SEARCH

SEARCH BY CITATION

Keywords:

  • ATP;
  • hemostasis;
  • inflammation;
  • P2X1;
  • phagocytes;
  • sepsis

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of Conflict of Interests
  10. References
  11. Supporting Information

Summary.  Background:  In sepsis, extracellular ATP, secreted by activated platelets and leukocytes, may contribute to the crosstalk between hemostasis and inflammation. Previously, we showed that, in addition to their role in platelet activation, ATP-gated P2X1 ion channels are involved in promoting neutrophil chemotaxis.

Objectives:  To elucidate the contribution of P2X1 ion channels to sepsis and the associated disturbance of hemostasis.

Methods:  We used P2X1−/− mice in a model of lipopolysaccharide (LPS)-induced sepsis. Hemostasis and inflammation parameters were analyzed together with outcome. Mechanisms were further studied ex vivo with mouse and human blood or isolated neutrophils and monocytes.

Results:  P2X1−/− mice were more susceptible to LPS-induced shock than wild-type mice, despite normal cytokine production. Plasma levels of thrombin–antithrombin complexes were higher, thrombocytopenia was worsened, and whole blood coagulation time was markedly reduced, pointing to aggravated hemostasis disturbance in the absence of P2X1. However, whole blood platelet aggregation occurred normally, and P2X1−/− macrophages displayed normal levels of total tissue factor activity. We found that P2X1−/− neutrophils produced higher amounts of reactive oxygen species. Increased amounts of myeloperoxidase were released in the blood of LPS-treated P2X1−/− mice, and circulating neutrophils and monocytes expressed higher levels of CD11b. Neutrophil accumulation in the lungs was also significantly augmented, as was lipid peroxidation in the liver. Desensitization of P2X1 ion channels led to increased activation of human neutrophils and enhanced formation of platelet–leukocyte aggregates.

Conclusions:  P2X1 ion channels play a protective role in endotoxemia by negatively regulating systemic neutrophil activation, thereby limiting the oxidative response, coagulation, and organ damage.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of Conflict of Interests
  10. References
  11. Supporting Information

Sepsis is a clinical syndrome characterized by the presence of both infection and systemic inflammation [1]. Severe sepsis is associated with organ failure, and in septic shock hypotension occurs, despite appropriate fluid resuscitation. Severe sepsis is often associated with disseminated intravascular coagulation characterized by microthrombus formation in the microvasculature, which is partially responsible for multiorgan damage [2].

Sepsis is a complex disease involving inflammation, immunity, and coagulation [3,4]. Neutrophils and monocytes/macrophages are key players in inflammation and innate immunity. These phagocytes display a wide array of antimicrobial activities, including the generation of reactive oxygen species (ROS) by the NADPH oxidase. In sepsis and endotoxemia, monocytes are the predominant cells expressing tissue factor (TF), which is responsible for activation of coagulation [5]. Inappropriate activation and positioning of neutrophils within the microvasculature is likely to contribute to the pathologic manifestations of multiple organ failure [6,7]. Moreover, neutrophil degranulation, nucleosome externalization and concomitant ROS production promote coagulation [8].

Beyond their central role in hemostasis and thrombosis [9], activated platelets facilitate immune cell recruitment and activation; they interact with a wide variety of microbial pathogens [10–12], and release antimicrobial peptides that kill pathogens and promote neutrophil extracellular trap formation [12] as well as nucleosome externalization from these cells [8]. It has recently been proposed that platelets and neutrophils contribute to the establishment of bidirectional communication between the host’s immune response and blood coagulation, enhancing the efficacy of two major host protection systems, hemostasis and innate immunity [13].

Over recent years, several studies have provided compelling evidence of hemostasis and immune regulation by purinergic signaling [14]. Upon secretion by activated blood platelets and leukocytes, extracellular nucleotides trigger major positive feedback signals to amplify initial activation [15–17]. These nucleotides act through two structurally distinct families of purinergic receptors: the G-protein-coupled P2Y receptors and the ATP-gated P2X non-selective ion channels. Platelet P2 receptors have been identified, and include the P2X1 ion channel and two receptors for ADP, the P2Y1 and P2Y12 receptors. Immune cells express multiple P2 receptors, but their respective functions in leukocyte biology are still poorly understood. The P2Y2 and P2X7 receptors are the best characterized, and were shown to promote, respectively, neutrophil chemotaxis and the oxidative response, and cytokine secretion by macrophages [15]. We recently demonstrated that neutrophils express P2X1 ion channels that facilitate neutrophil chemotaxis through activation of the RhoA–ROCK pathway [18].

P2X1 ion channels are widely distributed in the cardiovascular system [19]. On vascular smooth muscle cells, they are involved in the vasoconstriction of resistance arterioles [20]. Platelet P2X1 critically contributes to thrombus stability under high shear stress [21,22]. A recent study indicated a role for P2X1 ion channels in lymphocyte activation [23]. These channels are also present on mouse macrophages [24]. The newly discovered expression pattern of P2X1 suggests that these receptors have broader functions in the vascular compartment than previously expected and may be involved in the interplay between platelets and leukocytes.

In this study, we investigated whether P2X1 ion channels could contribute to endotoxin-induced sepsis and the associated disturbance of hemostasis. We found that inhibition or absence of P2X1 ion channels resulted in excessive neutrophil activation and oxidative responses. Lipopolysaccharide (LPS)-treated P2X1−/− mice exhibited exaggerated neutrophil sequestration in the lungs and aggravated oxidative tissue damage, associated with exacerbated thrombocytopenia and increased activation of coagulation, which translated into higher susceptibility to septic shock.

We propose that activation of P2X1 ion channels by ATP on neutrophils represents a new mechanism that dampens systemic inflammation and the associated disturbance of the normal hemostatic balance.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of Conflict of Interests
  10. References
  11. Supporting Information

Reagents and antibodies

All chemicals used for standard biochemical methods were either from Sigma-Aldrich (Bornem, Belgium) or Merck (Darmstadt, Germany). α,β-Methylene ATP (αβmeATP), the synthetic formylated peptide formyl-methionyl-leucyl-phenylalanine (fMLP), bovine serum albumin (BSA), dihydrorhodamine 123 (DHR), phorbol myristate acetate (PMA), zymosan and luminol were from Sigma-Aldrich. NF449 was from Tocris (Bristol, UK). W-peptide was from Innovagen (Lund, Sweden). Ficoll-paque PLUS was from GE Healthcare (Diegem, Belgium). Dextran T-500 was from VWR International (Leuven, Belgium). Hank’s balanced salt solution (HBSS), RPMI-1640 and phosphate-buffered saline (PBS) were from Lonza (Verviers, Belgium). Heparin was from LEO Pharma (Wilrijk, Belgium). HORM collagen was from Nycomed (München, Germany).

Rabbit polyclonal antibodies raised against p47phox and phosphoSer345-p47phox have been previously described [25].

Cangrelor, a competitive P2Y12 platelet receptor antagonist, was kindly provided by The Medicines Company (Parsippany, NJ, USA).

Mice

Homozygous P2X1−/− mice were generated by Mulryan et al. [26] and backcrossed with the C57Bl/6J strain for nine generations [21]. All animals used in this study were 8–12-week-old C57Bl/6J mice housed in specific pathogen-free animal facilities. All experiments were carried out following the guidelines of and in agreement with the local ethics committee.

Isolation of human neutrophils and monocytes

Peripheral blood was drawn by venipuncture from healthy volunteers in acid–citrate–dextrose anticoagulant (93 mm sodium citrate, 7 mm citric acid, 0.14 mm dextrose, pH 6.5). Institutional Review Board approval was obtained from the Centre Hospitalier Universitaire de Liège, and informed consent was obtained from volunteers in accordance with the Declaration of Helsinki. Ficoll-paque PLUS gradient was used to separate peripheral blood mononuclear cells from granulocytes. Granulocytes and monocytes were prepared according to standard procedures before being resuspended in HBSS supplemented with 1 mm CaCl2, 2 mm MgCl2 and 0.2% BSA at the desired concentrations.

Isolation of mouse phagocytes

Neutrophils were isolated from mouse bone marrow by positive magnetic selection with an anti-Gr1 antibody (Miltenyi Biotec, Utrecht, The Netherlands).

Mouse peritoneal macrophages were obtained from peritoneal lavage 3 days after injection of 3% thioglycollate, and cultured in RPMI complete medium.

Respiratory burst activity

The isolated neutrophils were incubated with 100 μm DHR for 5 min at 37 °C prior to induction of ROS production by the addition of various neutrophil-stimulating agents. After 15 min of incubation, the level of DHR oxidation was measured by flow cytometry (BD FACSCanto II; BD Biosciences, Erembodegem, Belgium).

Alternatively, ROS production was measured by the use of luminescence by adding 1 mm luminol to cell suspensions. An oxidative burst was triggered, and luminescence was recorded for 5 min in a Wallac VICTOR2 microplate reader (Perkin Elmer, Waltham, MA, USA). Data are expressed as relative luminescence units (RLU) measured after 2 min at the peak of ROS production.

Whole blood lumi-aggregation assays

ROS production and platelet aggregation were simultaneously measured in hirudinized whole blood samples with a lumi-aggregometer (Chrono-Log 700, Havertown, PA, USA). For ROS production, results are expressed as mean ± standard deviation (SD) of RLU. For ex vivo whole blood aggregation studies, impedance was measured.

Cell surface protein detection by flow cytometry

Levels of CD11b, CD18 and CD62L surface expression were determined on isolated human neutrophils or in hirudinized or citrated whole blood (human or mouse) with flow cytometry (FACSCanto II; BD Biosciences). The antigen-specific fluorochrome-conjugated antibodies were from BD Biosciences, except for the anti-CD11b (activation epitope) antibody (clone CBRM1/5), which was from eBioscience (Vienna, Austria). Platelet–leukocyte aggregates were analyzed with an anti-CD61 antibody (BD Biosciences) in combination with the anti-CD11b antibody.

Mouse model of endotoxemia

Mice were injected intraperitoneally with a sublethal dose of LPS (1 mg kg−1; Escherichia coli 0111:B4; Sigma-Aldrich). Blood was drawn under anesthesia from the retroorbital plexus, except for the study of coagulation parameters, for which blood was drawn from the inferior vena cava in citrate 3.2% anticoagulant. Platelet counts were determined on a CellDyn 3700 hematocytometer (Abbott Laboratories, Louvain-La-Neuve, Belgium). Lungs and liver were surgically removed, rinsed in ice-cold PBS, blotted dry, and snap-frozen until further use. For the septic shock model, the time course of mouse survival was monitored after intraperitoneal injection of 20 mg kg−1 LPS. Survival curves were compared by use of the log-rank test for statistical significance.

Measurement of plasma glutathione and total nitrates and nitrites

For the quantification of glutathione, blood drawn in heparin as anticoagulant was immediately mixed with four volumes of ice-cold 5% metaphosphoric acid. The concentrations of glutathione were measured with a colorimetric assay kit for total glutathione, reduced glutathione (GSH), and oxidized glutathione (GSSG) (Trevigen, Abingdon, UK), following the manufacturer’s instructions.

For the quantification of total nitrates, EDTA-anticoagulated plasma samples were ultrafiltered on Amicon Ultra centrifugal filter devices with a molecular mass cut-off of 10 kDa, and total nitrate + nitrite concentrations were determined with the nitrate/nitrite colorimetric assay kit, based on Griess reagents, following the manufacturer’s instructions (Cayman Chemical Company, Ann Harbor, MI, USA).

Myeloperoxidase (MPO) assays

MPO activity was measured in lung extracts with a method adapted from Bradley et al. [27]. See the expanded method in Data S1.

ELISA assays

Determination of MPO levels was performed in heparin-anticoagulated plasma samples with the mouse MPO ELISA kit from Hycult Biotech (Uden, The Netherlands).

Thrombin–antithrombin III (TAT) complexes were quantified in citrate-anticoagulated plasma samples with the Enzygnost TAT micro (Siemens Healthcare Diagnostics Products, Marburg, Germany).

Whole blood coagulation assay

In vitro whole blood coagulation time was measured with a KC4A coagulometer. See the expanded method in Data S1.

TF activity measurements

Peritoneal macrophages (1 × 106) were stimulated for 6 h either with 1 μg mL−1 LPS or with vehicle (saline), and then lysed with 15 mm n-octyl-glucopyranoside in 25 mm Hepes and 175 mm NaCl (HN buffer) for 15 min at 37 °C. TF activity in macrophage lysates was measured with a home-made activity assay based on factor Xa generation that has been described elsewhere [28]. See the expanded method in Data S1.

Immunohistochemistry

Lungs and liver were fixed in 4% paraformaldehyde in PBS overnight, and embedded in paraffin. Lung sections were stained with hematoxylin and eosin. Liver sections were stained with either hematoxylin and eosin or a polyclonal anti-4-hydroxy-2-nonenal antibody (1 : 1000 dilution; Alpha Diagnostic, San Antonio, TX, USA), with the Dako EnVision+ System-HRP (DAB) kit (Dako, Glostrup, Denmark). Bright-field images were taken with an FSX100 microscope (Olympus, Hamburg, Germany).

Statistics

Data are presented as mean ± SD of at least three independent experiments. Statistical analyses were performed with either Student’s t-test or, for multiple comparisons, one-way anova followed by the Bonferroni test.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of Conflict of Interests
  10. References
  11. Supporting Information

Absence of P2X1 increases susceptibility to endotoxin shock independently of disturbed cytokine production

P2X1−/− mice exhibited enhanced susceptibility to LPS-induced shock, and showed a significantly reduced median survival time as compared with wild-type mice (Fig. 1).

image

Figure 1.  P2X1−/− mice exhibit increased susceptibility to septic shock. Mice were injected with a lethal dose of lipopolysaccharide (LPS) (20 mg kg−1) and survival was monitored for 50 h. Data are expressed as percentage of survival. Results are from three independent experiments. ***P < 0.001 (log-rank test), n = 20 mice per group,.

Download figure to PowerPoint

An early trigger event elicited by LPS challenge consists of cytokine production. Nevertheless, serum levels of tumor necrosis factor-α (TNF-α), interleukin (IL)-1β, interferon (IFN)-γ and IL-10 in LPS-treated P2X1−/− mice were found to be similar to those in wild-type mice (Fig. S1), indicating that P2X1 deficiency did not result in dysregulated production of these major proinflammatory and anti-inflammatory cytokines.

Absence of P2X1 enhances LPS-induced activation of coagulation

We next investigated hemostasis parameters in endotoxemic P2X1−/− mice. Activation of coagulation was evaluated by measuring the levels of TAT complexes in the plasma of mice treated with a sublethal dose of LPS. TAT levels were more elevated for P2X1−/− mice (Fig. 2A), indicating that the absence of P2X1 enhanced LPS-induced thrombin generation.

image

Figure 2.  P2X1 deficiency is associated with enhanced lipopolysaccharide (LPS)-induced activation of coagulation. (A) Thrombin–antithrombin III (TAT) complex levels were measured in mouse plasma 8 h after treatment with 1 mg kg−1 LPS. Data are expressed as mean ± standard deviation (SD), and results are from three independent experiments. ***P < 0.001 vs. wild type, #P < 0.05, ###P < 0.001 vs. vehicle, n = 3–4 mice per group. (B) Platelet counts were determined in mouse whole blood 4 h after treatment with 1 mg kg−1 LPS. Data are expressed as mean ± SD. Results are from three independent experiments. *P < 0.05 vs. wild type, ##P < 0.01 vs. vehicle, n = 7 mice per group. (C) Coagulation times were measured in mouse whole blood 6 h after treatment with 1 mg kg−1 LPS. Data are expressed as mean ± SD. Results are from three independent experiments. *P < 0.05 vs. wild type, n = 5 mice per group. (D) Tissue factor (TF) activity of mouse peritoneal macrophages was measured in cell lysates after stimulation for 6 h with 1 μg mL−1 LPS or vehicle. Data are expressed as mean ± SD. n = 4 mice per group. (E) Aggregation was measured by impedance in mouse whole blood 4 h after treatment with 1 mg kg−1 LPS. Aggregation was triggered with 10 μm phorbol myristate acetate (PMA) or 20 μg mL−1 collagen. Representative traces of three independent experiments are shown. n = 4 mice per group. ND, not determined; NS, not significant.

Download figure to PowerPoint

Endotoxemia rapidly induces thrombocytopenia [29]. In the mouse model of LPS-induced sepsis, platelets mainly localize to the lung and liver microvasculature [30]. Platelet counts of P2X1−/− mice were significantly decreased at early stages after injection of a sublethal dose of LPS, when no reduction was observed in wild-type mice (Fig. 2B).

LPS-treated P2X1−/− mice exhibited a markedly reduced whole blood coagulation time as compared with wild-type mice (Fig. 2C), suggesting increased TF activity in response to LPS. As monocytes are the major source of TF in endotoxemia [31], we measured total TF activity in peritoneal macrophages upon ex vivo stimulation with LPS. However, P2X1−/− macrophages produced the same amount of TF activity in response to LPS as wild-type macrophages (Fig. 2D).

Whole blood platelet aggregation was then measured ex vivo to assess LPS-triggered platelet activation. Under control conditions, PMA-induced platelet aggregation in P2X1−/− blood was significantly delayed as compared to that in wild-type blood (lag time: wild type, 66 ± 2.8 s; knockout, 136 ± 19.8 s; = 0.04), whereas responses to collagen were not statistically different. After LPS injection, platelet responses to both agonists became similar in wild-type and P2X1−/− blood (Fig. 2E), indicating that increased coagulation in these mice is unlikely to depend on enhanced platelet activation.

Absence or inhibition of P2X1 ion channels increases ROS production by neutrophils

Because ROS production by neutrophils has been shown to affect coagulation [8,32], we hypothesized that the increased coagulation observed in LPS-treated P2X1−/− mice could rely on this mechanism. To address this question, neutrophils were isolated from the bone marrow of P2X1−/− mice, and the production of ROS in response to formylated peptide (W-peptide), phorbol esters (PMA) or serum-opsonized zymosan (SOZ) was measured. For each stimulus tested, ROS production by P2X1−/− neutrophils was significantly higher than that measured with wild-type cells (Fig. 3A), demonstrating that the absence of P2X1 ion channels increases the oxidative activity of neutrophils.

image

Figure 3.  P2X1 ion channels negatively regulate reactive oxygen species (ROS) production by phagocytes. (A) Bone marrow-derived neutrophils isolated from P2X1−/− and wild-type mice were stimulated with formylated peptide (W-peptide) (100 nm), phorbol myristate acetate (PMA) (1 μm), or serum-opsonized zymosan (SOZ) (300 μg mL−1) for 15 min to induce ROS generation. Dihydrorhodamine 123 (DHR) oxidation was measured by flow cytometry. Data are expressed as mean ± standard deviation (SD) of DHR mean fluorescence. Results are from at least three independent experiments. *P < 0.05 vs. WT. (B) Human neutrophils were treated for 15 min with 10 μmα,β-methylene ATP (αβmeATP), 10 nm NF449, or 10 μm ATP, or for 30 min with 1 U mL−1 apyrase or 200 U mL−1 tumor necrosis factor-α (TNF-α), before induction of ROS generation with 1 μm formyl-methionyl-leucyl-phenylalanine (fMLP). DHR oxidation was measured by flow cytometry. Data are expressed as mean ± SD of fold increase as compared with fMLP alone. Results are from five independent experiments. *P < 0.05, **P < 0.01 vs. fMLP. (C) αβmeATP (10 μm), NF449 (10 nm) or apyrase (1 U mL−1) was added to hirudinized human blood for 30 min before stimulation with fMLP (10 μm). Luminescence was recorded in a lumi-aggregometer. Data are expressed as mean ± SD of relative luminescence units (RLU). Results are from seven independent experiments. *P < 0.05, ***< 0.001 vs. fMLP. (D) Hirudinized human blood was incubated for 5 min with 10 μm NF449 in the presence of 100 μm luminol. Platelet activation was induced by addition of 0.5 μg mL−1 collagen (black arrows). Platelet aggregation and ROS production were recorded simultaneously in a lumi-aggregometer. The traces shown are representative of three independent experiments.

Download figure to PowerPoint

Similarly, pretreatment of isolated human peripheral neutrophils with the desensitizing P2X1 ion channel agonist αβmeATP, the selective P2X1 antagonist NF449 or the ATP/ADP scavenger apyrase significantly increased ROS production induced by the formylated peptide fMLP, as did preincubation with TNF-α [33] (Fig. 3B). This increase coincided with higher surface expression of flavocytochrome b558 and phosphorylation of p47phox on Ser345, two key events in NADPH oxidative priming [25] (Fig. S2A,B). In contrast, the P2X1 physiologic ligand ATP significantly inhibited fMLP-induced ROS production at doses compatible with channel activation (10 μm; Fig. 3B). However, as ATP is readily degraded, one cannot exclude the possibility that adenosine, the ATP metabolite, partly contributes to the inhibitory effect observed. Indeed, 50 μm adenosine efficiently reduced fMLP-triggered ROS production by 54.1% (± 26.7%, n = 4, P < 0.05).

Preincubation of neutrophils with αβmeATP or NF449 also significantly augmented ROS production induced by SOZ (Fig. S3) and phorbol ester (data not shown), indicating that P2X1 blockade or desensitization enhances the oxidative burst regardless of the initial trigger.

Autocrine activation of P2X1 ion channels would thus limit ROS production by neutrophils via stimulus-triggered ATP release.

Importantly, similar results were obtained when ROS production was measured in human whole blood following stimulation with fMLP (Fig. 3C) or with collagen to assess the involvement of paracrine P2X1 activation by ATP released from activated platelets (Fig. 3D). In both cases, P2X1 inhibition or ATP degradation led to increased ROS generation, further supporting the possible in vivo relevance of these observations. It is noteworthy that, upon stimulation with collagen, the NF449-induced increase in ROS production was not associated with significant changes in platelet aggregation (Fig. 3D).

Absence of P2X1 ion channels increases LPS-induced oxidative tissue damage

In agreement with these data, a 15-h treatment of P2X1−/− mice with a lethal dose of LPS caused more extensive lipid peroxidation in the liver than in wild-type mice (Fig. 4A), indicating increased oxidative tissue damage.

image

Figure 4.  P2X1−/− mice exhibit enhanced oxidative tissue damage and reactive oxygen species/nitric oxide imbalance. (A) Mice were injected with a lethal dose of lipopolysaccharide (LPS) (20 mg kg−1) or vehicle. Fifteen hours later, livers were removed, fixed, and embedded in paraffin. Lipid peroxidation was analyzed by staining liver sections with an anti-4-hydroxy-2-nonenal (HNE) antibody or with hematoxylin and eosin (H&E). Representative bright-field images from three independent experiments are shown. Scale bar: 32 μm. (B) Ratio of oxidized glutathione (GSSG) to reduced glutathione (GSH) and concentrations of total nitrates and nitrites in the plasma of untreated wild-type and P2X1−/− mice. Data are expressed as mean ± standard deviation (SD) of the GSSG/GSH ratio or mean ± SD of total nitrates (μm). Results are from three independent experiments. *P < 0.05 vs. wild type, n = 11 mice per group.

Download figure to PowerPoint

We then evaluated the redox status in the blood of untreated P2X1−/− mice. For this purpose, we measured plasma concentrations of GSSG and GSH, and calculated their ratio as an index of basal oxidative stress. The GSSG/GSH ratio was about 2.5-fold higher for P2X1−/− mice than for wild-type mice (Fig. 4B). In parallel, the plasma levels of nitrates and nitrites, products of nitric oxide (NO) metabolism, were also measured. Nitrate concentrations were significantly lower in the plasma of P2X1−/− mice than in that of wild-type mice (Fig. 4B).

Absence of P2X1 ion channels enhances systemic phagocyte activation and lung neutrophil infiltration in LPS-treated mice

Injection of P2X1−/− mice with a sublethal dose of LPS led to a strong increase in CD11b expression levels on the surfaces of circulating neutrophils and monocytes, whereas these levels remained unchanged for wild-type mice (Fig. 5A). Blood concentrations of MPO reached significantly higher levels for P2X1−/− mice than for wild-type mice (Fig. 5B). These results indicate higher systemic activation of P2X1−/− neutrophils and monocytes in response to LPS in vivo.

image

Figure 5.  P2X1−/− phagocytes are more responsive to lipopolysaccharide (LPS) in vivo. Wild-type and P2X1−/− mice were treated with 1 mg kg−1 LPS or vehicle. Four hours later, blood was drawn, and CD11b expression was measured at the surfaces of neutrophils and monocytes by flow cytometry (A). **P < 0.01 vs. wild type, ##P < 0.01 vs. vehicle, n = 4 mice per group. (B) Myeloperoxidase (MPO) levels were measured by ELISA in the plasma of untreated or LPS-treated mice. Data are expressed as mean ± standard deviation (SD), and results are from three independent experiments. *P < 0.05 vs. wild type, ###P < 0.001 vs. vehicle, n = 8 mice per group. (C) MPO activity was measured in mouse lung homogenates. Data are expressed as mean ± SD of MPO units. Results are from three independent experiments. *P < 0.05 vs. wild type, ###P < 0.001 vs. vehicle, n = 9–11 mice per group.

Download figure to PowerPoint

A main consequence of endotoxemia-induced neutrophil systemic activation is their relocalization and sequestration in the lung microvasculature [34]. After LPS treatment, MPO activity, an index of neutrophil infiltration, increased in wild-type and P2X1−/− lungs but reached significantly higher levels in the lungs of P2X1−/− mice, thus confirming exaggerated activation of P2X1−/− neutrophils (Fig. 5C).

Inhibition of P2X1 ion channels increases neutrophil activation and platelet–leukocyte aggregate formation

Prolonged exposure of isolated human neutrophils to αβmeATP, as for TNF-α, led to increased surface expression of CD11b and CD18 and to CD62L shedding (Fig. 6A–C). Inhibition of P2X1 ion channels therefore modifies the phenotype of human neutrophils towards an activated state.

image

Figure 6.  Inhibition of P2X1 causes neutrophil hyperactivation and enhances the formation of platelet–leukocyte aggregates. (A–C) Isolated human neutrophils were treated with 10 μmα,β-methylene ATP (αβmeATP) or with 200 U mL−1 tumor necrosis factor-α (TNF-α), either for 3 h to measure membrane expression of CD11b (A) and CD18 (B), or for 30 min to analyze shedding of CD62L (C). Surface expression levels were determined by flow cytometry. Data are expressed as mean ± standard deviation (SD) of relative mean fluorescence as compared with the control conditions. Results are from five independent experiments. **P < 0.01, ***P < 0.001 vs. control. (D–F) Hirudinized human whole blood was treated with 10 μmαβmeATP for 2 min or with 0.01 μm cangrelor for 15 min prior to stimulation with collagen 10 μg mL−1 (D) or 1 μg mL−1 (E, F) for 15 min. Percentages of platelet–neutrophil and platelet–monocyte aggregates (D) and total (E) and activated (F) CD11b expression levels in the neutrophil and monocyte populations are shown. Data are expressed as mean ± SD. Results are from four to seven independent experiments. #P < 0.05, ##P < 0.01, ###P < 0.001 vs. unstimulated control conditions; *P < 0.05, **P < 0.01 vs. collagen.

Download figure to PowerPoint

In collagen-activated hirudinized whole blood, αβmeATP pretreatment increased the percentages of platelet–neutrophil and platelet–monocyte aggregates. This is in contrast to the inhibitory effect of the P2Y12 receptor antagonist cangrelor (Fig. 6D). P2X1 ion channels may thus act to limit platelet-dependent neutrophil and monocyte activation. Accordingly, the collagen-induced increases in CD11b expression and activation of the CD11b/CD18 integrin on neutrophil and monocyte surfaces were augmented by αβmeATP pretreatment (Fig. 6E,F). These effects of αβmeATP were not observed when sodium citrate was used as an anticoagulant instead of hirudin, a condition under which P2X1 channels are not functional, owing to divalent cation chelation.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of Conflict of Interests
  10. References
  11. Supporting Information

In this study, we have identified a novel and unexpected protective role for the ATP-gated P2X1 ion channels in endotoxin-induced sepsis. Upon LPS injection in mice, the absence of P2X1 enhances systemic neutrophil activation, causes exaggerated oxidative tissue damage, and increases coagulation and platelet consumption, all of these events being associated with higher susceptibility to endotoxin shock. Ex vivo, the inhibition or absence of these channels augments NADPH oxidase-dependent ROS production by neutrophils; it increases the expression and activity of CD11b/CD18 integrin on the surfaces of neutrophils and monocytes, as well as the formation of platelet–leukocyte aggregates.

An early trigger event elicited by LPS challenge consists of cytokine production. Nevertheless, serum levels of TNF-α, IL-1β, IFN-γ and IL-10 in LPS-treated P2X1−/− mice were found to be similar to those in wild-type mice (Fig. S1), indicating that P2X1 deficiency did not result in dysregulated production of these major proinflammatory and anti-inflammatory cytokines. Accordingly, P2X1−/− macrophages produced normal levels of TNF-α following ex vivo stimulation with LPS (data not shown).

In the setting of sepsis-associated inflammation, ROS are produced mainly through inflammatory cell NADPH oxidase. We found that the absence or inhibition of P2X1 ion channels increases NADPH oxidase-dependent ROS production by neutrophils. Luminescence-based assays performed in human whole blood also revealed higher ROS generation following P2X1 desensitization by αβmeATP. During sepsis, large amounts of NO are produced by de novo expression of inducible NO synthase (iNOS). In our study, we observed identical LPS-triggered NO production in P2X1−/− and wild-type mice (160.4 ± 57.4 μm vs. 163.6 ± 77.0 μm, respectively, n = 9), indicating normal induction of iNOS in the absence of P2X1 ion channels. As NO reacts preferentially with O2 to form the cytotoxic molecule peroxynitrite, which is responsible for tissue damage [35], we can assume that, in the LPS-treated P2X1−/− mice, excessive ROS vs. NO production will lead to exaggerated oxidative damage as compared with wild-type mice. We did indeed observe increased lipid peroxidation in the livers of LPS-treated P2X1−/− mice. Interestingly, untreated P2X1−/− mice displayed lower nitrate levels in their plasma, paralleled by a higher GSSG/GSH ratio. As glutathione is the most abundant antioxidant, acting as a free radical scavenger and inhibitor of lipid peroxidation, these observations could point to the existence of a basal ROS/NO imbalance in the blood of P2X1−/− mice, possibly leading to the increased oxidative response observed upon endotoxin challenge.

We found that LPS-treated P2X1−/− mice display increased coagulation, as shown by shortened whole blood coagulation time and elevated levels of plasma TAT complexes. However, platelet aggregability measured in whole blood does not seem to be altered in P2X1−/− mice as compared with wild-type mice. Thus, the increased coagulation observed in P2X1−/− mice would not result from increased platelet activation. Similarly, total TF activity of LPS-treated P2X1−/− macrophages was not different from that of wild-type macrophages, suggesting that the reduced coagulation time observed was not attributable to higher induction of TF in the monocytes of P2X1−/− mice. However, we cannot exclude the possibility of a role for P2X1 in TF decryption on circulating monocytes. Such a mechanism has recently been described for P2X7, another member of the P2X ion channel family. In the work of Furlan-Freguia et al. [36], P2X7 appeared to mediate protein disulfide isomerase-regulated TF activation/decryption on mouse myeloid cells and the generation of procoagulant (TF-bearing) microparticles derived from myeloid cells and smooth muscle cells. However, P2X7 is unique among the P2X ion channels because of its ability to form a large pore upon prolonged activation, allowing massive K+ efflux [37]. To date, there is no evidence supporting a role for P2X1 in the formation of platelet or monocyte microparticles, as shown for P2X7 in monocytes. It is noteworthy that the respective affinities of these channels for ATP are very different: P2X1 is activated by concentrations below 1 μm, whereas the P2X7 EC50 is ∼ 500 μm. All of the studies published so far point to distinct functions and signaling pathways for these two ATP-gated channels [15].

Alternatively, increased TF activity in endotoxemic P2X1−/− mice could result from enhanced platelet–monocyte interactions (as discussed below).

LPS-triggered neutrophil hyperactivation might lead to increased ROS production in P2X1−/− mice, which would probably affect the anticoagulant properties of the endothelium [38]. In particular, the existence of a chronic basal ROS/NO imbalance in the blood of P2X1−/− mice might reflect pre-existing endothelial dysfunction that may prime coagulation. Oxidative stress, subsequent reduced NO bioavailability and cellular redox state all play important roles in the regulation of platelet activation and of the vascular phenotype [39,40]. Reactive oxygen and nitrogen species can induce modifications in the structure and functions of platelet proteins and coagulation factors, thus changing their hemostatic properties. Fibrinogen, FV, FVIII and FX were shown to be sensitive to phagocyte-derived oxidants [41]. Oxidation of a methionine in thrombomodulin inhibits its cofactor activity, resulting in decreased levels of activated protein C [32]. TF coagulant activity was shown to be inhibited in an NO-dependent manner, further linking the regulation of endothelial thrombogenicity to oxidative stress in the setting of inflammation [42]. Importantly, a recent study indicated that neutrophil activation can promote coagulation via TF pathway inhibitor degradation by elastase [8].

Our data indicate more profound thrombocytopenia in LPS-treated P2X1−/− mice than in wild-type mice. As platelet Toll-like receptor 4 (TLR4) contributes to this phenomenon through neutrophil-dependent pulmonary sequestration [43], it is likely that, in P2X1−/− mice, the increased systemic neutrophil activation and accumulation in the lungs contributes to the augmented LPS-induced thrombocytopenia. Accordingly, histologic analyses have suggested increased amounts of microthrombi in lung microvessels of P2X1−/− mice.

In sepsis, the CD11b/CD18 integrin mediates neutrophil adhesion and infiltration into the lungs and contributes to vascular injury [44]; on neutrophils, the increased expression of this integrin depends on LPS-triggered TLR4 signaling [45], and is directly linked to ROS production [46]. Our findings that neutrophils express elevated levels of total and activated CD11b/CD18 upon P2X1 channel inhibition or deficiency are thus in agreement with the augmented neutrophil infiltration into the lungs of LPS-treated P2X1−/− mice and oxidative tissue damage.

In human whole blood, P2X1 desensitization enhances the formation of platelet–neutrophil and platelet–monocyte aggregates in response to collagen. This result suggests that P2X1 ion channels contribute to the negative regulation of vascular inflammation by affecting platelet-dependent leukocyte activation and adhesion on inflamed endothelia [47], which would further support their protective function in the setting of sepsis. Furthermore, TF expression on monocyte surfaces, the major coagulation trigger in sepsis [48], depends on monocyte interactions with activated platelets. In view of our previous study on the role of P2X1 ion channels in facilitating neutrophil migration [18], we propose that these channels could act to prevent undesired leukocyte activation and the oxidative response in the blood compartment, therefore allowing efficient chemotaxis towards inflammation/infection sites. Strikingly, a defect of neutrophil migration has been described in human sepsis [49], and the reduction in neutrophil chemotaxis to several chemotactic mediators has been associated with illness severity and organ damage [50].

Thus, P2X1 ion channels protect against endotoxemia through negative regulation of neutrophil and monocyte activation and the oxidative burst, thereby limiting oxidative tissue damage and hemostatic imbalance. Our results point to intriguing discrepancies between studies indicating a prothrombotic function of platelet P2X1 channels in some experimental mouse models of thrombosis [21,51] and the increased coagulation observed in the absence of this channel under conditions of sepsis. In models of acute systemic thromboembolism triggered by injection of collagen and adrenalin, intravascular thrombosis is rapid (< 10 min) and depends solely on platelet activation, whereas in sepsis models, endotoxemia induces long-term (over hours) activation of the whole hemostatic system, including platelets and the coagulation cascade, as well as systemic activation of leukocytes and vascular cells. The involvement of hyperactivated P2X1-deficient leukocytes in thrombosis in the latter model might then overcome the antithrombotic effect of platelet P2X1 deficiency. Moreover, the effect of P2X1 deficiency on platelet activation, as it can be measured in vitro, is small and subtle, and could therefore easily be overwhelmed by more potent activators [52].

During inflammation-induced activation of coagulation, platelet activation is potentiated by endotoxin, resulting in enhanced platelet–neutrophil interactions and cross-activation [12,53]. Platelets can also be activated directly by proinflammatory mediators such as platelet-activating factor. Whereas the function of P2X1 ion channels in collagen-induced and shear stress-induced platelet activation is well documented and has led the consideration of P2X1 as a potential target for novel antiplatelet therapies [54], data on the role of these channels in the platelet response to inflammation-related agonists are lacking. Moreover, the role of P2X1 ion channels in intravascular coagulation has never been investigated. It is noteworthy that our study may reveal a dual role for P2X1 ion channels, which could act as positive or negative regulators of thrombosis, depending on the presence of inflammation and concomitant crosstalk between platelets and inflammatory cells. In the setting of sepsis, neutrophil hyperactivation would represent a major determinant of microvascular thrombosis. Further investigations are needed to address this issue. Nevertheless, as antagonists of P2X1 ion channels may not only target platelets but may also affect neutrophils, inhibiting these channels to prevent thrombosis in the highly inflammatory environment of severe sepsis or of acute coronary syndromes might be detrimental.

Addendum

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of Conflict of Interests
  10. References
  11. Supporting Information

C. Lecut: performed research, analyzed the data, and wrote and edited the manuscript; C. Faccinetto: performed research; C. Delierneux: performed research and analyzed data; R. van Oerle and H. M. H. Spronk: designed and performed the TF activity assay, and analyzed the data; R. J. Evans: generated the P2X1−/− mice; J. El Benna: contributed to the design of the study of neutrophil respiratory burst activity and revised the manuscript; V. Bours: revised the manuscript; C. Oury: designed research, performed research, analyzed the data, and wrote and edited the manuscript.

Disclosure of Conflict of Interests

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of Conflict of Interests
  10. References
  11. Supporting Information

This work was supported by the Belgian National Fund for Scientific Research (F.R.S-FNRS), the Belgian Science Policy (IAP6/18), the Fondation Leon Fredericq and Fonds speciaux pour la recherche of the University of Liege, and the Wellcome Trust (R. J. Evans). C. Lecut is a postdoctoral researcher at the F.R.S-FNRS, C. Faccinetto was supported by a Fond pour la recherche industrielle et agricole fellowship, and C. Oury is a research associate at the F.R.S-FNRS.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of Conflict of Interests
  10. References
  11. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of Conflict of Interests
  10. References
  11. Supporting Information

Data S1. Material and methods.

Fig. S1. Time course of proinflammatory and anti-inflammatory cytokine production in response to LPS challenge.

Fig. S2. Inhibition of P2X1 receptors primes neutrophil NADPH oxidase.

Fig. S3. P2X1 ion channels inhibit reactive oxygen species production by human neutrophils in response to serum-opsonized zymosan.

FilenameFormatSizeDescription
JTH_4606_sm_SupportingInformation-FigS1-S3.pdf111KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.