Platelet-Derived GARP Induces Peripheral Regulatory T Cells-Potential Impact on T Cell Suppression in Patients with Melanoma-Associated Thrombocytosis

Abstract

Platelets have been recently described as an important component of the innate and adaptive immunity through their interaction with immune cells. However, information on the platelet-T cell interaction in immune-mediated diseases remains limited. Glycoprotein A repetitions predominant (GARP) expressed on platelets and on activated regulatory T cells (Treg) is involved in the regulation of peripheral immune responses by modulating the bioavailability of transforming growth factor β (TGF-β). Soluble GARP (sGARP) exhibits strong regulatory and anti-inflammatory capacities both in vitro and in vivo, leading to the induction of peripheral Treg. Herein, we investigated the effect of platelet-derived GARP on the differentiation, phenotype, and function of T effector cells. CD4⁺CD25^- T cells cocultured with platelets upregulated FoxP3, the master transcription factor for Treg, were anergic, and were strongly suppressive. These effects were reversed by using a blocking anti-GARP antibody, indicating a dependency on GARP. Importantly, melanoma patients in different stages of disease showed a significant upregulation of GARP on the platelet surface, correlating to a reduced responsiveness to immunotherapy. In conclusion, our data indicate that platelets induce peripheral Treg via GARP. These findings might contribute to diseases such as cancer-associated thrombocytosis, wherein poor prognosis and metastasis are associated with high counts of circulating platelets.

Keywords: GARP; Treg; melanoma; platelets; thrombocytosis.

Figure A1

sGARP induced Treg. (A) GARP expression on Treg vs. Teff: GARP is an activation marker on Treg. 16h after stimulation, CD4⁺ CD25⁺ Treg showed a strong GARP expression compared to CD4⁺CD25⁻ Teff. One representative dot plot of four different experiments is shown. (B) Suppressive capacity of sGARP-conditioned T cells. CD4⁺CD25⁻ T cells were stimulated with 0.5 µg/mL anti-CD3 mAb and 1.0 µg/mL anti-CD28 mAb for 6 days in the presence of 10 µg/mL sGARP, resulting in Treg. Simultaneously, allogenic Teff were generated without the use of sGARP. After 6 days of stimulation, Teff and titrated Treg were cocultured and re-stimulated with anti-CD3 mAb and irradiated PBMC. After 4 days, proliferation was determined by tritiated thymidine (³H TdR) pulse. Results show the pooled data of three independent experiments. (C) sGARP induction of FoxP3, inhibition of proliferation, and inhibition of effector cytokine production in CD4⁺ T cells (n = 3, means ± SD, * p < 0.05, ** p ≤ 0.01, **** p ≤ 0.0001, and n.s. determined by one-way ANOVA).

Figure A2

Platelets bound to T cells upon activation in coculture. CD4⁺CD25⁻ T cells were cocultured with or without platelets at the ratio of 1:50 and stimulated with 0.5 µg/mL anti-CD3 mAb and 1.0 µg/mL anti-CD28 mAb for 6 days. At day 1 and 6, platelet–T cell conjugates (CD41a⁺CD4⁺ double-positive cells) were analyzed via flow cytometry. For assessment of GARP expression, only CD4⁺CD41a- cells were included in the analysis (indicated by the pregating on the upper left (UL)). GARP expression was normalized to the untreated (w/o) control. Dot plots show one representative result of five independent experiments. Isotype controls are shown (n = 5, box and whiskers, medians + min/max, ** p ≤ 0.01, *** p ≤ 0.001, and n.s. determined by one-way ANOVA).

Figure A3

Experimental setup of a platelet-conditioned T cell suppression assay. 1 × 10⁶ CD4⁺CD25⁻ T cells were stimulated with 0.5 µg/mL anti-CD3 mAb and 1.0 µg/mL anti-CD28 mAb with or without 50 × 10⁶ platelets for 6 days. T cells cocultured with platelets developed an iTreg-like phenotype, whereas T cells cultured without platelets displayed a Teff phenotype. Next, iTreg and Teff were cultured at different ratios (iTreg/Teff: 1:1, 1:2, 1:4, and the corresponding controls) and restimulated with 0.5 µg/mL anti-CD3 mAb and 90Gy irradiated PBMC. Proliferation (Ki-67) was measured via flow cytometry 3 days after restimulation.

Figure A4

Thrombin-activated platelets induced a regulatory phenotype in CD4⁺CD25⁻ T cells. 1 × 10⁶ CD4⁺CD25⁻ T cells were stimulated with 0.5 µg/mL anti-CD3 mAb and 1.0 µg/mL anti-CD28 mAb with or without 50 × 10⁶ platelets for 6 days and treated with or without 10 U/mL thrombin. (A) Foxp3 and GARP expression and proliferation were determined at day 3 via flow cytometry. (B) Using intracellular flow cytometry, we analyzed cytokine production of IL-2 and IFN-γ on day 6. Dot plots show one representative result of five independent experiments (n = 5, box and whiskers, medians ± min/max, * p < 0.05, ** p ≤ 0.01, and n.s. determined by one-way ANOVA).

Figure A5

TRAP-6-activated platelets induced a regulatory phenotype in CD4⁺CD25⁻ T cells. 1 × 10⁶ CD4⁺CD25⁻ T cells were stimulated with 0.5 µg/mL anti-CD3 mAb and 1.0 µg/mL anti-CD28 mAb with or without 50 × 10⁶ platelets for 6 days and treated with or without 5 µM TRAP-6. (A) Foxp3 and GARP expression and proliferation were determined at day 3 via flow cytometry. (B) Using intracellular flow cytometry, we analyzed cytokine production of IL-2 and IFN-γ on day 6. Dot plots show one representative result of five independent experiments (n = 5, box and whiskers, medians ± min/max, * p < 0.05, ** p ≤ 0.01, and n.s. determined by one-way ANOVA).

Figure A6

Flow cytometric analysis of T cell purity. CD4⁺CD25⁻ cells showed a purity of around 98% after isolation. CD4⁺CD25⁻ T cells were isolated as described in the manuscript. Shown are the unstained control, the CD25-positive staining control, and successful CD4⁺CD25⁻ isolations of three different healthy donors.

Figure A7

Gating strategy. Cells were pregated via forward- and side scatter (FSC/SSC) following a dead cell exclusion via fixable viability dye.

Figure 1

Glycoprotein A repetitions predominant (GARP) was expressed on the surface of platelets and was detectable in the supernatant of activated platelets. (A) Flow cytometric analysis of CD62P and GARP expression levels in resting and pre-activated platelets. Only singlets were used in the analysis. Isotype controls are shown. Bar diagrams of CD62P and GARP expression show pooled data of percentages (%) of positive cells and raw means (n = 3, means ± SD * p < 0.05, and n.s. determined by Student’s t-test). (B) Presence of soluble GARP (sGARP) in the supernatant of pre-activated platelets. sGARP content of the supernatant of 2 × 10⁹ activated platelets after 16 h compared to a negative medium control (n.d. = not detected). sGARP levels were determined by ELISA from three different healthy donors (HD).

Figure 2

Platelets induce peripheral regulatory T cells (iTreg). (A) CD4⁺ CD25⁻ T cells and platelets were cocultured as indicated. Herein, carboxyfluorescein succinimidyl ester (CFSE)-labeled CD4⁺CD25⁻ T cells were stimulated with 0.5 µg/mL anti-CD3 mAb and 1.0 µg/mL anti-CD28 mAb with or without different ratios of platelets. The expression of Foxp3, GARP and the proliferation of cells were analyzed via flow cytometry on day 3 after stimulation. (B) Effector cytokine production of Interferon(IFN)-γ and interleukin (IL)-2 was analyzed using intracellular flow cytometry on day 6. Only live CD4⁺CD25⁻ T cells were included into the analysis. Representative dot plots of 12 independent experiments are shown. The graphs show cells cultured in the presence of platelets normalized to CD4⁺CD25⁻ T cells without platelets (expression levels were normalized to 100) (n = 12, means ± SD, * p < 0.05, ** p ≤ 0.01, *** p ≤ 0.001, and n.s. determined by Kruskal–Wallis test).

Figure 3

Platelet-induced iTreg suppressed T effector cells (Teff) cells. To analyze iTreg induction by platelets, we expanded CD4⁺CD25⁻ T cells for 6 days in the presence of platelets at the ratio of 1:50, as described previously, and were subsequently incubated at various ratios with 0.5 × 10⁵ CD4⁺CD25⁻ T cells and restimulated with 1 × 10⁵ irradiated peripheral blood mononuclear cells (PBMC) and 0.5 µg/mL anti-CD3 mAb. Proliferation was determined on day 3 of culture by Ki-67 staining (n = 3, means ± SD, * p < 0.05, and n.s. determined by one-way ANOVA).

Figure 4

Platelets induced iTreg in a mainly GARP-dependent manner. (A) Blocking anti-GARP antibody inhibited platelet-induced effects on FoxP3 and GARP expression, as well as proliferation. CFSE-labeled CD4⁺CD25⁻ T cells were stimulated with 0.5 µg/mL anti-CD3 mAb and 1.0 µg/mL CD28 mAb with different ratios of platelets. In coculture, 10 µg/mL blocking anti-GARP Ab was added at day 0, as indicated. Foxp3 and GARP expression and proliferation is shown on day 3 after stimulation. (B) Cytokine production is shown on day 6. The graphs show cells cultured in the presence of platelets normalized to CD4⁺CD25⁻ T cells without platelets. For analysis, only live cells were included. Representative dot plots of eight independent experiments are shown (n = 9, box and whiskers, medians ± min/max, * p < 0.05, *** p ≤ 0.001, and n.s. determined by one-way ANOVA).

Figure 5

Blockade of transforming growth factor (TGF)-β I-III did in part prevent regulatory T cells (Treg) induction. (A) CFSE-labeled CD4⁺CD25^- T cells were cocultured with platelets in the ratio of 1:50 and were stimulated with anti-CD3 mAb (0.5 µg/mL) and anti-CD28 mAb (1.0 µg/mL) in the presence of either anti-TGF-β I-III (10 µg/mL) or anti-TGF-β receptor II (10 µg/mL) antibodies. Antibodies were added at day 0. The expression of Foxp3 and GARP and cell proliferation were determined on day 3 via flow cytometry. (B) Production of IL-2 and IFN-γ was assessed by intracellular flow cytometry on day 6. The graphs show cells cultured in the presence of platelets normalized to CD4⁺CD25⁻ T cells without platelets. Dot plots show one representative result of 10 independent experiments (n = 10, box and whiskers, medians ± min/max, * p < 0.05, ** p ≤ 0.01, *** p ≤ 0.001, and n.s. determined by one-way ANOVA).

Figure 6

Combining blockade of TGF-β signaling and GARP led to a complete inhibition of platelet effects. (A) CFSE-labeled CD4⁺CD25⁻ T cells were cocultured with platelets in the ratio of 1:50 and were stimulated with anti-CD3 mAb (0.5 µg/mL) and anti-CD28 mAb (1.0 µg/mL). CD4⁺CD25⁻ T cells were incubated for 15 min with TGF-β receptor II (10 µg/mL) antibody prior to coculture, as indicated. Excess antibody was removed. Pre-treated CD4⁺CD25⁻ T cells were cultured in the presence of either anti-TGF-β I–III (10 µg/mL) and/or anti-GARP Ab (10 µg/mL) antibodies. Antibodies were added at day 0. The expression of Foxp3, GARP and cell proliferation were determined on day 3 via flow cytometry. (B) Production of IL-2 and IFN-γ was assessed by intracellular flow cytometry on day 6. The graphs show cells cultured in the presence of platelets normalized to CD4⁺CD25⁻ T cells without platelets. Dot plots show 1 representative result of 10 independent experiments (n = 3, means ± SD, * p < 0.05, ** p ≤ 0.01, and n.s. determined by one-way ANOVA).

Figure 7

Platelet-conditioned medium (PCM) inhibited IFN-γ production, but failed to induce a Treg phenotype. (A) CD4⁺CD25⁻ T cells were cultured in X-Vivo 15 (Lonza, Basel, Switzerland) with 60% PCM content, with or without 10 µg/mL anti-GARP Ab and stimulated with 0.5 µg/mL anti-CD3 mAb and 1.0 µg/mL anti-CD28 mAb. Antibodies were added at day 0. The expression of Foxp3, GARP and cell proliferation were determined at day 3 with flow cytometry. (B) Cytokine production of IL-2 and IFN-γ was measured by intracellular flow cytometry on day 6. Dot plots show one representative result of five independent experiments (n = 5, box and whiskers, medians ± min/max, * p < 0.05, ** p ≤ 0.01, *** p ≤ 0.001, and n.s. determined by one-way ANOVA).

Figure 8

Increased platelet surface GARP levels in stage I/IV melanoma patient stage. (A) Comparison of platelet GARP and CD62P surface levels between HD (n = 8), stage I (n = 18), and stage IV (n = 17) melanoma patients assessed via flow cytometry. Means for GARP expression for HD = 1161, stage I = 1489, and stage IV = 1418. Means for CD62P expression are HD = 1470, stage I = 2611, and stage IV = 1977 (box and whiskers, medians ± min/max, * p < 0.05, ** p <.01, *** p <.001, and n.s. determined by unpaired t-test). (B) The platelet count in melanoma patients before starting immunotherapy using checkpoint inhibitors was measured as part of the routine blood tests. Platelet counts and the platelet–lymphocyte ratios of non-responders (n = 25) and responders (n = 11; staging results after 6 months of therapy) were compared. Means of platelet counts (PC) and the platelet–lymphocyte ratio (PLR) were as follows: progression: PC = 339.8, PLR = 269.0; stable disease: PC = 222.3, PLR = 177.5 (means ± SD, * p < 0.05, ** p ≤ 0.01, *** p ≤ 0.001, and n.s. determined by unpaired t-test).