Stent deployment to treat coronary artery disease causes damage and loss of endothelial cells (EC). Wall shear stress (WSS), a mechanical force exerted by flowing blood on the vessel wall, has been proposed as a major regulator of EC migration through downstream signalling events involving the Rho family of GTPases. However, the biomechanical mechanisms that control re-endothelialisation in stented arteries remain poorly understood. It was hypothesised that stent struts may impede repair of injured endothelium by inducing localised disturbances in WSS.
An in vitro platform was developed to assess the influence of surface features on EC migration under flow. To simulate a stented artery, a polydimethylsilozane-based flow chamber was fabricated with ridges (100 µm-high) positioned perpendicular to the flow direction. Confluent EC monolayers were seeded on one side of the ridged chamber (or on a non-ridged chamber as control) and were then exposed to flowing cell culture medium using the Ibidi® pump system. The migratory behaviour of EC was monitored with time-lapse imaging and analysed using ImageJ® software. Velocity and directional persistence (DP; a measure of migration efficiency; ratio of absolute distance to contour length) were calculated. Computational fluid dynamic (CFD) modelling and particle velocimetry were performed to determine flow patterns in the flow chamber.
CFD modelling and live cell imaging indicated that EC on the non-ridged chamber slide were exposed to a uniform WSS of 13 dyn/cm2 and migrated relatively uniformly in parallel with the flow direction (average velocity 1.13 ± 0.20 µm/min; DP 0.59 ± 0.14). By contrast, significant spatial differences in WSS were observed over the ridged slide in CFD, with significant spikes above physiological levels (>70 dyn/cm2) at the corners of the ridges and distinctive flow recirculation zone immediately upstream/ downstream from the ridge (-4 dyn/cm2). These features were verified by particle velocimetry using fluorescently labelled polystyrene beads (2 µm diameter). Time-lapse imaging further revealed interrupted EC migration at the ridges. Specifically, although EC could migrate over the ridges, those that reached the recirculation zone downstream form the ridge migrated with non-uniform directionality (DP 0.25 ± 0.06) and displayed a reduction in velocity (0.78 ± 0.18 µm/min). Inhibition of the RhoA/ROCK signalling pathway with ROCK inhibitors (Y27632 or HA1077) promoted EC forward migration within the recirculation zone by significantly elevating DP (0.41 ± 0.04, p = 0.01) and velocity (1.16 ± 0.09 µm/min, p = 0.01).
Disturbed WSS generated downstream from stent strut-like ridges prevented the forward migration of EC. Inhibition of the RhoA/ROCK signalling pathway promoted EC migration and endothelialisation of these sites. Our data suggest that treatment using a ROCK inhibitor may promote re-endothelialisation of stented arteries; a concept that is currently been tested using a porcine model.