Antagonistic Roles of Human Platelet Integrin αIIbβ3 and Chemokines in Regulating Neutrophil Activation and Fate on Arterial Thrombi Under Flow

Background: Platelets and neutrophils are the first blood cells accumulating at sites of arterial thrombus formation, and both cell types contribute to the pathology of thrombotic events. We aimed to identify key interaction mechanisms between these cells using microfluidic approaches. Methods: Whole-blood perfusion was performed over a collagen surface at arterial shear rate. Platelet and leukocyte (in majority neutrophil) activation were microscopically visualized using fluorescent markers. The contributions of platelet-adhesive receptors (integrin, P-selectin, CD40L) and chemokines were studied by using inhibitors or antibodies and using blood from patients with GT (Glanzmann thrombasthenia) lacking platelet-expressed αIIbβ3. Results: We observed (1) an unknown role of activated platelet integrin αIIbß3 preventing leukocyte adhesion, which was overcome by short-term flow disturbance provoking massive adhesion; (2) that platelet-expressed CD40L controls the crawling pattern and thrombus fidelity of the cells on a thrombus; (3) that continued secretion of platelet substances promotes activation of identified neutrophils, as assessed by (fMLP [N-formylmethionyl-leucyl-phenylalanine, a potent chemotactic agent and leukocyte activator] induced) [Ca2+]i rises and antigen expression; (4) and that platelet-released chemokines activate the adhered cells in the order of CXCL7>CCL5>CXCL4. Furthermore, postsilencing of the platelets in a thrombus suppressed the leukocyte activation. However, the leukocytes on thrombi did no more than limitedly form neutrophil extracellular traps, unless stimulated with phorbol ester or lipopolysaccharide. Conclusions: Together, these findings reveal a multifaceted regulation of adhesion and activation of neutrophils by platelets in a thrombus, with a balanced role of several platelet-adhesive receptors and a promoting role of platelet-released substances. This multivalent nature of neutrophil-thrombus interactions offers novel prospects for pharmacological intervention.


A Major Resources
included at the end of this document.

Blood Cell Preparation
Where indicated, polymorphonuclear leukocytes were enriched from heparinized blood using a histopaque-1077 and histopaque-1119 density gradient centrifugation method, essentially as indicated. 1 Briefly, a mixture of equal volumes of heparinized blood, histopaque-1077 and histopaque-1119 was centrifuged at 700g for 30 minutes without brake. The granulocyte-containing layer was removed, then washed twice with 10 mL isotonic phosphate-buffered saline, and recentrifuged at 350g for 10 minutes. The pelleted granulocytes, containing >90% neutrophils, were resuspended at a concentration of 1×10 6 cells/mL into Hepes buffer pH 7.45 (10 mmol/L Hepes, 136 mmol/L NaCl, 2.7 mmol/L KCl, 2 mmol/L MgCl2, 0.1% glucose and 0.1% bovine serum albumin). Cell suspensions were supplemented with 2 mmol/L CaCl2 before use. Purity and density of the cell suspensions were assessed with a Sysmex-XP300 hematology analyzer.

Flow Cytometry
Platelet integrin aIIbb3 expression levels (based on PE anti-CD41 mAb staining), and platelet-neutrophil conjugates in diluted, thrombin-stimulated whole blood (based on PerCP anti-CD42a mAb staining) were assessed, as described before. 2 Purified suspensions of granulocytes were evaluated for activation by flow cytometry, using an Accuri C6 flow cytometer and C-Flow Plus software (Becton-Dickinson Bioscience). Neutrophils were gated in forward/side scatter plots from the positive signal with FITC anti-CD66b mAb (1 µg/mL).
Activation was assessed from cells containing elevated CD66b expression, MPO expression and/or ROS staining using CellRox Deep Red. After gating, mean fluorescence intensities of events were recorded. Where indicated, the granulocytes were pre-stimulated with N-formyl-methionine leucyl phenylalanine (fMLP, 1-4 µmol/L) or CCL5 (500 nmol/L).

Parameter analysis of thrombus formation
Brightfield and fluorescence images, recorded from thrombi formed after whole-blood perfusion flow over indicated collagen surfaces, were analyzed by observers blinded to the condition using previously described scripts in Fiji/ImageJ (supplement of Ref. 3 ). In brief, surface area coverage % per image type was obtained by background correction, object recognition, and semi-automated (operator-dependent) intensity threshold settings. Thrombus morphological scores were compared to reference images and set as: 0, no adhesion; 1, single platelets; 2, platelet monolayer; 3, platelet aggregates; 4, large aggregates; and 5, end-stage aggregates.

Analysis of Leukocyte Ca 2+ Rises and Movement Patterns
For defined trace analyses of single-cell rises in [Ca 2+ ]i, the confocal time stacks were imported in Matlab with the OME Bio-Formats toolbox (https://www.openmicroscopy.org/), and then processed using DIPImage (http://www.diplib.org/). In short, the stacks were filtered by a 3D median filter and thresholded to a fixed value. Objects (cells) touching the image-border anytime during the acquisition, and objects that split into multiple objects were rejected from the binary mask. The 3D objects were then assigned in time frames of individual cells, where for each time frame the intensity was determined.
The single cell fluorescence traces were normalized to the resting level (F/Fo), and analyzed for responsiveness using GraphPad Prism software.
Confocal time series were also analyzed for movement patterns of individual cells, using Fiji software. Movement profiles were classified as: <20% contact with platelet thrombi, a temporary 20-80% contact, or a persistent >80% contact with the thrombi. Regarding directionality of the movement, contacting cells were classified as: low fidelity (switching between thrombi) or high fidelity (>80% stay on or return to same thrombus).

Figure S5. Activation and NET formation of neutrophils adhered to thrombi
Thrombi of type I-III with adhered leukocytes were formed in microfluidic chambers and incubated for 0-16 h at 37°C in incubation medium containing antibiotics, as for Figure S3.        Online video S3. Whole blood was flowed over collagen under coagulant conditions at 1000 s -1 for 4 min to form type III thrombi (see Figure S3). Thrombi were then post-perfused with AF647-labelled annexin A5, FITC-anti-CD62P mAb and AF568-anti-CD15, and imaged for 10 minutes. Image size 213 x 213 µm.