Ectosomes released by platelets induce differentiation of CD4+T cells into T regulatory cells.

1219

Platelets and Blood Cells

Ectosomes released by platelets induce differentiation of CD4+ T cells into T regulatory cells Salima Sadallah1*; Francesca Amicarella2*; Ceylan Eken1; Giandomenica Iezzi2+; Jürg A. Schifferli1+ 1Department

of Biomedicine, University Hospital Basel, Switzerland; 2Institute of Surgical Research and Department of Biomedicine, University Hospital Basel, Switzerland

Summary Accumulating evidence suggests an immune-modulatory role for platelets (PLT) and PLT-derived microvesicles. In particular, ectosomes, i.e. vesicles budding from PLT surface, have been shown to exert immunosuppressive activities on phagocytes. Here we investigated the effects mediated by PLT-derived ectosomes (PLT-Ecto) on CD4+ T cells. Exposure of activated CD4+ T cells to PLT-Ecto decreased their release of IFNγ, TNFα and IL-6, and increased the production of TGF-β1. Concomitantly, PLT-Ecto-exposed CD4+ T cells displayed increased frequencies of CD25high Foxp3+ cells. These phenomena were dose-dependent and PLT-Ecto specific, since they were not observed in the presence of polymorphonuclear- and erythrocyte-derived ectosomes. Analysis of specific T cell subsets revealed that PLT-Ecto induced differentiation of naïve T cells into Foxp3+ cells, but had no effect on pre-

Correspondence to: Salima Sadallah Immunonephrology Laboratory, Department of Biomedicine University Hospital Basel, Hebelstrasse 20 4031 Basel, Switzerland Tel.: +41 61 265 32 62, Fax: +41 61 265 23 50 E-mail: [email protected]

* These authors contributed equally as first authors of the manuscript. + These authors contributed equally as senior authors of the manuscript.

Introduction The goal of platelet (PLT) transfusion is to stop or prevent bleeding in patients with PLT decrease or dysfunction. Apheresis PLT have become the most commonly used product and an essential part of the treatment of solid cancers, haematological malignancies, and haematopoietic stem cell transplantation. Even though PLT transfusions are common, it is not a benign practice since it is quite often – up to 20 % – associated with transfusion reactions (1, 2). Platelet storage lesions are the major cause of transfusion reactions. During storage, the number of microvesicles released by PLT increases with time. Platelet microvesicles have potent procoagulant properties, which may be beneficial in bleeding patients, but also detrimental by inducing thrombosis. Experimental data suggest that beside enhancing coagulation, microvesicles have proinflammatory properties (3–7). The nature of the released microvesicles was analysed some years ago, with clear evidence that activated PLT release exosomes (< 100 nm diameter) originating from multivesicular bodies, in ad© Schattauer 2014

differentiated Foxp3+ regulatory T cells (Tregs). Importantly, PLT-Ectoinduced Foxp3+ cells were as effective as peripheral blood Tregs in suppressing CD8+ T cell proliferation. PLT-Ecto-mediated effects were partly dependent on PLT-derived TGF-β1, as they were to some extent inhibited by PLT-Ecto pretreatment with TGF-β1-neutralising antibodies. Interestingly, ectosome-derived TGF-β1 levels correlated with Foxp3+ T cell frequencies in blood of healthy donors. In conclusion, PLT-Ecto induce differentiation of CD4+ T cells towards functional Tregs. This may represent a mechanism by which PLT-Ecto enhance peripheral tolerance.

Keywords Microparticles, ectosomes, extracellular vesicles, platelet immunology, T cell differentiation

Financial support: This work was supported by the Swiss National Science Foundation (320030–146255/1). Received: March 27, 2014 Accepted after major revision: August 14, 2014 Epub ahead of print: September 11, 2014 http://dx.doi.org/10.1160/TH14-03-0281 Thromb Haemost 2014; 112: 1219–1229

dition to the bigger “shed vesicles/ectosomes” (in general > 100 nm diameter) that are formed by budding from the PLT surface (8). By contrast, quiescent PLT stored for transfusion, release mostly ectosomes (Ecto) (9), with less than 2.5 % being exosomes (10). In a recent review the difficulty to differentiate precisely between these two types of extracellular vesicles was emphasised (11). We have previously shown that PLT-Ecto derived from stored PLT were heterogeneous in size from 100 to 1,000 nm, expressed PLT proteins like CD61, CD36 and CD47, and were mostly negative for the PLT lysosomal protein CD63 (9). Importantly, they were largely positive for phosphatidylserine (PS) (> 80 %). The difference in PS expression between ectosomes and exosomes might be of relevance for their biological properties (11). In previous work, we have shown that PLT-Ecto share with polymorphonuclear- (PMN) and erythrocytes (ERY) -Ecto immune suppressive properties related to a large extent to the expression of PS (9, 12–14). These observations corresponded to the known biological effects of PS expression on apoptotic cells, which are not only “eaten in silence”, but also inhibit the phagocytes that Thrombosis and Haemostasis 112.6/2014

Downloaded from www.thrombosis-online.com on 2015-01-07 | ID: 1000467556 | IP: 147.140.20.32 For personal or educational use only. No other uses without permission. All rights reserved.

1220

Sadallah, Amicarella, et al. PLT-ectosomes are immunomodulators

have ingested them (15–17). Little is known about the direct interference or enhancing properties of ectosomes on other cells involved in inflammation and immunity. In particular, whether ectosomes may impact activation/differentiation of T cells, including effector and regulatory cells, has not been investigated so far. In this work, we focused on the effects of PLT-Ecto on peripheral blood CD4+ T cells. This encounter produced a shift of differentiation of naïve and, to a lower extent, memory CD4+ T cells towards functional Tregs, an effect partially dependent on PLT-Ecto derived transforming growth factor (TGF)-β1. Additionally, it has been shown that membrane-bound TGF-β have much greater efficacy than the soluble form probably due to its sustained signalling on responding cells (18). Interestingly, we observed that ectosomederived TGF-β1 levels correlated with frequencies of Tregs in peripheral blood of healthy donors. These results suggest a role for PLT-Ecto derived TGF-β1 in the modulation of T cells homeostasis in physiological conditions and, possibly, in transfusion related reactions.

Materials and methods Antibodies and reagents The following mouse anti-human monoclonal Abs were used: FITC- and/or APC-conjugated anti-CD4 and anti-CD8, PE-labelled anti-CD45RA and, PECy7-coupled anti-CD25 and pacific blue-coupled anti-CD62L Abs (all purchased from BD Biosciences, San Jose, CA, USA). PS expression was assessed by binding of FITC-labelled Annexin V (BD Biosciences). Intracellular staining for TGF-β1 and Foxp3 were respectively performed with a PE-labelled anti-human TGF-β1 (BD Biosciences) and Foxp3 (eBioscience, San Diego, CA, USA), according to the manufacturer’s instructions. T cell activation/proliferation was performed in the presence of anti-human CD3 (clone OKT3, eBioscience) and anti-human CD28 (clone CD28.2, BD Biosciences). rIL-2 was purchased from Roche. Multiplex measurements of cytokines in high-density formats using Multi-Array Technology (Meso Scale Discovery, Meso Scale Diagnostics) were used for simultaneous determination of interferon (IFN)γ, tumour necrosis factor (TNF)α, interleukin (IL)-10 and IL-2. IL-17 was measured by ELISA (eBioscience). TGF-β1 was measured using TGF-β1 Duoset ELISA (R&D Systems, Minneapolis, MN, USA). Total TGF-β1, comprising both the active and the latent form, was measured according to the manufacturer’s protocol, which includes acid activation of samples followed by their neutralisation.

PLT-Ecto preparation PLT concentrates were obtained by apheresis as previously described (9); briefly, donor blood was processed with a cell separator with an in-line centrifuge for PLT collection (Blood Transfusion Center Basel). The PLT were then transferred to a collection bag, whereas the other blood elements were returned to the donor. Leukocyte contamination per unit was less than 0.2 × 106. PLT

concentrates were stored at room temperature (RT) in motion for up to five days in a 30:70 plasma:T-Sol mixture. T-Sol contains sodium chloride, trisodium citrate and sodium acetate. The purification of PLT-Ecto from the PLT concentrates was done after five days of storage. PLT concentrates were subjected to several RT rounds of centrifugations to clear all residual cells. To remove the high-density cells (erythrocytes and leukocytes), the centrifugation was performed for 15 minutes (min) (300 g). Any residual erythrocytes and leukocytes were removed by recentrifugation for 15 min (500 g). The PLT were pelleted by centrifugation of the suspension for 20 min (800 g). Any residual PLT and low-density debris were removed by centrifugation for 20 min (3,000 g), and the supernatants were stored in aliquots at –80 °C until use. PLT-Ecto containing supernatants as well as citrated plasma of healthy donors were ultracentrifuged for 1 hour (h) at 200,000 g at RT. The pellets were washed through ultracentrifugation at the same speed and for the same time in 0.9 % NaCl. The morphological characteristics and the purity of the PLT-Ecto have been previously described (9) (Suppl. Figure 1, available online at thrombosis-online.com). For the analysis of plasma-derived ectosomes, citrated blood from healthy donors were centrifuged at 3,000 g for 20 min, then at 13,000 g for 2 min at RT, in order to obtain PLT-free plasma. PLT-free plasma was then ultracentrifuged at 200,000 g (the speed used to pellet Ecto from PLT concencentrates) for 1 h at 4 °C. The supernatant was kept aside to determine the concentration of soluble TGF-β1. The pellet containing Plasma-Ecto was washed by the same ultracentrifugation conditions in NaCl 0.9 % in order to remove an eventual contamination with soluble plasma TGF-β1.

PMN-Ecto preparation PMN were isolated from fresh buffy coats of normal donors, according to the technique described previously (19). Briefly, a fresh buffy coat was diluted 1/1 (vol/vol) with 2 mM PBS-EDTA, mixed gently with 0.25 vol 4 % Dextran T500, and left for 30 min for ERY sedimentation. The leukocyte-rich supernatant was aspirated and centrifuged for 10 min at 200 g. The pellet was resuspended for 1 min in 9 ml ultrapure water to lyse ERY. Isotonicity was restored by addition of 3 ml 0.6 M KCl and 40 ml 0.15 M NaCl. Cells were then centrifuged 10 min at 350 g and resuspended in 20 ml 2 mM PBS-EDTA. This suspension was layered over 20 ml Ficoll gradient, and centrifuged for 30 min at 350 g. The PMN-rich pellet was recovered and washed twice in 2 mM PBS-EDTA. All manipulations were performed at 4 °C. PMN-Ecto, which have inhibitory properties on macrophages and dendritic cells (12, 13), were prepared as previously described (20, 21). For stimulation, PMN (107 cells/ml) were diluted 1/1 (vol/vol) in prewarmed (37 °C) RPMI 1640 (Life Technologies, Basel, Switzerland) with 1 µM fMLP (Sigma, St. Louis, MO, USA) and incubated for 20 min at 37 °C. PMN were removed by centrifugation (4,000 g at 4 °C), and the supernatants were concentrated with Centriprep centrifugal filter devices (10,000 MW cutoff, Millipore Corp., Bedford, MA, USA) and stored in aliquots at -80 °C until use.

Thrombosis and Haemostasis 112.6/2014 Downloaded from www.thrombosis-online.com on 2015-01-07 | ID: 1000467556 | IP: 147.140.20.32 For personal or educational use only. No other uses without permission. All rights reserved.

© Schattauer 2014


ERY-Ecto preparation Blood units were obtained from healthy volunteer donors and submitted to standard procedure preparation and storage of packed erythrocytes. Briefly, whole blood (450 ml) was collected in plastic bags (triplicate bag system with an integrated whole blood filter Leucoflex Sang Total 1, Macopharma, Tourcoing, France) with 63 ml citrate phosphate dextrose in the primary bag. The bags were stored at RT, and filtration took place 3 h following donation. After centrifugation (10 min, 1,500 g, 20 °C), packed leuko-depleted erythrocytes (LD-E) were separated from plasma and transferred into the satellite bags containing 100 ml saline-adenine-glucose-mannitol. Storage time for packed LD-E before tests was 25 days. The supernatant of packed LD-E were obtained through two rounds of centrifugation (10 min, 1000 g, 4 °C) to clear all residual erythrocytes. The supernatants containing ERY-Ecto were concentrated with Centriprep centrifugal filter devices (10,000 MW cut-off, Millipore) and stored in aliquots at -80 °C until use (14).

A

Isolation of CD4+ and CD8+ T cell subsets PBMC were separated on a Ficoll gradient from fresh buffy coats of normal donors and washed in PBS. CD4+ or CD8+ T cells were enriched from PBMC of healthy donors using a CD4+ or CD8+ isolation kit (Miltenyi Biotec, Bergisch Gladbach, Germany) according to manufacturer’s procedure. The purity of CD4+ or CD8+ T cells was > 97 %, as confirmed by flow cytometry. In separate experiments, naïve T cells (CD4+CD45RA+CD62L+), memory T cells (CD4+CD45RA-CD62L±) and Tregs (CD4+CD25high) were isolated from enriched CD4+ T cells by FACS sorting (purity > 97 %).

Membrane labelling PLT-Ecto and CD4+ T cells were incubated with DyLight AmineReactive Dye 633 and 488, respectively, for 30 min at RT. Labelled PLT-Ecto were separated from the remaining unbound dye by ultracentrifugation (45 min 200,000 g at RT) and washed with 0.9 % NaCl. Labelled CD4+ T cells were washed by centrifugation (300 g) twice with NaCl before use.

Live-cell microscopy Labelled CD4+ T cells were incubated on 4-chambered #1.0 Borosilicate Coverglass System (Lab-Tek, Nunc, Thermo Fischer Scientific, Waltham, MA, USA), with medium alone or fluorescently labelled PLT-Ecto, and time-lapse microscopy was performed.

Confocal microscopy Confocal images were acquired on an Axiovert confocal laserscanning microscope (LSM 710) from Zeiss AG (Feldbach, Switzerland) using a 40× oil-immersed objective (Zeiss). For a given experiment, all the settings on the microscope were kept constant for all samples, including exposures, pinhole size (2 µm) and photomultiplier tube gain. Images were exported as JPEG files. © Schattauer 2014

B Figure 1: Interactions between PLT-Ecto and T cells. PLT-Ecto proteins were labelled with DyLight Amine-Reactive Dye 633 (red), and CD4+ T cells with DyLight Amine-Reactive Dye 488 (green). They were then co-incubated, and followed by time-lapse confocal microscopy for 20 h. A) Representative images of CD4+ T cells alone (no ecto) and CD4+ T cells (0.5 ×106 cells/ml) + PLT-Ecto (15 µg/ml) were taken at different time points (10 min, 30 min, 3 h, 6 h, 20 h). B) The contact between the same CD4+ T cell and PLT-Ecto was followed up to 20 h, and four different time points are shown.

CD4+ T cell cultures and intracellular staining Total CD4+ T cells, unlabelled or stained with CFSE (as previously described [22]) were cultured in 96-well plates pre-coated with anti-CD3/anti-CD28 Abs (1 µg/ml), at cell density of 5 × 104 per well, in RPMI 1640 medium containing 5 % of human AB serum and 150 IU/ml IL-2, in the presence or absence of PLT-Ecto (1, 5 or 15 µg/ml). For some experiments, intracellular staining for TGF-β1 was performed on day 5 after initiation of culture. CD4+ T cells were stimulated with PMA/Ionomycin (Sigma) in the presence of Brefeldin A (Sigma) for 5 h. After stimulation cells were Thrombosis and Haemostasis 112.6/2014


1221

1222


surface stained with anti-human CD4 Ab, then fixed and permeabilised and stained with anti-human TGF-β1. After three, five or eight days, frequencies of CD25high Foxp3+ Tregs and percentages of proliferating cells, as indicated by CFSE dilution, were assessed by flow cytometry. Viability of T cells cultured in the presence of PLT-Ecto (15 µg/ml) was monitored by propidium iodide (PI) staining. No significant difference (p=0.7) in frequencies of PI+ cells between CD4+ T cell cultures performed in the absence or presence of PLT-Ecto were detected. Consistently, comparable absolute numbers of T cells were monitored in initial experiments, to exclude possible PLT-Ecto-related toxicity on T cells. No significant difference (p=0.1) in numbers between T cell cultures performed in the absence or presence of PLT-Ecto were observed. When stated, culture supernatants containing PLT-Ecto were removed after 3 or 6 h culture, and replaced with fresh medium. The presence of PLT-Ecto in T cell cultures, before and after washing, was evaluated by flow cytometry, upon staining with specific Abs for CD4 and CD41a markers. After the washing step their frequency was reduced by 95 % (Suppl. Figure 2, available online at www.thrombosis-online.com). CD4+ T cells were then cultured for additional three days before analysis. When specified, total CD4+ T cells were cultured in the presence of ERY-Ecto or PMN-Ecto (1, 5 or 15 µg/ml). In separate experiments, sorted naïve T cells, memory T cells and Tregs were stained or not with CFSE, then cultured as described above in 96-well plates in absence or presence of the same concentrations of PLT-Ecto. In addition, in order to analyse the role of TGF-β1, PLT-Ecto were pre-treated with anti-TGF-β1 (10 µg/ml) Abs or isotype controls for 2 h. The amount of ectosomes used in our experiments was quantified by using a Bradford protein assay. Culture supernatants were collected and spin for 10 min at 1,000 g at 4 °C to remove cell debris, and cytokine contents were assessed.

CD8+ T cell suppression assay Tregs generated from naïve T cells exposed to PLT-Ecto for five days, were sorted by FACS based on CD25high expression, and cocultured with autologous CFSE-labelled CD8+ T cells in antiCD3/anti-CD28 (1 µg/ml) -coated 96-well plates. After 72 h, the proliferation of CD8+ T cells was analysed by FACS.

Statistical analysis Statistical significances were evaluated by Mann Whitney, Friedman, ANOVA tests and Spearman Correlation Coefficient (Rs) as appropriate, using Prism software (GraphPad Software, La Jolla, CA, USA).

Results PLT-Ecto interaction with CD4+ T cells In order to analyse the interactions between PLT-Ecto and CD4+ T cells, we labelled their respective surface proteins with two differ-

ent amine-reactive dyes, and coincubated them for 20 h. We performed time-lapse confocal microscopy. After 30 min of culture, short-time contacts (lasting from seconds to few minutes) between PLT-Ecto and CD4+ T cells could be detected (▶ Figure 1 A). Interestingly, starting from 3 h, longer contact periods for up to 17 h could be observed (▶ Figure 1 B).

PLT-Ecto modulate CD4+ T cell cytokine release To investigate the potential impact of these interactions on T cell function, we co-incubated CD4+ T cells, activated by means of anti-CD3/anti-CD28 Abs, with concentrations of PLT-Ecto similar to those of microparticles circulating in peripheral blood (24), i. e. 1, 5 and 15 µg/ml. After three, five or eight days, the release of specific T cell cytokines, including IFNγ, TNFα, IL-6, IL-17 and TGF-β1 in the supernatant was analysed. The presence of PLTEcto inhibited in a time- and dose-dependent manner the release of IFNγ already after three days (p=0.05). Although less pronounced and delayed as compared to IFNγ, TNFα and IL-6 release decreased after five days of exposure (p=0.05). The inhibition of IL-6 release was almost maximal with the lowest dose of PLT-Ecto, 1 µg/ml (▶ Figure 2 A). We observed no significant modifications of IL-17 release (Suppl. Figure 3, available online at www.thrombo sis-online.com). In contrast, they enhanced the secretion of TGF-β1 (▶ Figure 2 B, left panel). This was evident after exposing the T cells to 15 µg/ml PLT-Ecto, as shown by the time dependent increase of TGF-β1 above any possible contamination with TGF-β1 from the PLT-Ecto. Furthermore, TGF-β1 intracellular staining of CD4+ T cells exposed to different PLT-Ecto concentrations revealed a dose dependent increase of TGF-β1 production clearly induced by PLT-Ecto (▶ Figure 2 B, right panel). Taken together, the T cell secretome profile suggested that PLT-Ecto may shift the T cell effector function from a proinflammatory response to an immunosuppressive one. The next step was to define, whether PLT-Ecto have also an impact on CD4+ T cell proliferation and/or differentiation. Purified CD4+ T cells were activated by anti- CD3/anti-CD28 Abs in the absence or presence of PLT-Ecto, and their proliferation extent was evaluated by CFSE-dilution assay. In the presence of PLT-Ecto a modest decrease of T cell proliferation was observed (from 62 ± 6 % to 51 ± 8 % of proliferating cells; ▶ Figure 2 C, upper panel). Intriguingly, however, we observed an increase in the frequency of Foxp3+ cells within the compartment of proliferating cells (▶ Figure 2 C, lower panel). This phenomenon was time and dose dependent, and it was mostly evident upon exposure of T cells to 15 µg/ml PLT-Ecto up to eight days (▶ Figure 2 D). However, we observed that a period of exposure to PLT-Ecto as short as 3 h, was sufficient to induce a significant increase in CD25high Foxp3+ T cells population (p ≤ 0.001) comparable to that observed in CD4+ T cell treated with PLT-Ecto continuously for three days (▶ Figure 2 E). Thus PLT-Ecto had the capacity to act in a concentration and a time frame corresponding to physiological conditions. To investigate whether the observed effects were specific for PLT-Ecto, we performed the same experiments in the presence of PMN-Ecto or ERY-Ecto, known to downmodulate macrophages and dendritic


© Schattauer 2014


A

B Figure 2: Effects of PLT-Ecto on CD4+ T cells. A) PLT-Ecto modulate CD4+ T cells cytokine release. CD4+ T cells were incubated for three, five or eight days in medium alone (Ø) or medium with (1, 5, 15 µg/ml) PLT-Ecto. Supernatants were analysed for IFNγ, TNFα, IL-6. Assays were performed in triplicate, means ± SEM are shown. Results are representative of three independent experiments. Statistical analysis was done using Mann Whitney test. *, p ≤ 0.05. B) PLT-Ecto modulate CD4+ T cells TGF-β1 release. TGF-β1 released in the supernatants from: i) CD4+ T cells cultured alone; ii) CD4+ T cells cultured with PLT-Ecto (15 µg/ml); iii) PLT-Ecto (15 µg/ml) cultured alone, for three, five or eight days, was analysed by ELISA (left panel). Intracellular staining of TGF-β1 in CD4+ T cells incubated for five days with 1, 5, and 15 µg/ml PLTEcto, and stimulated PMA/Ionomycin in the presence of Brefeldin A for 5 h (right panel). One representative out of three independent experiments is shown. Assays were performed in triplicates, means ± SD are shown. Statistical significance was assessed by one way ANOVA test. C) PLT-Ecto modify CD4+ T cell proliferation and differentiation. CFSE labelled CD4+ T cells were cultured alone or with PLT-Ecto (15 µg/ml) for three days before staining for

© Schattauer 2014

Foxp3, and analysed by flow cytometry. A histogram (upper panel) represents the proliferation of CD4+ T cells and a dot plot (lower panel) represents the proliferation of Foxp3+ cells within the compartment of proliferating cells. One representative out of three independent experiments is shown. D) Dose and time dependency effect of PLT-Ecto. Dot plot (upper panel) shows the double staining CD25/Foxp3 of CD4+ T cells in medium alone (Ø) or with (1, 5, 15 µg/ml) PLT-Ecto. Histograms (lower panel) show the % of CD25high Foxp3+ cells within CD4+ T cells after three and eight days exposure to PLTEcto. Significance refers to the pool of the total corresponding experiments at 15 µg/ml of PLT-Ecto as assessed by Friedman test. E) Effect of PLT-Ecto start after 3 h coincubation. Histograms show the % of CD25high Foxp3+ cells within CD4+ T cells incubate for 3 h or three days in presence of PLT-Ecto (15 µg/ml). Culture supernatants containing PLT-Ecto were removed or not after 3 h, and replaced with fresh medium. CD4+ T cells were then cultured for additional three days before analysis. Assays were performed in triplicate, means ± SD are shown. Statistical significance was assessed by one way ANOVA test.


1223

1224


C

D

E

Figure 2: continued


© Schattauer 2014


Figure 3: Specific effects of PLT-Ecto on CD4+ T cells. Representative dot plots (upper panel) of the CD25high Foxp3+ CD4+ T cells cultured in medium alone (Ø), or with PLT-Ecto, ERYEcto, PMN-Ecto (15 µg/ml). Data from one out of three experiments are shown. Histogram (lower panel) representing the % of CD25high Foxp3+ CD4+ T cells at different doses of PLT-Ecto, ERY-Ecto, and PMN-Ecto. Data pooled from three independent experiments are displayed. The significance of the result was assessed by Friedman test.

cells, similarly to PLT-Ecto. Neither PMN-Ecto nor ERY-Ecto had any effect on CD25high Foxp3+ T cells frequency (▶ Figure 3). In addition, no modification of T cell cytokine release was observed with any of these two ectosome types (Suppl. Figure 4, available online at www.thrombosis-online.com).

Identification of the CD4+ T cell population modified by PLT-Ecto In order to determine whether the increased frequency of CD25high Foxp3+ T cells resulted from a preferential expansion of pre-differentiated Tregs in the presence of PLT-Ecto, sorted Tregs (as CD25high), and, as controls, naïve (CD45RA+CD62L+) and memory (CD45RA-CD62L±) CD4+ T cells, were stained with CSFE and exposed to different PLT-Ecto concentrations. Unexpectedly, PLT-Ecto had no effect on Treg proliferation (▶ Figure 4 A). In contrast, on naïve T cells, PLT-Ecto induced a dose dependent increase in the frequency of proliferating Foxp3+ cells (from 3.4 % ± 1.4 % to 27.8 % ± 12.7 %; p< 0.0001; ▶ Figure 4 B). Furthermore increased release of IL-10 and decreased IL-2 concentrations were detected in culture supernatants (Suppl. Figure 5, available online at www.thrombosis-online.com). The addition of PLT-Ecto to memory T cells also enhanced, although to low extent, the frequency of proliferating Foxp3+ T cells (from 2.9 % ± 1.9 to 6.6 %± 3 %; p< 0.0001; ▶ Figure 4 C). Thus, PLT-Ecto-induced increase in Foxp3+ T cells is due to differentiation of naïve T cells and, to low extent, to conversion of memory T cells.

© Schattauer 2014

Suppressive activity of PLT-Ecto generated Foxp3+ cells To evaluate whether PLT-Ecto generated Foxp3+ cells were functional Tregs, we tested their inhibitory activity in a suppression assay. CD25high cells, sorted from cultures of PLT-Ecto-exposed naïve T cells, were incubated with CFSE-labelled CD8+ T cells, as responders, at a 1:1 ratio. Strikingly, PLT-Ecto-induced Foxp3+ cells suppressed CD8+ T cell proliferation in a comparable manner to that induced by Tregs isolated from peripheral blood (▶ Figure 5 A, B).

PLT-Ecto-derived TGF-β1 plays a role in the induction of Foxp3+ regulatory T cells Since TGF-β1 is an essential factor for the differentiation of CD4+ T cells into T cells endowed with regulatory ability, we analysed whether TGF-β1 expressed by PLT-Ecto was involved in the observed expansion of Foxp3+ T cells. We co-cultured naïve CD4+ T cells with different concentration of PLT-Ecto preincubated with TGF-β1 neutralising Abs. In three individual experiments, preincubation of PLT-Ecto with neutralising TGF-β1 Abs significantly reduced the capacity of PLT-Ecto to expand Tregs from naïve T cells as compared to their isotype control Abs (▶ Figure 6 A). Thus, PLT-Ecto derived TGF-β1 appeared to be a main mediator in the induction of Foxp3+ regulatory T cells. Since PLT-Ecto are the most abundant extracellular vesicles in the peripheral blood, we measured the concentration of soluble or ectosome-derived TGF-β1 and the Foxp3+ T cells frequencies in Thrombosis and Haemostasis 112.6/2014


1225

1226


blood samples of healthy donors. Frequencies of Foxp3+ T cells positively correlated only with ectosome-derived TGF-β1 (Rs=0.5086; p< 0.005; ▶ Figure 6 B).

Discussion

A

B

C Figure 4: CD4+ T cell populations modified by PLT-Ecto. A) Regulatory T cells (CD4+CD25highFoxp3+) n=3, B) naïve T cells (CD4+CD45RA+CD62L+) n=5, C) memory T cell (CD4+CD45RA-CD62L±) n=5, were stained with CFSE, and cultured alone or with different concentration of PLT-Ecto (1, 5, 15 µg/ml) for five days. Frequencies of Foxp3+ proliferating T cells, based on CFSE dilution, were analysed by FACS. In all graphs, each line corresponds to an independent experiment, i. e. n=3 in A), n=5 in B), and n=5 in C). Significance refers to cumulative data from all experiments and was assessed by Friedman test.

Many clinical studies suggest that transfusions might be immunosuppressive (1, 2). Different transfusion components might be involved in modulating immunity. Beside cells and soluble factors, blood products contain extracellular vesicles released by different cell types (25), capable of dampening inflammatory responses (15–17). In particular, ectosomes shed from the cell surface of PMN, ERY and PLT have been shown to inhibit cytokine release by activated macrophages and to block dendritic cell maturation (9, 12–14). The biological effects of ectosomes on T cells have not been studied until now. Here we found that the interactions between PLT-Ecto and CD4+ T cells occured after 30 minutes. The contact periods increased with time. Some of the large sized PLT-Ecto may represent aggregates. Indeed, Vasina et al have shown that PLT-Ecto released during storage period (ageing) express activated αIIbβ3 integrins and tended to assemble into aggregates (23), thus explaining our observation. PLT-Ecto, but not PMN- and ERY-Ecto, diminish the release of inflammatory cytokines by activated CD4+ T cells and promote their differentiation into functional Tregs, in a TGF-β1-dependent manner. Upon three to eight days contact with PLT-Ecto, T cells exhibited an almost complete suppression of IFNγ production and a lower release of TNFα and IL-6. In contrast, the release of TGF-β1 by T cells increased over time, suggesting that upon exposure to PLT-Ecto a skew from a proinflammatory to a suppressive T cell response had occurred. Most interestingly, modifications of T cell cytokine response induced by PLT-Ecto were accompanied by an expansion of Foxp3+ T cells, capable of suppressing T cell proliferation as effectively as peripheral blood Tregs. Notably, this phenomenon did not result from an increased proliferation of pre-differentiated Tregs, but it was rather due to the de novo differentiation of naïve T cells and, to a smaller extent, to the conversion of memory T cells into Foxp3+ cells. Thus, PLT-Ecto act on T cells by modifying their differentiation pathway. PLT-Ecto-associated TGF-β1 appeared to be a critical mediator of the phenomena observed. Indeed, ectosomes not expressing TGF-β1, such as those derived from PMN and ERY did not induce any increase in Treg numbers. Strikingly, the addition of TGF-β1 neutralising Abs reverted significantly and in a dose dependent manner the differentiation of naïve T cells into Tregs. The ability of TGF-β1 to promote differentiation of CD4+ T cells into a regulatory phenotype and to dampen T cell-mediated effector responses is well documented by previous work (26–28). Furthermore, the apparent lack of effect on pre-differentiated Tregs is also consistent with previous studies reporting no enhancement by TGF-β1 on Foxp3 expression on pre-existing Tregs


© Schattauer 2014


A

Figure 5: Suppressive activity of the PLT-Ecto generated CD25high CD4+ T cells. Sorted naïve CD4+ T cells were activated in the absence (control CD4+ T cells) or presence of PLT-Ecto (15 µg/ml) for five days. CD25high cells from cultures in the presence of PLT-Ecto were then sorted by FACS (PLT-Ecto CD25high T cells). Autologous CFSElabelled CD8+ T cells were activated by anti-CD3/anti-CD28 Abs and cultured (i) alone, with (ii) control CD4+ T cells, with (iii) PLT-Ecto CD25high T cells, or with (iv) autologous Tregs isolated from peripheral blood. After three days, the proliferation of CD8+ T cells was analysed by FACS. A) One Representative FACS analysis. B) Percentage of nonproliferating cells are shown as cumulative data from three independent experiments (means ± SD).

B

(29, 30). TGF-β1 is also been reported to inhibit cytokine release from effector T cells. Whether in our model the observed reduction of cytokine response is due to its direct effect on T cells or is consequent to Tregs induction remains to be elucidated. Notably, we detected an increase in IL-10 and a decrease in IL-2 concentrations in supernatants from cultures of naïve T cells differentiated into Tregs. Thus, PLT-Ecto-induced Tregs may also contribute to dampen effector T cell responses through the production of suppressive cytokines, such as TGF-β1 and IL-10, and/or through the consumption of IL-2, as described for peripheral blood Tregs (31). Beside TGF-β1, other molecules present in PLT-Ecto may contribute to modulate T cell responses. For instance, proteomic studies have shown that microvesicles released by PLT expressed PLT Factor 4, a member of chemokine CXC family, that strongly inhibits T cell proliferation and IFNγ release (32). An additional molecule possibly involved in the observed effect could be (15-d-PGJ(2)), a key metabolite of Prostaglandin D(2) which is abundantly produced by PLT. This metabolite is a known agonist of the peroxisome proliferator activated receptor-gamma. Peroxisome proliferator activated receptor-gamma agonists together with TGF-β have been described to induce potent and stable © Schattauer 2014

Foxp3 expression, resulting in the generation of functional induced Tregs (33, 34). In contrast, PS, although highly expressed by PLT-Ecto, is unlikely to be involved in the phenomena observed, since PMN- and ERY-Ecto, also expressing PS at high levels, had no effect on activated T cells (12, 14). Notably PS expression on vesicles/ectosomes, has been shown to be responsible for suppression of macrophages and dendritic cells (12–14, 16). Thus, PLT-Ecto might mediate a spectrum of immunosuppressive effects by targeting, through different mechanisms, responding effector T cells as well as antigen presenting cells. It is attractive to suggest that the transfusion of PLT preparations containing ectosomes, in addition to prevent bleeding may exert several immunomodulatory activities, such as those reported here, since their half-life is around 6 h (10) a time period sufficient to interact with various cells, before being cleared. Indeed, a 3 h exposure of CD4+ T cells to PLT-Ecto was sufficient to promote differentiation of proliferating cells into Foxp3+ Tregs. Microvesicles originating from PLT are also found in the circulation of normal individuals (35, 36). It is intriguing to hypothesise that these vesicles may also be endowed with immunomodulatory properties. Indeed, we observed a positive correlation between the Thrombosis and Haemostasis 112.6/2014


1227

1228


A

B Figure 6: PLT-Ecto TGF-β1 mediates functional Foxp3+ regulatory T cells. A) PLT-Ecto associated TGF-β1 plays a role in the induction of Foxp3+ regulatory T cells. CFSE-labelled naïve CD4+ T cells exposed to the indicated concentrations of PLT-Ecto pretreated with neutralising anti-TGF-β1 Abs or with an isotype control Ab. After four days, % of Foxp3+ proliferating cells were analysed by FACS. One representative out of three independent experi-

ments is shown. Statistical significance was assessed by Mann Whitney test. B) Ectosome derived TGF-β correlates with Foxp3+ T cells frequencies in peripheral blood of healthy donors. Correlation analysis between PLT-Ectoderived (left) or soluble (right) TGF-β1 and Foxp3+ T cell frequencies in peripheral blood of healthy donors (n=30). Spearman Rank Correlation Coefficient (Rs) is indicated.

concentration of microvesicle-associated TGF-β1 and Treg frequencies in peripheral blood of healthy donors. Moreover, PLTEcto have been shown to accumulate at inflammatory sites such as

joints of patients with rheumatoid arthritis (37, 38). Within inflamed tissues, they may act on effector memory T cells and favour acquisition of immunosupressive phenotypes. Activated PLT have previously been reported to increase the frequency of Foxp3+ T cells upon coculture with total CD4+ T cells (39, 40). However, they observed in addition to the inhibition of T cell proliferation, a concomitant increase of IFNγ, TNFα and IL-2 secretion contrasting with our results. Thus, the effects mediated by entire PLT on CD4+ T cells appear to differ from those of PLT-Ecto. In addition, it has been reported that PLT “microparticles” (ectosomes) were more efficient compared to the entire PLT for coagulation. Finally, PLT-Ecto but not entire PLT may enter the lymphatic vessels thus targeting distant lymphoid tissues. In fact, considering the smallest PLT-Ecto population which size is around 100 nm and charge is negative (PS expression), it is tempting to speculate that PLT-Ecto might enter the lymphatic vessels, traffic to lymph nodes, as reported for mast cells derived microvesicles displaying similar size and charge (41). Once there, they could promote the differentiation of naïve T cells into Tregs.

What is known about this topic? • Platelet transfusions are associated with immune reactions. Platelet storage lesions represent a major cause to these reactions. Platelet-derived extracellular vesicles/ectosomes are a part of the storage lesions. What does this paper add? • Platelet-, but not PMN- and Erythrocyte-Ectosomes, have the capacity to shift the T cell effector function from a proinflammatory to an immunosuppressive response in vitro. • Platelet-Ectosomes promote differentiation of CD4+ T cells into functional T regulatory cells in a TGF-b1-dependent manner in vitro. • Platelet-Ectosomes could therefore participate in the platelet transfusion related immune reactions. Thrombosis and Haemostasis 112.6/2014


© Schattauer 2014


Thus, circulating PLT-derived microvesicles may participate in the maintenance of Treg homeostasis, influencing therefore the level of immune tolerance. Acknowledgements

We thank the Blutspendezentrum Basel for their blood support.

18. 19. 20. 21.

Authors’ contributions

Salima Sadallah designed the project, performed the experiments, analysed the results, and wrote the manuscript. Francesca Amicarella designed the project, performed the experiments, analysed the results, and wrote the manuscript. Ceylan Eken performed the experiments, wrote the manuscript. Giandomenica Iezzi supervised the project, wrote the manuscript. Jürg. A. Schifferli supervised the project, wrote the manuscript.

22.

23. 24. 25.

Conflicts of interest

None declared.

26.

References

27.

1. Stroncek DF, Rebulla P. Platelet transfusions. Lancet 2007; 370: 427–438. 2. Vamvakas EC. Platelet transfusion and adverse outcomes. Lancet 2004; 364: 1736–1738. 3. Shrivastava M. The platelet storage lesion. Transfus Apher Sci 2009; 41: 105–113. 4. Thon JN, Schubert P, Devine DV. Platelet storage lesion: a new understanding from a proteomic perspective. Transfus Med Rev 2008; 22: 268–279. 5. von Hundelshausen P, Weber C. Platelets as immune cells: bridging inflammation and cardiovascular disease. Circ Res 2007; 100: 27–40. 6. Sinauridze EI, Kireev DA, Popenko NY, et al. Platelet microparticle membranes have 50– to 100-fold higher specific procoagulant activity than activated platelets. Thromb Haemost 2007; 97: 425–434. 7. Morel O, Morel N, Jesel L, et al. Microparticles: a critical component in the nexus between inflammation, immunity, and thrombosis. Semin Immunopathol 2011; 33: 469–486. 8. Heijnen HF, Schiel AE, Fijnheer R, et al. Activated platelets release two types of membrane vesicles: microvesicles by surface shedding and exosomes derived from exocytosis of multivesicular bodies and alpha-granules. Blood 1999; 94: 3791–3799. 9. Sadallah S, Eken C, Martin PJ, et al. Microparticles (ectosomes) shed by stored human platelets downregulate macrophages and modify the development of dendritic cells. J Immunol 2011; 186: 6543–6552. 10. Rank A, Nieuwland R, Crispin A, et al. Clearance of platelet microparticles in vivo. Platelets 2011; 22: 111–116. 11. van der Pol E, Boing AN, Harrison P, et al. Classification, functions, and clinical relevance of extracellular vesicles. Pharmacol Rev 2012; 64: 676–705. 12. Gasser O, Schifferli JA. Activated polymorphonuclear neutrophils disseminate anti-inflammatory microparticles by ectocytosis. Blood 2004; 104: 2543–2548. 13. Eken C, Gasser O, Zenhaeusern G, et al. Polymorphonuclear neutrophil-derived ectosomes interfere with the maturation of monocyte-derived dendritic cells. J Immunol 2008; 180: 817–824. 14. Sadallah S, Eken C, Schifferli JA. Erythrocyte-derived ectosomes have immunosuppressive properties. J Leukoc Biol 2008; 84: 1316–1325. 15. Chen X, Doffek K, Sugg SL, et al. Phosphatidylserine regulates the maturation of human dendritic cells. J Immunol 2004; 173: 2985–2994. 16. Wu Y, Tibrewal N, Birge RB. Phosphatidylserine recognition by phagocytes: a view to a kill. Trends Cell Biol 2006; 16: 189–197. 17. Shi D, Fu M, Fan P, et al. Artificial phosphatidylserine liposome mimics apoptotic cells in inhibiting maturation and immunostimulatory function of murine

© Schattauer 2014

28. 29. 30.

31. 32. 33.

34. 35. 36. 37. 38. 39. 40. 41.

myeloid dendritic cells in response to 1-chloro-2,4-dinitrobenze in vitro. Arch Dermatol Res 2007; 299: 327–336. Pap E, Pallinger E, Falus A. The role of membrane vesicles in tumorigenesis. Crit Rev Oncol Hematol 2011; 79: 213–223. de Almeida MC, Silva AC, Barral A, et al. A simple method for human peripheral blood monocyte isolation. Mem Inst Oswaldo Cruz 2000; 95(2): 221–3. 20. Hess C, Sadallah S, Hefti A, et al. Ectosomes released by human neutrophils are specialized functional units. J Immunol 1999; 163: 4564–4573. Gasser O, Hess C, Miot S, et al. Characterisation and properties of ectosomes released by human polymorphonuclear neutrophils. Exp Cell Res 2003; 285: 243–257. Rosenthal R, Groeper C, Bracci L, et al. Differential responsiveness to IL-2, IL-7, and IL-15 common receptor gamma chain cytokines by antigen-specific peripheral blood naive or memory cytotoxic CD8+ T cells from healthy donors and melanoma patients. J Immunother 2009; 32: 252–261. Vasina EM, Cauwenberghs S, Staudt M, et al. Aging- and activation-induced platelet microparticles suppress apoptosis in monocytic cells and differentially signal to proinflammatory mediator release. Am J Blood Res 2013; 3: 107–123. Hunter MP, Ismail N, Zhang X, et al. Detection of microRNA expression in human peripheral blood microvesicles. PLoS One 2008; 3: e3694. Bode AP, Orton SM, Frye MJ, et al. Vesiculation of platelets during in vitro aging. Blood 1991; 77: 887–895. Chen W, Jin W, Hardegen N, et al. Conversion of peripheral CD4+CD25– naive T cells to CD4+CD25+ regulatory T cells by TGF-beta induction of transcription factor Foxp3. J Exp Med 2003; 198: 1875–1886. Fantini MC, Becker C, Monteleone G, et al. Cutting edge: TGF-beta induces a regulatory phenotype in CD4+CD25– T cells through Foxp3 induction and down-regulation of Smad7. J Immunol 2004; 172: 5149–5153. Dons EM, Raimondi G, Cooper DK, et al. Induced regulatory T cells: mechanisms of conversion and suppressive potential. Hum Immunol 2012; 73: 328–334. Hori S, Nomura T, Sakaguchi S. Control of regulatory T cell development by the transcription factor Foxp3. Science 2003; 299: 1057–1061. Tran DQ, Ramsey H, Shevach EM. Induction of FOXP3 expression in naive human CD4+FOXP3 T cells by T-cell receptor stimulation is transforming growth factor-beta dependent but does not confer a regulatory phenotype. Blood 2007; 110: 2983–2990. Vercoulen Y, Wehrens EJ, van Teijlingen NH, et al. Human regulatory T cell suppressive function is independent of apoptosis induction in activated effector T cells. PLoS One 2009; 4: e7183. Fleischer J, Grage-Griebenow E, Kasper B, et al. Platelet factor 4 inhibits proliferation and cytokine release of activated human T cells. J Immunol 2002; 169: 770–777. Lei J, Hasegawa H, Matsumoto T, et al. Peroxisome proliferator-activated receptor alpha and gamma agonists together with TGF-beta convert human CD4+CD25– T cells into functional Foxp3+ regulatory T cells. J Immunol 2010; 185: 7186–7198. Barry OP, Pratico D, Lawson JA, et al. Transcellular activation of platelets and endothelial cells by bioactive lipids in platelet microparticles. J Clin Invest 1997; 99: 2118–2127. Italiano JE, Jr., Mairuhu AT, Flaumenhaft R. Clinical relevance of microparticles from platelets and megakaryocytes. Curr Opin Hematol 2010; 17: 578–584. Flaumenhaft R, Mairuhu AT, Italiano JE. Platelet- and megakaryocyte-derived microparticles. Semin Thromb Hemost 2010; 36: 881–887. Boilard E, Nigrovic PA, Larabee K, et al. Platelets amplify inflammation in arthritis via collagen-dependent microparticle production. Science 2010; 327: 580–583. Boilard E, Blanco P, Nigrovic PA. Platelets: active players in the pathogenesis of arthritis and SLE. Nature Rev Rheumatol 2012; 8: 534–542. Gerdes N, Zhu L, Ersoy M, et al. Platelets regulate CD4(+) T-cell differentiation via multiple chemokines in humans. Thromb Haemost 2011; 106: 353–362. Li N. CD4+ T cells in atherosclerosis: regulation by platelets. Thromb Haemost 2013; 109: 980–990. Kunder CA, St John AL, Li G, et al. Mast cell-derived particles deliver peripheral signals to remote lymph nodes. J Exp Med 2009; 206: 2455–2467.


1229

Regulatory T cells inhibit CD34+ cell differentiation into NK cells by blocking their proliferation.

progenitor cells induce conversion of encephalitogenic T cells into CD4+-CD25+- FOXP3+ regulatory T cells.

Hepatocytes induce Foxp3⁺ regulatory T cells by Notch signaling.

Microbe-specific unconventional T cells induce human neutrophil differentiation into antigen cross-presenting cells.

Eosinophils from Murine Lamina Propria Induce Differentiation of Naïve T Cells into Regulatory T Cells via TGF-β1 and Retinoic Acid.

Regulatory T Cells: Differentiation and Function.

Activated CD8+ T cells induce expansion of Vβ5+ regulatory T cells via TNFR2 signaling.

TGF-β-producing regulatory B cells induce regulatory T cells and promote transplantation tolerance.

Adipose tissue macrophages induce PPARγ-high FOXP3(+) regulatory T cells.

IL-17+Foxp3+ T cells: an intermediate differentiation stage between Th17 cells and regulatory T cells.

PDGF upregulates CLEC-2 to induce T regulatory cells.

Hypercholesterolemia Induces Differentiation of Regulatory T Cells in the Liver.

Liver X receptor activation promotes differentiation of regulatory T cells.

Mutant KRAS Conversion of Conventional T Cells into Regulatory T Cells.

BDC12-4.1 T-cell receptor transgenic insulin-specific CD4 T cells are resistant to in vitro differentiation into functional Foxp3+ T regulatory cells.

Embryonic trophoblasts induce decidual regulatory T cell differentiation and maternal-fetal tolerance through thymic stromal lymphopoietin instructing dendritic cells.

Human adipose tissue-derived mesenchymal stem cells abrogate plasmablast formation and induce regulatory B cells independently of T helper cells.

YY1 inhibits differentiation and function of regulatory T cells by blocking Foxp3 expression and activity.

Immunometabolism of regulatory T cells.

EZH2 is crucial for both differentiation of regulatory T cells and T effector cell expansion.

T Cells: Soldiers and Spies--The Surveillance and Control of Effector T Cells by Regulatory T Cells.

Platelet-Released Growth Factors Induce Differentiation of Primary Keratinocytes.

iPSC-Derived Regulatory Dendritic Cells Inhibit Allograft Rejection by Generating Alloantigen-Specific Regulatory T Cells.

Regulatory T cells and COPD.