Inhibition of neutrophil apoptosis by ATP is mediated by the P2Y₁₁ receptor

Abstract

Neutrophils undergo rapid constitutive apoptosis that is delayed by a range of pathogen and host derived inflammatory mediators. We have investigated the ability of the nucleotide ATP, to which neutrophils are exposed both in the circulation and at sites of inflammation, to modulate the lifespan of human neutrophils. We found physiologically relevant concentrations of ATP cause a concentration-dependent delay of neutrophil apoptosis (assessed by morphology, Annexin V/To-Pro3 staining and mitochondrial membrane permeabilisation). We found even brief exposure to ATP (10 mins) was sufficient to cause a long-lasting delay of apoptosis and showed the effects were not mediated by ATP breakdown to adenosine. The P2 receptor mediating the anti-apoptotic actions of ATP was identified using a combination of more selective ATP analogues, receptor expression studies and study of downstream signalling pathways. Neutrophils were shown to express the P2Y₁₁ receptor and inhibition of P2Y₁₁ signalling using the antagonist NF157 abrogated the ATP-mediated delay of neutrophil apoptosis, as did inhibition of type-I cAMP-dependent protein kinases activated downstream of P2Y₁₁, without effects upon constitutive apoptosis. Specific targeting of P2Y₁₁ could retain key immune functions of neutrophils but reduce the injurious effects of increased neutrophil longevity during inflammation.

Introduction

Apoptosis (programmed cell death) of neutrophils is an important control point in the physiological resolution of innate immune responses. In this context, neutrophil apoptosis must be delayed until essential host functions such as pathogen clearance are completed but then proceed promptly to abrogate inflammation and avoid tissue damage (1). Neutrophils are typically short-lived cells that can engage a constitutive programme of apoptosis (2), leading to down-regulation of neutrophil pro-inflammatory functions and clearance by phagocytes (2, 3). There is abundant evidence the balance between neutrophil survival and death by apoptosis is exquisitely regulated by a range of extracellular factors, including host-derived cytokines and pathogen-derived molecules to which neutrophils are exposed in the circulation and in tissues (4, 5). The potential for extracellular nucleotides, particularly adenosine 5′-triphosphate (ATP), to modulate neutrophil survival has not been investigated.

Neutrophils are exposed to ATP in a variety of in vivo situations and its role as a signalling molecule in pathophysiological situations is increasingly recognised (6). ATP is released into the circulation following activation of platelets and endothelial cells (7, 8), for example in acute coronary syndromes (7), potentially exposing circulating neutrophils to high local concentrations. Within 3 minutes following vessel wall injury, ATP concentrations of 20μM can be detected (9) and 1 × 10⁷ platelets can release > 55μM ATP (8). ATP is also released from dying cells (10), notably in chronic inflammatory conditions such as cystic fibrosis (11, 12). The effects of ATP are mediated via P2 receptors (13), which are further divided into P2X and P2Y subfamilies (14). Both are widely expressed in tissues and implicated in diverse cellular functions. ATP has been shown to modulate neutrophil pro-inflammatory functions, including chemotaxis (15), NADPH oxidase-dependent superoxide anion generation (16), and secretion of granule contents (17, 18).

We hypothesised that extracellular ATP may be a critical regulator of neutrophil apoptosis. We found that even short exposures to ATP delay neutrophil apoptosis, an effect that is independent of increases in [Ca²⁺]_i but dependent upon type-I cAMP-dependent protein kinases. Studies of receptor expression and use of P2 subtype inhibitors and agonists identified P2Y₁₁ as the purinergic receptor mediating the anti-apoptotic effect. These studies identify a novel potential therapeutic target for the amelioration of neutrophilic inflammation in a wide range of inflammatory diseases.

Materials and Methods

Materials

All chemicals were from Sigma-Aldrich (Poole, UK) unless otherwise stated. The phospholipase C inhibitor U73122 and its inactive analogue U73343, Rp-8-Br-cAMPS and pluronic acid were purchased from Calbiochem (Beeston, UK). Fluo-4 AM was obtained from Molecular Probes (Oregon, USA). The P2Y₁₁ and P2X₇ (APR-008) receptor antibodies for western blotting were from Alomone Labs (Jerusalem, Israel). The P2X₇ antibody for flow cytometry was a kind gift from GlaxoSmithKline (19) and the IgG2b isotype control was from Sigma-Aldrich (Poole, UK). The P2Y₁₁ antagonist NF157 was synthesised as previously described (20).

Neutrophil isolation and culture

Human neutrophils were isolated by dextran sedimentation and plasma-Percoll gradient centrifugation from whole blood of normal volunteers (21) with written informed consent and the approval of the South Sheffield Research Ethics Committee. Purity (>94%) was assessed by counting >300 cells on duplicate cytospins, with contaminating cells being mostly eosinophils. In some experiments neutrophils were further purified by negative magnetic selection, increasing neutrophil purity to >99.9% by removal of the low level of contaminating cells, especially PBMCs, that can affect the responses of neutrophils (22). These neutrophils are referred to in text as highly-purified neutrophils. Neutrophils were suspended at 2.5 × 10⁶/ml in RPMI with 1% penicillin/streptomycin and 10% FCS (all Invitrogen, Paisley, UK) and cultured in 96 well Falcon “Flexiwell” plates (Becton Dickinson). Freshly isolated neutrophils were designated as time 0.

ATP and ATP analogues were added at time 0 and incubated for 5 hours unless otherwise stated. Neutrophils treated with Rp-8-Br-cAMPS or NF157 were pre-incubated for 30 mins before addition of ATP or ATP analogues.

Assessment of Cell Viability and Apoptosis

Nuclear morphology was assessed on Giemsa-stained cytospins, with blinded observers counting >300 cells per slide on duplicate cytospins. Hemocytometer counts were performed at all time points, both to assess cell retrieval and for trypan blue exclusion. No significant differences in cell retrieval were noted with different treatments and necrosis was <2% throughout. In some experiments, neutrophils were washed in PBS and stained with PE-labelled Annexin V (BD Biosciences) and To-Pro®-3 iodide (Molecular Probes) to identify apoptotic (Annexin V+) and necrotic (To-Pro3+) cells (23). Controls included EDTA treatment (to determine Annexin V-population), freeze-thawed, necrotic cells (To-Pro3+ cells) and constitutively aged neutrophils (Annexin V+), to set appropriate gating. To detect loss of mitochondrial membrane potential (ΔΨ_m), neutrophils were incubated with 10μM JC-1 (Molecular Probes) at 37°C and loss of ΔΨ_m was assayed by gain in FL-1 fluorescence (24). Samples were analysed by FACSCalibur flow cytometer (Becton Dickinson, San Jose, CA), with 10,000 events recorded and analysed by CellQuest software (BD Biosciences).

Measurement of Neutrophil [Ca²⁺]_i

Ca²⁺ transients were measured in populations of human neutrophils loaded with the Ca²⁺ indicator Fluo-4 AM using the magnetically stirred Cairn Integra fluorimeter (Cairn Research, Kent, UK) and a previously described method (25). Briefly, neutrophils were incubated with 2μM Fluo-4 AM and 0.1% pluronic acid for 30 min at 37°C in buffer containing 136mM NaCl, 1.8mM KCl, 1.2mM KH₂PO₄, 1.2mM MgSO₄, 5mM NaHCO₃, 1.2mM CaCl₂, 5.5mM glucose, 20mM HEPES and 5mM EGTA. Cells were washed once in the same buffer with or without EGTA, and then 5×10⁶ cells/ml were placed in a 1ml cuvette for calcium measurements using the Cairn Integra fluorimeter (Cairn Research, Kent, UK). Cumulative agonist concentration-response curves were obtained and results normalised to maximum fluorescence induced by digitonin (2mM) applied at the end of the experiment. Digitonin-induced fluorescence varied by <10% in a single set of experiments (six cuvettes/experiment, data not shown). Curves shown are representative of the whole cell population.

Analysis of mRNA Expression

Total RNA was extracted from highly-pure populations of neutrophils using TRI-reagent according to the manufacturer’s instructions. To obtain cDNA, 1μg of total RNA was primed with oligo-dT primers and reverse transcribed with AMV reverse transcriptase (Promega, Southampton, UK). Primers for human P2X receptors were designed based on published sequence data. The primers used (forward, reverse) were as follows: P2X₁: TTTCATCGTGACCCCGAAGCAG, TCAAAGCGAATCCCAAACACC; P2X₂: ACCTGCCCCGAGAGCATAAG, AATGACCCCGATGACACCACCC; P2X₃: CACCTCGGTCTTTGTCATCATCAC, TGTTGAACTTGCCAGCATTCC; P2X₄: ACAGCAACGGAGTCTCAACAGG, CCTTCCCAAACACAATGATGTCG; P2X₅: TCCTGGCGTACCTGGTCGTATGG, CAGTCGCTGTCCTTGGAGCACG; P2X₆: GGTGACCAACTTCCTTGTGACG, CCCAGTGAACTCTGATGCCTACAG; P2X₇: TGCGATGGACTTCACAGATTTG, TGCCCTTCACTCTTCGGAAAC. PCR settings were 94°C for 1 minute, 55°C for 2 minutes, followed by extension at 72°C for 3 minutes. Each PCR reaction was performed with 35 cycles and the PCR products were resolved by electrophoresis on a 1.5% agarose gel.

Western Blotting

Highly-pure populations of neutrophils were washed with ice-cold PBS and lysed with 2x SDS lysis buffer (100mM Tris-HCl (pH 6.8), 100mM dithiothreitol, 20% SDS, 20% glycerol, and 0.2% bromophenol blue, made up in water) containing EDTA-free protease inhibitor cocktail, at 90°C for 10 minutes. The protein was subjected to 8% SDS-PAGE then electrophoretically transferred from the gel onto a 0.45μM polyvinylidene fluoride (PVDF) microporous membrane (Biorad, Hercules, CA). The membrane was probed with a rabbit polyclonal antibody directed against the human P2Y₁₁ or P2X₇ receptor, with or without the control peptide antigen (at 1μg peptide per 1μg antibody), overnight at 4°C. After washing, the membrane was incubated with HRP-conjugated goat anti-rabbit antibody (DAKO, Denmark), and visualised using the ECL system (Amersham Biosciences, Buckinghamshire, UK) followed by autoradiography.

Flow Cytometry for P2X₇ expression

Highly-purified neutrophils, HEK-293 cells (negative control) and HEK-293 cells transfected with human P2X₇ (positive control) (26), were used to detect the presence of P2X₇. Extracellular expression was detected using fixed cells and total P2X₇, both extracellular and intracellular, using cells that were fixed then permeabilised. Fixation and permeabilisation buffers were from eBioscience (San Diego, CA). Cells were then incubated with the P2X₇ receptor antibody (1.25μg/μl) or control IgG2b antibody (1.5ng/ml) for 30 minutes, followed by an anti-mouse IgG FITC-conjugated secondary antibody (10μl/100μl) (Sigma). Analysis of P2X₇ receptor expression was by flow cytometry, using a dual-laser FACSCalibur (BD Biosciences, Mountain View, CA) and CellQuest software (BD Biosciences), with appropriate single-stained samples for setting of compensation. We confirmed we were able to detect increased expression of CD11b, a protein present intracellularly, following permeabilisation (data not shown).

Measurement of Neutrophil [cAMP]_i

Neutrophils have low concentrations of cAMP and so were cultured at 3 × 10⁶/ml, as previously described (27), and pre-incubated with 3-isobutyl-1-methylxanthine (IBMX) (1mM) for 20 minutes to inhibit breakdown of cAMP, prior to stimulation with ATP, ATP-γ-S or BzATP (all at 100 μM) for 8 minutes. Cells were lysed with 0.1M HCl then spun at 850g for 10 minutes. Supernatants were acetylated to detect intracellular cAMP, using a direct immunoassay kit (Sigma, measuring sensitivity 0.039 pmol/ml) according to the manufacturer’s instructions. Data are expressed as fold increase in intracellular cAMP compared with unstimulated cells.

Statistical Analysis

All data are expressed as mean±SEM. Data were analysed as appropriate by students t-test or ANOVA with either Dunnett’s or Bonferroni’s (selected pairs) post-test using the Prism 4.0 program (GraphPad, San Diego, CA). Results were considered to be statistically significant where p < 0.05. Statistically significant differences from controls are indicated by *p<0.05, **p<0.01 and ***p<0.001. Differences between treated populations are indicated by #p<0.05, ##p<0.01 and ###p<0.001.

Results

ATP delays neutrophil apoptosis in a concentration-dependent manner

Incubation of neutrophils with ATP resulted in concentration-dependent reductions in neutrophil apoptosis at 5 hours that were significant at ATP concentrations of 1μM and above (Fig. 1A). Such concentrations are physiological and readily achieved in vivo (8, 9). This delay of apoptosis was maintained over a prolonged time course (Fig. 1B). Inhibition of apoptosis was assessed by light microscopy using morphological features (2) and, in further experiments, these changes were correlated with evidence that ATP also delayed cell membrane changes of apoptosis (Annexin V binding, Fig. 1C) and loss of mitochondrial membrane potential (JC-1 staining, Fig. 1D). There was no evidence of necrotic cell death on trypan blue exclusion or To-Pro3 staining (data not shown), nor of differences in cell retrieval on hemocytometer counts with ATP treatment compared with controls. We have previously shown that the anti-apoptotic effects of a prototypic proinflammatory mediator, LPS, are principally dependent upon the small numbers of mononuclear cells present in neutrophil populations prepared by gradient centrifugation (22). We therefore compared the anti-apoptotic effects of ATP upon neutrophils prepared by the standard method or with an additional purification step based on immunomagnetic bead purification to achieve >98% purity (22). There were no differences in the anti-apoptotic effects of ATP over a range of concentrations (Table 1).

Figure 1. ATP inhibition of neutrophil apoptosis is concentration and time-dependent.

Neutrophils were cultured for varying periods of time in the presence or absence of various concentrations of ATP. Apoptosis was assessed by morphologic criteria and data are expressed as mean±SEM % apoptosis from at least 3 independent experiments, each from a separate donor. (A) Neutrophils were cultured either alone (squares) or with various concentrations of ATP [0.1-1000μM] for 5 hours. (B) Neutrophils were incubated in the presence (circles) or absence (squares) of ATP (100μM) for 5, 9, 16, and 22 hours prior to assessment of apoptosis. (C) Neutrophils were cultured in the presence or absence of ATP (100μM) for 5 hours and apoptosis and necrosis were assessed by staining with Annexin V/To-Pro3 and flow cytometry. Representative dot plots show Annexin V and To-Pro3 staining in untreated cells (negative control), following ATP treatment or constitutively aged, 24 hour neutrophils (positive control). Results are representative of 3 independent experiments in which ATP delayed the onset of apoptosis compared to untreated cells (mean±SEM 5.2±0.8% compared with 9.1±0.8% in control cells, p < 0.05). (D) Neutrophils were cultured for 4 hours with ATP (100μM) or staurosporine (5μM) then stained with the mitochondrial dye, JC-1 and assessed by flow cytometry. Representative dot plots show JC-1 fluorescence in untreated cells (negative control) and following ATP or staurosporine (positive control) treatment. An increase in FL-1 (green) fluorescence indicates loss of mitochondrial membrane potential in neutrophils. Results are representative of 3 independent experiments in which ATP significantly maintained membrane potential compared to untreated cells (10.0±1.1% apoptosis in ATP-treated cells compared with 21.4±2.5% in control cells, p < 0.05).

Table I. Presence of Contaminating Cells Does Not Influence Inhibition of Neutrophil Apoptosis by ATP.

Neutrophils were isolated by plasma/Percoll gradients and further purified by immunomagnetic selection to remove contaminating eosinophils and peripheral blood mononuclear cells (PBMCs). Mean±SEM neutrophil purity from each isolation method is given in brackets *. Neutrophils were treated with ATP [0.1-1000μM] for 5 hours. Apoptosis was assessed by morphologic criteria and data is expressed as mean±SEM % apoptosis from 3 independent experiments, each from a separate donor. For both isolation methods ATP significantly inhibited apoptosis in a concentration-dependent manner († p = < 0.01, # p = < 0.001) and there was no significant difference between the two isolation methods (p = 0.6614)

[ATP] (μM)	% Apoptosis
	Percoll Purified (94.2±1.6%)*	Highly Purified (98.8±0.8%)*
0	12.9 ( ± 0.8)	15.2 ( ± 1.6)
0.1	11.1 ( ± 0.4)	10.7 ( ± 1.3) †
1	7.8 ( ± 0.7) #	5.8 ( ± 1.1) #
10	4.6 ( ± 0.4) #	5.1 ( ± 0.7) #
100	4.7 ( ± 1.0) #	4.1 ( ± 0.6) #
1000	4.3 ( ± 0.5) #	3.7 ( ± 0.6) #

The exposure of neutrophils to significant concentrations of ATP may be of relatively short duration in vivo, since ATP is readily broken down (28), yet the survival effect of ATP treatment of neutrophils was long-lasting. We therefore determined whether prolonged neutrophil survival could be initiated by short exposures to ATP. In Figure 2A, neutrophils were cultured for 5 hours and apoptosis measured. For the first 10, 30, 60 or 180 minutes of this period neutrophils were treated with ATP (or media only as a control), after which time they were washed and returned to fresh media. Other neutrophils were cultured with media alone or ATP throughout the time course, without additional washing. These data show that a 10 minute exposure to ATP (100 μM) at the beginning of the experiment induced similar levels of neutrophil survival as when cells were exposed to ATP throughout the time course, demonstrating rapid engagement of an extremely effective survival mechanism. We also determined whether neutrophils exposed to ATP after they had “aged” in culture could still have a measurable increase in survival following ATP exposure. We found that ATP had a significant survival effect upon neutrophils when added up to 180 minutes after culture (Fig. 2B).

Figure 2. Delay of neutrophil apoptosis is mediated by short exposures to ATP.

Neutrophils were cultured in the presence (filled bars) or absence (open bars) of ATP. Data are expressed as mean±SEM % apoptosis, assessed by morphologic criteria, from at least 3 independent experiments, each from a separate donor. (A) Neutrophils were cultured with or without ATP (100μM) for varying lengths of time as indicated, after which cells were washed and resuspended in fresh media to remove any remaining extracellular ATP. Cells were cultured for a total of 5 hours prior to assessment of apoptosis. (B) Neutrophils were cultured with or without ATP (10μM) for 5 hours, with ATP added at the time points indicated. Apoptosis was then assessed at 5 hours. ATP significantly delayed apoptosis to similar levels at each time point. (C) Neutrophils were incubated in the presence (circles) or absence (square) of varying concentrations of the stable ATP analogue, ATP-γ-S, for 5 hours and apoptosis assessed.

The “lifespan” of ATP in the tissues and circulation is tightly regulated by the actions of ecto-nucleotidases that hydrolyse ATP to adenosine, which can itself delay neutrophil apoptosis (29). To confirm the effects of ATP on neutrophil apoptosis were not due to breakdown to adenosine, we employed a stable ATP analogue, ATP-γ-S, that is resistant to hydrolysis by ecto-ATPases or phosphatases (30). ATP-γ-S also effectively inhibited neutrophil apoptosis as shown in Figure 2C.

Effect of different ATP analogues upon neutrophil apoptosis

To determine the P2 receptor subtypes responsible for ATP-mediated delay of apoptosis, a variety of approaches were needed because of the lack of a comprehensive battery of specific agonists and antagonists. Initially, ATP analogues with differing preferences for P2 receptor subtypes were studied for their effect upon neutrophil apoptosis (Fig. 3). BzATP, a typical P2X₇ agonist, delayed neutrophil apoptosis at concentrations of 1μM and above. In contrast, αβMeATP (which preferentially activates P2X₁ and P2X₃ receptors), 2MeSATP (a P2Y₁, P2X₁ and P2X₅ agonist) and UTP (a P2Y₂ and P2Y₄ agonist) did not alter the rate of neutrophil apoptosis at any tested concentration.

Figure 3. Effects of ATP analogues on neutrophil apoptosis.

Neutrophils were cultured in the presence (circles) or absence (squares) of varying concentrations of ATP analogues for 5 hours and apoptosis assessed by morphologic criteria. Data are expressed as mean±SEM % apoptosis from at least 3 independent experiments, each from a separate donor. (A) BzATP, (B) αβMeATP, (C) 2MeSATP and (D) UTP.

P2X₇ receptor subtype is not expressed by human neutrophils

ATP-γ-S and BzATP are the only agonists, other than ATP, which are known to activate P2X₇, the purine receptor most often associated with apoptosis in immune cells (31, 32). Therefore we initially examined the possibility that P2X₇ receptors may mediate the anti-apoptotic effects of ATP on neutrophils. However, it is unclear whether human neutrophils do (16) or do not (33) express P2X₇ receptors, We therefore sought evidence of mRNA expression in highly-purified neutrophil populations. As shown in Figure 4A, only mRNAs for P2X₁ and P2X₅ were detected in these highly pure populations. If the samples were “spiked” with 5% PBMCs, however, transcripts for P2X₄, P2X₆ and P2X₇ were also intermittently detected (Fig. 4B). Since P2X₄ and P2X₇ are known to be highly expressed in monocytes (34), it is possible that their detection in previous studies was due to the presence of contaminating cells. There are, however, clear precedents for human neutrophils to express proteins but for the corresponding mRNA to be undetectable. Examples include lactoferrin and myeloperoxidase, major constituents of neutrophil granules (35). Since P2X₇ has been thought to mediate functional effects in neutrophils in a number of recent studies, we wanted to exclude its possible expression as rigorously as possible. We found that P2X₇ protein could not be detected by flow cytometry either intra or extracellularly in neutrophils, although the protein was readily detected in HEK-293 cells transfected with hP2X₇ (Fig. 4C). P2X₇ was also not detected by western blotting, and LPS treatment, which up-regulates P2X₇ expression in other cell types (36), did not induce its expression in neutrophils (Fig. 4D).

Figure 4. Expression of P2 receptors by human neutrophils.

(A) PCRs were performed for each of the P2X receptors according to the protocols described in Methods and bands of the appropriate sizes identified as indicated. Three samples of cDNA from highly-purified human neutrophils, each from a different donor (Lanes 3 - 5), were tested for the presence of each P2X receptor mRNA. Lanes 1 and 2 represent negative and positive controls, respectively. Positive controls are vectors containing the receptor subtype cDNA. (B) The studies were repeated using samples of cDNA from highly-purified human neutrophils “spiked” with 5% PBMCs, each from a different donor (Lanes 3-5). Lanes 1 and 2 represent negative and positive controls respectively. P2X₄, P2X₆ and P2X₇ mRNAs were detected in these populations but not in the samples from highly-purified neutrophils. (C) Cells were fixed and/or permeabilised and P2X₇ receptor expression was detected by flow cytometry. Fold-increase in P2X₇ receptor expression (mean±SEM of 3 independent experiments) is plotted against cell type. HEK-X₇ cells showed significant expression of P2X₇ both extracellularly and intracellularly (p < 0.01 and p < 0.05 respectively) compared with both untransfected HEK cells and neutrophils. Neutrophils do not express P2X₇ receptors either intra-(black bars) or extra-cellularly (open bars). (D) Highly-purified neutrophils were cultured with or without 10ng/ml LPS for 16 hours, then western blotted to detect P2X₇ receptor protein. Neither untreated nor LPS-stimulated neutrophils expressed detectable P2X₇ receptor protein (n = 3). In parallel samples, LPS inhibited apoptosis of neutrophils isolated from each donor population used for western blotting (mean apoptosis in control cells was 74.4±5.7% and in LPS-treated cells was 35.9±4.0%, p < 0.01).

ATP causes elevations of intracellular calcium levels ([Ca²⁺]_i) in neutrophils independently of the delay of apoptosis

To explore further the possibility that the effects of ATP were mediated via a P2Y receptor we measured intracellular calcium levels, since the G protein-coupled seven transmembrane P2Y receptors are known to cause elevations of cytosolic calcium levels via phospholipase C activation and inositol-1,4,5-triphosphate-(IP₃-) mediated release of calcium from intracellular stores. Moreover, we have previously shown elevations of [Ca²⁺]_i delay neutrophil apoptosis (37) and thus speculated this might be the mechanism of ATP-mediated delay of apoptosis. We found ATP increased neutrophil [Ca²⁺]_i in a concentration-dependent manner, in the presence or absence of extracellular calcium, implying the increase in [Ca²⁺]_i is mediated from intracellular stores. Addition of a PLC inhibitor, U73122, completely abolished the ATP-mediated [Ca²⁺]_i effect (Fig. 5A). The P2Y agonist UTP increased [Ca²⁺]_i to similar levels to ATP, again a PLC dependent effect (Fig. 5B). Use of the different structural analogues of ATP showed that only the stable analogue, ATP-γ-S, also mediated a calcium response (Fig. 5C). These data suggest the ATP-mediated inhibition of apoptosis and elevation of [Ca²⁺]_i are separate events. UTP increases [Ca²⁺]_i but has no effect upon apoptosis, whereas ATP analogues that increase neutrophil survival, such as BzATP, do not increase [Ca²⁺]_i. This data contrasts with our previous studies linking increases in [Ca²⁺]_i with delay of neutrophil apoptosis (37), but in those studies delay of apoptosis was mediated by influx of extracellular calcium rather than release of calcium from intracellular stores mediated by ATP.

Figure 5. ATP and some analogues increase intracellular Ca²⁺ via activation of phospholipase C.

Neutrophils were incubated with Fluo-4 AM for 30 mins, then washed and resuspended in buffer in the presence (circles) or absence (squares) of calcium. Cells were then treated with ATP or an ATP analogue as indicated, in the absence (filled symbols) or presence (open symbols) of a PLC inhibitor U73122 (10μM). Increases in intracellular [Ca²⁺] were measured as a percentage of the maximum Ca²⁺ released by addition of 2mM digitonin. All data shown are mean±SEM % maximum Ca²⁺ increase from at least 3 independent experiments, each from a separate donor. Neutrophils were treated with increasing concentrations of (A) ATP (1-1000μM) and (B) UTP (1-1000μM). Both ATP and UTP increase intracellular [Ca²⁺] from intracellular stores via a PLC dependent mechanism. (C) 100μM concentrations of ATP, UTP, ATP-γ-S, BzATP, αβMeATP and 2MeSATP were added to neutrophils in Ca²⁺ containing (open bars) or EGTA containing buffer (filled bars). Of these analogues, only ATP-γ-S caused an increase in intracellular [Ca²⁺] similar to the changes observed with ATP and UTP.

Combining the data from the effects of ATP analogues, expression data and calcium studies allowed us to narrow the range of possible receptors mediating the delay of neutrophil apoptosis by ATP. Since two P2Y agonists (2MeSATP and UTP) did not delay neutrophil apoptosis and the delay of apoptosis was independent of [Ca²⁺]_i, the anti-apoptotic effect of ATP was not typical for a P2Y receptor-mediated effect. Of the P2X receptors identified in neutrophils, P2X₁ was an unlikely candidate, since two recognised P2X₁ agonists, αβMeATP and 2MeSATP (38), did not inhibit neutrophil apoptosis. There is evidence human P2X₅ may not be a functional receptor (38) and 2MeSATP, which is also an agonist at this receptor, was without effect on neutrophil apoptosis. Involvement of P2X₇ was effectively excluded by its lack of expression in neutrophils. We also found that a P2X₇ antagonist, 1-[N,O-bis(5-isoquinolinesulfonyl)-N-methyl-_L-yrosyl]-4-phenylpiperazine (KN62), did not inhibit BzATP-mediated delay of apoptosis, further evidence this effect was independent of P2X₇ (data not shown). BzATP is, however, also a potent agonist of the P2Y₁₁ receptor, which is coupled to both phosphoinositide and cyclic AMP signalling pathways (39). BzATP exerts effects upon both phosphoinositide and cAMP signalling pathways with EC₅₀’s in the low micromolar range (39, 40). This is in keeping with a significant effect of BzATP on neutrophil apoptosis at a concentration of 10μM, and contrasts with an EC₅₀ of >50μM for BzATP at the P2X₇ receptor (41). Moreover, the rank order of agonist potency at the P2Y₁₁ receptor reveals that ATP-γ-S is at least equivalently potent to BzATP (40), again in keeping with our data and suggesting P2Y₁₁ as the receptor mediating anti-apoptotic signalling.

ATP mediated delay of neutrophil apoptosis is P2Y₁₁ dependent

If ATP and its analogues were exerting their anti-apoptotic effects via the P2Y₁₁ receptor, they would be predicted to cause increases in intracellular cAMP. We showed ATP, ATP-γ-S and BzATP all increased [cAMP]_i compared with untreated control cells (Fig. 6A). A recent study showed NAD⁺ exerted its pro-inflammatory effects upon human neutrophils via the P2Y₁₁ receptor (42) and we found NAD⁺ also caused a concentration-dependent delay of apoptosis that was significant at concentrations of 1μM and above (e.g. mean±SEM apoptosis 4.2±0.5% following 1μM NAD⁺ compared with 7.9±0.5% in control cells, n = 3, p < 0.01 at 5 hours). During these studies, a potent antagonist at the P2Y₁₁ receptor, NF157, became available, which is derived from suramin, a compound with antagonist activity at most P2 receptors. NF157 was shown to have an IC₅₀ of approximately 0.5μM at the P2Y₁₁ receptor and high selectivity for P2Y₁₁ over all P2X and P2Y receptors tested other than P2X₁ (20). We found NF157 blocked ATP-γ-S – mediated delay of neutrophil apoptosis (Fig. 6B) but was without effect upon constitutive neutrophil apoptosis. The inhibitory effect of NF157 was overcome at higher concentrations of ATP-γ-S, confirming the competitive nature of its P2Y₁₁ antagonism (Fig. 6C). The anti-apoptotic effect of ATP was also abrogated by NF157 (Fig. 6D). Moreover, we were able to detect P2Y₁₁ protein in highly-purified neutrophils (Fig. 6E), with a band detected at 45kDa, slightly smaller than the band detected in platelets but within the expected size range for P2Y₁₁ (43).

Figure 6. P2Y₁₁ receptors mediate ATP-induced delay of neutrophil apoptosis.

(A) Neutrophils were pre-treated with IBMX (1mM) then with ATP, ATP-γ-S or BzATP (100μM) or buffer control for 8 minutes prior to lysis and detection of intracellular cAMP. Data is from 3 independent experiments, each assessed in duplicate, and is expressed as fold increase in [cAMP]_i compared to buffer control. All 3 compounds caused a significant increase in [cAMP]_i. In further experiments, neutrophils were pre-treated for 30 minutes with the P2Y₁₁ receptor antagonist, NF157, at varying concentrations or buffer control, then treated with ATP-γ-S, ATP or buffer for a total incubation time of 5 hours. Apoptosis was then assessed by morphologic criteria and data are expressed as mean±SEM % apoptosis from a minimum of 3 independent experiments, each from a separate donor. (B) Neutrophils were pre-treated with NF157 (0.01-0.5μM) prior to stimulation with ATP-γ-S (1μM). The survival effect of ATP-γ-S was significantly reduced by 0.5μM NF157 (### p <0.001). (C) Neutrophils were pre-treated with NF157 (0.5μM) prior to stimulation with increasing concentrations of ATP-γ-S (0.1-100μM). ATP-γ-S significantly inhibited neutrophil apoptosis at all concentrations compared to control cells (*** p < 0.001). NF157 abrogated the inhibition of apoptosis by ATP-γ-S (# p <0.05, ### p <0.001). (D) Neutrophils were pre-incubated with buffer (open bars) or NF157 (0.5μM) (filled bars) followed by stimulation with ATP (0.5μM) or buffer control. ATP reduced neutrophil apoptosis (*) and this effect was significantly abrogated by NF157 (#). (E) Detection of P2Y₁₁ protein expression by highly-purified neutrophils, with human platelets used as a positive control. In all four donors, P2Y₁₁ is detected as an approximately 45kDa protein.

ATP mediated delay of neutrophil apoptosis is dependent upon cAMP-dependent kinases

To further confirm a role for P2Y₁₁ in delay of neutrophil apoptosis, we determined the effects of modulating cyclic AMP signalling pathways downstream of P2Y₁₁. Reagents that mediate increases in intracellular cAMP ([cAMP]_i) delay neutrophil apoptosis (44). Krakstad et al. (45) have shown the anti-apoptotic effects of cAMP upon neutrophils are mediated by activation of cAMP-dependent protein kinases (cA-PKs), as opposed to the recently discovered cAMP regulated guanosine 5′-triphosphate exchange proteins directly activated by cAMP (Epac). ATP has also been shown to increase [cAMP]_i in neutrophils (46). We confirmed the results of Krakstad et al. (45), showing the ability of a cA-PK type I agonist (N⁶-MB-cAMP) but not an Epac activator (8-pCPT-2′-O-Me-cAMP) to cause delay of neutrophil apoptosis equivalent to that seen with the stable cAMP analogue, dibutyryl cAMP (data not shown). We then showed that a specific inhibitor of cA-PK type I (Rp-8-Br-cAMPS) was able to completely abrogate the ATP-mediated survival of neutrophils (Fig. 7A). Moreover, the anti-apoptotic effects of all ATP analogues found to delay neutrophil apoptosis were abolished by the inhibitor, whilst the compound was without effect upon constitutive neutrophil apoptosis (Fig. 7B). We used a 16 hour time point for these experiments, where there would be a sufficiently large inhibition of neutrophil apoptosis by ATP (Fig. 1B) for abrogation of these anti-apoptotic effects to be clearly demonstrated. These data confirm that P2Y₁₁-mediated delay of apoptosis is dependent upon activation of cAMP signalling pathways, an effect that is unique to P2Y₁₁ amongst P2 receptors (47).

Figure 7. ATP inhibits neutrophil apoptosis via activation of cAMP-dependent protein kinase.

(A) Neutrophils were cultured in the presence or absence of the cA-PK type I agonist N⁶-MB-cAMP (1mM) or ATP (100μM), alone (open bars) or in combination with a specific cA-PK type I inhibitor, Rp-8-Br-cAMPS (0.7mM) (filled bars), for 16 hours. Apoptosis was assessed by morphologic criteria and data are expressed as mean±SEM % apoptosis from at least 3 independent experiments, each from a separate donor. Rp-8-Br-cAMPS abrogated the enhanced neutrophil survival caused by these treatments (## p <0.01 and ### p <0.001). (B) Similar experiments were performed for two ATP analogues shown to delay neutrophil apoptosis, ATP-γ-S and BzATP (both at 100μM), again using N⁶-MB-cAMP (1mM) as a positive control. Data are from at least 3 independent experiments, each from a separate donor. The ATP analogues significantly inhibited neutrophil apoptosis and these effects were blocked by Rp-8-Br-cAMPS.

Discussion

Extracellular ATP is a crucial mediator of the inflammatory response, in part through its non-redundant role in activation of P2X₇ and thus IL-1 release from macrophages (6, 48). Here, we show that ATP regulates the lifespan of the inflammatory response by a further mechanism: preservation of neutrophil survival. Together with previous work that has shown that ATP can enhance neutrophil functions including degranulation (18) and superoxide production (16), these data place ATP as a multifunctional central regulator of inflammation, whose effects can considerably outlast the period of time for which it is present in the extracellular environment.

We showed that ATP mediates effects upon neutrophil apoptosis at concentrations readily achieved when platelets are activated, or cells are damaged (8). Significant effects were achieved at 1μM ATP and ATP concentrations of up to 250μM have been detected on lytic release from cells (49). Our findings are supported by a previous study showing delay of neutrophil apoptosis by a single concentration of ATP at a single time-point (50). Since ATP is rapidly metabolised by ectonucleases (CD39 and CD73) to adenosine, which is also anti-apoptotic to neutrophils via actions on A_2A receptors (29), we confirmed the effects were seen with a non-hydrolysable ATP analogue, ATP-γ-S. Neutrophils themselves have been shown to release ATP, particularly under conditions of hypoxia (51) or hypertonic stress (52), suggesting that ATP may act as an autocrine or paracrine survival factor, prolonging neutrophil lifespan at sites of inflammation.

To determine which receptor mediates the anti-apoptotic effects of ATP, we used a combination of more selective ATP analogues, receptor expression studies and examination of downstream signalling pathways. Of the known P2X receptors, we found neutrophils expressed mRNA for P2X₁ and P2X₅ but agonist studies, combined with evidence P2X₅ may not be functional in humans (38), again made them unlikely candidates. The combination of agonist studies and the lack of requirement for calcium signalling to delay apoptosis made classical P2Y receptors unlikely candidates to mediate the delay of apoptosis. We therefore considered the possibility the anti-apoptotic effect was mediated via P2Y₁₁, which is the only P2 receptor known to mediate cAMP dependent signalling (47). We showed ATP and other analogues causing delay of neutrophil apoptosis also caused increases in intracellular cAMP levels. Moreover, a recently synthesised selective P2Y₁₁ inhibitor, NF157, abrogated the anti-apoptotic effect of ATP and the effects of ATP were also abolished by inhibition of signalling via cAMP-dependent protein kinases, further implicating P2Y₁₁. The delay of apoptosis by ATP analogues was achieved at concentrations in keeping with the described EC₅₀′s for BzATP and ATP-γ-S as agonists of cAMP production in P2Y₁₁ transfected cells (7.2 and 3.4 μM respectively) in complete media (40), although a ten-fold lower EC₅₀ for BzATP has been reported in divalent free media (53). We did not detect a significant rise in intracellular calcium following BzATP treatment of neutrophils. Previous studies have shown increases in HL-60 cells (16) but there are no reports of a calcium rise in neutrophils in response to BzATP. We were able to detect a rise in intracellular calcium in response to ATP, ATP-γ-S and UTP, which will activate P2Y₂ receptors at the concentrations used, but did not detect a calcium rise in response to BzATP nor to 2MeSATP which, as a P2Y₁ agonist, would also be expected to cause an increase in intracellular calcium via activation of phospholipase C. This suggests the calcium response we detected was predominantly P2Y₂ mediated and there is evidence BzATP is only a weak partial agonist at human P2Y₂ receptors (54). We have not been able to further demonstrate the role of P2Y₁₁ by study of P2Y₁₁ deficient neutrophils, since, to our knowledge, a P2Y₁₁ deficient mouse has not been developed and because genetic manipulation of primary human neutrophils is notoriously difficult.

P2Y₁₁ is unique as a P2 receptor coupling both to phospholipase C and to adenylate cyclase (39). P2Y₁₁ is expressed on several leukocyte populations, including neutrophils (52), lymphocytes (55), monocytes (34) and monocyte-derived dendritic cells (56), and is associated with maturation of bone marrow precursor populations (57). More recently, P2Y₁₁ was found to be abundantly expressed in the endothelium (58).

In the course of our experiments we found that neutrophils express a limited repertoire of P2X receptors and, in particular, do not express P2X₇ at mRNA or protein level in resting or LPS-stimulated cells, in agreement with Chen et al. (15), who also failed to detect P2X₇ mRNA. This contrasts with previous studies showing P2X₇ expression in human neutrophils at the mRNA level (16, 52) In both these studies, neutrophils were purified using dextran sedimentation and gradient centrifugation, with neutrophil purities of approximately 97%. As shown in Figure 4B, neutrophil populations contaminated with PBMCs can have detectable mRNA for P2X₇ as well as P2X₄ and P2X₆ and this may explain the discrepant results. Gu et al. (59) detected intracellular P2X₇ protein expression in human neutrophils. We were unable to detect P2X₇ in neutrophils by flow cytometry or by western blotting under conditions where the protein was detected in other cell types. Thus, P2X₇ expression may be present at low levels in neutrophils, but it is likely that contamination of neutrophils with PBMC has contributed to a confused picture in which the role of P2X₇ may have become overstated. Our data are supported by the finding of Labasi et al. (60) that granulocytes from P2X₇ deficient mice showed no lack of shape change or L-selectin shedding in response to ATP, in contrast to monocytes and lymphocytes, which the authors speculated reflected lack of surface expression of P2X₇ on neutrophils.

In a number of studies, functional effects of BzATP upon neutrophils have been attributed to P2X₇-mediated actions (16, 61). BzATP, however, is also an agonist at other receptors, notably P2Y₁₁ but also P2X₁ (38), and the potency of BzATP in delay of neutrophil apoptosis is more in keeping with effects at the P2Y₁₁ receptor (40). We also considered the possibility that ATP may exert its effects on neutrophil apoptosis indirectly, by activation of P2X₇ receptors expressed upon mononuclear cells. We have previously shown using highly-purified LPS, a TLR4 agonist, that the anti-apoptotic effect of LPS on human neutrophils is principally mediated via indirect effects upon the small numbers of monocytes present in the cell preparations, inducing release of neutrophil survival factors (22, 62). Importantly, in our studies, the anti-apoptotic effects of ATP were the same in gradient-purified and highly-purified neutrophil populations, supporting the view that ATP and ATP analogues were mediating their anti-apoptotic effects directly upon neutrophils but via a receptor other than P2X₇.

Extracellular ATP can cause induction of cell death, either apoptotic or cytolytic, in many cell types e.g. endothelial cells (32), lymphocytes (31), hepatocytes (63). In neutrophils, in contrast, ATP mediates a significant delay of apoptosis. Thus, an environment associated with cell damage and death mediates prolongation of effect of an important cell of the innate immune system, potentially facilitating resolution of tissue injury. The effects of ATP upon neutrophils are, however, dependent upon activation of cA-PK via elevation of intracellular cAMP. cAMP also induces apoptosis in many cell types, including thymocytes (64) and B lymphocytes (65), yet is anti-apoptotic to both neutrophils (44) and eosinophils (66). Krakstad et al. (45) showed cAMP inhibited neutrophil apoptosis by activation of cA-PK type I, the dominant isoform in neutrophils (67). Elevation of cAMP stabilises the anti-apoptotic Bcl-2 protein, Mcl-1, and phosphorylates Bad, providing potential downstream mechanisms of its anti-apoptotic actions (68).

P2X₇ has profoundly pro-inflammatory effects upon monocyte-macrophage populations and we now propose that enhancement of neutrophil lifespan, and thus proinflammatory functions since these are regulated by onset of apoptosis (3), is separately regulated via P2Y₁₁. The lack of effect of P2Y₁₁ antagonism on constitutive apoptosis suggests specific targeting of P2Y₁₁ could provide a mechanism that retains key immune functions of neutrophils but reduces the injurious effects of neutrophil longevity at inflamed sites. Further experiments are underway to determine the full consequences of early P2Y₁₁ antagonism on neutrophil phenotypes generated by proinflammatory stimuli. Strategies to selectively “drive” neutrophil apoptosis using cyclin-dependent kinase (CDK) inhibitors (69) have recently shown the potential of this strategy to reduce inflammation in a range of animal models. Importantly, our data suggest that encounters of the neutrophil with ATP in the microvasculature at sites of inflammation, clotting, and tissue damage may induce a prolonged survival phenotype that can be rapidly initiated and subsequently carried into tissues. These data further suggest that P2Y₁₁ antagonists will be most efficacious when administered systemically, and may conceivably deprive the neutrophil of a mechanism important in the generation of a fully-activated tissue neutrophil.

Abbreviations Used in this Paper

αβMeATP

αβ-methylene-adenosine 5′-triphosphate

ATP

adenosine 5′-triphosphate

ATP-γ-S

adenosine 5′-[γ-thio]-triphosphate

BzATP

2′-3′-O-(4-benzoylbenzoyl)adenosine 5′-triphosphate

Ca²⁺_i

intracellular calcium

cAMP_i

intracellular cyclic adenosine-3′,5′-monophosphate

cA-PK

cAMP-dependent protein kinases

HEK

human embryonic kidney

JC-1

5,5′,6,6′-tetrachloro-1,1′,3,3′-tetraethylbenzimidazolyl-carbocyanine iodide

2MeSATP

2-(methylthio)adenosine 5′-triphosphate

N⁶-MB-cAMP

N⁶-monobutyryladenosine-3′,5′-cyclic monophosphate

NAD

nicotinamide adenine dinucleotide

PBMC

peripheral blood mononuclear cell

8-pCPT-2′-O-Me-cAMP

8-(4-chlorophenylthio)-2′-O-methyladenosine-3′,5′-cyclic monophosphate

PLC

phospholipase C

Rp-8-Br-cAMP

Rp-isomer, 8-Bromo-adenosine-3′,5′-cyclicmonophosphorothioate

U73122

1-(6-((17β-3-methoxyestra-1,3,4(10)-trien-17-yl)amino)hexyl)-1H-pyrrole-2,5-dione

U73343

1-(6-((17β-3-methoxyestra-1,3,5(10)-trien-17-yl)amino)hexyl)-2,5-pyrrolidinedione

UTP

uridine 5′-triphosphate.