Platelet Adhesion to Podoplanin Under Flow is Mediated by the Receptor CLEC-2 and Stabilised by Src/Syk-Dependent Platelet Signalling

Summary Platelet-specific deletion of CLEC-2, which signals through Src and Syk kinases, or global deletion of its ligand podoplanin results in blood-filled lymphatics during mouse development. Platelet-specific Syk deficiency phenocopies this defect, indicating that platelet activation is required for lymphatic development. In the present study, we investigated whether CLEC-2-podoplanin interactions could support platelet arrest from blood flow and whether platelet signalling is required for stable platelet adhesion to lymphatic endothelial cells (LECs) and recombinant podoplanin under flow. Perfusion of human or mouse blood over human LEC monolayers led to platelet adhesion and aggregation. Following αIIbβ3 blockade, individual platelets still adhered. Platelet binding occurred at venous but not arterial shear rates. There was no adhesion using CLEC-2-deficient blood or to vascular endothelial cells (which lack podoplanin). Perfusion of human blood over human Fc-podoplanin (hFcPDPN) in the presence of monoclonal antibody IV.3 to block FcγRIIA receptors led to platelet arrest at similar shear rates to those used on LECs. Src and Syk inhibitors significantly reduced global adhesion of human or mouse platelets to LECs and hFcPDPN. A similar result was seen using Syk-deficient mouse platelets. Reduced platelet adhesion was due to a decrease in the stability of binding. In conclusion, our data reveal that CLEC-2 is an adhesive receptor that supports platelet arrest to podoplanin under venous shear. Src/Syk-dependent signalling stabilises platelet adhesion to podoplanin, providing a possible molecular mechanism contributing to the lymphatic defects of Syk-deficient mice.

Human LECs were purchased from Promocell (Heidelberg, Germany) and cultured in the supplier's recommended medium (MV2) containing penicillin, streptomycin and gentamycin. Human umbilical vein endothelial cells (HUVECs) were isolated from umbilical cords obtained after informed consent from the University of Birmingham Human Biomaterials Resource Centre (ethical approval reference 09/H1010/75), and cultured as described elsewhere (3).
The extracellular domain of human podoplanin (amino acids 23-131) was amplified from human dermal microvascular endothelial cells cDNA with primers hPodoFHindIII (GTCAGCAGGAAGCTTCCAGGAGAGCAACAACTCAAC) and hPodoRBamHI (TCGGCTCCGGATCCACTGTTGACAAACCATCTTTCTC). After digestion with HindIII and BamHI the product was cloned into IgFc vector pcDNA3Ig to yield a construct encoding the podoplanin extracellular domain fused at the COOH terminus to the Fc region of human IgG1. Recombinant protein was expressed in human 293T cells and purified as previously described (4).

Whole blood impedance aggregometry
Heparinized human blood was diluted 1:1 using saline (0.9%) and stirred for 3 minutes at 37°C. PRT060318 at the stated concentrations or DMSO as control were preincubated at 37°C for 1 minutes. Following collagen stimulation (10 g/ml), aggregation was monitored at 37°C for 6 minutes, calculated by impedance across two electrode pairs (Multiplate analyser, Instrumentation Laboratory UK Ltd, Warrington, UK), and reported in arbitrary impedance units (AU).
Flow cytometry: Mouse CLEC-2 was detected in blood diluted 1:10 with Tyrode's buffer using Alexa Fluor 488-conjugated rat anti-mouse CLEC-2 or rat IgG2b isotype control antibodies. For the detection of Syk-dependent P-selectin exposure, mouse whole blood was stimulated with 100 nM rhodocytin for 5 minutes at 37°C in the presence of 5 l FITClabelled anti-mouse P-selectin. Non-stimulated platelets were used as negative control. Expression of human podoplanin was assessed in HUVECs and LECs using a phycoerythrin-conjugated rat anti-human podoplanin antibody. Endothelial cells incubated with a phycoerythrin-conjugated rat IgG2a isotype control were used as negative control. Cells were identified by their FSC/SSC profile in a FACScalibur flow cytometer using Cell-Quest software (BD Biosciences, Oxford, UK).

Suppl. Figure 1: Analysis of podoplanin expression in LECs and HUVECs by flow cytometry. Human podoplanin expression was assessed in human LECs and
HUVECs by flow cytometry using a phycoerythrin-conjugated rat anti-human podoplanin antibody (green traces). Endothelial cells incubated with a phycoerythrinconjugated rat IgG2a isotype control were used as negative control (violet traces). Cells were identified by their FSC/SSC profile. A representative example is shown.
Suppl. Figure 2: CD31 or CD61 blocking antibodies do not affect human platelet binding to LECs. Confluent human LECs were treated for 10 minutes with either control IgG, anti-CD31 or anti-3 integrin blocking antibodies (10 g/ml), washed and then perfused for 8 minutes at 50 s -1 with heparinized whole blood in the presence of integrilin. Channels were washed, fixed and stained with anti-CD41/anti-mouse Alexa-488 antibodies. LEC monolayer integrity was confirmed by immunostaining for podoplanin. (A) Representative pictures are shown. Bars represent 20 μm. (B) Quantification of percentage of surface coverage from 3 independent experiments. Error bars represent SEM. No significant differences between treatments were found by one-way ANOVA test. Figure 3: Verification of CLEC-2 genotypes by flow cytometry. Platelet CLEC-2 expression in CLEC-2 floxed/floxed; CreERT2 mice injected with tamoxifen to induce CLEC-2 excision or with corn oil as control was analysed in whole blood by flow cytometry using an Alexa Fluor 488-conjugated rat anti-mouse CLEC-2 (green traces). Platelets incubated with a rat IgG2b isotype control were used as negative control (filled violet traces). Platelets were identified by their FSC/SSC profile. A representative example from 3 independent experiments is shown.

Suppl.
Suppl. Figure 4: PLC inhibition does not affect human platelet binding to LECs. Confluent human LECs were treated for 10 minutes with either vehicle control (DMSO) or the non-specific inhibitor of phospholipase C gamma (PLC) U73122 (10 M) and then perfused for 8 minutes at 50 s -1 with heparinized whole blood in the presence of 9M integrilin. Channels were washed, fixed and stained with anti-CD41/anti-mouse Alexa-488 antibodies. LEC monolayer integrity was confirmed by immunostaining for podoplanin. (A) Representative pictures are shown. Bars represent 20 μm. (B) Quantification of percentage of surface coverage from 3 independent experiments. Error bars represent SEM. No significant differences between treatments were found by two-tailed paired t-test.
Suppl. Figure 5: Inhibition of collagen-induced whole blood platelet aggregation by PRT060318. Heparinized human blood was incubated with PRT060318 at the stated concentrations or DMSO vehicle control and then stimulated with 10 g/ml collagen. Aggregation was monitored at 37°C for 6 min, calculated by impedance across two electrode pairs (red and blue lines) using a Multiplate analyser, and reported in arbitrary impedance units (AU).
Suppl. Figure 6: Fixation of platelets but not LECs abrogates platelet binding to LEC monolayers. Confluent human LECs were treated for 15 minutes with either (A & C) PBS as control or (B) 1% formalin, washed and then perfused for 8 minutes at 50 s -1 with (A & B) unfixed or (C) 1% formalin-fixed heparinized blood in the presence of integrilin. Phase contrast images were obtained using a 20x lens, bars represent 100 μm. Fluorescence images were obtained in a different field of view using a 63x lens. Bars represent 20 μm. Each experiment was repeated twice.
Suppl. Figure 7: Verification of chimeric mice genotypes by flow cytometry. Sykdependent platelet P-selectin exposure was analysed in whole blood from WT or Sykdeficient chimeric mice. Blood was stimulated with 100 nM rhodocytin in the presence of FITC-labelled anti-mouse P-selectin antibody (green traces). Non-stimulated platelets were used as negative control (filled violet traces). Platelets were identified by their FSC/SSC profile. A representative example from 5 independent experiments is shown.