Mutation of a Dibasic Amino Acid Motif Within the C Terminus of the P2X7 Nucleotide Receptor Results in Trafficking Defects and Impaired Function 1

Activation of the P2X7 receptor by extracellular nucleotides modulates multiple immune functions, including inflammatory mediator production, membrane fusion events, and apoptosis. Previous studies have revealed that the C terminus of this multimeric cation channel possesses a lipid-interaction motif that has been proposed to regulate receptor function. This domain is homologous to the LPS binding region of the LPS binding protein, and we demonstrated that two basic residues (Arg578, Lys579) within this motif are essential for LPS binding to P2X7 in vitro. Because P2X7 can influence LPS action, and because lipid interaction motifs modulate the trafficking of other ion channel-linked receptors, we hypothesized that this motif of P2X7 is critical for receptor function and trafficking. In these studies we mutated Arg578 and Lys579 of P2X7, and the expression profile, channel activity, and pore formation of the mutant were characterized in transfected human embryonic kidney 293 cells. In contrast with the wild-type receptor, the P2X7-R578E/K579E mutant fails to demonstrate surface immunoreactivity despite normal levels of total protein expression. This effect on the mutant receptor is unlikely to result from widespread defects in protein folding, because surface localization, determined using conformation-specific Abs, can be restored by growing the cells at 25°C, conditions that slow receptor recycling. Despite surface expression at reduced temperatures, at 25°C the P2X7-R578E/K579E mutant still exhibits greatly reduced sodium, potassium, and calcium channel activity when compared with the wild-type receptor, and cannot induce pore formation. These data suggest that the lipid interaction motif of the P2X7 C terminus controls receptor trafficking and modulates channel activity.

The P2 receptor family has been divided into the P2X and P2Y subfamilies according to whether the individual member acts as a ligand-gated ion channel (P2X) or a G protein coupled receptor (P2Y). P2X 7 belongs to the P2X subfamily on the basis of struc-tural similarity with the six other members, each possessing two predicted membrane-spanning domains (11,12). Several features of P2X 7 make it unique within the family of purinergic receptors. For example, various P2X receptors are thought to form heteroand/or homotrimers; however, P2X 7 has only been observed to form homotrimers (12). Additionally, the P2X 7 C terminus is Ն119 aa longer than those of other family members, which is a property that has been correlated with several of the unique functions of the receptor (11).
The reversible formation of pores is also a characteristic feature of P2X 7 (1). This receptor normally functions as a nonselective cation channel that allows agonist-dependent conductance of sodium, potassium, and calcium (11). After brief stimulations (Յ1 s), removal of the agonist is associated with cessation of these nonselective currents with minimal desensitization, such that repetitive brief agonist applications do not attenuate the maximum achievable current amplitude (11). Longer applications of agonist allow passage of cations with progressively larger diameters. The pattern of time constants associated with the passage of increasingly large cations is not consistent with a model of simple diffusion, suggesting that the channel diameter dilates with chronic administration of agonist (13). Pore activity is arbitrarily defined as the passage of larger molecules, including fluorescent dyes with an upper size limit of 900 Da, probably as a result of modulation of the selectivity filter of the channel (14). The process associated with pore dilation requires at least a few seconds of agonist administration, is reversible upon agonist removal, and is modulated by temperature. The process is also modulated by the concentrations of sodium and divalent cations in the extracellular solution (11,13).
There are several lines of evidence demonstrating that P2X 7 expression is tightly regulated at the level of protein localization (2). Diverse cell types including PBMC, tissue macrophages, platelets, fibroblasts, neurons, and epithelial cells express P2X 7 mRNA. However, P2X 7 protein expression has a much more restrictive pattern under normal conditions (1). For example, monocytes express similar levels of P2X 7 mRNA, but they have much lower levels of receptor surface expression than macrophages (15). Additionally, P2X 7 function in monocytes is relatively attenuated when measured using physiological agonists and buffers (16). However, the receptor can still induce pore formation in response to cell treatment with potent pharmacological agonists (4) or by altering the ionic environment using buffers that do not contain NaCl (16). Multimerization also appears important for receptor function, although there are tissue-specific differences in this process. For example, microglia largely express the receptor as monomers, whereas peritoneal macrophages contain P2X 7 multimers (17). This is likely to be specific to the cell type and/or differentiation state in that transfected fibroblasts do not appear to undergo P2X 7 -enhanced green fluorescence protein clustering in response to agonist administration (14). Furthermore, analogous to other P2X family members, P2X 7 receptor trafficking is crucial to its functionality (18 -20). Specifically, there appears to be an intracellular pool of P2X 7 in monocytes that is recruited to the cell surface during macrophage differentiation (15,20). Thus, there is differential expression, trafficking, and localization of P2X 7 in various cell types, and elucidation of the details governing these processes will likely contribute to a better understanding of its physiological actions and functional diversity.
The unique P2X 7 C terminus (residues 353-595) is thought to mediate many receptor biological activities (2). Early data supporting this notion included the observation that truncation of the last 177 aa of the rat receptor abolished P2X 7 pore-inducing activity while preserving channel function (11). From these data, channel and pore activities appear distinct, and there may be additional C-terminal features that direct pore formation. In addition to participation in channel and pore activity, the P2X 7 C terminus may direct receptor trafficking, consistent with observations showing that P2X 4 C-terminal mutations can lead to constitutive surface localization (21). However, it is unclear whether P2X 7 also contains stabilization domains that interact with cell surface constituents or retention signals that prevent surface expression under certain conditions. Because of the differential control of P2X 7 expression and localization in various cell types and differentiation states, it is likely that there are multiple domains controlling this process.
Several protein-and lipid-interaction motifs within the C terminus may contribute to channel-independent signaling and other receptor functions (22). Of note, P2X 7 contains a conserved LPS/ lipid-binding motif, and mutation of 2 aa (Arg 578 , Lys 579 ) within this domain abolishes LPS binding in vitro (22), analogous to the effects of similar mutations in LPS binding protein (23). Interestingly, highly basic amino acid motifs have been shown in other ion channel systems to regulate receptor trafficking (24) and promote lipid-dependent endocytosis (25). Thus, we hypothesized that the lipid-interaction motif of P2X 7 is critical to receptor trafficking and/or function. In this study, we demonstrate that point mutations within this domain promote a temperature-sensitive defect in surface localization, and that alterations within this domain are associated with attenuation of channel activities. Thus, the dibasic amino acid motif within the distal C-terminal portion of P2X 7 contributes to receptor trafficking as well as ligand binding or channel gating.

Cell culture
Human embryonic kidney (HEK) 293 cells were obtained from the American Type Culture Collection (ATCC, Manassas, VA) and were maintained in DMEM supplemented with 10% FBS (HyClone Laboratories, Logan, UT), 2 mM L-glutamine, 2 mM sodium pyruvate, and 100 U/ml penicillin/ streptomycin. The cells were grown in 10-cm tissue culture dishes at 37°C or at room temperature (24 -27°C) in a humidified atmosphere containing 5% CO 2 . Cells were lifted with a solution of 0.1% trypsin in PBS and split 1:5 every 2-3 days.

Construction of the human P2X 7 -R578E/K579E mutant
Two point mutations in the P2X 7 cDNA that are necessary for the desired 2 aa change were introduced by PCR using the following primers (Operon Technologies, Alameda, CA): BsrGI sense: 5Ј-AATTTGACATTATCCA GCTGGTGGTGTACA; BsrGI antisense: 5Ј-GAAACTCTTCCTCGATC CTCCAGCGGCA; NotI sense: 5Ј-GAGGATCGAGGAAGAGTTTCCGA AGAG; NotI antisense: 5Ј-ATAGCGGCCGCAAAAGGTGGGATTACA. The P2X 7 cDNA from the pcDNA3/human P2X 7 receptor expression vector was cleaved out with NotI and ligated the pIREShyg expression vector (Clontech Laboratories, Palo Alto, CA). Using the pIREShyg/P2X 7 expression vector as a template, two separate primary PCR sets were run with the BsrGI primer pairs and the NotI primer pairs using Pyrococcus furiosus DNA polymerase. The PCR products were gel purified using the Qiaex II gel purification kit (Qiagen, Valencia, CA) and were used as both template and megaprimers for the secondary PCR process in the presence of the antisense NotI and the sense BsrGI primers. The resulting PCR product was gel purified, digested with NotI and BsrGI, and ligated into the pIREShyg/hP2X 7 expression vector that was also digested with NotI and BsrGI. The correct insertion of the P2X 7 -R578E/K579E cDNA into the expression vector was verified by a restriction digest analysis. The PCRgenerated portion of the P2X 7 -R578E/K579E cDNA was sequenced in both directions. Sequencing reactions were performed using Big Dye-labeled dideoxy nucleotides and resolved by electrophoresis at the University of Wisconsin Biotechnology Center (Madison, WI). The P2X 7 -S342F, S342Y, and S342A mutants were created by a similar strategy using different primers.

Qualitative analysis of mRNA expression
Total RNA was prepared from confluent 10-cm plates of HEK 293 cells using RNA-STAT-60 (Tel-Test, Friendswood, TX). The resulting RNA (5 g/reverse transcriptase reaction) was annealed with random hexamer primers (Operon Technologies), and cDNA was synthesized using Superscript Reverse Transcriptase (Life Technologies). Using control buffer (20 mM Tris-HCl (pH 8.0), 10 mM EDTA) or HEK 293 transfectant cDNA as templates, PCR was performed with the rat P2X 7 -specific primers that have been described previously (27). The predicted size of the PCR product was 939 bp. PCR products were resolved on a 1% agarose gel, stained with ethidium bromide, and visualized under UV illumination.

Determination of total P2X 7 protein expression levels
Total cell membranes were isolated by differential centrifugation from parallel cultures of cells, and the protein content was determined by the Micro-BCA method (Pierce, Rockford IL). Equal amounts of protein (25 g) were resolved on 10% SDS-PAGE gels and transferred to Immobilon polyvinylidene difluoride membranes (Millipore, Bedford, MA) as previously described (28). Membranes were blocked in TBST (10 mM Tris-HCl (pH 8.0), 150 mM NaCl, 0.05% Tween 20) containing 5% nonfat dry milk. The polyclonal anti-P2X 7 Ab was used at a 1/1000 dilution, and HRP-conjugated anti-rabbit IgG was used at a 1/5000 dilution. Proteins were visualized by LumiGlo-ECL (Kirkegaard & Perry Laboratories, Gaithersburg, MD).

Functional screening of P2X 7 mutants with a low-divalent cation cell-killing assay
At times 18 -24 h before each experiment, HEK 293 transfectants (2 ϫ 10 5 cells per well) were plated in 200 l of medium using 96-well tissue culture plates. Before treatment the medium was removed, and 50 l of the indicated concentration of the potent P2X 7 agonist BzATP, prepared in low divalent cation bathing solution (96 mM NaCl; 2 mM KCl; 0.1 mM CaCl 2 ; 5 mM HEPES, pH 7.55), was added to each well. These conditions are similar to those previously shown to facilitate agonist-dependent cell lysis in the presence of functional P2X 7 receptors (13,29,30). After a 6-h incubation at 37°C, 50 l of 2 mg/ml MTT in distilled water was added to each well, and the cells were incubated for 2 h at 37°C. A 100-l volume of solubilization buffer (10% SDS in 50% dimethylfluoride) was added to each well, and the plates were incubated for 12-18 h at 37°C before OD were measured at 562 nm. The data were normalized to that obtained using untreated cells within the same experiment, which were considered to be 100% viable. Each treatment was performed in triplicate.

Confirmation of P2X 7 surface localization
Approximately 2 ϫ 10 5 transfected HEK 293 cells were plated in 6-well plates or 35-mm dishes 2 days before experimentation. After overnight attachment at 37°C, half of these cultures were shifted to room temperature (24 -27°C) for another 18 h. Cells were harvested on the day of experimentation by brief exposure to 0.1% trypsin, and the cultures were divided in half and transferred to tubes (Falcon; Fisher Scientific, Chicago, IL). For cell surface P2X 7 expression experiments, the cells were washed by centrifugation at 300 ϫ g in Dulbecco's PBS containing Ca 2ϩ and Mg 2ϩ . Labeling was performed for 1 h at 4°C with 0.5 g/tube mouse IgG2b (Sigma) or with 0.5 g/tube conformation-specific, mAb directed against an extracellular epitope of P2X 7 (26). After additional washes, 2 l/tube FITC-conjugated goat anti-mouse IgG was added for secondary staining using 30 min of incubation at 4°C. Cells were resuspended in 250 l of PBS containing 5 g/ml propidium iodide after final washes. Flow cytometry was performed on a BD Biosciences FACSCan flow cytometer (Mountain View, CA), and 10,000 events were collected. Data were analyzed with the exclusion of dead cells as determined by propidium iodide uptake.

Electrophysiological analysis of ATP-gated ion channels expressed in HEK 293 cells
HEK 293 cells stably expressing either the wild-type human P2X 7 or the human P2X 7 -R578E/K579E mutant were plated at low density (10 5 cells/ ml) on 35-mm dishes. Cells were incubated at room temperature for at least 18 h before patch-clamp analysis. After transfer of the culture dish to the recording chamber, the cells were detached by incubation in Ca 2ϩ -and Mg 2ϩ -free Hank's saline. Individual detached cells were sealed to patch electrodes (Garner Glass TW-150) with resistances of 2-10 M⍀. The bathing medium was then switched to a Ca 2ϩ -and apyrase-containing balanced saline (140 mM NaCl, 5 mM KCl, 2.5 mM CaCl 2 , 0.5 mM MgCl 2 , 10 mM D-glucose, 10 mM NaHEPES (pH 7.5), 0.1 U/ml potato apyrase) to reduce the background level extracellular nucleotides and minimize their contribution to receptor activation and/or desensitization. The whole-cell configuration was then established by suction. Cells were dialyzed internally with an intracellular saline containing 120 mM CsCl, 20 mM tetraethylamonium chloride, 10 mM EGTA, 5 mM MgCl 2 , and 10 mM tetraethylamonium-HEPES (pH 7.3) for several minutes, during which time they rounded up and detached from the chamber floor. Cells were continuously superfused at room temperature (22-25°C) with an extracellular saline solution containing 140 mM NaCl, 5 mM KCl, 1.5 mM CaCl 2 , 0.5 mM MgCl 2 , 10 mM D-glucose, and 10 mM NaHEPES (pH 7.5). All cells were held at Ϫ40 mV and were dialyzed internally for at least 5 min before being challenged with ATP. Rapidly activated inward Na ϩ currents were recorded by computer using the pClamp series of programs and an Axoclamp 200A patch clamp with a Digidata 1200 Series interface. ATP and its analogs (dissolved in the following low divalent cation-containing solution: 140 mM NaCl, 10 mM HEPES, 10 mM glucose, 5 mM KCl, 0.5 mM CaCl 2 ) were applied through a gravity-fed superfusion system with solution switching controlled by parallel pinch valves (Cell MicroControls, Norfolk, VA). The patch-clamped cell was directly placed in the stream of the superfusing solution.

Measurement of cell K ϩ content
HEK 293 cells that stably express the wild-type human P2X 7 or the human P2X 7 -R578E/K579E mutant were plated at a density of 10 6 cells/well in 12-well dishes and grown at room temperature for 24 h before the experiment. Cells were treated at room temperature (24 -27°C) in NaCl or sodium gluconate solutions (25 mM Na-HEPES, 130 mM sodium gluconate or NaCl, 5 mM KCl, 1.5 mM CaCl 2 , 1.0 mM MgCl 2 , 0.5% BSA, 10 mM glucose (pH 7.4)). The cells were stimulated with 3 mM ATP or with 0.3 mM BzATP for 10 min. The potassium content of the cells was assayed by atomic absorbance.

Fluorometric analysis of cytosolic Ca 2ϩ
HEK 293 cells stably expressing either the wild-type human P2X 7 receptor or the human P2X 7 -R578E/K579E mutant receptor were plated at low density (10 6 cells/ml) on 100-mm dishes. Cells were incubated at room temperature for at least 18 h before fluorometric analysis of agonist-induced changes in cytosolic Ca 2ϩ concentration. HEK cells expressing either the wild-type or mutant receptor were removed from the culture dishes by trypsinization. The suspended cells were washed, resuspended to 10 6 cells/ml in a balanced saline containing 140 mM NaCl, 5 mM KCl, 1 mM CaCl 2 , 1 mM MgCl 2 , 10 mM D-glucose, 1 mg/ml BSA, and 10 mM Na-HEPES (pH 7.5). The cell suspension was supplemented with 1 M fura 2-acetoxymethyl ester (fura 2-AM) (Molecular Probes) and incubated at room temperature for 45 min. The cells were pelleted, washed once, and resuspended in a fresh balanced saline solution to a density of 10 6 cells/ml. Cytosolic Ca 2ϩ levels were fluorometrically assayed at 24°C using equipment and calibration protocols that have previously been described (31). HEK 293 cells were first challenged with 30 M ADP and 30 M UTP to activate and desensitize the endogenously expressed Ca 2ϩ -mobilizing P2Y 1 and P2Y 2 receptors before stimulation with 3 mM ATP to activate the heterologously expressed P2X 7 receptors. The cells were then permeabilized with 50 g/ml digitonin to facilitate calibration of fura 2-AM as a function of extracellular Ca 2ϩ levels.

Detection of nucleotide-induced pore activity
Transfected HEK 293 cells were plated, shifted to the indicated temperatures, and harvested as described previously. After washing in HEPESbuffered saline (HBS: 130 mM NaCl, 5 mM KCl, 20 mM HEPES (pH 7.4), 0.1% BSA, 10 mM glucose, and 0.5 mM CaCl 2 ), the cells were resuspended in 450 l of fresh HBS. The dye uptake reaction was started by adding 50 l of 10ϫ stock solutions to provide a final concentration of 2 M YO-PRO-1 with and without 300 M BzATP. After incubating at room temperature for 5 min, the reactions were equilibrated to 10 mM MgCl 2 , followed by washing and resuspension in 250 l of HBS containing 5 g/ml propidium iodide. Flow cytometry and data analysis with the exclusion of dead cells was performed as described previously.

P2X 7 mutant design, generation, and screening
To test the hypothesis that the distal and potential LPS/lipid-binding portion of the P2X 7 C terminus (residues 571-595) is critical for receptor function, we mutated this region within the cDNA corresponding to the human P2X 7 gene. Specifically, 2 aa (Arg 578 , Lys 579 ) were targeted for point mutation because of previous work demonstrating their importance in the binding of LPS in vitro (22) and because of recent evidence suggesting that basic residues within lipid interaction motifs are necessary for the trafficking of certain receptors and other proteins (24,25,32). Stable HEK 293 cell transfectants containing a human P2X 7 -R578E/K579E double point mutant exhibited normal levels of P2X 7 mRNA and protein as assessed by RT-PCR and immunoblotting, relative to HEK cells expressing the wild-type P2X 7 control (Fig. 1, A and B). To screen for receptor function, we treated these transfected HEK cells with increasing concentrations of a P2X 7 agonist, Bz-ATP, and evaluated the capacity of this treatment to initiate a rapid low-divalent cation killing response, using conditions similar to those previously described to facilitate pore formation and cell lysis (13,29,30). Under these conditions, Bz-ATP treatment for 6 h caused a reduction in the viability of HEK cells expressing wild-type P2X 7 with an EC 50 between 30 and 100 M Bz-ATP (Fig. 1C). By contrast, HEK cells transfected with the control vector or the P2X 7 -R578E/K579E mutant are resistant to Bz-ATP-induced death in this low-divalent cation killing assay, using agonist concentrations as high as 1 mM (Fig. 1C). Thus, HEK cells require P2X 7 expression for Bz-ATP-mediated toxicity, and mutation of the distal C-terminal domain of this receptor does not allow the transfected cells to undergo agonist-induced cell death, despite the observation that the transfected cells have normal levels of total P2X 7 protein expression.

Surface localization of P2X 7 protein
Because mutation of the P2X 7 C-terminal region does not allow for Bz-ATP-induced cell death, it was postulated that either the mutant receptor was not effectively expressed on the cell surface and/or that this mutation abrogates receptor function. To delineate whether the distal C-terminal domain contributes to receptor trafficking and membrane localization, agonist binding, or channel gating, we first analyzed P2X 7 cell surface expression by flow cytometry using a neutralizing Ab reactive against a conformationdependent epitope (26). This Ab has been shown to block agonistdependent channel and pore activity in P2X 7 -transfected HEK cells as well as IL-1␤ processing in THP-1 monocytes (24). Rel-ative to the isotype control, HEK cells transfected with wild-type P2X 7 exhibit a 4-to 10-fold increase in the geometric mean of the intensity of P2X 7 cell surface immunostaining ( Fig. 2A). By contrast, the P2X 7 -R578E/K579E mutant cells exhibit a staining profile identical to the isotype control (Fig. 2B), suggesting a lack of surface expression. Given that another P2X family member, P2X 4 , has been shown to undergo rapid agonist-dependent endocytosis and recycling in transfected HEK cells and neurons (19), we evaluated the surface expression of P2X 7 when the HEK cells were grown at room temperature to slow down endocytic pathways. Relative to growth at 37°C, room temperature cultures of HEK cells expressing wild-type P2X 7 showed a trend toward a modest increase in levels of surface expression (7.3 Ϯ 2.0-fold increase in the staining relative to the isotype control at 37°C, and 11.4 Ϯ 4.2-fold increase in staining at room temperature; Fig. 2, A and C). This increase in surface P2X 7 immunostaining at room temperature is more substantial with HEK cells transfected with P2X 7 -R578E/K579E (1.2 Ϯ 0.1-fold at 37°C vs 3.2 Ϯ 0.7-fold at room temperature; Fig. 2, B and D), suggesting that slowing down endocytic pathways can partially restore surface expression of this mutant. Consistent with these data are the observations that permeabilized HEK cells expressing the P2X 7 -R578E/K579E mutant exhibit P2X 7 immunostaining at 37°C that is not detected with intact cells (Fig. 3). Collectively, these data regarding immunoreactivity with a conformational-specific Ab and the temperaturedependent restoration of mutant protein surface localization support the concept that the R578E/K579E mutation does not induce widespread deformation of P2X 7 structure, but rather that the mutation affects proper receptor trafficking, membrane localization, and/or endocytosis.
As an additional control for nonspecific effects of our mutagenesis and transfection methods, we evaluated the expression profiles of other P2X 7 mutants. One such mutant is P2X 7 -S342F, which we  created as a result of a partial sequence analysis of the P2X 7 gene in a spontaneous murine RAW 264.7 macrophage clone (F6-2) that was selected for resistance to lethal doses of Bz-ATP. Because mutations introducing bulky residues in transmembrane domains have been shown to influence protein folding of other receptor systems (33), we also created the tyrosine and alanine mutants, P2X 7 -S342Y and P2X 7 -S342A. In this regard, the bulky point mutants (S342F and S342Y) are not expressed on the surface and have reduced levels of total protein expression (Fig. 4). By contrast, the S342A mutant is robustly expressed and is effectively routed to the cell surface (Fig. 4, A and B, left panel). Preliminary data suggest that the S342A mutant is functional (29.7-fold stimulation of Bz-ATP-induced YO-PRO-1 uptake vs 1.0-and 1.0-fold for the S342F and S342Y mutants, respectively), and that growth at room temperature is not sufficient to substantially restore the pore activities of the S342F and S342Y mutants (2.3-and 0.9-fold, respectively). Taken together, the protein expression profiles in HEK cells of P2X 7 -S342F and P2X 7 -S342Y, which are mutations that are predicted to exhibit disturbances in receptor folding, are different from the pattern observed with the P2X 7 -R578E/K579E mutant. Specifically, receptors with a transmembrane mutation likely to cause misfolding or inappropriate membrane insertion are expressed at very low levels (probably because of elevated degradation), whereas the C-terminal mutation allows for normal levels of total receptor expression. These data together with receptor reactivity of conformationally directed Abs suggest that the temperature-dependent surface immunoreactivity of the P2X 7 -R578E/ K579E mutant is caused by alterations in receptor trafficking events, and not simply a factor of receptor misfolding or defective membrane insertion.

Channel and pore activity of the P2X 7 -R578E/K579E mutant
To determine whether surface expression is sufficient for P2X 7 agonist binding capacity and inducible ion channel activity, we performed patch clamping experiments evaluating inward sodium currents of ATP-treated HEK cells grown at room temperature (Fig. 5). Untransfected HEK cells do not have ATP-inducible sodium currents (Ref. 11 and data not shown); however, P2X 7 expression confers ATP-stimulated responses that do not desensitize until agonist removal (Ref. 11 and Fig. 5A). The P2X 7 -R578E/ K579E-transfected cells are also responsive to ATP with respect to inward sodium currents; however, these cells require higher concentrations of ATP to generate smaller sodium currents (Fig. 5B). Additionally, fewer of the P2X 7 -R578E/K579E-transfected cells are responsive at all (Fig. 5C), despite growth at the permissive temperature, which is consistent with a subpopulation within the culture that has very low levels of surface expression (Fig. 2D). Furthermore, the peak ATP-induced currents are reduced in the mutant cells ϳ60% relative to those found in the wild-type HEK/ P2X 7 cells, even when only measuring cells that are responsive to ATP. Thus, mutation of a dibasic amino acid motif of the distal C-terminal domain of P2X 7 also attenuates inward sodium channel activity.
Potassium efflux by P2X 7 has been linked to ATP-stimulated IL-1␤ processing in monocytes (34). Therefore, we measured this channel function by atomic absorption spectrophotometry after treatment of the transfected HEK cells with either buffer, 3 mM ATP, or 0.3 mM Bz-ATP (Fig. 6). Treatment of HEK cells expressing wild-type P2X 7 with either ATP or Bz-ATP for 10 min results in depletion of ϳ50 -60% of the cellular potassium stores relative to the buffer control, whereas this reduction was only ϳ10% after treatment of the mutant HEK cells. Because chloride ions have been shown to negatively regulate channel activity (35), we also measured potassium efflux in a sodium gluconate buffer system. The removal of chloride ions reduced the total potassium FIGURE 3. P2X 7 immunostaining of intact and permeabilized HEK cells expressing P2X 7 -R578E/K579E at 37°C. Transfected HEK cells expressing the double-point mutant P2X 7 were grown at 37°C before treatment with saline without (A) or with (B) digitonin for permeabilization. Immunostaining with the IgG control or an anti-human P2X 7 Ab and a FITC-conjugated anti-IgG secondary label was performed at 4°C followed by paraformaldehyde fixation and flow cytometry. This experiment was performed three times with identical results. content of both the mutant and wild-type P2X 7 -expressing HEK cells and facilitated a small increase in the agonist-stimulated potassium efflux regulated by the wild-type receptor (ϳ70 -80%). However, depending upon the agonist, the P2X 7 -R578E/K579E mutant cells exhibited a variable effect of chloride removal on potassium efflux (ϳ0 or 25% for ATP and Bz-ATP, respectively). In sum, there is reduced potassium conductance of the channels formed by P2X 7 -R578E/K579E subunits.
Calcium fluxes are also regulated by P2X 7 channels and appear to be important for the nucleotide-stimulated expression of inducible nitric oxide synthase in LPS-treated RAW cells (28). Thus, we evaluated the contribution of the dibasic amino acid motif to this P2X 7 channel function by monitoring changes in fura 2-AM fluorescence in transfected HEK cells grown at room temperature, which is a condition that allows for cell surface expression of wild-type and mutant P2X 7 -R578E/K579E receptors (Fig. 2). HEK cells expressing either receptors respond to both ADP and UTP, which are agonists for the P2Y 1 and P2Y 2 receptor systems expressed endogenously by HEK cells (36). By contrast, ATP treatment causes a response only in HEK cells expressing wildtype P2X 7 , and does not promote a Ca 2ϩ flux in cells containing the P2X 7 -R578E/K579E mutant (Fig. 7). This response of the FIGURE 5. ATP-induced sodium currents in transfected HEK cells grown at room temperature. HEK cells expressing either wild-type (A) or R578E/K579E mutant (B) human P2X 7 were grown at room temperature for 18 h before patch clamping of single cells treated with increasing doses of ATP. C, Whole-cell inward sodium currents triggered by 3 mM ATP were recorded from 12 cells expressing wild-type P2X 7 and 19 cells expressing the P2X 7 -R578E/K579E mutant. Cells that exhibited ATP-gated inward currents that could be readily distinguished from background current noise and slow fluctuations in the holding were scored as ATP-responsive cells (11/12 P2X 7 wild-type cells vs 10/19 P2X 7 mutant cells). D, Differences in maximal ATP-gated inward current in single HEK 293 cells stably transfected with wild-type or mutant receptors. The peak inward current triggered by 3 mM ATP was normalized to cell surface area (membrane capacitance: 13.7 Ϯ 1.84 F in HEK/P2X 7 cells vs 8.8 Ϯ 1.24 F in HEK/ P2X 7 -R578E/K579E cells) in 10 ATP-responsive cells in each case. FIGURE 6. P2X 7 agonist-stimulated potassium efflux in transfected HEK cells grown at room temperature. HEK cells expressing either wildtype or R578E/K579E mutant human P2X 7 were grown at room temperature for at least 18 h before trypsinization and resuspension in NaCl saline or sodium gluconate solutions at 10 6 cells/ml. Cells were stimulated for 10 min with the indicated concentrations of ATP or BzATP, and cell-associated potassium content was assayed by atomic absorbance spectrophotometry.  wild-type P2X 7 -expressing cells appears to be maximal in that subsequent treatment with digitonin does not cause further increases in fura 2-AM fluorescence. Removal of NaCl in the buffer attenuates the fura 2-AM fluorescence induced by ADP treatment of either mutant or wild-type P2X 7 -expressing cells, but not that stimulated by the addition of UTP. Interestingly, the potassium gluconate buffer system also causes a partial restoration of ATPinduced calcium flux in the P2X 7 -R578E/K579E mutant. These data suggest that the dibasic amino acid motif of the distal P2X 7 C-terminal domain also regulates calcium channel activity.
In addition to its capacity to mediate ion fluxes, the P2X 7 pore activity has been linked to IL-1␤ processing, membrane fusion events, and apoptosis (2). Therefore, we evaluated the pore-forming capacity of the P2X 7 -R578E/K579E mutant cells to determine whether the channel and pore activities can be dissociated. Transfected HEK cells expressing wild-type P2X 7 have a 6.4 Ϯ 0.5-fold increase in YO-PRO-1 uptake in response to Bz-ATP (Fig. 8A). Similar to the trend in protein expression, cell growth at ambient temperatures slightly increases the total amount of YO-PRO-1 uptake after 5 min (14.5 Ϯ 3.2-fold; Fig. 8C), a time sufficiently long to overcome the temperature-induced delay in initial formation of the pore (13). By contrast, the P2X 7 -R578E/K579E mutant is unable to form pores at ambient temperatures, despite surface expression and partial channel activity (Fig. 8, B and D). Collectively, point mutations within the lipid-interaction motif of the distal C terminus of P2X 7 affect surface immunoreactivity as well as the formation of agonist-regulated channels and pores. These data suggest that this domain is critical to receptor function and trafficking events.

Discussion
The data presented in this study are supportive of a model wherein residues within the lipid interaction domain of P2X 7 , specifically Arg 578 and Lys 579 , are required for normal trafficking of the receptor. Point mutation of these two residues in P2X 7 prevents receptor expression on the cell surface as assessed by flow cytometry. This defect is partially restored at lower temperatures (24 -27°C), which is a condition known to slow down receptor recycling events and has been used previously to demonstrate different P2X 7 channel permeability states (37). Because the P2X 7 -R578E/K579E mutant is recognized by conformation-sensitive Abs, and because its cell surface expression is detectable at 24 -27°C, it is likely that this mutation does not promote large-scale receptor deformation, misfolding, or improper membrane insertion. In addition, the total protein expression level of the doublepoint mutant is not diminished, which contrasts with the reductions in protein expression that occur after disruption of conserved cysteine residues in the human P2X 1 receptor that contribute to conformation stabilization of the extracellular domain (38). Additionally, two P2X 7 mutants with bulky amino acid substitutions within the second transmembrane domain (S342F and S342Y, but not S342A) have reduced protein expression (Fig. 4). Similar mutations in other systems have been shown to affect protein folding (33). Thus, the dibasic amino acid motif within the distal C-terminal domain of P2X 7 appears to regulate receptor trafficking rather than specify long-range conformational effects.
Previous studies have shown that all leukocytes contain a large intracellular pool of P2X 7 , which is recruited by monocytes to the surface during macrophage differentiation (16,20). It is unclear whether recombinant P2X 7 trafficking in HEK cells uses similar pathways, compartments, and cofactors; however, the trafficking of P2X 1 , P2X 2 , and P2X 4 has been studied in these cells (19,39) and has been shown to exhibit similarities to the patterns observed in neurons and smooth muscle. Interestingly, a YXXGL motif in the distal C terminus of P2X 4 has been shown to direct agonistdependent endocytosis, and mutation of this domain results in constitutive surface localization (21). That this motif is not present in P2X 7 , combined with the observations that leukocytes have a large intracellular pool of the receptor, suggests that P2X 7 has compartmental retention and/or surface stabilization domains. A P2X 7 retention domain might exist between residues 418 and 578, given that a 177-aa C-terminal truncation of the rat receptor allows for receptor expression on the surface, whereas the P2X 7 -R578E/ K579E mutant is not stabilized on the cell surface unless the growing temperature is reduced. This observation suggests that Arg 578 and/or Lys 579 of P2X 7 are part of a surface stabilization domain, such that mutation of these residues traps or routes the receptor to intracellular compartments.
We also observed that P2X 7 channel activity is attenuated and that pore activity is completely abolished following mutation for Arg 578 /Lys 579 , despite the partial restoration of cell surface expression of P2X 7 -R578E/K579E at ambient growing temperatures. One interpretation is that this may simply be caused by a threshold effect of receptor number, such that there is a requirement for higher receptor expression to achieve efficient channel and pore activity. In this regard, we have observed a correlation between receptor expression level and the relative capacity to form pores in sorted populations of HEK cells expressing P2X 7 -S342A receptors (data not shown). This notion of a threshold receptor number has also been observed when comparing monocytes and macrophages (16). However, select P2X 7 truncation mutants (P2X 7 -1-540, P2X 7 -1-550) have recently been shown to have reasonably normal levels of channel activity, despite low levels of surface expression that are undetectable by flow cytometry (40), thus making the threshold model less likely. Other possibilities include the scenario that the distal C terminus may be involved in modulating ligand binding and/or channel gating interacting with accessory proteins and lipids, perhaps by folding back toward the membrane. The present data would also be consistent with a role of the distal C terminus in P2X 7 multimerization, although this function has been suggested to be largely controlled by the second transmembrane domain in the rat P2X 2 receptor system (12). In sum, HEK cells expressing P2X 7 -R578E/K579E may be a useful tool to elucidate signaling events or other receptor activities, such as association with effector molecules or the cytoskeleton, that are independent of channel and pore formation.
The composition of the lipid microdomain within the membrane may also be a factor regulating P2X 7 trafficking events and/or activity (41). We previously reported that P2X 7 residues 574 -589 represent a lipid interaction motif, and that mutation of Arg 578 and Lys 579 abolished the ability of synthetic peptides to bind LPS in vitro (22). Because highly basic C-terminal amino acid motifs have been shown in other ion channel systems to regulate receptor trafficking (24) and promote lipid-dependent endocytosis (25), one model is that this P2X 7 domain interacts with endogenous lipids that influence receptor localization. Another possibility is that during infection LPS released within the cell from phagosomes, and this may promote local P2X 7 recruitment, which is an event that has been shown to facilitate phagolysosomal maturation needed for microbial killing (8).
In conclusion, our data support a model wherein the lipid interaction motif of P2X 7 is involved in receptor trafficking. Because mutations within this motif allow for cell surface expression at lower temperatures, which is a condition that slows down receptor endocytosis, it appears likely that the C-terminal domain functions to control surface stabilization, consistent with recent observations by Smart et al. (40). Together, these data suggest that the binding of endogenous lipids or LPS to this region of P2X 7 may serve to modulate receptor availability to the extracellular milieu and thereby regulate its contribution to the inflammatory response.