Pannexin-1 hemichannel–mediated ATP release together with P2X1 and P2X4 receptors regulate T-cell activation at the immune synapse

T-cell activation requires a sustained elevation of intracellular Ca²⁺ levels, which is accomplished by Ca²⁺ entry through calcium release-activated calcium (CRAC) channels that are composed of stromal interaction molecule 1 (STIM1) and Orai1 proteins.^1-3  Both proteins translocate to the immune synapse upon T-cell activation, where they mediate localized influx of extracellular Ca²⁺.⁴  Ca²⁺ entry contributes to the activation of nuclear factors of activated T cells (NFATs) that induce interleukin-2 (IL-2) gene expression and subsequent signaling events that lead to T-cell proliferation.^5-7 

Recent studies have shown that extracellular adenosine-5′-triphosphate (ATP) regulates T-cell activation.^8-10  T cells release ATP in a controlled manner, as do other leukocytes, thereby facilitating intercellular communication and autocrine feedback regulation of cell function.^8-13  Stimulation of T cells by T-cell receptor (TCR) ligation, mechanical stimulation, membrane deformation, or osmotic stress induces the release of cellular ATP.^9,10,13-16  T cells express the gap junction hemichannels pannexin-1, which can mediate ATP release and T-cell activation.^8,10  T-cell activation has been shown to involve P2X receptor subtypes.^8,10  The 7 mammalian P2X receptor family members (P2X1-7) are ATP-gated ion channels.^17,18  All these receptors, with the exception of P2X5, can facilitate entry of Ca²⁺ in response to stimulation by extracellular ATP,^18-20  thus suggesting that P2X receptors regulate T-cell activation by mediating Ca²⁺ entry.

T-cell activation during antigen recognition requires the formation of an immune synapse between T cells and antigen-presenting cells. The immune synapse is a complex structure with a limited number of TCRs, implying that TCR stimulation requires signal amplification.²¹  Here we examined the role of ATP release and P2X receptor activation at the immune synapse. Our findings reveal that pannexin-1 and P2X1 and P2X4 receptors translocate to the immune synapse, where they contribute to Ca²⁺ entry and T-cell activation. The results thus indicate that pannexin-1–induced ATP release and autocrine activation of P2X1 and P2X4 receptors have important functions in TCR signal amplification and intercellular communication between T cells and antigen-presenting cells at the immune synapse.

Jurkat cells (E6-1 clone, ATCC) were maintained and human CD4⁺ T cells were isolated as previously described.¹⁴  All human and animal studies were approved by the Institutional Review Board and Institutional Animal Care and Use Committee of Beth Israel Deaconess Medical Center.

Real-time polymerase chain reaction (PCR), performed as previously described (supplemental Figure 1A-B, available on the Blood Web site; see the Supplemental Materials link at the top of the online article),^12,14  was used to assess the gene expression of P2X1 and P2X4 receptors in resting or stimulated Jurkat and/or human CD4⁺ T cells. IL-2 gene expression was measured in human peripheral blood mononuclear cells (PBMCs) and mouse splenocytes (BALB/c) after stimulation with anti-CD3/CD28 coated Dynabeads for 4 hours in the presence or absence of antagonists (carbenoxolone, A438079,¹⁰panx-1, NF023, TNP-ATP, and suramin). IL-2 primers were purchased from QIAGEN. IL-2 mRNA expression was also measured in CD3/CD28-bead stimulated Jurkat cells after siRNA silencing of P2X1 and P2X4 receptors (described in “P2X receptor silencing by siRNA”). ATP release upon stimulation with anti-CD3/CD28-coated Dynabeads was assessed using the ATP Bioluminescence Assay Kit HSII (Roche), as previously described.⁸ 

Immunocytochemistry of Jurkat cells and human CD4⁺ T cells with goat anti–pannexin-1 (Santa Cruz Biotechnology), rabbit anti-P2X1 or anti-P2X4 receptor antibody (Alomone Labs) was performed as described.⁸  For receptor redistribution experiments, primary CD4⁺ T cells were stimulated with anti-CD3/CD28 coated Dynabeads for 0-30 minutes in complete medium (RPMI 1640, 10% fetal bovine serum [FBS], 100 U/mL penicillin, and 100 μg/mL streptomycin] before fixation. Immunofluorescence and brightfield images were captured using a confocal laser scanning microscope Zeiss LSM510 META.

Jurkat cells (5 × 10⁷) were stimulated with phytohemagglutinin (PHA; 50 ng/mL) and phorbol 12-myristate 13-acetate (PMA; 5 ng/mL) for various times, resuspended in low salt buffer, sonicated on ice, and centrifuged. Bicinchoninic acid protein assays (Pierce) were performed. Samples were prepared in Novex 2× Tris glycine sodium dodecyl sulfate loading buffer (Invitrogen) containing 100μM dithiothreitol and boiled. Equal amounts of protein were separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis. Western blotting was performed using standard procedures and rabbit anti-P2X1 or anti-P2X4 receptor antibodies (1:200, Alomone Labs).

Transfections were carried out using an Eppendorf multiporator with cells suspended in hypo-osmolar electroporation buffer. Electroporation was performed with 10 μg of the respective plasmid using the following settings: 260 V, 70 microseconds, and 2-mm path length. After transfer to complete media, cells were cultured for 24 hours.

Plasmids containing cDNAs of the wild-type P2X1 and P2X4 receptor were purchased from Origene. The NFAT-luciferase reporter plasmid was a gift from Dr A. Altman (La Jolla Institute of Allergy and Immunology), and the β-galactosidase control plasmid was purchased from Roche. STIM1-mCFP and Orai1-mYFP constructs were kindly provided by Dr L. E. Samelson (Laboratory of Cellular and Molecular Biology, Center for Cancer Research, National Institutes of Health). Mutated and fluorescent P2X1 and P2X4 receptor fusion constructs were generated as described in the following section.

Enhanced green fluorescent protein (EGFP)–tagged P2X1 and P2X4 receptors were synthesized as follows: the P2X1 and P2X4 receptor cDNAs (Origene) were amplified by PCR using the following primers: P2X1 sense: 5′ATATAGAAGCTTGCCACCATGGCACGGCGGTTCCAGGAG3′ and antisense: 5′ATATGAATTCTAGCGTAGTCTGGGACGTCGTATGGGTAGGATGTCCTCATGTTCTC3′. P2X4 sense: 5′ATATAGAAGCTTGCCACCATGGCGGGCTGCTGCGCCGCG3′ and antisense: 5′ATATGAATTCTAGCGTAGTCTGGGACGTCGTATGGGTACTGGTCCAGCTCACTAGC3′. The primers introduce a HindIII restriction site upstream of the start codon, remove the stop codon, and add an in-frame EcoRI restriction site at the 3′ end. The PCR products and the pEGFP-N1 vector (Clontech) were digested, purified and then ligated together. Plasmids were transformed into competent DHα Escherichia coli cells, screened, and sequences were confirmed by DNA sequencing (Seqxcel).

The T18A and L352W mutations, which impair receptor function, were introduced into the P2X1, and P2X4 receptors, respectively.^22,23  Site-directed mutagenesis was performed with the Quick-Change mutagenesis kit (Stratagene), according to the manufacturer's instructions, using the following primers: P2X1 sense: 5′ CTTCGAGTATGACGCTCCCCGCATGGTGC 3′; and antisense: 5′ GCACCATGCGGGGAGCGTCATACTCGAAG 3′, or P2X4 sense: 5′GGCATGGCGACCGTGTGGTGTGACATCATAGTC 3′; and antisense: 5′ GACTATGATGTCACACCACACGGTCGCCATGCC 3′. Successful introduction of the mutations was confirmed by DNA sequencing.

The NFAT-luciferase reporter gene assay was performed as described.^8,24  To examine the effect of P2X1 and P2X4 receptor silencing on NFAT activation, cells were electroporated with 6 μg of the NFAT-luciferase reporter plasmid, 8 μg of the β-galactosidase reporter plasmid and 3 μg of P2X1 or P2X4 receptor siRNA. After 72 hours, cells were stimulated with anti-CD3/CD28 antibody-loaded Dynabeads. Luciferase and β-galactosidase activity were measured 8 hours later.

siRNA constructs targeting the mRNA of P2X1 (sense: 5′GGCCGAGAACUUCACUCUUtt3′; antisense: 5′AAGAGUGAAGUUCUCGGCCtc3′), P2X4 (sense: 5′GUACUACAGAGACCUGGCUtt3′; antisense: 5′AGCCAGGUCUCUGUAGUACtt3′) and P2X7 (sense: 5′GGUGAAAGAGGAGAUCGUGtt3′; antisense: 5′CACGAUCUCCUCUUUCACCtc3′) receptors were purchased from Ambion and a nonsense siRNA control was purchased from QIAGEN. Silencing efficiency was tested and Jurkat cells were transfected with 3 μg siRNA by electroporation as described (supplemental Figure 1C-F).⁸  Triple silencing was carried out using one-third of the siRNA per gene compared with silencing of individual genes.

Ca²⁺ signaling was assessed before and after P2X receptor silencing of Jurkat cells using the Ca²⁺ indicator Fluo-3-AM (Molecular Probes), and of Fluo-4-AM (Invitrogen)–loaded primary CD4⁺ T cells after treatment with P2X antagonists: 10μM NF023, 30μM 2′,3′-O-(2,4,6-Trinitrophenyl)adenosine-5′-triphosphate tetra(triethylammonium) salt (TNP-ATP), 200μM suramin, or 200μM or 400μM of the pannexin-1 mimetic inhibitory peptide ¹⁰panx-1 or the scrambled control peptide ^scpanx-1 (TOCRIS), and a FACSCalibur flow cytometry system (Becton Dickinson). Changes in intracellular Ca²⁺ levels in response to stimulation with anti-CD3 antibodies (0.5 μg/mL) were assessed as previously described.⁸  We observed donor-to-donor differences in the response of PBMC preparations to CD3 stimulation compared with controls treated with ionomycin. However, each PBMC preparation showed consistent patterns of inhibition of Ca²⁺ signaling by P2X receptor antagonists and by ¹⁰panx-1. Data representative of at least 3 individual experiments are shown.

For microscopy experiments measuring Ca²⁺ influx, 5 × 10⁶ cells were washed in pre-warmed Ca²⁺ -free Hanks balanced salt solution (Thermo Scientific), and incubated with 4μM Fluo-4 (Invitrogen) and 10μM EGTA (ethyleneglycoltetraacetic acid) for 1 hour. Cells were washed in Ca²⁺ -free Hanks balanced salt solution, seeded on fibronectin-coated glass bottom dishes (MatTek), and incubated with anti-CD3/CD28–coated beads for 30 minutes. Ca²⁺ was added to a final concentration of 2mM. Cells were scanned with a Zeiss LSM 510 META in 200-millisecond intervals.

Jurkat cells were transfected with EGFP–tagged P2X1 and P2X4 receptor constructs, placed into sterile fibronectin-coated glass bottom dishes, and stimulated with anti-CD3/CD28 antibody–coated beads. Brightfield and time-lapse images were captured using a Zeiss LSM 510 META.

Jurkat cells were transfected with 10 μg of STIM1-mCFP or Orai1-mYFP plasmids, and 5 μg of EGFP-tagged P2X-receptor constructs, and cultured for 24 hours. Cells were seeded on fibronectin-coated glass bottom dishes, and stimulated with anti-CD3/CD28 coated Dynabeads. Brightfield and confocal fluorescence images were captured using a Zeiss LSM 510 META.

Unless otherwise stated, all data are expressed as mean ± SEM of ≥ 4 experiments performed in triplicate. Statistical analyses were performed with GraphPad Prism 4.0 software program (GraphPad). For comparisons between groups, 2-tailed unpaired Student t tests were used. A P value of ≤ .05 was considered significant.

We have previously shown that human lymphocytes express P2X7 receptors.⁸  Real-time reverse transcription PCR analysis revealed that Jurkat T cells and human primary CD4⁺ T cells also express P2X1, P2X4, and P2X5 receptors (Figure 1A). Previous studies have demonstrated lower Ca²⁺ permeation for P2X5 receptors compared with P2X1 and P2X4 receptors.^25,26  Thus, we focused our study on P2X1 and P2X4 receptors. Immunocytochemistry indicated that Jurkat cells and human CD4⁺ T cells express P2X1 and P2X4 receptors on the cell surface (Figure 1B). T-cell activation with PHA and PMA alters the expression of these receptors, increasing P2X1 receptor mRNA expression by approximately 15-fold and decreasing P2X4 mRNA expression by 50% within 4 hours after cell stimulation (Figure 1C). Immunoblots show corresponding changes in expression of P2X1 and P2X4 protein (Figure 1D). T-cell stimulation induces ATP release and extracellular ATP enhances IL-2 production of activated T cells.¹⁵  Such findings and the current data implicate P2X1 and P2X4 receptors in mediating T-cell activation in response to extracellular ATP.

To study the potential role of P2X1 and P2X4 receptors in T-cell activation, we expressed EGFP-tagged receptor constructs in Jurkat cells. Upon stimulation of cells with microbeads bearing antibodies to CD3 and CD28 coreceptors, EGFP-tagged P2X1 and P2X4 receptors aggregate in clusters and translocate to the contact sites with the microbeads. A large portion of P2X4 receptors on the cell surface translocate to these contact sites within 2 minutes after cell stimulation (Figure 2). Cell stimulation reduced the amount of fluorescent submembranous vesicles, suggesting the integration of receptor-loaded vesicles into the cell membrane (Figure 2). While some P2X1 receptor redistribution was evident at 2-6 minutes (Figure 2), this translocation process is less rapid than that of P2X4 and requires ≥ 15 minutes for significant receptor accumulation at the contact site (Figure 3). Although P2X1 and P2X4 receptors differ in their kinetics of translocation to the immune synapse, both receptor subtypes remain closely associated with the synapse for at least 1 hour after cell stimulation (Figure 2). P2X1 and P2X4 receptors also translocate to the immune synapse of primary T cells with kinetics of redistribution similar to those observed in Jurkat cells (Figure 3A-B). By contrast, P2X7 receptors do not translocate to the immune synapse (Figure 3C).

These results suggest that translocation of P2X1 and P2X4 receptors to the immune synapse facilitates autocrine/paracrine signaling through these receptors. Thus, P2X1 and P2X4 receptors may amplify TCR signaling in response to ATP released at the synapse by T cells or antigen-presenting cells. By contrast, P2X7 receptors, which do not accumulate at the immune synapse, appear less restricted in terms of the source of ATP to which they respond.

Ca²⁺ entry is a key event in T-cell activation.^5-7  We suspended Fluo-4-loaded Jurkat cells in Ca²⁺-free medium and then stimulated the cells for 30 minutes with CD3/CD28-antibody–bearing beads to allow synapse formation. After returning extracellular Ca²⁺ to the culture medium, we observed Ca²⁺ entry that originates at the immune synapses (supplemental Figure 2). STIM1 and Orai1 are components of CRAC channels that play a critical role in T-cell activation.^1,27,28  STIM1 and Orai1 translocate to the immune synapse where they facilitate localized Ca²⁺ entry in response to TCR stimulation.^4,29  Using confocal laser scanning microscopy, we found that STIM1 and Orai1 reside with P2X1 and P2X4 receptors at the immune synapse of Jurkat cells (Figure 4). These findings suggest that P2X1 and P2X4 receptors along with STIM1 and Orai1 contribute to localized Ca²⁺ entry at the immune synapse.

While the roles of STIM1 and Orai1 in Ca²⁺ entry and T-cell activation are well established, the possible involvement of P2X receptors is less clear.^1-4,7  We found that silencing of P2X1, P2X4, and P2X7 receptors inhibits Ca²⁺ signaling in response to TCR stimulation (Figure 5A), thus implicating all 3 P2X receptor subtypes in Ca²⁺ entry during T-cell activation.

We used pharmacologic inhibitors to evaluate the roles of P2X1 and P2X4 receptors in Ca²⁺ signaling of human primary CD4⁺ T cells. Similar to our findings with Jurkat T cells, we observed that receptor blockade reduces Ca²⁺ influx in response to TCR stimulation (Figure 5B). TNP-ATP, an antagonist of P2X1 and P2X4 receptors, suramin, a general P2 receptor antagonist, and NF023, a P2X1 receptor antagonist, block Ca²⁺ influx in a manner that suggests a predominant role for P2X4 receptors in facilitating Ca²⁺ influx in primary CD4⁺ T cells.

Taken together, these findings indicate that P2X1, P2X4, and P2X7 receptors contribute to Ca²⁺ signaling in T cells. While P2X4 and P2X1 receptors associate with the immune synapse, P2X7 receptors remain uniformly distributed (Figure 3), which suggests that the cellular localization of these receptor subtypes helps determine their roles in T-cell activation.

Sustained intracellular Ca²⁺ levels lead to the activation of NFAT family signaling molecules that control cytokine expression.⁶  Using NFAT-luciferase reporter assays,²⁴  we found that overexpression of P2X1 receptors augments NFAT activation in response to TCR/CD28 stimulation (Figure 6A). Conversely, overexpression of mutated P2X1 or P2X4 receptors, which were rendered dysfunctional by point mutations at T18A and L352W, respectively, or silencing of P2X1 or P2X4 receptors, inhibits NFAT activation (Figure 6B-C). NFAT is required to induce IL-2 gene expression and to promote T-cell activation events.⁶  Silencing of both P2X1 and P2X4 receptors decreased IL-2 mRNA transcription (Figure 6D). Because the effect of silencing of P2X4 receptors was more pronounced than that of P2X1 receptors, P2X4 receptors may have a more important role in activation processes induced by TCR stimulation at the immune synapse. We previously showed that silencing of P2X7 receptors inhibits Ca²⁺ influx and IL-2 gene transcription.^8,10  Here we show that the combined silencing of P2X1, P2X4, and P2X7 receptors results in a stronger suppression of IL-2 transcription than silencing of either receptor subtype alone (Figure 6D-E), which suggests that all 3 P2X receptor subtypes contribute to TCR induced T-cell activation.

To test the involvement of P2X1 and P2X4 receptors in the activation of primary T cells, we used P2 receptor antagonists in experiments with human and mouse lymphocytes. We found that the P2X1/P2X4 receptor antagonist TNP-ATP and the non-specific P2X receptor antagonist suramin inhibit IL-2 mRNA expression of TCR/CD28-stimulated primary human T cells by approximately 50%, while P2X1- and P2X7-selective antagonists (NF023 and A438079, respectively) produce an approximately 25% inhibition (Figure 6F). All 4 antagonists strongly suppressed IL-2 mRNA expression of mouse T cells stimulated with microbeads carrying anti–mouse CD3/CD28 antibodies. These findings implicate autocrine stimulation of P2X1, P2X4, and P2X7 receptors in T-cell activation. The differences in the responses of mouse versus human T cells to these P2 antagonists suggest inter-species variation in P2X receptor expression patterns of mouse and human T cells or differences in the inhibitory efficiency of the purinergic receptor antagonists.^18,30-32  Moreover, P2X1, P2X4, and P2X7 receptors form heteromeric structures that have different combinations of the individual P2X receptor subunits and the composition of these heteromeric ion channels may differ in mouse versus human T cells.^32,33 

Pannexin-1 hemichannels have been identified as gap junction molecules involved in the release of ATP from activated T cells.^10,34  We found that pannexin-1 is uniformly distributed across the cell surface of unstimulated primary T cells but translocates to the immune synapse within 5 minutes in response to TCR/CD28 stimulation (Figure 7A). This rapid translocation suggests that pannexin-1 may facilitate ATP release and contribute to autocrine feedback of TCR stimulation at the immune synapse. Indeed, inhibition of pannexin-1 with the gap junction inhibitor carbenoxolone blocks TCR-mediated ATP release and IL-2 gene transcription (Figure 7B). Treatment of T cells with a more specific pannexin-1 hemichannel blocker, ¹⁰panx-1, reduces Ca²⁺ entry in response to TCR stimulation (Figure 7C) and inhibits IL-2 transcription of human and mouse primary T cells (Figure 7D), while treatment with the scrambled control peptide ^scpanx-1 had no significant effect on Ca²⁺ entry. Pannexin-1 hemichannels thus appear to contribute to ATP release at the immune synapse.

Extracellular ATP can enhance IL-2 production of activated T cells via P2X7 receptors and ATP release from T cells plays an important role in T-cell activation.^8-10  This study demonstrates novel roles for the ATP release channel pannexin-1 and for P2X1 and P2X4 receptors in events that control T-cell activation at the immune synapse, a membrane domain that facilitates T cell–accessory cell communication and antigen recognition.^35-38  We find that P2X1, P2X4, and P2X7 receptors contribute to calcium influx and that pannexin-1, and P2X1 and P2X4, but not P2X7, receptors rapidly redistribute to the immune synapse upon TCR stimulation. Antagonists of pannexin-1 or of P2X1 and P2X4 receptors inhibit IL-2 mRNA expression of activated T cells. Taken together with the release of ATP upon T-cell activation,^8,9,13  the current results indicate that pannexin-1–mediated ATP release and autocrine feedback through P2X1 and P2X4 receptors contribute to T-cell activation at the immune synapse. Furthermore, T-cell activation markedly changes P2X1 and P2X4 receptor expression, thus implicating such changes in receptor expression as additional mechanisms that may produce fine-tuning of T-cell responses.

The translocation of P2X1 and P2X4, but not P2X7 receptors to the immune synapse may relate to their localization in lipid rafts.^39-41  Lipid rafts in T cells form the backbone of the immune synapse^37,38  where, as implied by data here, P2X1 and P2X4 receptors and their ligand, ATP, contribute to Ca²⁺ entry and T-cell activation. The kinetics of translocation of pannexin-1 and P2X4 receptors to the immune synapse imply that these molecules are involved in T-cell activation processes that take place in the first few minutes after cell stimulation, while the slower translocation of P2X1 receptors suggest that these receptors contribute to subsequent activation events. These purinergic activation events may help amplify TCR-induced signals through localized autocrine and paracrine feedback mechanisms. Our findings thus provide an explanation as to how the limited number of TCR molecules at the synapse can trigger robust downstream signaling responses that are needed for T-cell activation.

Our results indicate that P2X1 and P2X4 receptors are redistributed to the immune synapse along with STIM1 and Orai1, critical components of CRAC channels that facilitate Ca²⁺ entry and activation of T cells.^{1,27,28,42-44}  In addition to Orai1 and STIM1, additional proteins such as transient receptor potential–like channels, may also be involved in Ca²⁺ entry.^7,42,45,46  Orai1 and STIM1 can translocate to the immune synapse, where they facilitate localized Ca²⁺ entry.^4,29  Our results here indicate that P2X1 and P2X4 receptors also contribute to Ca²⁺ entry at the immune synapse, with P2X4 receptors contributing to the early phase of this process. Inhibition of purinergic response at the immune synapse, (by blocking pannexin-1–induced ATP release or targeting P2X1 or P2X4 receptors) inhibits Ca²⁺ entry. Combined silencing of P2X1, P2X4, and P2X7 receptors suppresses Ca²⁺ entry and downstream activation events to a greater extent than that of any one receptor alone, suggesting that all 3 P2X receptors regulate T-cell activation. Perhaps, as in other cell types, P2 receptor heteromers composed of P2X1, P2X4, and P2X7 receptor subunits contribute to Ca²⁺ entry at the T-cell immune synapse.^18,33,47-49 

We find that TCR stimulation induces translocation of pannexin-1 hemichannels to the immune synapse and that inhibition of pannexin-1 hemichannels suppresses TCR-induced ATP release, Ca²⁺ entry, and IL-2 expression. These findings imply that pannexin-1 along with P2X1 and P2X4 receptors form a purinergic feedback system that promotes T-cell activation at the immune synapse. This conclusion is supported by results indicating that removal of extracellular ATP by addition of apyrase and other enzymes that hydrolyze extracellular ATP blocks TCR-induced Ca²⁺ signaling and IL-2 production.⁸ 

The current results thus emphasize a key role for purinergic signaling, and pannexin-1 and P2X4 receptors in particular, in modulating TCR signaling at the immune synapse. These purinergic signaling components may serve to amplify TCR-induced activation signals at the immune synapse, thereby facilitating efficient T-cell responses despite the limited number of TCR molecules that are engaged by antigen-presenting cells.

This study was supported in part by National Institutes of Health grants R01 GM-51 477, GM-60 475, AI-072287, and R01 AI-080582 (W.G.J.), and Department of Defense Congressionally Directed Medical Research Programs grant PR043034 (W.G.J.), GM-66 232, and a grant from the Leukemia & Lymphoma Society (P.A.I.), and National Institutes of Health General Clinical Research Center grant 5MO1-RR-00 827-25 (University of California San Diego, San Diego, CA).

National Institutes of Health

Contribution: T.W. performed laser scanning microscopy, immuno-cytochemistry, and functional assays; L.Y. was responsible for generation of P2X receptor plasmids and functional assays; A.E., Y.S., Y.C., and Y.Y. performed functional assays and provided advice; P.A.I. provided advice and assistance with certain experiments and with manuscript preparation; and W.G.J. provided overall direction, supervised project planning and execution, and finalized the manuscript for publication.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Wolfgang G. Junger, Beth Israel Deaconess Medical Center, Harvard Medical School, Department of Surgery, 330 Brookline Ave, Dana 811, Boston, MA 02215; e-mail: wjunger@bidmc.harvard.edu.