Assessment of whole blood thrombosis in a microfluidic device lined by fixed human endothelium

Abstract

The vascular endothelium and shear stress are critical determinants of physiological hemostasis and platelet function in vivo, yet current diagnostic and monitoring devices do not fully incorporate endothelial function under flow in their assessment and, therefore, they can be unreliable and inaccurate. It is challenging to include the endothelium in assays for clinical laboratories or point-of-care settings because living cell cultures are not sufficiently robust. Here, we describe a microfluidic device that is lined by a human endothelium that is chemically fixed, but still retains its ability to modulate hemostasis under continuous flow in vitro even after few days of storage. This device lined with a fixed endothelium supports formation of platelet-rich thrombi in the presence of physiological shear, similar to a living arterial vessel. We demonstrate the potential clinical value of this device by showing that thrombus formation and platelet function can be measured within minutes using a small volume (0.5 mL) of whole blood taken from subjects receiving antiplatelet medications. The inclusion of a fixed endothelial microvessel will lead to biomimetic analytical devices that can potentially be used for diagnostics and point-of-care applications.

Keywords: Biomedical technology; Hemostasis; Lab-on-a-Chip; Platelet function tests; Thrombosis; Vascular endothelium.

Fig. 1

Engineering of the microfluidic device containing a chemically preserved endothelium. a Schematic of the microfluidic device showing a blood inlet reservoir, followed by the straight microchannel that ends at the outlet, where the syringe pump is attached to pull the blood, and b a photograph of a single polydimethylsiloxane (PDMS) microfluidic chip containing six independent microfluidic devices (bar, 15 mm) fabricated on a glass slide. c A fluorescence micrograph showing the entire microchannel covered with human umbilical vein endothelial cells (HUVECs) immunostained against the cell adhesion molecule, VE-cadherin (bar, 1 mm). d Confocal immunofluorescence microscopic images showing a section of the microchannel with HUVECs when viewed from above (‘xy’) and reconstruction of cross-sectional views from the front and the side (‘yz’ and ‘xz’, respectively) demonstrating full coverage of all microfluidic channel walls. (bottom; green, VE-Cadherin; blue, nuclear DAPI; bar, 200 μm). e–h Graph showing fluorescence (normalized by the untreated endothelium) measured after immunostaining the fixed endothelium with ICAM-1 (e), VCAM-1 (f), VWF (g), or tissue factor (h). *P < 0.05 versus untreated; n = 3

Fig. 2

Platelet coverage and fibrin formation on the fixed endothelium in the microdevice. a Representative maximum intensity projection micrographs showing fluorescently labeled platelets adhering to the fixed endothelium in a tumor necrosis factor (TNF-α) dose-dependent manner (bar, 100 μm). b Graph showing platelet coverage when blood is perfused inside the microchannel lined with a living or fixed endothelium that was stimulated by TNF-α before fixation. Comparison of living vs. fixed microchannels was not significantly different (P > 0.05) at each TNF-α concentration. (n = 4; *P < 0.05). c Fluorescent micrograph shows fibrin (green) is formed along with platelet aggregates (red) on a fixed endothelium that was pretreated with TNF-α and perfused with recalcified citrated whole blood (left, bar, 200 μm; right, bar, 20 μm)

Fig. 3

Assessment of antiplatelet therapy with the microfluidic device. a Platelet coverage on the fixed endothelium pretreated with TNF-α when blood samples containing different doses of the drug abciximab were perfused through the microfluidic device (n = 4) b Light transmission aggregometry of blood samples containing different doses of abciximab using either ADP or collagen as an agonist (n = 4). c Platelet coverage when blood samples containing different doses of the drug abciximab were perfused through collagen-coated microfluidic devices (n = 4). d Platelet coverage on the fixed endothelium pretreated with TNF-α when blood samples from healthy donors versus subjects treated with antiplatelet drugs were perfused through microfluidic devices (n = 11). e Light transmission aggregometry of healthy versus antiplatelet treated blood samples using ADP or collagen as an agonist (n = 11). f Platelet coverage when healthy versus subject blood samples were perfused through collagen-coated microfluidic devices (n = 11). *P < 0.05