Research ArticleLupus

Platelet CD154 Potentiates Interferon-α Secretion by Plasmacytoid Dendritic Cells in Systemic Lupus Erythematosus

See allHide authors and affiliations

Science Translational Medicine  01 Sep 2010:
Vol. 2, Issue 47, pp. 47ra63
DOI: 10.1126/scitranslmed.3001001

Abstract

Systemic lupus erythematosus (SLE) is a systemic inflammatory autoimmune disease characterized by the involvement of multiple organs and an immune response against nuclear components. Although its pathogenesis remains poorly understood, type I interferon (IFN) and CD40 ligand (CD154) are known to contribute. Because platelets are involved in inflammatory processes and represent a major reservoir of CD154, we hypothesized that they participate in SLE pathogenesis. Here, we have shown that in SLE patients, platelets were activated by circulating immune complexes composed of autoantibodies bound to self-antigens through an Fc-γ receptor IIa (CD32)–dependent mechanism. Further, platelet activation correlated with severity of the disease and activated platelets formed aggregates with antigen-presenting cells, including monocytes and plasmacytoid dendritic cells. In vitro, activated platelets enhanced IFN-α secretion by immune complex–stimulated plasmacytoid dendritic cells through a CD154-CD40 interaction. Finally, in lupus-prone mice, depletion of platelets or administration of the P2Y(12) receptor antagonist (clopidogrel) improved all measures of disease and overall survival; transfusion of activated platelets worsened the disease course. Together, these data identify platelet activation as an important contributor to SLE pathogenesis and suggest that this process and its sequelae may provide a new therapeutic target.

Introduction

Systemic lupus erythematosus (SLE) is a systemic inflammatory autoimmune disease characterized by an immune response against cellular nuclear components. Patients with SLE experience a waxing and waning disease course and a wide array of clinical manifestations, reflecting the protean nature of the disease (1). In addition, SLE patients prematurely develop atherosclerosis independent of standard cardiovascular risk factors, which represents a leading cause of morbidity and mortality in SLE (2). The cause of SLE remains largely unknown, but environmental triggers such as viruses, in the context of susceptibility genes, may contribute.

The immunological mediators type I interferons (IFNs) also play a key role in SLE pathogenesis (3, 4), because activation of the type I IFN system is a prominent feature of early and active disease (5, 6). Plasmacytoid dendritic cells (pDCs) are the major type of IFN-producing cells during viral infections and the major source of IFN-α in SLE patients. Indeed, immune complexes (ICs) containing self nucleic acid are internalized by pDCs through Fc-γ receptor IIA (FcγRIIA) (CD32), reach the endosomal compartment, and activate IFN-α secretion downstream of Toll-like receptor 7 (TLR7) and TLR9 pathways (7, 8). Naturally occurring or induced differences in the expression of the TLR7/9 gene (9), including by CD154 (CD40 ligand), could therefore contribute to SLE pathogenesis.

CD154 is a type II transmembrane protein originally identified on activated T cells as the ligand for CD40, a molecule that promotes B cell and DC activation. CD154 is also expressed by activated platelets. Platelets are a major source of soluble CD154 (sCD40L) (10), and platelet-derived CD154 modulates innate immunity and inflammatory disease states such as atherosclerosis (11, 12). Recent studies in mice have shown that platelet CD154 also acts on the adaptive immune response by directly influencing CD8+ T and B cell responses during viral infection (1315). Although its origin is not known, CD154 concentrations in plasma are increased in SLE patients, promoting the differentiation of autoantibody-secreting cells (16).

Here, we demonstrate that platelet CD154 contributes to the activation of the type I IFN system in human SLE. In addition, we show that inhibition of activated platelets decreases disease severity in two murine models of SLE.

Results

Activation of platelets in SLE patients and correlation with disease activity

Platelets isolated from SLE patients with a wide range of disease severity (table S1) spontaneously show CD154 and CD62P (P-selectin) on their membrane surface, two markers that are hallmarks of platelet activation (14, 17). The basal levels of CD62P and CD154 on the platelet surface were significantly higher in SLE patients than in healthy blood donors, both before and after thrombin activation (Fig. 1, A and B). Because membrane-bound CD154 on activated platelets is cleaved and released in a soluble form (sCD154), we hypothesized that the increased sCD154 in the serum of SLE patients (18) (Fig. 1C, left upper panel) could result from platelet activation. Platelet lysates from SLE patients were assayed for their CD154 content and compared with those from normal donors. SLE lysates contained significantly less CD154 than did lysates from healthy individuals (Fig. 1C, left lower panel). Furthermore, platelet lysates from SLE patients with active disease contained less CD154 than did platelet lysates from patients with quiescent SLE, indicating that disease severity was associated with depletion of platelet-associated CD154 (Fig. 1C, right lower panel). This result is consistent with the observation that serum concentrations of sCD154, mean fluorescence intensity (MFI) of membrane-bound CD154, and amounts of CD154 in platelet lysates correlated with both SLE disease severity, as measured by the SLE Disease Activity Index (SLEDAI), and serum C4 concentrations (Fig. 1D and fig. S1). Conversely, in SLE patients, an inverse correlation was found between the concentration of CD154 in platelet lysates and the concentration of serum sCD154 (Fig. 1E), suggesting that increased serum sCD154 may be a result of the cleavage of platelet membrane–bound CD154 (fig. S2, A and B).

Fig. 1

Platelets are activated in SLE patients and are a major source of elevated sCD154 in SLE sera. (A) Flow cytometric analysis of CD62P and CD154 expression on freshly isolated platelets from an SLE patient and a healthy volunteer (Control) before and after thrombin activation. (B) Platelet surface CD62P and CD154 expression was evaluated in 20 SLE patients and 20 healthy individuals (Control). MFI corresponds to the expression levels of surface platelet CD62P and CD154. Boxes, 25th and 75th percentiles; error bars, 10th and 90th percentiles. (C) sCD154 was evaluated in 30 SLE sera samples, 10 control sera, and within platelet lysates for each patient or control. Boxes, 25th and 75th percentiles; error bars, 10th and 90th percentiles. (D) Serum sCD154, membrane expression (MFI) of CD154 (mCD154), and the amount of CD154 in platelet lysates were correlated with SLEDAI. (E) sCD154 in SLE sera was inversely correlated with CD154 assayed in SLE platelet lysates. *P < 0.05, **P < 0.025, ***P < 0.001, ****P < 0.0001.

Mechanism of platelet activation in SLE patients

To investigate the mechanisms of platelet activation in SLE patients, we exposed platelets from healthy controls to serum from SLE patients. Contrary to the results with allogeneic normal serum, serum from SLE patients induced platelet activation in a dose-dependent manner, as assessed by the amount of CD62P and CD154 on the membrane (Fig. 2A). We hypothesized that ICs contribute to platelet activation because (i) circulating ICs are present in large amounts in blood from SLE patients (19) and (ii) concentrations of serum sCD154 correlated with circulating ICs (Fig. 2B). On testing this, we found that platelet activation after incubation with serum from SLE patients was inhibited by preincubation of platelets with a blocking monoclonal antibody (mAb) to CD32 or by depleting the serum of immunoglobulin G (IgG) (Fig. 2C, left panel). In contrast, blocking IFN-α, CD154, and interleukin-6 (IL-6) or heat-inactivating the serum did not alter the capacity of SLE serum to induce platelet activation. Moreover, the addition of DNA and ribonucleoprotein (RNP) containing ICs to normal serum up-regulated the membrane expression levels of CD62P and CD154 in a CD32-dependent manner (Fig. 2C, right panel, and fig. S3, A and B). To assess whether ICs are found in association with platelets in SLE patients, we analyzed platelet lysates from SLE patients by probing Western blots with antibody to human IgG. Whereas IgG could be detected in SLE lysates, barely detectable amounts of IgG were found in platelet lysates from patients with rheumatoid arthritis (RA) or from healthy controls (Fig. 2D, left panel). Antinuclear antibodies (ANAs) were found among the IgGs in platelet lysates from SLE patients but were lacking in those from RA or normal patients even when RA patients had detectable serum ANA (Fig. 2D, middle panel). In addition, after the migration of lysates from purified platelets (>99%) on a 1% agarose gel, cleaved DNA sensitive to deoxyribonuclease (DNase) could be detected in platelet lysates from active SLE patients but not in platelet lysates from quiescent SLE patients (Fig. 2D, right panel, and fig. S4). Together, these data support the conclusion that in SLE patients, platelets are activated by ICs in a CD32-dependent manner.

Fig. 2

Platelets are activated by ICs through a CD32-dependent mechanism. (A) Serum from patients with active SLE–activated control platelets in a dose-dependent manner as assessed by membrane CD154 and CD62P surface levels. Mean values (±SD) of four independent experiments are shown. (B) sCD154 was correlated with serum ICs in SLE sera (n = 14). (C) Left: Preincubation of control platelets with blocking mAb to CD32, or depleting SLE sera of IgG, inhibited the ability of SLE sera to induce platelet activation. Mean values (±SD) of the inhibitory effects obtained in three different experiments with different SLE sera are shown. Right: Platelets from control individuals incubated with different SLE sera (n = 10) were significantly more activated than platelets incubated with control sera (n = 6). Addition of ICs to control serum before incubation with platelets promoted platelet activation similar to that observed with SLE sera. (D) Left: Platelets from SLE patients interact with ICs. Platelets from SLE patients, RA patients, or control individuals were purified, and their corresponding lysates were obtained. Western blot analysis of the lysates revealed that binding of IgG by platelets was much higher in SLE patients than in control or RA patients. Middle: IgG within the platelet lysates from SLE patients was directed against nuclear components (ANA) as assessed by an indirect immunofluorescence assay on Hep2 cells. Right: Platelet lysate from an active SLE patient, but not from a healthy control (Control −) or quiescent SLE patient, contained cleaved DNA. Platelet lysates were electrophoresed in 1% agarose gel before staining with ethidium bromide (one representative experiment of five). Apoptotic Jurkat cell line was used as a positive control (Control +) to detect cleaved DNA.

Consequences of platelet activation on antigen-presenting cells in SLE patients

Once activated, platelets may interact with antigen-presenting cells (APCs) and form aggregates (20). Monocytes [known to act as myeloid DCs in SLE (21)] and pDCs may also form aggregates with activated platelets, as assessed by flow cytometry and confocal microscopy analysis (Fig. 3, A and B). The relative abundance of monocyte-platelet and pDC-platelet aggregates was significantly higher in the blood of SLE patients than in healthy individuals (Fig. 3C). pDCs interact with circulating ICs from SLE patients and secrete type I IFN in a CD32- and TLR9/7-dependent manner (7, 22). For this reason, we examined the effects of activated platelets on the secretion of IFN-α by pDCs by adding control-activated platelets to purified pDCs from healthy donors. Although adding activated platelets or ICs alone triggered some IFN-α secretion, their simultaneous addition in vitro increased IFN-α production by a factor of 5 (Fig. 3D). Addition of CD154-transfected L cells potentiated IC-mediated IFN-α production. The increase in IFN-α secretion was down-regulated by the addition of blocking mAb to CD40. This demonstrates that the interaction between CD154 on activated platelets and CD40 on pDCs is a prerequisite for the enhancement of IFN-α production (Fig. 3D). In addition, platelets previously activated by SLE serum enhanced IFN-α secretion by pDCs upon TLR9 and, to a lesser extent, TLR7 stimulation (fig. S5). Supernatant from activated platelets, acting as a source of sCD154, did not have any effect on IFN-α secretion by pDCs (fig. S6). Together, these findings demonstrate a key role for platelets in SLE pathogenesis through their ability to interact with pDCs and potentiate IFN-α production, confirming that the CD154-CD40 pathway contributes to SLE disease (16).

Fig. 3

Activated platelets form aggregates with circulating APCs and modulate pDC-induced IFN-α production. (A) Platelet aggregates counted with flow cytometry. Fresh whole SLE blood was incubated with a mixture of antibodies to CD62P and CD42b to stain platelets and with an antibody to BDCA2 as a specific marker for pDCs. The proportion of triple-positive cells (BDCA2+CD42b+CD62P+) represents the occurrence of platelet-pDC aggregates. (B) Analysis by confocal microscopy confirmed that double-positive cells represented platelet-pDC aggregates. pDCs obtained by negative selection from SLE patients are stained on a slide with mAb to BDCA2 (green) and mAb to CD42b (red) and counterstained with DAPI (blue). The upper panels show CD42b staining on nonactivated platelets (left) and activated platelets (right). The left lower panel shows the BDCA2 staining on a pDC. A platelet-pDC aggregate is shown in the right lower panel. (C) The proportion of platelet-pDC and platelet-monocyte aggregates was evaluated in 15 SLE patients and 15 controls by flow cytometry. The percentage of aggregates represents the proportion of cells expressing CD42a and BDCA2 or CD14 among the total population of pDCs or monocytes. The proportion of blood platelet–pDC and platelet-monocyte aggregates was significantly higher in SLE patients than in healthy individuals. *P < 0.005, **P < 0.001. (D) Purified pDCs were incubated with ICs in the presence or absence of activated platelets for 24 hours. In one condition, blocking mAb to CD40 was added at culture onset. IFN-α in supernatants from pDCs was measured by ELISA after 24 hours of culture. Mean values (±SD) of three different independent experiments are shown.

Platelets as a therapeutic target in SLE mouse models

NZBxNZW(F1) mice with ongoing lupus-like disease exhibited increased concentrations of serum sCD154 and an increased occurrence of platelet-DC aggregates (Fig. 4A, left panels). In addition, incubation of murine normal platelets with murine SLE sera induced platelet activation, which enhanced IFN-α secretion by CpG-A–stimulated pDCs in a CD154-dependent manner (Fig. 4A, right panels). Because these features of the NZBxNZW(F1) mouse seem to recapitulate the human disease, we assessed the involvement of platelets by administering a platelet-depleting rat mAb to 6-month-old mice once a week for 6 weeks (fig. S7). Another group of animals was transfused with platelets previously activated in vitro. We found that, when compared to controls, mice depleted of platelets had significant decreased concentrations of serum antibody to double-stranded DNA (dsDNA) IgG, whereas IgM against dsDNA and blood urea concentrations remained unchanged (Fig. 4B). We found no difference between the groups in total IgG and IgM serum concentrations (fig. S8). After 6 weeks, all mice were killed and their kidneys were examined under light microscopy. Inflammatory infiltrates in the interstitium and signs of glomerulonephritis were reduced in kidneys from platelet-depleted mice compared to animals treated with the isotype control antibody. This was in agreement with an improvement of immunohistological and histological scores (Fig. 4C). In contrast, the kidneys of animals transfused with activated platelets displayed denser inflammatory infiltrates and IC deposits (Fig. 4C). Daily administration of the P2Y(12) receptor antagonist clopidogrel, which is known to down-regulate CD154 platelet expression, to 2-month-old NZBxNZW(F1) mice for 5 months blunted the expression of platelet activation markers, including serum sCD154, and also muted the interaction between platelets and leukocytes (23) (Fig. 4A, left panels). Clopidogrel administration also decreased antibody to dsDNA, proteinuria, and activation of T cells while improving kidney histological score (Fig. 4D). In addition, clopidogrel improved the survival of NZBxNZW(F1) mice and MRL.lpr lupus-prone mice (Fig. 4E and fig. S9). Together, these data reinforce the key role of platelets in SLE pathogenesis.

Fig. 4

In vivo involvement of platelets in SLE pathogenesis in NZBxNZW(F1) and MRL.lpr mouse models of SLE. (A) Sera sCD154 concentrations and occurrence of platelet-DC aggregates were evaluated in 7-month-old NZBxNZW(F1) SLE mice and wild-type (WT) mice (n = 6 in each group) (left panels). Lupus-prone NZBxNZW(F1) sera activated normal platelets, which then enhanced IFN-α production by murine pDCs upon TLR9 CpG-A triggering (right panels). Mean values (±SD) of four different independent experiments are shown. (B) Six-month-old NZBxNZW(F1) mice (n = 6 per group) were injected intravenously with a mAb to CD42b p0p3/4 (3 μg/g) (platelet depletion) or with an isotype control mAb (3 μg/g) (isotype control) on day 0 and once a week for 6 weeks. A third group was composed of six NZBxNZW(F1) mice given 2 × 107 in vitro–activated platelets (platelet transfusion) on day 0 and once a week. Serum harvested on day 0 and 6 weeks later was assayed for antibodies to dsDNA IgG and blood urea nitrogen concentrations. Data are expressed as mean values of six different mice. (C) On week 6, mice were killed and kidneys were analyzed by light microscopy. IC deposits within the glomeruli (upper panels), immunohistologic (middle panel), and histologic scores (lower panel) are represented. (D) Daily administration of the P2Y(12) receptor antagonist clopidogrel (2.5 mg/kg) to NZBxNZW(F1) mice (n = 8 clopidogrel-treated mice and n = 10 untreated controls) significantly decreases DNA IgG antibody, T cell activation, proteinuria, and kidney histological score. (E) Daily administration of clopidogrel significantly improved survival of NZBxNZW(F1) mice (n = 7 per group, 2.5 mg/kg) and MRL.lpr mice (n = 5 per group, 25 mg/kg), respectively. *P < 0.05, **P < 0.01. NS, nonsignificant.

Discussion

The best-known function of platelets is hemostasis (24). However, platelets can also be potent effectors of the innate immune response through their expression of TLRs and CD32 and through their ability to secrete bioactive molecules, including CD154 (25). Platelet activation also forms a link between the innate and adaptive immune compartments, especially through CD154 expression. Platelet-derived CD154 is sufficient to induce DC activation, trigger B cell isotype switching, and enhance T lymphocyte function (13).

SLE is the prototypic systemic autoimmune disease and is characterized by activation of the adaptive immune system, including B and T cells, as well as DCs, through the unabated secretion of type I IFNs (3). The role of CD154 in SLE pathogenesis has been extensively studied. mAbs to CD154 injected into lupus-prone mouse models prolong survival, ameliorate or prevent kidney disease, and decrease antibodies to DNA (26). Transgenic mice expressing B cell CD154 develop a lupus-like disease with age (27). In human SLE, CD154 is overexpressed on CD4+ T and CD8+ T cells and is ectopically expressed on monocytes and B cells (28, 29). sCD154 is increased in SLE sera and correlates with disease activity (30). Different mAbs to CD154 have been used in clinical trials and show promising therapeutic effects in SLE patients. Unfortunately, trials were stopped early because of thrombotic events (31). mAb bound to CD154 on the platelet surface may engage FcγRIIA, causing aggregation, an important step in thrombosis. CD154 may also stabilize arterial thrombi by binding platelet integrins.

Our study demonstrates that platelets from SLE patients are activated by circulating ICs in a CD32-dependent mechanism because, as we found, SLE serum–containing ICs induced platelet activation that could be inhibited by blocking the signal through CD32 or by depleting the serum of IgG. Platelet activation was responsible for the increased sCD154 in SLE sera, and platelet activation correlated with disease activity assessed by SLEDAI. The engagement of ICs with CD32 on pDC cell surfaces is critical for the endocytosis of self nucleic acid. Once internalized by pDCs, ICs containing self nucleic acid reach the endosomal compartment and activate TLR7 and TLR9, leading to the downstream secretion of IFN-α (7, 22, 32). We now provide additional evidence that ICs also act on platelets, leading to higher CD154 expression on platelets. This result reinforces the role of CD32 in SLE pathogenesis. The interaction between ICs and platelets could also be relevant in other inflammatory diseases in which serum sCD154 is elevated.

Among the consequences of platelet activation in active SLE, we found a significantly increased proportion of platelets aggregated with APCs, including monocytes and pDCs. This demonstrates a link between APCs and platelets in SLE patients. We showed that this interaction activated APCs via a CD154-CD40 interaction. Indeed, we demonstrated here that platelet CD154, in combination with ICs, enhanced the ability of pDCs to secrete IFN-α. Therefore, activated platelets are a major contributor to the activation of the APC network in SLE patients. The close interaction between platelets and APCs could also favor the uptake of platelet autoantigens, thus explaining the increased prevalence of immune thrombocytopenia in SLE patients. Moreover, thrombocytopenia induces platelet activation, which may explain why thrombocytopenia is a risk factor for a poor prognosis in SLE patients (33).

Depletion of platelets and megakaryocytes or administration of a P2Y(12) receptor antagonist (clopidogrel) in two models of lupus-prone mice improved the disease outcome, highlighting the value of platelets as targets for SLE treatment and establishing a rationale for antiplatelet therapy in SLE patients who are free of thrombocytopenia. Moreover, because sCD154 is increased in other autoimmune diseases including RA and inflammatory bowel disease, platelets may also be involved in the chronic inflammatory process in those patients, and similar therapy may be beneficial.

Finally, SLE patients are prone to premature cardiovascular disease, and several SLE-associated risk factors have been identified in this context, such as prothrombotic antiphospholipid syndrome, accelerated endothelial cell apoptosis, and impaired repair of damaged endothelium (34, 35). An increase in systemic platelet activation occurs in a variety of atherosclerotic diseases, including coronary artery disease, transplant vasculopathy, and carotid artery disease (25). Enhanced systemic platelet activation correlates with the increase in the intima-media thickness of the carotid in humans, a feature frequently observed in SLE patients. Thus, long-term antiplatelet strategy may help not only the immune aspects of SLE pathogenesis but also the cardiovascular problems. Consistent with this idea is the fact that statin therapy, which acts on platelet activation and reduces plasma sCD154 in patients with coronary artery disease (36), reduces serum urea, proteinuria, autoantibody production, and glomerular Ig deposition in NZB/W F1 female mice with established lupus disease (37).

Our results extend for human autoimmune disease the demonstration that activated platelets amplify the inflammatory response in a murine model of RA (38). Together, these observations open a possible therapeutic avenue for human inflammatory autoimmune diseases—the long-term utilization of antiplatelet therapy.

Materials and Methods

Patients

SLE patients (n = 37) who fulfilled four or more of the 1982 revised American College of Rheumatology criteria for the disease were enrolled in this study (39). The group of healthy subjects (controls) comprised our medical staff who were free of any autoimmune or infectious diseases. Patients with phospholipid antibodies and/or thrombopenia were excluded from the study. Table S1 summarizes demographic and clinical characteristics. An active patient is defined as a patient at diagnosis or a patient with a three-point increase in the SLEDAI when compared to the previous one. Quiescent patients are characterized by SLEDAI less than or equal to 3.

Platelets

Platelet-rich plasma. Platelet-rich plasma (PRP) was prepared by centrifugation of citrated blood samples at 1500 rpm for 10 min. Platelets were counted with a Cell-Dyn 3500 hemocytometer, and PRP was adjusted to the desired concentration by addition of Tyrode’s buffer [134 mM NaCl, 2.9 mM KCl, 0.34 mM Na2PO4, 12 mM NaHCO3, 20 mM Hepes, 1 mM MgCl2, 5 mM glucose (40)].

Washed platelets. PRP was added to 1:10 (v/v) ACD-A [tris–sodium citrate (25 g/liter), d-glucose (20 g/liter), and citric acid (15 g/liter)] and spun at 1500 rpm for 10 min. The platelet pellet was resuspended in Tyrode’s buffer and 1 μM prostaglandin E1 (PGE1), and apyrase (80 U/ml) (Sigma) was added. Platelets were counted and adjusted to the desired concentration by addition of Tyrode’s buffer.

Platelet activation. To activate platelets, we added thrombin (0.5 IU/ml) (Sigma) to platelets in Tyrode’s buffer. To obtain activated platelet supernatants, we centrifuged activated thrombin platelets at 1500 rpm for 10 min at room temperature; supernatants were harvested and used fresh.

To evaluate the percentage of activated platelets by flow cytometry, we gated platelets by forward light and side scatter profiles and by CD61 fluorescein isothiocyanate (FITC) fluorescence. Platelets (50,000 events) were analyzed at a flow rate of <1000 per second. To determine the percentage of activated platelets in each sample, we set the negative cutoff for each antibody (CD154 and CD62P) using appropriate isotype control antibodies.

Platelet lysate. Platelets (1 × 108), activated or not by thrombin, were lysed in Tyrode’s buffer supplemented with 1% Triton X-100 during 30 min at room temperature in the presence of leupeptin (20 μg/ml) (Sigma). The lysate was then centrifuged at 10,000 rpm for 10 min at room temperature. Supernatants were harvested and frozen at −20°C until used.

Analysis of DNA fragmentation. Jurkat cells (108) were incubated for 4 hours with human recombinant Fas ligand IgG (100 ng/ml) (Alexis). Platelets (6 × 108 per condition) or apoptotic Jurkat cells were lysed, and DNA was extracted by ethanol precipitation before migration in 1% agarose gel as described previously (41). In some cases, DNA fragments were digested by DNase I (20 U/ml) (Sigma).

Western blot analysis. Platelet lysates (108) were separated by SDS–polyacrylamide gel electrophoresis on 10% gels in reducing conditions and transferred to a polyvinylidene difluoride membrane, as described previously (42). Ig was revealed by the use of peroxidase-labeled human polyclonal antibody to Ig and a luminol fluorescence kit (enhanced chemiluminesence).

Determination of ANA autoantibodies. ANAs were detected by indirect immunofluorescence on Hep2 cell lines (Institut Jacques).

F(ab′)2 preparation

IgG fractions from SLE sera were purified by means of an Immunopure IgG Preparation kit (Pierce). F(ab′)2 fragments were obtained with a F(ab′)2 Preparation kit according to the manufacturer’s instructions. The purity of the resulting F(ab′)2 fragments was tested by electrophoresis.

Immune complexes

ICs were generated as published elsewhere (43). Briefly, serum samples from SLE patients were obtained and anti-dsDNA and anti-RNP concentrations were measured with a commercially available enzyme-linked immunosorbent assay (ELISA) kit (Fidis, BMD). All samples were filtered through a 0.45-μm polyvinylidene fluoride syringe before purification as described previously. The T cell line Jurkat cells were treated with Fas ligand IgG (100 ng/ml) (Alexis) for 4 hours at 37°C. Cells were then washed once before use. Apoptotic bodies were then incubated with sera from SLE patients or healthy volunteers (no ICs) for 2 hours at 37°C and then centrifuged at 10,000 rpm. The pellet was harvested and diluted before use.

Cell preparation

Human peripheral blood mononuclear cells (PBMCs) were isolated from buffy coats by Ficoll gradient centrifugation. Primary human pDCs were purified from PBMCs by immunomagnetic bead positive selection with BDCA4 microbeads following the manufacturer’s instructions (Miltenyi Biotec). Briefly, PBMCs were stained with antibody to BDCA4 coupled to colloidal paramagnetic microbeads and passed through a magnetic separation column (LS and RS column; Miltenyi Biotec). In some experiments, pDCs were then cell-sorted by flow cytometry (FACSAria cell sorting system, BD Biosciences) to reach a purity of >95%.

Media and reagents. All cells were cultured in complete medium [RPMI 1640, 2 mM l-glutamine, penicillin-streptomycin (100 U/ml), and 8% low-endotoxin fetal calf serum (FCS)] and maintained at 37°C and 5% CO2.

Antibody to CD40 mAb89 IgG1 was purchased from Schering Plough and used at a concentration of 1 μg/ml. Antibodies to P-selectin glycoprotein ligand 1 (PSGL-1) (clone IM 2091) and to IL-6 were purchased from BD Biosciences. Gardiquimod (1 μg/ml), CpG type A (ODN 2216, 5 μM), and chloroquine (100 μM) were purchased from InvivoGen. IL-3 was purchased from R&D Systems and used at 10 ng/ml.

Stimulating conditions. Isolated pDCs were cultured in 96-well round-bottom plates (Nunc) (7.5 × 104 cells in 200-μl medium per well) with platelets (5 × 106) in the presence of the different stimuli. In some experiments, CD40L-transfected L cells (or nontransfected control L cells) (at a 1:4 ratio) were cocultured with pDCs.

Flow cytometry

Washed or plasma-rich platelets (104/μl) were incubated in the dark for 15 min at room temperature with 20 μl of FITC-conjugated mAb specific for CD61 [glycoprotein IIIa (GPIIIa)], 20 μl of phycoerythrin (PE)–conjugated mAb specific for CD154, or 20 μl of allophycocyanin-conjugated mAb specific for CD62P (P-selectin) (BD Biosciences). All readings were carried out with the same cell cytometer (FACSCalibur, BD Biosciences) settings. PE-conjugated mAb specific for CD42b was used for the monitoring of aggregates.

Confocal microscopy

PBMCs were isolated from SLE patients by Ficoll gradient separation, and pDCs were enriched by negative selection. Cells (105) were allowed to adhere for 10 min at room temperature onto poly-l-lysine–coated slides (Kindler). Then, cells were fixed by dipping slides in phosphate-buffered saline (PBS) containing 4% formaldehyde for 15 min, washed twice in PBS–1% bovine serum albumin (BSA), and finally stained with a primary mouse antibody to human GPIb (Wm23 antibody to GPIb, given by M. Berndt) in PBS–1% BSA for 1 hour at room temperature. Slides were washed twice with PBS and stained with Alexa Fluor 568–conjugated goat secondary antibody to mouse in PBS–1% BSA during 1 hour at room temperature. Slides were washed with PBS and stained with BDCA2-FITC antibody and 4′,6-diamidino-2-phenylindole (DAPI) during 1 hour at room temperature. After washing with PBS, slides were dried and mounted with mounting medium (DakoCytomation). Images were acquired and processed on a confocal microscope (Leica) with a 63× objective.

ELISA

Commercially available kits were used according to the manufacturer’s instructions to quantify human IFN-α, and sCD154 for human and mice (Bender MedSystems). ICs were evaluated with a commercially available kit (Binding Site).

Mice

Isolation and activation of murine platelets. Mice were anesthetized and bled for cardiac puncture. Blood was collected into syringes containing 0.1 ml of ACD [sodium citrate (12.5 g/liter), d-glucose (10.0 g/liter), and citric acid (6.85 g/liter)], added to 1 ml of Pipes [150 mM NaCl and 20 mM Pipes (pH 6.5)], and centrifuged at 1000 rpm for 20 min. The PRP was collected, and 1 μM PGE1 was added and centrifuged at 2000 rpm for 10 min. The platelet pellet was then resuspended in 2 ml of Tyrode’s buffer [134 mM NaCl, 2.9 mM KCl, 0.34 mM Na2PO4, 12 mM NaHCO3, 20 mM Hepes, 1 mM MgCl2, 5 mM glucose, and 0.5 mg/ml BSA (pH 6.5)] and counted. To activate platelets, we added thrombin (0.5 IU/ml) (Hyphen Biome) to platelets.

Flow cytometry. Platelets (3 × 106) were incubated for 15 min at room temperature in the dark with 10 μl of PBS buffer for controls or 10 μl of sera with 5 μl of PE-conjugated mAbs specific for CD61 (BD Pharmingen) and 5 μl of FITC-conjugated mAbs specific for CD62P (BD Pharmingen) and then washed twice with PBS (0.5% BSA and 2 mM EDTA). All readings were carried out with the same cell cytometer (FACSCanto, BD Biosciences) settings.

DC generation and activation. Bone marrow cells were seeded at 1.5 × 106 cells/ml in complete RPMI 1640 [10% fetal bovine serum, 2 mM l-glutamine, 50 mM 2-mercaptoethanol, penicillin (100 U/ml), and streptomycin (100 mg/ml)] supplemented with 7.5% conditioned medium from Fms-like tyrosine kinase 3 ligand (FL)–transfected B16 cells. The FL B16 cells were originally made by H. Chapman (and provided by U. Von Andrian). The cells were used for experiments after 8 days, and >90% were CD11c+, of which 15 to 40% displayed a pDC phenotype (CD11c+CD45RAhighB220highCD11blow) and the remainder displayed a conventional DC (cDC) phenotype (CD11c+CD45RAlowB220lowCD11bhigh). Collectively, the cells obtained on day 8 containing both pDC and cDC populations are referred to as FL-DCs. Before setting up each assay, FL-DCs were routinely checked by flow cytometry for the relative percentages of pDCs and cDCs. The following were used: antibody to CD11c-PE; biotinylated antibody to CD45RA, followed by streptavidin-PE-Cy5; antibody to B220-FITC; antibody to CD11b-FITC; and antibody to CD62L-FITC (BD Biosciences).

Platelets (6 × 106) were seeded in 96-well round-bottom plates in the presence or absence of a blocking mAb to CD154 (MR-1). FL-DCs were immediately added and cultured in complete RPMI 1640 with the appropriate TLR ligands in a total well volume of 200 μl. After 24 hours, the supernatants were collected for cytokine measurement. After collection of the supernatants, the FL-DC activation status was assessed by flow cytometry with CD86-PE antibody staining (BD Biosciences). All the data were analyzed with FlowJo software (Tree Star).

Mouse IFN-α measurement. IFN-α concentrations were measured by an in-house ELISA with commercially available antibodies as described (44).

Platelet-DC aggregates. Blood was collected by intracardiac puncture with 3.8% citrate buffer prefilled syringes and immediately transferred to a 3.8% citrate buffer tube. Blood samples (100 μl) were incubated with 20 μl of FITC-labeled antibody to mouse CD62P (BD Pharmingen), with 20 μl of PE-labeled antibody to mouse CD11c (BD Pharmingen), or with the same amounts of FITC- and PE-labeled relevant control isotype for 30 min at room temperature. Red blood cells were then lysed with fluorescence-activated cell sorting (FACS) lysing solution (Becton-Dickinson) for 3 min at room temperature. After two washes with PBS–2% FCS, samples were acquired on a FACSCanto II cytometer and analyzed with FlowJo software.

This study was conducted in compliance with approved animal experimental procedures by the Animal Care and Use Committee at Paris Descartes University. Six-month-old NZBxNZW(F1) mice (n = 6) (Harlan) were injected intravenously with a mAb to CD42b p0p3/4 (3 μg/g) (platelet depletion) or with the isotype control on day 0 and once a week for 6 weeks. The platelet-depleting antibody (a mixture of mAb clones p0p3 and p0p4) was generated as previously described (45). A third group of six NZBxNZW(F1) mice was given 1 ml of 2 × 107 activated platelets/ml on day 0 and once a week. Serum was harvested on day 0 and once a week and tested for anti-dsDNA IgG and IgM blood concentrations. Concentrations of total antibodies to dsDNA IgG and IgM were measured with standard ELISA as described previously (46). Blood urea nitrogen concentrations were measured on day 0 and week 6 with standard techniques for laboratory determinations. On week 6, mice were killed and kidneys were removed for histological and immunohistological studies. The left kidney was fixed in 10% neutral buffered formalin, and the right kidney was snap-frozen in optimal cutting temperature compound and stored at −70°C. The formalin-fixed tissue was sectioned and treated with hematoxylin and eosin. The degree of glomerular damage was assessed with the National Institutes of Health activity score system (47) by an assessor who was blinded with regard to the source of the sample. Briefly, each sample was assessed on the basis of cellular proliferation, leukocyte infiltration, cellular crescents, and wire loop formation. Each of these elements was scored 0 (normal), 1 (mild), 2 (moderate), or 3 (severe abnormality). The maximal possible score was 12 points. Immunofluorescence staining of cryosectioned kidneys was used to evaluate IgG deposition. Cryosections were fixed in acetone and blocked with 5% skim milk in PBS. Samples were stained with FITC-conjugated goat antibody to mouse IgG (10 μg/ml) (Sigma-Aldrich) in PBS–0.1% BSA for 1 hour at room temperature. Images were viewed and captured with an Olympus IX50 image system (Olympus). IgG deposition was scored on a 0 to 3 scale as described previously (48): 0 = undetectable, 1 = detectable, 2 = moderate intensity of staining or >50% of glomeruli with IgG deposits, and 3 = severe or >75% of glomeruli with IgG deposits. Blood was collected in EDTA-coated tubes upon retro-orbital puncture and platelet count was performed with an automated hematology analyzer (Beckman Coulter LH 750 TM). Two-month-old NZBxNZW(F1) mice were treated 5 days a week with either PBS (n = 8) or clopidogrel (n = 10; 2.5 mg/kg by gavage; Plavix, Sanofi-Aventis) for 5 months. At the time of killing, sera were collected and tested for anti-dsDNA IgG concentrations by standard ELISA, for blood urea nitrogen concentrations as described above, or for sCD40L (Bender MedSystems). Spleens were collected, cells were isolated by gentle disruption of the tissues, and erythrocytes were lysed by hypotonic shock in potassium acetate solution. Antibodies for the following mouse antigens were used for flow cytometry analyses of the splenic T cell population: CD4-PerCP (peridinin chlorophyll protein), CD8-PE-Cy7, CD44-FITC, and CD62L-PE (eBioscience). Samples were Fc-blocked with the 2.4G2 antibody (BD Biosciences). At least 5 × 104 cells were acquired in the live gate, as defined by size and granularity. Samples were acquired on a FACSCanto II cytometer and analyzed with FlowJo software. Twenty-four–hour urine samples were collected from individual mice the day before the killing and tested for protein content by the Coomassie Plus (Bradford) protein assay (Pierce). For the survival time courses, 2-month-old NZBxNZW(F1) mice (n = 7 per group) or 2-month-old MRL.lpr mice (n = 5 per group) (Harlan) were treated 5 days a week with either PBS or clopidogrel by gavage (Plavix) (2.5 and 25 mg/kg, respectively). Mice were euthanized when they appeared moribund to two observers who were not aware of their regimen.

Statistical analysis

Pearson’s correlation analysis was used to measure correlation between two parameters. Comparisons between groups were performed with the nonparametric Mann-Whitney U test with a level of significance at P = 0.05. We used the Spearman test to determine the correlations. The tests were performed with the statistical software Statistica (Statsoft).

Supplementary Material

www.sciencetranslationalmedicine.org/cgi/content/full/2/47/47ra63/DC1

Fig. S1. Platelet content of CD154 correlates with disease activity.

Fig. S2. Platelet release of CD154 upon activation is metalloproteinase-dependent.

Fig. S3. Immune complexes directly activate platelets and induce the release of sCD154.

Fig. S4. Cleaved DNA is not detected in platelets from RA patient or immune thrombopenic purpura patient.

Fig. S5. Platelets enhance IFN-α secretion by TLR7 and TLR9 agonist–stimulated pDCs.

Fig. S6. Soluble CD154 from activated platelets has no effect on IFN-α secretion by pDCs upon TLR9 activation.

Fig. S7. Administration of the rat mAb to CD42b p0p3/4 efficiently induces platelet depletion in vivo.

Fig. S8. In vivo depletion of platelets does not change total serum IgG and IgM.

Fig. S9. Long-term clopidogrel treatment improves survival of NZBxNZW(F1) and MRL.lpr lupus mouse models in a dose-dependent manner.

Table S1. Summary of the demographic, clinical, and biological characteristics and treatments of SLE patients enrolled in the study.

Footnotes

  • * These authors contributed equally to this work.

  • Citation: P. Duffau, J. Seneschal, C. Nicco, C. Richez, E. Lazaro, I. Douchet, C. Bordes, J.-F. Viallard, C. Goulvestre, J.-L. Pellegrin, B. Weil, J.-F. Moreau, F. Batteux, P. Blanco, Platelet CD154 potentiates interferon-α secretion by plasmacytoid dendritic cells in systemic lupus erythematosus. Sci. Transl. Med. 2, 47ra63 (2010).

References and Notes

  1. Acknowledgments: We thank J. C. Carron, M. Garcie, D. Aufrere, V. Pitard, F. Mouginot, and all members of the Laboratory of Clinical Immunology at the Centre Hospitalier Régional de Bordeaux who contributed to this study. We thank A. Taïeb and T. Aprahamian for critically reading the manuscript and for their helpful comments. Funding: This work was supported by CNRS. Author contributions: P.D., J.S., and I.D. did the experiments on human SLE. P.B., J.-F.M., and C.B. wrote the paper. C.N., F.B., C.R., and B.W. did the experiment on mice. E.L., J.-F.V., and J.-L.P. conducted the sampling and clinical study of SLE patients. P.D., J.S., and F.B. participated in the interpretation of the results and the editing of the manuscript. Competing interests: The authors declare that they have no competing interests.
View Abstract

Stay Connected to Science Translational Medicine

Navigate This Article