Blockade of Human P2X₇ Receptor Function With a Monoclonal Antibody

P2X RECEPTORS ARE ligand-gated ion channels that are activated by extracellular ATP. Their activation results in the opening of a cationic channel with significant permeability to calcium and intracellular depolarization.1,2 In contrast to other P2X receptors (P2X_1-63), P2X₇ is uniquely bifunctional. When stimulated briefly by low concentrations of agonist, the receptor acts as a nonselective cation channel. However, repeated or prolonged application of higher agonist concentrations, especially in solutions containing low concentrations of extracellular divalent cations, creates a much larger aqueous pore. Formation of this pore allows entry of fluorescent DNA binding dyes such as YO-PRO-1 (629 Daltons) and eventually leads to cell lysis. These same responses to ATP have been shown for native P2 receptors expressed by mast cells, macrophages, or microglia and were previously referred to as P2Z receptors.4 Among all ligand-gated ion channels, P2X₇ receptors thus have two remarkable features: they appear only in the immune system, and there they can mediate ATP-induced cell death.

Investigation of P2X₇ receptors in the immune system has suggested a potentially important role in immune responses. In either macrophages or microglial cells, P2X₇ receptors are functionally upregulated by lipopolysaccharide (LPS) or interferon-γ (γ-IFN).5-7 Stimulation of P2X₇ receptors leads to the release of mature interleukin-1β (IL-1β) in macrophages8 and microglial cells9 and the induction of phospholipase D activity as demonstrated in THP-1 cells.7 Recent evidence in microglial cells has shown an unusual p65 homomeric form of NFκB produced by P2X₇activation, suggesting a unique transcriptional activation pathway.10 Finally, the extracellular ATP-induced killing of mycobacteria in infected human macrophages has been shown to be mediated by P2X₇ receptors.11 Whether this killing of vesicle encapsulated bacteria is related to the cell fusion observed in macrophage cultures expressing high levels of P2X₇ receptors is currently unknown.12 

The functional study of P2X receptors has been hindered by the relative absence of good subtype specific antagonists (see discussion). We describe here a monoclonal antibody (MoAb) to human P2X₇that is both species and subtype specific and that has been found, unexpectedly, to functionally antagonize the activation of both recombinant and endogenous P2X₇ receptors by extracellular ATP.

Human P2X₇2 was expressed in a Balb/c mouse myeloma cell line, XS63 (ATCC TIB-17), by stable transfection. Balb/c mice were immunized on days 0, 7, and 28 subcutaneously in the limbs and behind the neck with 10⁷ transfected cells per injection in MPL+TDM emulsion (RIBI; Inotech, Dottikon, Switzerland). Three days after the final injection, the draining lymph nodes were obtained and the tissue was digested using a DNase and collagenase cocktail as reported elsewhere.13 The resulting cell suspension was resuspended at 10⁶ cells/mL and fused with Sp2 myeloma cells using a standard protocol.14 The hybridomas were selected in HAT medium, and 7 to 10 days after fusion, the supernatants were harvested for differential screening by flow cytometry on the transfected and nontransfected XS63 cells. Briefly, cells were washed with FACS buffer (1% bovine serum albumin [BSA] and 0.01% Na Azide in phosphate-buffered saline [PBS]) and successively incubated for 30 minutes with 50 μL supernatant, followed by washing, and a fluorescein isothiocyanate (FITC)-labeled sheep antimouse F(ab′)₂ fragment (Silenius Laboratories, Hawthorn, Australia) diluted 1/100 in FACS buffer. Mean fluorescence intensity was measured using a FACSCalibur (Becton Dickinson, Erembodeggen, Belgium). A similar method was used to investigate the effects of the antibody on HEK-293 cells transfected with human P2X₁15 or human P2X₄.16 Antibodies were purified by chromatography on Protein A Sepharose Fast Flow in PBS and eluted in 0.1 mol/L citrate, pH 4.5. Eluates were then subjected to gel filtration on Superdex-200 (Pharmacia, Uppsala, Sweden) equilibrated in PBS.

In some experiments, FITC-labeled MoAb was used to investigate the specificity of the MoAb. Briefly, wild-type HEK293 cells (5 × 10⁵ cells per well) or HEK293 cells expressing hP2X₃ or hP2X₇ receptors were incubated with the FITC-labeled MoAb for 6 hours in PBS containing 1% BSA. After 3 washes in PBS, cell-associated fluorescence was measured using spectrofluorimeter. Specific fluorescence signals were obtained by subtraction of relative fluorescence units (RFU) obtained in wild-type cells from those determined from the transfected cells.

Human monocytes were isolated by one-step Ficoll gradient separation6 and selected by forward and side scatter profiles by flow cytometry (FACSVantage; Becton Dickinson). The purified monocytes were then stimulated for 1, 2, and 3 days with LPS (1 μg/mL) or γ-IFN (10 ng/mL).

XS63 cells (5 × 10⁶), transfected with hP2X₇ or vector alone, were resuspended in PBS at 4°C and 40 μL biotinylation reagent (Amersham, Buckingham, UK) was added for 20 minutes with mixing. Cells were washed three times with PBS and lysed by the addition of cold extraction buffer (1% Triton X-100, 20 mmol/L Tris-HCl, pH 7.4, 150 mmol/L NaCl, 1 mmol/L CaCl₂, and 1 mmol/L MgCl₂) in the presence of protease inhibitors (4 μmol/L phenylmethylsulfonyl fluoride, 2 μg/mL pepstatin, 2 μg/mL leupeptin, 2 μg/mL trypsin inhibitor, and 2 μg/mL aprotinin; Sigma, St Louis, MO). After 20 minutes, the extract was centrifuged at 14,000 rpm for 10 minutes at 4°C. The resulting supernatant was incubated with 6 μg MoAb plus a mixture of protein A and protein G beads (Pharmacia, Uppsala, Sweden) on a roller mixer for 16 hours at 4°C. The beads were recovered by centrifugation and washed with extraction buffer, and the immune complexes were eluted by boiling for 2 minutes in Laemmli sample buffer.17 Biotinylated proteins were visualized by 8% sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis under reducing conditions, transfer to nitrocellulose membranes, incubation with peroxidase-coupled streptavidin, and development with the ECL system (Amersham).

Human tonsils were embedded in Tissue-Tec (Miles Inc, Naperville, IL) and frozen on dry ice. Seven-micron to 10-μm thick sections were air-dried for 1 hour and then fixed in acetone for 10 minutes before storage at −70°C. For immunohistochemistry, the sections were incubated with 10 μg/mL primary antibody for 30 minutes at room temperature (RT), followed by incubation with biotin-conjugated rat antimouse IgG-specific F(ab′)₂fragments (Jackson Immunoresearch Laboratories, West Grove, PA) for 30 minutes at RT. The labeling was shown using the ABC kit and diaminobenzidine substrate (as described in the manufacturer’s protocol; Vector Laboratories, Burlingame, CA). The primary antibodies were either mouse antihuman CD3 (Leu-4), mouse antihuman CD20 (Leu 16; both purchased from Becton Dickinson), or the anti-P2X₇ MoAb. The sections were counterstained with May-Grunwald-Giemsa.

Whole cell recordings18 were made to investigate the effects of the MoAb on nucleotide-evoked inward currents from a variety of P2X receptor subtypes. For all experiments, HEK293 cells transfected stably with the indicated receptors were used. Recordings were made essentially as described elsewhere.19 Briefly, cells were perfused with either a normal (consisting of 145 mmol/L NaCl, 2 mmol/L KCl, 2 mmol/L CaCl₂, 1 mmol/L MgCl₂, 10 mmol/L HEPES, 10 mmol/L D-glucose, pH 7.3; osmolarity, 300 mOsm) or low-divalent cation containing (as described above, but with only 0.5 mmol/L Ca²⁺ and without added Mg²⁺) solution.

Agonists were applied using a computer-controlled fast-flow U-tube system20 modified to include an extra solenoid valve. The following agonists were used to evoke inward currents: 2′,3′-(4-benzoyl)-benzoyl-ATP (BzATP; human, rat, and mouse P2X₇1,2,19) at 300 μmol/L, ATP (human P2X₄21) at 10 μmol/L, and α,β-methylene-ATP (αβmeATP; human P2X₃16) at 3 μmol/L. Nucleotides were obtained from Sigma (Poole, UK). All experiments were performed at room temperature (22°C to 24°C).

THP-1 cells (ECAC, Porton Down, UK) were grown in a humidified atmosphere (95% air 5% CO₂) at 37°C as a suspension culture in RPMI 1640 with 10% heat-inactivated fetal bovine serum (FBS; GIBCO, Paisley, UK). To measure release of IL-1β, cells were resuspended at 1 × 10⁶ cells/mL in fresh media containing 10 μg/mL LPS for 18 hours at 37°C. Cells were centrifuged at 200g for 5 minutes and resuspended in assay buffer composed of 140 mmol/L NaCl, 5 mmol/L KCl, 10 mmol/L glucose, 1 mmol/L CaCl₂, 10 mmol/L HEPES, and 10 mmol/L N-methyl-D-glucamine and supplemented with 0.1% BSA (pH 7.4 at 37°C). After a second centrifugation as described above, the cells were resuspended in assay buffer at 37°C and 100 μL of cell suspension was added to the wells of a 96-well V-bottom plate (150,000 cells per well) containing 100 μL of antibody. After preincubation for 30 minutes at 37°C, BzATP or ATP was added and incubations were continued for 30 minutes. The plates were subsequently centrifuged at 200g for 5 minutes and 10 μL aliquots of the supernatants were removed for determination of IL-1β release using a reporter bioassay.22 In this assay, supernatants from the THP-1 cells were added to an A549 cell line that expresses the human IL-1β receptor and that has been genetically modified to secrete soluble placental alkaline phosphatase (SPAP) in response to IL-1β. These cells were the generous gift of Dr Keith Ray (Glaxo Wellcome, Stevenage, UK). The A549 cells were cultured in Dulbecco’s modified Eagle’s medium containing 10% FBS at 37°C in a humidified atmosphere (95% air, 5% CO₂) for 16 hours, during which time the cells released SPAP. To quantify the SPAP release evoked by IL-1β, aliquots of the A549 cell supernatants were transferred to fresh plates and heated at 60°C for 30 minutes to inactivate nonspecific phosphatase activity. After cooling to 22°C, 200 μL of 5 mmol/L para-nitrophenol phosphate (pNpp) substrate, dissolved in 1 mol/L diethanolamine, 0.28 mol/L NaCl, and 0.5 mmol/L MgCl₂(pH adjusted to 9.85 with HCl), was added to the wells. SPAP activity was determined by absorbance at 405 nm with time. The release of SPAP from A549 cells was proportional to the concentration of IL-1β applied to the cells, and the IL-1β released from the THP-1 cells was determined by a calibration curve based on human recombinant IL-1β (R&D Systems, Minneapolis, MN). The reporter assay was validated by showing that 0.2 μg/mL of a neutralizing MoAb against human IL-1β (R&D Systems) was able to eliminate the effect of both human recombinant IL-1β and the THP-1 cell supernatants on the release of SPAP from A549 cells (data not shown).

The human P2X₇ receptor was expressed in XS63 cells, a Balb/c myeloma. These stably transfected cells reacted with a previously characterized low-titer rabbit polyclonal antisera generated to the C-terminal peptide of rat P2X₇.23Application of BzATP, a selective agonist for P2X₇receptors, resulted in cell swelling and permeability to the propidium dye, YOPRO-1 (data not shown). HEK293 cells transfected with either the rat or human P2X₇ receptor cDNAs displayed the same properties.1 2 

Hybridomas were generated with lymph node cells from Balb/c mice immunized with the human P2X₇ receptor-expressing cells. Screening of the hybridoma supernatants by FACS analysis on the transfected XS63 cells yielded an IgG2b MoAb that reacted strongly with cells expressing human P2X₇ receptor (Fig 1). The MoAb bound to the surface of HEK293 cells bearing hP2X₇ receptors (Fig 1) but not to HEK293 cells that expressed human P2X₁ or human P2X₄ (Fig 1A and B). In studies using the FITC-labeled MoAb for hP2X₇ transfected HEK293 cells, the cell-associated specific fluorescence approached saturation at an MoAb concentration of 1 μg/mL (K_D = 58 ± 18 ng/mL; B_max = 1,349 ± 12 specific RFU). In contrast, there was no detectable specific binding of the MoAb binding to hP2X₃transfected HEK293 cells (the RFU value of 101 ± 19 at 1 μg/mL of the MoAb was not significantly different from the value of 130 ± 10 RFU determined in wild-type HEK293 cells).

The MoAb also recognized the native P2X₇ receptor on human monocytes and macrophages (Fig 2). Previous work had functionally demonstrated P2X₇ receptors on this cell type and shown an enhanced activity during the monocyte to macrophage transition.5 6 Human monocytes were derived from peripheral blood mononuclear cells and cultured for 1, 2, or 3 days in media alone or in media supplemented with LPS or γ-IFN. Compared with the isotype control, the MoAb to hP2X₇ receptor showed significant reactivity with cells cultured in media alone at all three time periods. This reactivity was enhanced by addition of either LPS or γ-IFN. Interestingly, whereas P2X₇ expression was augmented within 24 hours of γ-IFN treatment and continued to be high, LPS showed a more marginal upregulation, needing 48 hours for augmentation (Fig 2A and B).

Next, the ability of the MoAb to immunoprecipitate the human P2X₇ receptor protein was tested. XS63 cells transfected with human P2X₇ receptor cDNA or mock transfected (vector alone) were surface labeled by biotinylation and lysed to generate membrane protein extracts. These were subjected to immunoprecipitation with the MoAb, followed by polyacrylamide gel electrophoresis (PAGE) analysis (Fig 3). As detected by labeled streptavidin, the MoAb was able to immunoprecipitate a major protein of approximately 72 kD from the hP2X₇ receptor containing extract, but not from mock-transfected cells. Minor bands of 74 and 56 kD were also visible.

Using serial cryostat sections of tonsils, the expression of P2X₇ in human lymphoid tissue was evaluated (Fig 4). Numerous distinct cells (Fig 4C and D) with dendritic morphology (Fig 4D, insert) labeled strongly in the area of the marginal zone. In addition, a more diffuse lighter labeling was detected within the light zone of the germinal center (GC, Fig 4C) and in the marginal zone. These observations suggest that macrophages as well as certain dendritic cells express P2X₇within tonsils.

Inward currents evoked by BzATP in HEK293 cells transfected with human P2X₇ receptor were inhibited by incubation of the cells with the MoAb. This inhibition was concentration-dependent, and currents were reduced to approximately half maximal with 200 ng/mL antibody (Fig 5). Assuming a molecular weight of 150 kD, the estimated IC₅₀ value for the antibody was about 5 nmol/L. The effects of the antibody were highly specific for the human P2X₇ receptor; currents evoked in HEK293 cells transfected with the mouse or rat orthologues of the hP2X₇ receptor or in cells transfected with hP2X₄21 or hP2X₃16 were unaffected by the MoAb (Fig 5) applied at concentrations that caused greater than 80% inhibition of the currents observed in the human P2X₇-expressing cells. Application of the antibody alone to human P2X₇-expressing cells produced no inward currents. Blockade of the human P2X₇ receptor by the MoAb was only slowly reversible, such that after 30 minutes of washing, agonist-evoked inward currents were still inhibited by approximately 70% of their control values (Fig 5A).

BzATP or ATP evoked a concentration- and time-dependent release of IL-1β from LPS-treated human monocytic cells (THP-1). The ability of both γ-IFN and LPS to induce P2X₇ receptor activity in THP-1 cells has been previously shown.7 IL-1β release, measured by bioassay (see Materials and Methods), increased more that 30-fold with maximum agonist stimulation (512 μmol/L BzATP for 30 minutes; 16.0 ± 1.0 ng IL-1β per 150,000 cells was releasedv 0.4 ± 0.05 ng from control cells). Incubation of THP-1 cells with the MoAb caused a concentration-dependent inhibition of IL-1β release, such that significant inhibition of the BzATP-induced release could be obtained with the MoAb at a concentration of 38.3 ng/mL (Fig 6).

The aim of this study was to obtain an antibody directed against the external domain of the human P2X₇ receptor. This was achieved by immunizing Balb/c mice with mouse cells expressing recombinant hP2X₇ receptor to maximize the potential for raising antibodies to the intact protein. The hybridomas obtained were then screened by differential flow cytometry using nonpermeabilized cells expressing the human P2X₇ receptor to obtain an antibody that recognized the external domains of the receptor.

The isolated antibody was highly selective for human P2X₇receptors and did not recognize human P2X₁ and human P2X₄ receptors by flow cytometry. These receptors are the only P2X receptors so far localized to immune cells. The MoAb also failed to label or affect responses at the human P2X₃receptor, and functional data suggest that the MoAb does not recognize rat or mouse P2X₇ orthologues.

The antibody was found to be suitable for quantitative analysis of cell surface receptor expression and could be used to detect P2X₇ receptors in THP-1 cells and to confirm the results of earlier functional studies that have suggested that the human P2X₇ receptor is upregulated by γ-IFN and to a lesser extent by LPS.7 Functionally, P2X₇ receptors have been primarily localized to myeloid lineages and we noted by flow cytometry that the MoAb reacted with another human myeloid cell line, U937 (not shown). However, there have also been reports of ATP-activated channels with similar operational characteristics to the P2X₇ receptor, but that apparently lack the ability to form the large pore in Epstein-Barr virus (EBV)-transformed lymphoblasts24 and lymphocytes from chronic lymphocytic leukemia lymphocytes,25 and this MoAb should provide a useful tool for identification of P2X₇ receptor-containing assemblies. The finding that the antibody was also effective at immunoprecipitating the P2X₇ receptor from cells known to express the receptor will also enable it to be used for the detection of other subunits or proteins that may interact with the P2X₇ receptor.

Previous in situ hybridization studies have shown that rat P2X₇ receptor is most abundant in bone marrow; in the brain, it has only been observed in microglial cells.23 The localization of distinct cell populations within the tonsil is itself an interesting finding and demonstrates the utility of this MoAb to investigate P2X₇ protein expression using immunohistochemical methods. The MoAb labeled a cellular subpopulation of the marginal zone and T-cell area that presented a distinct morphology from lymphocytes (inset, Fig 4D). Because macrophages and dendritic cells are found in these areas, respectively, and both can be derived from a common myeloid precursor26 known to express P2X₇ receptors,27 we hypothesize that certain antigen-presenting cells express the P2X₇ receptor during immune responses. Future studies will be aimed at defining these populations and characterizing the functional significance of their P2X₇ receptor expression.

The MoAb was selected by flow cytometry for surface binding to nonpermeabilized cells that express the hP2X₇ receptor (Fig1). The epitope recognized is presumably present in the extracellular loop of the native receptor, because the MoAb demonstrated functional antagonism (see below). However, it is likely that contributions to the epitope are made from the tertiary or quaternary structures of the receptor, because all attempts to use the MoAb in Western blot experiments with denatured proteins failed, and yet immunoprecipitation of native protein was successful.

The study of ATP as an extracellular modulator in the immune system has been largely hampered by lack of pharmacological tools. The commonly used P2 receptor antagonists, such as PPADS and suramin, are relatively weak at the hP2X₇ receptor2 (but see Chessell et al28) and are unable to differentiate between P2X receptor subtypes. In addition, many of these antagonists also show affinity for P2Y receptors or have a highly nonspecific profile.29 Oxidized ATP, an irreversible antagonist, has been used to characterize P2X₇ receptor responses, but seems unlikely to be specific, because this compound binds to several ATP-binding proteins.30 Furthermore, we have found in functional studies that oxidized ATP is as effective at blocking the rat P2X₂ receptor as it is in blocking the hP2X₇ receptor (A.D. Michel, unpublished observation). One of the major findings of this study is the demonstration that the MoAb is an effective antagonist of BzATP-induced P2X₇ channel activation. The binding of the MoAb to human P2X₇ receptors occurred with relatively high affinity, giving an estimated IC₅₀ of 5 nmol/L to inhibit inward currents, and was only slowly reversible. Because the MoAb did not cross-react with P2X₁ or P2X₄ receptors and was inactive as an antagonist of responses mediated by P2X₁, P2X₃, or P2X₄ receptors, it represents a unique tool for the identification of putative P2X₇ receptors and for determining the potential coassembly of P2X₇ with other P2X subunits. Considerable evidence, both biochemical31 and functional,32 exists for the heteromeric assembly of other P2X channels (P2X₂ + P2X₃) in the peripheral nervous system, and disparate findings of the characteristics of P2Z receptors in differing systems may be explained by heteropolymeric combination of P2X₇with other P2X subunits.

ATP-induced IL-1β release has been previously reported from activated macrophage and microglial cells.9,33 Our observation of inhibition of this release from the myeloid cell line, THP-1, by the MoAb (Fig 6) is important, demonstrating unequivocally that this effect of ATP is mediated by a native receptor (P2Z) containing P2X₇ subunits, even though THP-1 cells express other ATP receptors, including P2U7 (now known as P2Y₂) receptors. Other phenomena distal to the activation of P2Z receptors, such the induction of NF-κB and caspase10 or the killing of intracellular mycobacteria,11 can now be investigated by functional antagonism using the MoAb.

The demonstration of functional receptor antagonism of a neurotransmitter ligand with an MoAb is not unique and has been previously described for the nicotinic receptor.34 35However, in the absence of specific P2X₇ antagonists, the MoAb described in the present study represents a unique tool with which to explore the function of the P2X₇ receptor. Further studies will be required to determine if the receptor blocking activities of the antibody are a consequence of a direct interaction with the ATP-binding site or accessory sites or are simply a consequence of steric hindrance of either ligand-binding or channel opening.