Lymphocytes from P2X7 -deficient mice exhibit enhanced ... - CiteSeerX

7 downloads 100276 Views 994KB Size Report
A central role for P2X7 in IL-1 secretion via the Nacht domain- .... figures prepared with Adobe Illustrator 13.0 software (Adobe Systems, San. Jose, CA, USA).
Lymphocytes from P2X7-deficient mice exhibit enhanced P2X7 responses Simon R. J. Taylor,* Mireya Gonzalez-Begne,† Dorothy K. Sojka,‡ Jill C. Richardson,§ Steven A. Sheardown,ⱍⱍ Stephen M. Harrison,ⱍⱍ Charles D. Pusey,¶ Frederick W. K. Tam,¶ and James I. Elliott**,1 *MRC Clinical Sciences Centre and **Department of Molecular Genetics and Rheumatology, Division of Medicine, and ¶Imperial College Kidney and Transplant Institute, Imperial College London, London, United Kingdom; Centers for †Oral Biology and ‡Vaccine Biology and Immunology, Department of Microbiology and Immunology, University of Rochester Medical Center, Rochester, New York, USA; and §Neurosciences Centre of Excellence for Drug Discovery and ⱍⱍTransgenic Technologies, Molecular Discovery Research, GlaxoSmithKline R & D Limited, Harlow, Essex, United Kingdom

Abstract: The purinergic receptor P2X7 is expressed on immune cells, and its stimulation results in the release of IL-1␤ from macrophages. Its absence, as evidenced from the analysis of two independent strains of P2X7-deficient mice, results in reduced susceptibility to inflammatory disease, and the molecule is an important, potential therapeutic target in autoimmunity. However, P2X7 has also been detected in several neuronal cell types, although its function and even its presence in these cells are highly contested, with anti-P2X7 antibodies staining brain tissue from both strains of P2X7ⴚ/ⴚ mice identically to wild-type mice. It has therefore been suggested that neurons express a distinct “P2X7-like” protein that has similar antibody recognition epitopes to P2X7 and some properties of the genuine receptor. In this study, we show that whereas P2X7 activity is absent from macrophages and dendritic cells in P2X7ⴚ/ⴚ animals, T cells from one gene-deficient strain unexpectedly exhibit higher levels of P2X7 activity than that found in cells from control, unmanipulated C57BL/6 mice. A potential mechanism for this tissue-specific P2X7 expression in P2X7ⴚ/ⴚ animals is discussed, as is the implication that the immune and indeed neuronal functions of P2X7 may have been underestimated. J. Leukoc. Biol. 85: 978 –986; 2009. Key Words: inflammation knock-out mice



P2 receptors



rodent



transgenic/

INTRODUCTION The P2X7 receptor is a cation channel activated by high concentrations of ATP and its analog 3⬘-O-(4-benzoyl)benzoyl ATP (BzATP) [1]. Its stimulation is proinflammatory, and activation results in the release of cytokines (notably IL-1␤), changes in plasma membrane lipid distribution, and cell death [2– 4]. A central role for P2X7 in IL-1␤ secretion via the Nacht domain-, leucine-rich repeat-, and PYD-containing protein 3 978

Journal of Leukocyte Biology Volume 85, June 2009

(NALP3) inflammasome [5] has been confirmed in gene-deficient mice generated by GlaxoSmithKline (GSK; Harlow, Essex, UK) [6] and Pfizer (Groton, CT, USA) [7]. Indeed, P2X7 has become an important, potential therapeutic target, and antagonists are currently in Phase II clinical trials for treatment of rheumatoid arthritis. However, the tissue distribution and nonimmune functions of P2X7 remain unclear. In particular, although P2X7 has been detected frequently in several neuronal cell types, its function in these cells is highly contested [8]. Notably, immunohistochemical studies of P2X7 expression, in which P2X7⫺/⫺ mice serve as controls for wild-type staining, have proved problematic. Indeed, P2X7 antibodies appear to stain brain tissue from both strains of P2X7⫺/⫺ mice identically to their wild-type littermates [9, 10], although P2X7 reactivity is abolished in other tissues, such as submandibular gland and lung. Similarly, antibodies identify the 75-kDa band expected for P2X7 in brain tissue of wild-type and knockout animals by Western blotting, although this too is abolished in the submandibular gland of P2X7⫺/⫺ animals [9]. Given the staining pattern exhibited in P2X7⫺/⫺ mice, it has been suggested that neurons express a “P2X7-like” protein that has similar antibody-recognition epitopes to P2X7 and that may have some properties of the genuine receptor [9, 11, 12]. Consistent with this suggestion, cerebellar granule neurons (CGN) cultured from P2X7⫺/⫺ animals exhibit Ca2⫹ uptake in response to BzATP [13]. However, as BzATP is also an agonist for receptors other than P2X7, its target in this assay is unclear. We have shown previously that responses to BzATP are absent from T cells obtained from Pfizer P2X7⫺/⫺ mice [14], showing not only that the functional activity of P2X7 has been removed efficiently, but also that at least in T cells, responses to BzATP are entirely P2X7-dependent. To our surprise, therefore, in conducting equivalent experiments using lymphocytes

1 Correspondence: Department of Molecular Genetics and Rheumatology, Division of Medicine, Imperial College, Hammersmith Hospital, Du Cane Rd., London W12 0NN, UK. E-mail: [email protected] Received April 18, 2008; revised February 4, 2009; accepted February 14, 2009. doi: 10.1189/jlb.0408251

0741-5400/09/0085-978 © Society for Leukocyte Biology

from GSK P2X7⫺/⫺ mice, we have found that T cells in this strain exhibit high levels of P2X7 activity. Moreover, unlike previous studies, the activity we report here is unambiguously that of the genuine P2X7 receptor, reflecting the unique ability of this channel to evoke, when stimulated, the rapid translocation of phosphatidylserine (PS) to the cell surface and cell death [2, 15], as well as shedding of the lymphocyte homing receptor CD62 ligand (CD62L) [16]. Our data, therefore, clearly show that pronounced tissuespecific expression of functional P2X7 protein occurs in the GSK P2X7⫺/⫺ strains. In contrast, although Pfizer P2X7⫺/⫺ also continues to express a P2X7 protein, this is nonfunctional, at least in immune cells. The implications of this finding for the study of P2X7 biology are discussed.

MATERIALS AND METHODS GSK P2X7⫺/⫺ mice have been described elsewhere [6] and were obtained from GSK. Pfizer P2X7⫺/⫺ mice have also been described elsewhere [7] and obtained from Jackson Laboratories (Bar Harbor, ME, USA). Both strains were bred onto a C57BL/6 background. BALB/c and C57BL/6 mice were obtained from Harlan/OLAC (Bicester, UK). All Home Office and local ethical guidelines for the care of laboratory animals were followed.

Generation of the GSK P2X7⫺/⫺ mice Gene targeting was performed in 129 embryonic stem (ES) cells. A targeting vector was assembled from the P2X7 genomic DNA sequence; the 5⬘ and 3⬘ arms of homology are ⬃1.8 kb BglII-NaeI and ⬃7 kb NaeI-SpeI restriction fragments, respectively, and flank a cassette containing the LacZ reporter gene and neomycin-resistance cassette. The same NaeI site was used for both homology arms and is located immediately downstream of the P2X7 translation initiation codon in exon 1. The targeting event therefore results in the disruption of exon 1 by insertion of the LacZ gene (in-frame with the P2X7 ATG) and neomycin selection cassette without deletion of any of the P2X7 sequence.

Southern blotting in the GSK P2X7⫺/⫺ mice Homologous recombination in neomycin-resistant ES cells at the 5⬘ end of the target locus was determined by Southern blot of XbaI-digested ES cell genomic DNA with a 377-bp 5⬘ external probe generated by PCR. The primers used were 5⬘-TCTGGCTCCTTTCACGTCT-3⬘ and 5⬘-CACCAAAGATGAATGTAA3⬘, which detect ⬃3.7 kb and ⬃11 kb at the wild-type and targeted locus, respectively.

formed using FlowJo software (Treestar Inc., Ashland, OR, USA). All results are representative of a minimum of three independent experiments, except where stated.

Cytokine analysis Macrophages adherent to six-well plates were stimulated with 150 ␮M BzATP and supernatants removed at 30 min. Concentrations of IL-1␤ were obtained using a DuoSet ELISA (R&D Systems, Minneapolis, MN, USA), according to the manufacturer’s instructions. The mAb supplied with this kit does not cross-react with pro-IL-1␤.

Real-time live-cell imaging Macrophages adherent to glass-bottomed Petri dishes were maintained in DMEM/10% FCS at 37°C in a humidified environmental chamber and exposed to 150 ␮M BzATP. Cells were imaged with a DeltaVision inverted fluorescence wide-field microscope (Applied Precision, Issaquah, WA, USA) equipped with a 100⫻/1.35 UPlanApo oil-immersion objective lens (Leica, Wetzler, Germany). Sequential bright-field images were taken every 20 s for 20 –30 min. Images were analyzed with SoftWoRx, Version 3.3.5 (Applied Precision), and figures prepared with Adobe Illustrator 13.0 software (Adobe Systems, San Jose, CA, USA).

Immunostaining Macrophages and mesenteric lymph node cells were adhered to glass coverslips and fixed in 4% formaldehyde/4% sucrose in PBS. They were stained with monoclonal anti-CD11bPE or anti-Thy1PE (Becton Dickinson), respectively, followed by rabbit anti-␤-galactosidaseFITC (Acris Antibodies, Germany). These antibodies do not cross-react. Cells were imaged with the DeltaVision inverted fluorescence wide-field microscope.

PCR PCR amplification of the C57BL/6 mouse genomic sequence encompassing the T1352C polymorphism [17] was performed using the forward and reverse primers CCTGTCTAGGCTGTCCCTAT and GCTTATGGAAGAGCTTGGAG for 30 cycles [18]. PCR products were purified using the Qiagen PCR clean-up kit (Qiagen, Hilden, Germany), sequenced using an ABI PRISM Big Dye terminator ready-reaction kit (Applied Biosystems, Warrington, UK), and analyzed on a 3730 ⫻ l DNA analyzer (Applied Biosystems).

RT-PCR

Lymphocyte preparations were obtained as described previously [14]. Peritoneal macrophages were lavaged using DMEM media and maintained in sixwell plates or glass-bottomed Petri dishes. Except where indicated, macrophages were treated with 1 ␮g/ml LPS (Sigma Chemical Co., St. Louis, MO, USA) for 2 h and washed immediately prior to experimentation. Dendritic cells (DCs) were propagated from bone marrow precursors for 7 days in RPMI-1640 culture medium (Gibco, Grand Island, NY, USA) containing 10% FCS, 2% mouse GM-CSF, and 5 ng/ml IL-4 (PeproTech, Rocky Hill, NJ, USA). They were harvested on Day 7 and purity-assessed by flow cytometry.

RNA was extracted from freshly collected mouse tissues using TRIZOL reagent (Sigma Chemical Co., Poole, UK). RT was performed with Superscript III using 1 ␮g template/reaction, except for dorsal root ganglion, where 500 ng was used. PCRs were performed with Qiagen Multiplex mastermix for the following protocol: 95ºC for 10 min, followed by 30 cycles of 95ºC for 30 s, 66ºC for 30 s, and 72ºC for 30 s. For quantitative RT-PCR (qRT-PCR), RT from total RNA was effected using a first-strand cDNA synthesis kit (Roche Applied Science) according to the manufacturer’s instructions, and qRT-PCR was then performed using the 2⫻ Sensimix DNA kit with SYBR-green probes (Quantace), together with a 72-well Rotogene-6000TM machine and Rotor-Gene 6000 series software 1.7 (Corbett Research, Mortlanke, NSW, Australia). Specific primers for P2X7 and the housekeeping genes ribosomal phosphoprotein PO (RPLPO) and hypoxanthine-guanine phosphoribosyltransferase (HPRT) are summarized in Table 1.

Real-time flow cytometry

Western blot analysis

Mesenteric lymph node cells suspended in DMEM were stained with antiCD4FITC, anti-CD4PerCP, anti-CD8APC, anti-CD8PerCP, anti-CD19PerCP, and anti-CD62LPE antibodies (Becton Dickinson, San Jose, CA, USA), as indicated. Cells were washed and resuspended in DMEM and where indicated, equilibrated with Annexin V (AV)FITC or propidium iodide (PI; Becton Dickinson). Where cells were stimulated, baseline fluorescence was established for 30 – 60 s prior to addition of 150 ␮M BzATP (Sigma Chemical Co., St. Louis, MO, USA). Cells were monitored continuously in real-time for PS exposure and PI uptake by flow cytometry on a FACSCalibur machine. Analysis was per-

Isolated lymph nodes were homogenized, and 60 ␮g protein was then separated on a 10% SDS-PAGE Tris-glycine mini-gel (Bio-Rad, Hercules, CA, USA). Proteins were transferred onto polyvinylidene difluoride membrane (Invitrogen, Carlsbad, CA, USA) and incubated with primary anti-P2X4 antibody (Chemicon/Millipore, Temecula, CA, USA) at a dilution of 1:300 in 3% BSA blocking solution. After washing, the membrane was incubated with HRP-conjugated goat anti-rabbit IgG secondary antibody (Pierce, Rockford, IL, USA) at a dilution of 1:2500. Labeled proteins were visualized using an ECL detection kit (GE-Amersham Biosciences, Piscataway, NJ, USA).

Cellular preparations

Taylor et al. P2X7–/– T cells exhibit enhanced P2X7 responses

979

TABLE 1.

Specific P2X7 Primer Pairs Used Forward primer

Exon 1–2a (insert-spanning) Exon 1–2b (downstream of insert) Exon 3–4a Exon 3–4b HPRT RPLPO

CTGTGGTCTAGCCTGGGAAG GACAAACAAAGTCACCCGGA CTATGTCAAGTCAGAAGGCCAAG GATACAGTTCAGTCTTCCGGTTC ATTATGCCGAGGATTTGGAA GGCGACCTGGAAGTCCAACT

Statistics The unpaired Student’s t-test was used to calculate P values and determine statistical significance.

RESULTS P2X7⫺/⫺ CD4⫹ T cells exhibit enhanced P2X7 activity We have shown previously that consistent with expectations, responses to BzATP diagnostic of P2X7 activity (i.e., PS exposure, cell shrinkage, cell death, and CD62L shedding) are completely absent in T cells obtained from Pfizer P2X7⫺/⫺ mice [14]. To verify the phenotype of the P2X7⫺/⫺ mice that we obtained from GSK, we compared the kinetics of P2X7-stimulated PS exposure and cell death in CD4⫹ lymphocytes from GSK P2X7⫺/⫺ and C57BL/6 control mice. CD4⫹ lymphocytes were labeled with anti-CD4PerCP, and after basal binding of AVFITC was established, cells were treated with 150 ␮M BzATP. AV binding and PI uptake were then measured simultaneously to measure exposure of PS and breakdown of the cell membrane, respectively. Unexpectedly, P2X7⫺/⫺ CD4⫹ lymphocytes not only responded to BzATP but also exhibited a greater response to BzATP stimulation than those from P2X7⫹/⫹ mice, and PS exposure occurred faster and in a higher proportion of T cells (Fig. 1A). The proportion of cells undergoing cell death within 15 min of BzATP stimulation was also higher in the P2X7⫺/⫺ CD4⫹ cells than in P2X7⫹/⫹ cells, and it remained so for the rest of the time period over which the cells were analyzed (Fig. 1B). To confirm the unexpected presence of P2X7 activity in P2X7⫺/⫺ mice, we looked at another characteristic feature of P2X7 stimulation: the shedding of CD62L [20]. Consistent with the results above, shedding of CD62L occurred more rapidly from P2X7⫺/⫺ cells (Fig. 1C) than P2X7⫹/⫹ cells. We next assessed the pharmacological characteristics of the response. Consistent with the response being a result of P2X7 activity, the EC50 of BzATP was ⬃30 ␮M, whereas that of ATP was in excess of tenfold greater at ⬃2 mM (Fig. 1D). Finally, we compared the responses of Pfizer and GSK P2X7⫺/⫺ CD4⫹ T cells. GSK P2X7⫺/⫺ CD4⫹ lymphocytes responded to BzATP stimulation with PS exposure and uptake of the dye YO-PRO-1, whereas Pfizer P2X7⫺/⫺ CD4⫹ cells in the same tube failed to respond (Fig. 1, E and F).

GSK P2X7⫺/⫺ mice express the P451 allele Although some leakage of expression in gene-targeted mice might not be remarkable, the enhanced P2X7 responses exhib980 Journal of Leukocyte Biology Volume 85, June 2009

Reverse primer TCTTTCCGCTGGTACAGCTT TCTTTCCGCTGGTACAGCTT TCCATCCACCCCTTTTTACAAC TCCATCCACCCCTTTTTACAAC CCCATCTCCTTCATGACATCT CCATCAGCACCACAGCCTTC

ited by GSK P2X7⫺/⫺ T cells were surprising. We considered the possibility that the paradoxical, enhanced responses reflected the construction of the P2X7 strain. Common strains of laboratory mice express one of two P2X7 alleles, encoding high- or low-activity forms of the receptor as a result of a P451L polymorphism [17]. As is standard in the generation of gene-targeted mice, the P2X7⫺/⫺ mutation was generated in the 129/Sv mouse strain, which carries the high-activity 451P allele. However, the mutation was then backcrossed onto the C57BL/6 background, which carries the low-activity 451L allele. C57BL/6 mice are used as the control strain in comparisons between P2X7⫺/⫺ and wild-type mice. Therefore, we hypothesized that the enhanced responses of GSK P2X7⫺/⫺ T cells might reflect T cell expression of the high-activity P2X7 allele, which is then compared with C57BL/6 mice that have normal expression of the low-activity allele. If P2X7 expression were only eliminated partially, the resulting mouse carrying the high-activity P2X7 allele might reasonably be expected to exhibit significant P2X7 responses. Indeed, when we compared the responses of GSK P2X7⫺/⫺ CD4⫹ T cells with those from a high-activity receptor strain, BALB/c, the kinetics of PS exposure in the GSK P2X7⫺/⫺ animals appeared intermediate in character between low- and high-activity P2X7 allelic responses (Fig. 1A). Similarly, although the rate of BzATP-induced cell death was similar in GSK P2X7⫺/⫺ and BALB/c mice, both were higher than that observed in C57BL/6 mice (Fig. 1B). To support this hypothesis, we sequenced DNA obtained from GSK P2X7⫺/⫺ and control C57BL/6 mice. As expected, GSK P2X7⫺/⫺ mice proved to be homozygous for the P2X7-P allele of P2X7 [17], which is associated with high sensitivity to stimulation, whereas C57BL/6 mice were homozygous for the low-sensitivity allele P2X7-L (Fig. 1G).

P2X7⫺/⫺ macrophages and DCs do not respond to P2X7 stimulation P2X7 plays a key role in the secretion of the leaderless cytokine IL-1␤. Indeed, loss of P2X7-stimulated IL-1␤ release from macrophages was used originally to assess the veracity of the P2X7⫺/⫺ phenotype in GSK and Pfizer strains [7, 21]. Given the apparently strong P2X7 responses in GSK P2X7⫺/⫺ T cells, we attempted to confirm the original findings of loss of macrophage IL-1␤ secretion in our colony. Consistent with previous results, macrophages from GSK P2X7⫺/⫺ animals failed to respond to stimulation with 150 ␮M BzATP, in contrast to those from C57BL/6 control animals, which released a significantly greater amount of IL-1␤ (P⬍ 0.01; Fig. 2A). http://www.jleukbio.org

Fig. 1. T cells from GSK P2X7⫺/⫺ mice exhibit enhanced responses to P2X7 stimulation. (A and B) Mesenteric lymph node cells from P2X7⫺/⫺, C57BL/6, and BALB/c mice were labeled with anti-CD4APC, maintained at 37°C, and incubated with AVFITC and PI for 2 min; only cells staining positively for CD4 were analyzed thereafter. (C) Mesenteric lymph node cells from P2X7⫺/⫺ and C57BL/6 mice were labeled with anti-CD4APC and antiCD19PerCP and maintained at 37°C. (D and E) Mesenteric lymph node cells from GSK P2X7⫺/⫺, C57BL/6, and Pfizer P2X7⫺/⫺ mice were labeled with anti-CD8APC, anti-CD8PerCP, and anti-CD8PE, respectively, before being mixed and incubated with AVFITC or YO-PRO-1 for 2 min. In all panels, cells were stimulated with the P2X7 agonist BzATP (150 ␮M) at the time indicated by arrows and monitored in real-time by flow cytometry. (A) The proportion of cells with externalized PS as measured by binding of AVFITC. The graph consists of the superimposition of traces for P2X7⫺/⫺ (n⫽4), C57BL/6 (n⫽3), and BALB/c (n⫽2) mice, as indicated. Notably, the kinetics of the response to P2X7 stimulation in GSK P2X7⫺/⫺ CD4⫹ T cells is intermediate between that observed in wild-type C57BL/6 P2X7⫹/⫹ CD4⫹ cells and BALB/c P2X7⫹/⫹ CD4⫹ cells. (B) The rate of cell death (cells taking up PI) following P2X7 stimulation. A single representative trace for BALB/c (blue), GSK P2X7⫺/⫺ (red), and C57BL/6 (green) CD4⫹ T cells is shown, as indicated. There is an increased rate of cell death in GSK P2X7⫺/⫺ CD4⫹ compared with wild-type C57BL/6 P2X7⫹/⫹ CD4⫹ cells. This rate of cell death is similar but slightly less than that observed in BALB/c P2X7⫹/⫹ CD4⫹ cells. (C) Shedding of CD62L in response to P2X7 stimulation with BzATP in P2X7⫺/⫺ (blue) and C57BL/6 (red) mice. Cells were labeled with anti-CD62LFITC and stimulated with 150 ␮M BzATP. Fluorescence then decreases as CD62L is shed with its associated antibody. CD4⫹ T cells from both animals respond by shedding CD62L rapidly from the cell surface, and the GSK P2X7⫺/⫺ CD4⫹ T cells once again responded more rapidly. CD19⫹ B cells, which do not express P2X7 [19] and thus, do not respond, are shown as an internal control. (D) T cell dose-response curves to P2X7 stimulation with ATP (upper) and BzATP (lower) for BALB/c (high activity), C57BL/6 (low activity), and GSK P2X7⫺/⫺ mice, as measured by YO-PRO-1 uptake. The similar EC50 values for ATP and BzATP between strains and the finding that EC50 (ATP) ⬎⬎ EC50 (BzATP) support the contention that this assay represents P2X7 activity and that this activity is retained in the GSK P2X7⫺/⫺ mice. (E) The proportion of cells with externalized PS as measured by binding of AVFITC. The graph consists of the superimposition of traces for GSK P2X7⫺/⫺ (n⫽3), C57BL/6 (n⫽3), and Pfizer P2X7⫺/⫺ (n⫽3) mice, as indicated. (F) Similarly shows uptake of the dye YO-PRO-1. (G) The results of gene sequencing in the GSK P2X7⫺/⫺ mice. BALB/c mice possess the high-activity 451-P allele and C57BL/6 mice the low-activity 451-L allele. GSK P2X7⫺/⫺ mice were generated on the C57BL/6 background and should possess the 451-L form but in fact, have the high-activity 451-P form, explaining the higher levels of P2X7 activity seen in T cells from this strain. KO, Knockout.

Thus, it appeared that GSK P2X7⫺/⫺ T cells, but not macrophages, retain P2X7 activity. However, it was also possible that the differences in the two sets of results reflected the assays used rather than the cell type; that is, P2X7-dependent IL-1␤ secretion might be ablated but not other consequences of P2X7 activation such as PS translocation and cell death. Therefore, we performed real-time microscopy of macrophages to further assess their responses to P2X7 stimulation. Peritoneal macrophages cultured from wild-type animals demonstrated marked blebbing in response to BzATP stimulation, followed by cell swelling and death over a 20-min period. In contrast,

those from GSK P2X7⫺/⫺ animals did not respond to stimulation (Fig. 2B). Loss of P2X7 activity was also apparent in DCs, and those propagated from C57BL/6 animals retracted dendritic processes and shrunk in response to BzATP stimulation; those from P2X7⫺/⫺ animals failed to respond (Fig. 2C). Thus, GSK P2X7⫺/⫺ macrophages and DCs but not T cells lack P2X7 activity.

P2X7 and LacZ transgene expression in immune cells of GSK P2X7⫺/⫺ mice We next looked at expression of P2X7 protein in GSK P2X7⫺/⫺ mice by immunohistochemistry. We elected to examine splenic

Taylor et al. P2X7–/– T cells exhibit enhanced P2X7 responses

981

Fig. 2. Macrophages and DCs from GSK P2X7⫺/⫺ mice are unresponsive to P2X7 stimulation. (A) Mature IL-1␤ release from peritoneal macrophages cultured from wildtype C57BL/6 and GSK P2X7⫺/⫺ mice. Addition of the agonist BzATP caused a release of IL-1␤ from LPS-primed C57BL/6 but not GSK P2X7⫺/⫺ macrophages. Data shown are mean ⫾ SEM for wild-type (n⫽3) and P2X7⫺/⫺ (n⫽3) mice. **, P ⬍ 0.01. (B) Peritoneal macrophages from wild-type C57BL/6 and GSK P2X7⫺/⫺ mice were cultured, primed with 1 ␮g/ml LPS for 2 h, and then treated with 150 ␮M BzATP to stimulate P2X7. Macrophages from C57BL/6 control animals demonstrate marked blebbing of the cell membrane within 10 min of P2X7 stimulation, followed by cell swelling within 20 min of P2X7 stimulation. In contrast, those from GSK P2X7⫺/⫺ animals fail to respond. (C) DCs from wild-type C57BL/6 and GSK P2X7⫺/⫺ mice were cultured, primed with 1 ␮g/ml LPS for 2 h, and then treated with 150 ␮M BzATP to stimulate P2X7. DCs from C57BL/6 control animals demonstrate retraction of cell processes and cell shrinkage over a period of 30 min, in contrast to those from GSK P2X7⫺/⫺ animals, which fail to respond.

follicles, as murine B cells have been reported previously not to express P2X7 [19], and the B cell core of each follicle would thus provide an internal control for the T cell mantle when a polyclonal antibody was applied. In wild-type animals, antiP2X7 antibody efficiently stained the T cell mantle but not the

B cell core of splenic follicles. This pattern was maintained in GSK P2X7⫺/⫺ follicles (Fig. 3A), suggesting retention of protein and supporting our functional results. Interestingly, examination of Pfizer P2X7⫺/⫺ mice produced similar results, suggesting retention of nonfunctional protein expression in T cells

Fig. 3. P2X7 expression is retained in T cells of Pfizer and GSK P2X7⫺/⫺ mice. (A) Splenic follicles stained for P2X7 using the Alomone anti-P2X7 antibody are shown for wild-type C57BL/6, GSK P2X7⫺/⫺, and Pfizer P2X7⫺/⫺ mice. Efficient antibody staining occurs in the T cell mantle zone in each strain, whereas the B cell core, where P2X7 is known not to be expressed [19], is not stained. (B) Western blotting of lymph nodes from Pfizer P2X7⫺/⫺ and wild-type (WT) mice demonstrates retention of the expected 68-kDa band in the gene-targeted animals, although the protein appears slightly smaller in the gene-targeted animals, which would be consistent with the C-terminal deletion used in these animals.

982

Journal of Leukocyte Biology Volume 85, June 2009

http://www.jleukbio.org

of this strain. In agreement with these results, a Western blot of lymph nodes taken from Pfizer P2X7⫺/⫺ and wild-type mice, probed with the extracellular Alomone anti-P2X7 antibody, confirmed the presence of the expected 68-kDa band in the gene-targeted animals, although the protein appeared slightly smaller in the gene-targeted animals, which would be consistent with the C-terminal deletion used in these animals (Fig. 3B). We next examined expression of the transgene ␤-galactosidase in GSK P2X7⫺/⫺ animals. ␤-Galactosidase expression has been shown previously to be present in the submandibular gland but not CNS neurons of GSK P2X7⫺/⫺ animals [9], and we wanted to determine whether changes in its expression would match the differences in P2X7 activity discovered between T cells and macrophages. In line with our functional results, examination of lymphocyte and macrophage preparations demonstrated the presence of ␤-galactosidase in macrophages but not in T cells—the simplest hypothesis being that the latter splices out the disruptive LacZ insertion to generate functional protein (Fig. 4).

RT-PCR demonstrates alternative splicing of P2X7 in P2X7⫺/⫺ lymphocytes but abolished expression in P2X7⫺/⫺ macrophages Next, we analyzed levels of P2X7 mRNA expression using RT-PCR. The P2X7⫺/⫺ construct was created by inserting a LacZ gene into the beginning of exon 1 [21] (Fig. 5A). Homologous recombination was determined by Southern blot of XbaI-digested ES cell genomic DNA, using an external probe generated by PCR, which detects ⬃3.7 kb and ⬃11 kb at the wild-type and targeted locus, respectively (Fig. 5B). We considered it possible that P2X7 splice variants lacking exon 1 might exist in P2X7⫺/⫺ T cells and therefore, designed primer pairs to examine for the presence of exon 1 and exon 3 mRNA, in case alternative splicing was allowing mRNA transcription to occur downstream of exon 1 in the P2X7⫺/⫺ mice. We performed RT-PCR analysis on several tissues obtained from GSK P2X7⫺/⫺ and control C57BL/6 mice. This demonstrated the absence of mRNA encoded from exon 1 in all GSK

P2X7⫺/⫺ tissue tested, in contrast to C57BL/6 control mice. However, mRNA encoded from exon 3 was found to be present in all GSK P2X7⫺/⫺ tissue tested, suggesting that alternative splicing of P2X7 is indeed occurring in these animals (Fig. 6A). We next performed qRT-PCR on preparations of immune cells from GSK P2X7⫺/⫺ and C57BL/6 animals using the exon 3 primer pair. Consistent with the results of our functional studies, significant, albeit diminished, P2X7 expression remained in the P2X7⫺/⫺ T cells, in contrast to P2X7⫺/⫺ macrophages and DCs, in which P2X7 mRNA expression was abolished (P⬍0.05; Fig. 6B).

Changes in P2X4 expression do not underlie loss of P2X7 function in Pfizer P2X7⫺/⫺ mice P2X receptors usually function as homo- or heteromers, and P2X7 is shown recently to associate with P2X4 [22], with which it is coexpressed in lymphocytes [23], although this remains controversial [24]. The role of P2X4 in P2X7 activity remains unknown. Given the proximity of the murine loci for P2X4 and P2X7 (⬍100 kb), we considered the possibility that the loss of functional P2X7 responses in Pfizer P2X7⫺/⫺, although not GSK P2X7⫺/⫺, T cells might be a result of unexpected alterations in P2X4 expression in the former. Uptake of YO-PRO-1, for example, although characteristic of P2X7 activity, has also been reported for P2X4 [25]. The GSK P2X7⫺/⫺ construct was made by inserting a LacZ coding sequence into exon 1, creating a frameshift mutation [21], whereas the Pfizer P2X7⫺/⫺ construct was created by an exon 5 deletion and neomycin cassette insertion [7], so potential effects on P2X4 expression may differ between strains. In particular, it appeared possible that if P2X4 plays a role in P2X7-mediated PS exposure, loss of responses to BzATP in Pfizer P2X7⫺/⫺ mice might reflect unintended down-regulation of P2X4 in this strain. Therefore, we analyzed levels of P2X4 expression in T cells isolated from Pfizer P2X7⫺/⫺ and C57BL/6 P2X7⫹/⫹ mice by Western blotting using an anti-P2X4 antibody. There was no detectable difference in P2X4 expression between P2X7⫺/⫺ and wild-type mice, suggesting that alterations in P2X4 expression do

Fig. 4. ␤-Galactosidase expression occurs in GSK P2X7⫺/⫺ lymphocytes but not in macrophages. ␤-Galactosidase staining is shown for lymphocyte and macrophage preparations from GSK P2X7⫺/⫺ and C57BL/6 control animals. As expected, wild-type animals do not show any ␤-galactosidase expression. In contrast, P2X7⫺/⫺ lymphocytes, but not macrophages, are positive for its presence.

Taylor et al. P2X7–/– T cells exhibit enhanced P2X7 responses

983

Fig. 5. Strategies used to create P2X7 gene-targeted mice. (A) The genomic structure of the murine P2X7 gene. Exons are numbered and indicated by boxes, and introns are indicated by horizontal lines between the exons. Below the genomic structure are the gene targeting strategies used in the GSK and Pfizer P2X7⫺/⫺ mice; the GSK construct has a LacZ gene inserted 2 bp after the initial ATG at the 5⬘ end of exon 1, whereas the Pfizer mice contains a deletion of nucleotides 1527–1607. TM1, Transmembrane domain 1; TM2, transmembrane domain 2; Neo, neomycin. (B) Southern analysis was performed as per Materials and Methods and used to verify gene targeting in the GSK P2X7⫺/⫺ mice.

not underlie the loss of P2X7 function in Pfizer P2X7⫺/⫺ T cells (Fig. 7A). To explore whether altered P2X4 activity might underlie the apparent increase in P2X7 responses in GSK P2X7⫺/⫺ mice, we examined the effect of ivermectin, a potentiator of P2X4 channels [26]. Interestingly, the addition of 3 ␮M ivermectin reduced the rate of PS exposure and uptake of the dye YO-PRO-1 in GSK P2X7⫺/⫺ and C57BL/6 animals, suggesting a possible role for P2X4 in down-regulating P2X7 activity. However, this ivermectin-dependent down-regulation of P2X7 activity was similar in Pfizer P2X7⫺/⫺, GSK P2X7⫺/⫺, and C57BL/6 strains, indicating that differences in P2X4 activity do not underlie loss of P2X7 activity in the former (Fig. 7, B and C). Nevertheless, as ivermectin has targets other than P2X4, clarification of the role of P2X4 in regulation of P2X7 responses awaits studies to be conducted in P2X4⫺/⫺ mice.

DISCUSSION P2X7⫺/⫺ mice are an important resource in the study of the role of this receptor in neuronal and immune functioning. Not

only is P2X7 activation by extracellular ATP an important signal in the release of IL-1␤ by the NALP3 inflammasome [27], but also, the P2X7 protein has been found in synaptic terminals of the CNS, cerebellum, and hippocampus [11, 28], as well as in neuromuscular junctions in the peripheral nervous system [29]. However, the function and pattern of expression of P2X7 in neurons remain highly controversial. In particular, comparisons of P2X7⫺/⫺ and wild-type brain tissue show equivalent immunohistochemical staining [9, 10]. Moreover, CGN from P2X7⫺/⫺ mice appear to retain some responses to BzATP, a common, although not entirely specific, P2X7 agonist [13]. It has, therefore, been argued that neurons express a protein that cross-reacts with anti-P2X7 antibodies and has some P2X7-like activity [8, 9, 12, 13] and perhaps express relatively little P2X7 protein itself. The work reported here demonstrates the presence of fully functional P2X7 responses in T cells from GSK P2X7⫺/⫺ animals but not in macrophages or DCs. This is a significant finding for several reasons. Importantly, the assumption that neuronal staining of P2X7⫺/⫺ mice with anti-P2X7 antibodies

Fig. 6. P2X7 mRNA expression is knocked-down in T cells but abolished in macrophages and DCs from GSK P2X7⫺/⫺ mice. (A) P2X7 mRNA expression, as measured by RT-PCR in GSK P2X7⫺/⫺ and C57BL/6 wild-type mice, is shown. Tissues examined are: 1, testis; 2, large intestine; 3, cortex; 4, lung; 5, spinal cord; 6, duodenum; 7, striatum; 8, dorsal root ganglion; 9, lymph node; 10, spleen; 11, liver. Each pair consists of the wild-type followed by a gene-deficient animal. Notably, mRNA encoded from exons 1–2 (primer pairs exon 1–2a and exon 1–2b) is abolished in P2X7⫺/⫺ mice, whereas that encoded from exons 3– 4 (primer pair exon 3– 4a) is preserved. (B) P2X7 mRNA expression as measured by qRT-PCR in wild-type (filled bars) and GSK P2X7⫺/⫺ (hatched bars) mice is shown. Lymph node (LN; n⫽5), macrophage (Mac; n⫽3), and DC (n⫽3) samples were obtained as described in Materials and Methods. Expression level of mRNA encoded from exons 3– 4 (primer pair exon 3– 4b) is indicated relative to that of the control transcript RPLPO. *, P ⬍ 0.05; **, P ⬍ 0.01.

984

Journal of Leukocyte Biology Volume 85, June 2009

http://www.jleukbio.org

Fig. 7. P2X4 expression is similar in wildtype C57BL/6 and Pfizer P2X7⫺/⫺ mice. (A) Western blot of mesenteric lymph node cells taken from C57BL/6 and Pfizer P2X7⫺/⫺ mice probed with anti-P2X4 antibody demonstrates the presence of the expected 65-kDa band in both strains, with no detectable difference in the level of expression. (B and C) The effects of 3 ␮M ivermectin on YO-PRO-1 uptake and externalization of PS following P2X7 stimulation are shown for GSK P2X7⫺/⫺ and C57BL/6 control animals. Red traces indicate the addition of ivermectin in each case.

could not reflect the presence of P2X7 was inferred from the belief that as P2X7 was lacking from P2X7⫺/⫺ macrophages, P2X7 must have been removed from all tissues; this is now clearly not the case. Notably, the protein active in GSK P2X7⫺/⫺ T cells is functionally indistinguishable from wildtype P2X7, eliciting highly characteristic responses, including rapid cell-surface exposure of PS, cell death, and shedding of CD62L. Moreover, the complete absence of these responses from Pfizer P2X7⫺/⫺ mice [14] strongly argues that these activities are a result of P2X7 itself. Together with the retention of significant levels of P2X7 mRNA in GSK P2X7⫺/⫺ T cells and efficient T cell staining with anti-P2X7 antibodies, the responses in GSK P2X7⫺/⫺ T cells therefore reflect the activity of the genuine receptor and not that of a distinct P2X7-like protein. We cannot conclude from our studies that the neuronal P2X7-like protein is also the genuine P2X7 receptor. Indeed, the neuronal P2X7-like protein is found in GSK P2X7⫺/⫺ and Pfizer P2X7⫺/⫺ strains, whereas we only find T cell activity in GSK P2X7⫺/⫺ mice. Nevertheless, the argument that lack of P2X7 in macrophages implies lack of P2X7 in other tissues is clearly incorrect. Indeed, the neuronal protein appeared identical to genuine P2X7 by immunohistochemistry and Western blotting [9]. Therefore, we suggest that the simplest interpretation of the available data is that the protein detected in neuronal tissue of P2X7⫺/⫺ mice is a splice variant of P2X7 itself and not a distinct, P2X7-like protein. Our data show that GSK P2X7⫺/⫺ mice are, in effect, inadvertent, tissue-specific gene-knockout animals. It is unknown how common this phenomenon is among other genetargeted mice, but we suggest that the likelihood of its occurrence is probably increased for proteins such as P2X7, which

possess multiple splice variants [30]. Alternative splicing is highly abundant in the brain relative to other tissues [31, 32], and the likelihood that leaky expression of a protein will occur in a gene-targeted animal should be related to the selective pressure for its activity. Thus, if P2X7 plays an essential function in neurons, and expression of splice variants is possible in P2X7 gene-targeted mice, then such expression will occur. Perhaps the most unusual of our findings is that P2X7 activity in GSK P2X7⫺/⫺ T cells is not only retained but also, in fact, higher than in control animals. Indeed, it was this finding that alerted us to the perhaps more common phenomenon of tissue-specific leakage of protein expression in genetargeted mice. The generation of both strains of P2X7⫺/⫺ mice involved isolating the P2X7 gene from a genomic library drawn from 129/Sv mice, a strain that possesses the high-activity P2X7 451P polymorphism. In the GSK P2X7⫺/⫺ mice, a LacZcoding sequence was inserted at the 5⬘ end of P2X7 exon 1, creating a frame-shift mutation [21], whereas the construct for the Pfizer P2X7⫺/⫺ had nucleotides 1527–1607 deleted from exon 13 and a neomycin cassette inserted in the 5⬘-to-3⬘ direction [9]. In both strains, the P2X7 mutation was subsequently backcrossed onto a C57BL/6 background, which possesses the low-activity P2X7 451L allele. Thus, P2X7⫺/⫺ animals are always compared with wild-type C57BL/6 animals, which have P2X7 function dependent on normal levels of the low-activity 451L allele. Therefore, we suggest that the explanation for the increased activity of P2X7 in GSK P2X7⫺/⫺ T lymphocytes is that these cells express somewhat reduced levels of P2X7 protein but that this is of the high-activity 451P P2X7 allele.

Taylor et al. P2X7–/– T cells exhibit enhanced P2X7 responses

985

Our data suggest that at least some studies of immune responses using P2X7⫺/⫺ mice should be treated with caution, particularly where T cell expression of P2X7 in GSK P2X7⫺/⫺ mice may have influenced the findings. For example, although P2X7 deficiency was shown to protect GSK P2X7⫺/⫺ mice from adjuvant arthritis [21], it can be envisaged that a greater effect would have been detected had P2X7 not been expressed in T cells in these mice. In conclusion, we have shown that T cells from the widely used GSK P2X7⫺/⫺ mouse strain exhibit higher levels of P2X7 activity than those from the wild-type strain. In contrast, Pfizer P2X7⫺/⫺ do continue to express a P2X7 protein but one that is nonfunctional, at least in immune cells. The data thus suggest, but do not prove, that the activity of a P2X7-like protein in neurons from P2X7⫺/⫺ mice may also reflect that of the genuine receptor, albeit a splice variant. Together, our findings indicate that the role of P2X7 may have been underestimated, not only in neuronal functioning but also in the immune system.

ACKNOWLEDGMENTS S. R. J. T. was supported by the Medical Research Council of the UK. M. G-B. was supported in part by National Institutes of Health (NIH) grant DE016960, and D. K. S. was supported by NIH grant T32-AI0728. C. D. P. and F. W. K. T. were supported by a Wellcome Trust project grant. We thank Iain Chessell and Jon Hatcher (GSK, Harlow, UK) for the use of P2X7-deficient mice and Judy Latcham and Mick Fulleylove (GSK) for mouse breeding and tissue supply for RT-PCR. We also thank Prof. Robert Unwin for reading this manuscript and for helpful comments. We thank Dr. Selina Raguz for help with qRT-PCR and Dr. John McDaid for help with ELISAs.

REFERENCES 1. North, R. A. (2002) Molecular physiology of P2X receptors. Physiol. Rev. 82, 1013–1067. 2. Bulanova, E., Budagian, V., Orinska, Z., Hein, M., Petersen, F., Thon, L., Adam, D., Bulfone-Paus, S. (2005) Extracellular ATP induces cytokine expression and apoptosis through P2X7 receptor in murine mast cells. J. Immunol. 174, 3880 –3890. 3. Gu, B. J., Wiley, J. S. (2006) Rapid ATP-induced release of matrix metalloproteinase 9 is mediated by the P2X7 receptor. Blood 107, 4946 – 4953. 4. MacKenzie, A., Wilson, H. L., Kiss-Toth, E., Dower, S. K., North, R. A., Surprenant, A. (2001) Rapid secretion of interleukin-1␤ by microvesicle shedding. Immunity 15, 825– 835. 5. Mariathasan, S., Weiss, D. S., Newton, K., McBride, J., O’Rourke, K., Roose-Girma, M., Lee, W. P., Weinrauch, Y., Monack, D. M., Dixit, V. M. (2006) Cryopyrin activates the inflammasome in response to toxins and ATP. Nature 440, 228 –232. 6. Sikora, A., Liu, J., Brosnan, C., Buell, G., Chessel, I., Bloom, B. R. (1999) Cutting edge: purinergic signaling regulates radical-mediated bacterial killing mechanisms in macrophages through a P2X7-independent mechanism. J. Immunol. 163, 558 –561. 7. Solle, M., Labasi, J., Perregaux, D. G., Stam, E., Petrushova, N., Koller, B. H., Griffiths, R. J., Gabel, C. A. (2001) Altered cytokine production in mice lacking P2X(7) receptors. J. Biol. Chem. 276, 125–132. 8. Anderson, C. M., Nedergaard, M. (2006) Emerging challenges of assigning P2X7 receptor function and immunoreactivity in neurons. Trends Neurosci. 29, 257–262. 9. Sim, J. A., Young, M. T., Sung, H. Y., North, R. A., Surprenant, A. (2004) Reanalysis of P2X7 receptor expression in rodent brain. J. Neurosci. 24, 6307– 6314.

986

Journal of Leukocyte Biology Volume 85, June 2009

10. Kukley, M., Stausberg, P., Adelmann, G., Chessell, I. P., Dietrich, D. (2004) Ecto-nucleotidases and nucleoside transporters mediate activation of adenosine receptors on hippocampal mossy fibers by P2X7 receptor agonist 2⬘-3⬘-O-(4-benzoylbenzoyl)-ATP. J. Neurosci. 24, 7128 –7139. 11. Miras-Portugal, M. T., Diaz-Hernandez, M., Giraldez, L., Hervas, C., Gomez-Villafuertes, R., Sen, R. P., Gualix, J., Pintor, J. (2003) P2X7 receptors in rat brain: presence in synaptic terminals and granule cells. Neurochem. Res. 28, 1597–1605. 12. Sperlagh, B., Vizi, E. S., Wirkner, K., Illes, P. (2006) P2X7 receptors in the nervous system. Prog. Neurobiol. 78, 327–346. 13. Sanchez-Nogueiro, J., Marin-Garcia, P., Miras-Portugal, M. T. (2005) Characterization of a functional P2X(7)-like receptor in cerebellar granule neurons from P2X(7) knockout mice. FEBS Lett. 579, 3783–3788. 14. Taylor, S. R., Gonzalez-Begne, M., Dewhurst, S., Chimini, G., Higgins, C. F., Melvin, J. E., Elliott, J. I. (2008) Sequential shrinkage and swelling underlie P2X7-stimulated lymphocyte phosphatidylserine exposure and death. J. Immunol. 180, 300 –308. 15. Mackenzie, A. B., Young, M. T., Adinolfi, E., Surprenant, A. (2005) Pseudoapoptosis induced by brief activation of ATP-gated P2X7 receptors. J. Biol. Chem. 280, 33968 –33976. 16. Gu, B., Bendall, L. J., Wiley, J. S. (1998) Adenosine triphosphate-induced shedding of CD23 and L-selectin (CD62L) from lymphocytes is mediated by the same receptor but different metalloproteases. Blood 92, 946 –951. 17. Adriouch, S., Dox, C., Welge, V., Seman, M., Koch-Nolte, F., Haag, F. (2002) Cutting edge: a natural P451L mutation in the cytoplasmic domain impairs the function of the mouse P2X7 receptor. J. Immunol. 169, 4108 – 4112. 18. Elliott, J. I., McVey, J. H., Higgins, C. F. (2005) The P2X7 receptor is a candidate product of murine and human lupus susceptibility loci: a hypothesis and comparison of murine allelic products. Arthritis Res. Ther. 7, R468–R475. 19. Chused, T. M., Apasov, S., Sitkovsky, M. (1996) Murine T lymphocytes modulate activity of an ATP-activated P2Z-type purinoceptor during differentiation. J. Immunol. 157, 1371–1380. 20. Labasi, J. M., Petrushova, N., Donovan, C., McCurdy, S., Lira, P., Payette, M. M., Brissette, W., Wicks, J. R., Audoly, L., Gabel, C. A. (2002) Absence of the P2X7 receptor alters leukocyte function and attenuates an inflammatory response. J. Immunol. 168, 6436 – 6445. 21. Chessell, I. P., Hatcher, J. P., Bountra, C., Michel, A. D., Hughes, J. P., Green, P., Egerton, J., Murfin, M., Richardson, J., Peck, W. L., Grahames, C. B., Casula, M. A., Yiangou, Y., Birch, R., Anand, P., Buell, G. N. (2005) Disruption of the P2X7 purinoceptor gene abolishes chronic inflammatory and neuropathic pain. Pain 114, 386 –396. 22. Guo, C., Masin, M., Qureshi, O. S., Murrell-Lagnado, R. D. (2007) Evidence for functional P2X4/P2X7 heteromeric receptors. Mol. Pharmacol. 72, 1447–1456. 23. Xiang, Z., Burnstock, G. (2005) Expression of P2X receptors on rat microglial cells during early development. Glia 52, 119 –126. 24. Nicke, A. (2008) Homotrimeric complexes are the dominant assembly state of native P2X7 subunits. Biochem. Biophys. Res. Commun. 377, 803– 808. 25. Virginio, C., MacKenzie, A., North, R. A., Surprenant, A. (1999) Kinetics of cell lysis, dye uptake and permeability changes in cells expressing the rat P2X7 receptor. J. Physiol. 519, 335–346. 26. Khakh, B. S., Proctor, W. R., Dunwiddie, T. V., Labarca, C., Lester, H. A. (1999) Allosteric control of gating and kinetics at P2X(4) receptor channels. J. Neurosci. 19, 7289 –7299. 27. Qu, Y., Franchi, L., Nunez, G., Dubyak, G. R. (2007) Nonclassical IL-1 ␤ secretion stimulated by P2X7 receptors is dependent on inflammasome activation and correlated with exosome release in murine macrophages. J. Immunol. 179, 1913–1925. 28. Hervas, C., Perez-Sen, R., Miras-Portugal, M. T. (2003) Coexpression of functional P2X and P2Y nucleotide receptors in single cerebellar granule cells. J. Neurosci. Res. 73, 384 –399. 29. Deuchars, S. A., Atkinson, L., Brooke, R. E., Musa, H., Milligan, C. J., Batten, T. F., Buckley, N. J., Parson, S. H., Deuchars, J. (2001) Neuronal P2X7 receptors are targeted to presynaptic terminals in the central and peripheral nervous systems. J. Neurosci. 21, 7143–7152. 30. Cheewatrakoolpong, B., Gilchrest, H., Anthes, J. C., Greenfeder, S. (2005) Identification and characterization of splice variants of the human P2X7 ATP channel. Biochem. Biophys. Res. Commun. 332, 17–27. 31. Blencowe, B. J. (2006) Alternative splicing: new insights from global analyses. Cell 126, 37– 47. 32. Yeo, G., Holste, D., Kreiman, G., Burge, C. B. (2004) Variation in alternative splicing across human tissues. Genome Biol. 5, R74.

http://www.jleukbio.org