A. Microfluidic channel system
The fluid dynamic conditions of the human microvasculature, was mimicked in a pneumatically driven microfluidic channel system (BioFlux, San Francisco, California, USA). The BioFlux system is a bench-top instrument, which allows long-term temperature-controlled flow cell assays. Its pressure interface connects a high precision electropneumatic pump to the well plates to initiate controlled flow rates with a nominal shear rate precision of 36 s−1. The channel geometry is a straight rectangular duct with a width of 350 µm and a height of 75 µm. Biofunctionalisation succeeded by coating the channels with either wildtype von Willebrand factor (wt VWF) or a deletion mutant VWF in a concentration of 50 µg/ml each. A homogenous coating, representing the immobilised VWF at the extracellular matrix of a subendothelial vessel wall, was achieved after incubation over night at 37 °C in a moisture rich environment as previously published (20).
B. Microscopic setup
The microfluidic channel system was mounted onto an inverted microscope (Zeiss Axio Observer Z.1, Zeiss AG, Oberkochen, Germany) operated in either fluorescence- or RICM-mode, respectively. RICM is especially beneficial for studying both dynamic and static biological phenomena taking place in vicinity of a transparent substrate, especially under whole blood experimental conditions. This microscopy mode is used to study the interference pattern of polarised incoming light being reflected at an object in order to reconstruct the height profile of the object at an interface as previously published (20). Briefly, the interference of the object beam and the reference beams provides an image depicting sample’s height profile with high resolution. The reduction of scattered light succeeded in utilising the antiflex technique, in which the incident light passes a linear polariser. Therefore, the RICM technique assures detection of the platelet membrane coming in direct contact with the observed surface in a high spatial resolution (54). This specific microscopy technique succeeded by an adapted utilisation of a 63fold antiflex EC Plan-neofluar objective, a reflector module Pol ACR P&C for HBO and a LED module Colibri illumination at 470 nm (all Zeiss AG, Jena, Germany). A sketch of this combined microfluidic/microscopic setting is shown in Figure 1. Image acquisition was performed using a CCD camera (AxioCam MrM) and ZEN software (both Zeiss AG, Jena, Germany).
C. Cell culture and expression of recombinant wildtype and deletion mutant VWF
For recombinant VWF expression, we used HEK293 cells (Deutsche Sammlung von Mikroorganismen und Zellkulturen, Braunschweig, Germany) as previously published (55). In brief, cells were cultured in Dulbecco Modified Eagle Medium (DMEM, Invitrogen, Karlsruhe, Germany) supplemented with 10% foetal bovine serum and 1% penicillin/streptavidin at 37°C and transfected with Lipofectamine 2000 (Invitrogen, Karlsruhe, Germany) and VWF-plasmid-constructs in vector pIRESneo2. Recombinant expression of VWF variants was performed as previously described (56). In brief, HEK293 cells stably express wt VWF or indicated deletion mutants. Samples of the supernatant were taken after 72 hours, centrifuged (5 min at 270 g, 4°C) and concentrated with Amicon Ultrafree-15. The concentration of wt VWF, del-A1 (p.Glu1260_1480del) VWF, del-A2 (p.Asn1493_1673del) and del-A3 (p.Gly1672_1874del) VWF was determined by a polyclonal rabbit anti-human VWF:Ag-ELISA (DAKO, Hamburg, Germany). For a detailed description of its binding epitopes refer to Tan et al. (57).
D. Preparation of the perfusion media
For preparation of the perfusion media, blood was smoothly collected from healthy volunteers after informed consent. We utilised sodium citrated monovettes with manual syringe stamps to avoid platelet preactivation and further inhibit the activity of inherent degradation enzymes (e.g. ADAMTS-13). This study, approved by the Ethics Committee II of the Heidelberg University (Mannheim, Germany), was conducted in conformity to the Declaration of Helsinki (58), to The International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH) Guidelines and to the Convention for the Protection of Human Rights and Dignity of the Human Being with regard to the Application of Biology and Medicine: Convention on Human Rights and Biomedicine (Oviedo, 4 April 1997).
Platelets were isolated from blood samples, washed and fluorescently stained as previously published (35). These platelets were resuspended in divalent cation free phosphate-buffered saline solution and used for perfusion in a concentration of 200,000 per µl supplemented with 45% haematocrit. Alternatively, the blood samples were natively perfused as citrate-anticoagulated whole blood supplemented with wt VWF or deletion mutant VWF as indicated.
E. Experimental procedure
Under physiological conditions, VWF comes into effect in both the extracellular matrix of the subendothelial vessel wall (immobilised) and the circulating plasma (soluble) (4, 6-8). Three sets of experiments were designed to discretely mimic the complex physiology of VWF mediated platelet adhesion and VWF-platelet aggregate formation in vitro:
1. In order to investigate the impact of deletions in the VWF A-domains on platelet adhesion, microfluidic channels were biofunctionalised with indicated deletion mutant VWF and perfused with washed platelets supplemented with 45% haematocrit as described above at a shear rate of 500 s-1. After 1, 5 and 10 minutes of perfusion, platelet adhesion was studied by fluorescence microscopy using a Zeiss 10fold objective compared to platelet adhesion on wt VWF biofunctionalisation. Note that no soluble VWF was present in these experiments.
2. Addressing the impact of immobilised and soluble VWF on platelet adhesion under physiological and pathological flow conditions, channels were biofunctionalised with wt VWF. Then, we perfused the channels with the washed platelet solution supplemented with 45% haematocrit as described above with or without addition of 10 µg/ml wt VWF. Perfusion was performed at distinct shear rates in the range of 1,000 s-1 to 10,000 s-1 for 5 minutes. Live cell fluorescence videos were recorded with four frames per second using a Zeiss 20fold objective and analysed as described.
3. To study VWF-platelet aggregate formation, channels were biofunctionalised with wt VWF as described above. The biofunctionalised channels were perfused with native citrate-anticoagulated whole blood, additionally supplemented with indicated A-domain deleted VWF thus raising the collective VWF concentration to 50 µg/ml. The aggregation behaviour, namely the critical shear rate necessary for the formation of VWF-platelet aggregates, was then monitored as previously published (20). Briefly, the shear rate was consecutively increased from 1,000 s-1 to 5,000 s-1 in discrete steps for 30 seconds each, and RICM movies were recorded with two frames per second using a specialised Zeiss 63fold antiflex objective. VWF-platelet aggregates consisting of at least 15 platelets rolling at the channel footprint were considered rolling aggregates. We determined the critical shear rate for formation of these whole blood/A-domain deleted VWF aggregates compared to those of whole blood supplemented with wt VWF. For each experiment of the aforementioned settings at least four independent experiments were performed.
F. Image analysis and statistical computation
For image analysis, we used ZEN software (Zeiss AG, Jena, Germany). Calculation and quantification of the platelet SC succeeded using the open-source software ImageJ (V. 1.46r, National Institute of Health, Bethesda, Maryland, USA) analysing five randomly chosen contrast-normalised fields of view at each indicated point in time of each independent experiment. Quantification of the RICM signal intensity was also performed using ImageJ analysing contrast-normalised fields of view at each indicated point in time of each independent experiment, plotted against the time. Mean data of experiments are given with standard deviation (SD). Statistical computation was performed with SAS 9.2 (SAS Institute Inc., Cary, North Carolina, USA). Statistical significance was tested by the unpaired Student’s t-test. Significant differences of compared values are indicated by * (P < 0.05).