ex subunit (n = 32), P2X2b/X3ex (n = 48), or empty vector (n = 50) and sequentially stimulated with 20 µM -meATP and 100 µM ATP. Representative tracings were presented from average of five to nine single cell responses from five separate experiments.

When both chimeric subunits were co-transfected, a new pattern of [Ca²⁺]_i signaling was observed. Stimulation of the co-transfected cells with 20 µM -meATP resulted in full receptor activation that was independent of apyrase pretreatment, and the subsequent application of 100 µM ATP was ineffective (Fig. 7C). Therefore, the co-expression of two extracellular chimeras, P2X_2a/X₃ex and P2X₃/X₂ex, resulted in [Ca²⁺]_i signals similar to those of heteromultimeric P2X₂ and P2X₃ channels (17, 18). In addition, the subunits with spliced C termini accelerated the desensitization rates in heteromeric channels formed by P2X_2b/X₃ex + P2X₃/X₂ex (Fig. 7, B and C) and by P2X_2b + P2X₃ (Fig. 8A). As shown in Fig. 8, B and C, the biotinylated subunits were co-immunoprecipitated from cell lysate, indicating physical associations between P2X₃ and either P2X_2a or P2X_2b at the cell surface.

View larger version (19K): [in this window] [in a new window]   Fig. 8.   Heteromeric assembly of P2X2 and P2X3 subunits. A, the P2X3 subunit was co-expressed with either P2X2a or P2X2b subunit, and cells were stimulated with 10 µM -meATP. Decay curves of the [Ca2+]i signaling were fitted to single exponential function. Horizontal bars show the calculated time constants (in 103/s). *, p < 0.05. B, HA-tagged P2X3 subunit was co-immunoprecipitated with FL-tagged full-length or spliced P2X2 subunits, and blots were probed with anti-HA (upper) and anti-FL (lower) antibodies. C, co-immunoprecipitation of cell surface P2X2 and P2X3 subunits. GT1 cells co-expressing FL-tagged-P2X2 and HA-tagged P2X3 subunits were treated with amine-reactive and plasma membrane-impermeable biotinylation reagents, immunoprecipitated with the anti-FL antibody, separated on SDS-PAGE, blotted, and probed with peroxidase-conjugated streptavidin. Kd, molecular mass markers.

Photoaffinity Cross-linking of P2XR by [-³²P]8-Azido-ATP-- In further experiments, we characterized agonist potency of 8-azido-ATP and its ability to label recombinant P2XRs by photoaffinity cross-linking. When stimulated with equimolar (100 µM) concentration, P2X₃R-expressing cells showed preference for 8-azido-ATP over ATP, whereas P2X_2aR-expressing cells showed the opposite preference (Fig. 9A). In cells co-transfected with P2X_2aR and P2X₃R, application of 10 µM 8-azido-ATP without apyrase treatment resulted in no detectable changes in [Ca²⁺]_i, and the subsequent application of 10 µM -meATP induced a rapid and large [Ca²⁺]_i response (Fig. 9B). These results indicate that 8-azido-ATP is a preferred agonist for homomeric P2X₃Rs over heteromeric channels formed by the P2X_2a and P2X₃ subunits, although -meATP effectively activates both channels.

View larger version (14K): [in this window] [in a new window]   Fig. 9.   Agonist activity of and photoaffinity labeling by 8-azido-ATP in P2XR-expressing cell. A, cells expressing P2X3Rs (upper trace) and P2X2aRs (lower trace) were stimulated with 100 µM 8-azido-ATP (arrows) and 100 µM ATP (open triangles) after preincubation with apyrase. B, cells co-expressing P2X2a and P2X3 subunits were stimulated with 10 µM 8-azido-ATP (arrows), 10 µM -meATP (filled triangles), and 100 µM ATP (open triangles). In this experiment, cells were not treated with apyrase. C, FL-tagged P2X2a and HA-tagged P2X3 subunits expressed were subjected to photoaffinity labeling with [-32P]8-azido-ATP as described under "Experimental Procedures." Notice that homomeric P2X3Rs were efficiently labeled, whereas homomeric P2X2aRs or heteromeric channels were less efficient substrates for labeling.

This property of 8-azido-ATP was further confirmed by the effective photoaffinity labeling of P2X₃Rs (Fig. 9C). Interestingly, in cells co-transfected with P2X_2a and P2X₃, P2X₃ subunits joining a heteromeric channel and co-immunoprecipitated with P2X_2a, as shown in Fig. 7B, were not labeled efficiently with [-³²P]8-azido-ATP compared with the signal from homomeric P2X₃ subunits (Fig. 9C). This indicates that 8-azido-ATP can distinguish conformation of the ATP-binding pocket formed by homomeric P2X₃Rs from those formed by heteromeric P2X_2a + P2X₃ receptors. Preincubation of cells with 100-fold excess concentration of ATP resulted in a total inhibition of photolabeling (data not shown).

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In this study, we evaluated the combined effects of structural changes introduced into the extracellular and C-terminal regions of the P2XR subunit on agonist selectivity, receptor desensitization, and recovery from desensitization. ATP is a common agonist for all homomeric and heteromeric P2XRs, and modification at the triphosphate moiety of this molecule serves as an important determinant for P2X subtype selectivity. The substitution of bridging oxygen between - and -phosphorous with a methylene group resulted in an agonist (-meATP) that is equipotent for ATP for P2X₃R and P2X₁R and practically ineffective for other channels. On the other hand, modification in the 2'- and 3'-positions of ribose in trinitrophenyl-ATP made a selective antagonist against these subunits (33). High sensitivity of P2X₃R to -meATP can be transferred to P2X_2aRs and P2X_2bRs by generating the extracellular chimeras. Therefore, the responsible counterpart in P2X₃ subunit that functionally interacts with the polyphosphate chain in ATP is of importance for understanding the structure-activity relationship between ligand and receptor. We found that the N-terminal half of P2X₃ ectodomain, from Val⁶⁰ to Arg¹⁸⁰, was necessary for receptor activation by -meATP. The attempt to further narrow this region was hampered, presumably due to highly vulnerable feature of extracellular domains to modification by site-directed mutations. Thus, despite the conserved amino acid sequences in the extracellular domains of P2XR subunits, agonist potency is highly susceptible to subtle structural modifications.

In addition to altered preference for agonists, the desensitization rates were significantly accelerated in chimeric P2XRs. A comparison of time constants for desensitization between chimeric and wild-type P2X₂Rs revealed that extracellular substitution and C-terminal splicing at P2X₂ subunit had additive effects. Among the subunits examined, the P2X_2b/X₃ex channel showed the strongest and wild-type P2X_2a the weakest desensitization. Therefore, extracellular and C-terminal domains might have separate molecular mechanisms for controlling the desensitization rate. Furthermore, the rate of P2X_2b/X₃ex desensitization did not reach that of native P2X₃Rs. As shown previously (23, 29), the pore-forming transmembrane domain and the cytoplasmic N terminus of P2X_2aRs also contributed to the desensitization process. In particular, the N-terminally located Thr¹⁸ in the consensus sequence for phosphorylation by protein kinase C is constitutively phosphorylated in P2X_2aRs, and removal of this residue by site-directed mutagenesis converted a lasting receptor activity to fast desensitization. Therefore, intracellular events, in addition to primary structure of P2X subunits, can influence the desensitization process. The summation of modulatory effects from each subdomain can also determine the rate of receptor desensitization and the duration of P2XR-derived currents and [Ca²⁺]_i signals.

The somatotroph population of secretory anterior pituitary cells expresses both P2X_2a and P2X_2b subunits (10, 34, 35). The P2X_2b subunits form homomeric channels, which desensitize faster than P2X_2aRs. Reducing the amount of mRNA available for full-length P2X_2a by splicing is probably one of the mechanisms by which cells limit the number of mature P2X_2aRs and thus the excessive Ca²⁺ entry during receptor activation. The other post-transcriptional mechanism to achieve the same goal is heteromultimerization of P2X_2a and P2X_2b subunits. Because homomeric and heteromeric P2X₂Rs are indistinguishable from each other in pharmacological terms and both channels are present in single cells, it is difficult to quantitate the effect of spliced C terminus through the pattern of Ca²⁺ influx.

Here we used the extracellular region of the -meATP-sensitive P2X₃ as a marker for studies on the impact of heteromultimerization on [Ca²⁺]_i signaling pattern. A series of heteromers formed between wild-type and extracellular chimeric subunits enabled us to quantitate the contribution of spliced C terminus to the desensitization rates of receptors in heteromeric configuration. As the number of subunits with spliced C termini increases in co-transfection experiments, the -meATP-induced [Ca²⁺]_i signals became shorter (Fig. 5, A and B). Therefore, it is likely that heteromers formed by P2X_2a and P2X_2b would desensitize faster than homomeric P2X_2aRs and slower than homomeric P2X_2bRs. The actual association of these two isoform subunits in the plasma membrane was also demonstrated.

The transfer of P2X₂ extracellular domain to the corresponding part of P2X₃ resulted in P2X₃/X₂ex subunits that were less efficiently expressed as homomeric channels. However, the heteromers between two chimeric subunits, in which the extracellular domains were mutually exchanged, regained agonist sensitivity and preference to -meATP. Such heteromers desensitized with moderate rates and rapidly recovered from desensitization, as judged from responsiveness independent of ecto-ATPase treatment. These new characters of heteromeric P2X_2a/X₃ex and P2X₃/X₂ex channels were similar to those of the naturally occurring heteromeric channel between P2X_2a and P2X₃. Interestingly, the pore-forming transmembrane domains of mutant heteromers are exactly same as that of naturally occurring heteromers between P2X_2a and P2X₃, whereas relative positions of the -meATP-sensitive extracellular domains to the transmembrane region were exchanged. We may speculate that the effect of -meATP stimulation, leading to the allosteric conformational change required for activation and desensitization of channels, might pursue the same course in heteromeric channels made by P2X₃ + P2X_2a and by P2X_2a/X₃ex + P2X₃/X₂ex.

Our results indicate that heteromeric channels can be composed from any combinations of two subunits among P2X_2a, P2X_2b, and P2X₃ when expressed in a single cell and can exhibit particular desensitization kinetics determined by the nature of participating subunits. In accordance with this, P2X_2a and P2X₃ subunits are also able to form functional heteromers in small diameter sensory neurons (17, 18). As shown here, the C-terminal splicing of P2X₂ did not affect the ability of this subunit to form heteromers with P2X₃ or P2X_2a. Pituitary cells express comparable levels of P2X_2a and P2X_2b transcripts, and the calcium-signaling pattern by native channels resembles that observed in cells co-transfected with P2X_2a and P2X_2b (10). On the other hand, the transcripts for P2X_2b subunit produced in human spinal cord are hardly detectable by RT-PCR analysis (36) and at the level of detection in rat dorsal root ganglion (12), indicating that not all cells use C-terminal splicing leading to heteromeric channels.

In heteromeric configuration, the contribution of P2X₃ subunit to the ATP-binding pocket seems different from that in homomeric channels. We demonstrated efficient photoaffinity labeling of homomeric P2X₃ channels by radiolabeled 8-azido-ATP. Because 8-azido-ATP is an agonist for P2X₃Rs, photoaffinity labeled channels are likely to be in a desensitized state. Agonist potency examined in single cell [Ca²⁺]_i measurements indicated that preincubation of 10 µM 8-azido-ATP with cells expressing heteromeric P2X_2a and P2X₃ channels did not alter the apparent amplitude induced by -meATP. These results suggested that heteromeric channels did not undergo detectable levels of desensitization during incubation with 10 µM 8-azido-ATP. Therefore, it is likely that 8-azido-ATP can detect structural differences at the ATP-binding pockets formed by the homomeric and heteromeric channels.

In summary, the extracellular domain of P2X₃ and the spliced C terminus of P2X_2b can additively accelerate channel desensitization when these subunits form heteromeric channels. The relative position of these two modulatory receptor domains to the cation-permeable pore formed by transmembrane domains is exchangeable between channel-forming subunits, suggesting a symmetrical configuration of subunits around the pore. In cells expressing wild-type and spliced P2X₂ subunits together, regulated Ca²⁺ influx is accomplished by the two kinds of post-transcriptional modification mechanisms, C-terminal splicing and heteromeric assembly between full-length and spliced subunits. Channel function can be further regulated by phosphorylation and alteration of receptor localization. The consequence of more than one regulatory factor present at same time seems to be a subject for the future P2XR studies.

	ACKNOWLEDGEMENT

We thank Dr. Melanija Tomic for initial [Ca²⁺]_i measurements.

	FOOTNOTES

^* The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^§ Supported by Japan Health Sciences Foundation.

^** To whom correspondence should be addressed: Dept. of Molecular, Cell Pharmacology, National Research Institute for Child Health and Development, 3-35-31, Taishido, Setagaya-Ku, Tokyo 154, Japan. Tel.: 81-3-3419-2476; Fax: 81-3-3419-1252; E-mail: gtsujimoto@nch.go.jp.

Published, JBC Papers in Press, October 1, 2002, DOI 10.1074/jbc.M205274200

	ABBREVIATIONS

The abbreviations used are: P2XRs, ligand-gated purinergic receptor-channels; -meATP, -methylene ATP; GFP, green fluorescent protein; HA, hemagglutinin; [Ca²⁺]_i, intracellular free calcium concentration: FL, FLAG.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES