The development of 3-D, in vitro, endothelial culture models for the study of coronary artery disease

Abstract

The response of the vascular endothelium to wall shear stress plays a central role in the development and progression of atherosclerosis. Current studies have investigated endothelial response using idealized in vitro flow chambers. Such cell culture models are unable to accurately replicate the complex in vivo wall shear stress patterns arising from anatomical geometries. To better understand this implication, we have created both simplified/tubular and anatomically realistic in vitro endothelial flow models of the human right coronary artery. A post-mortem vascular cast of the human left ventricular outflow tract was used to create geometrically accurate silicone elastomer models. Straight, tubular models were created using a custom made mold. Following the culture of human abdominal aortic endothelial cells within the inner lumen, cells were exposed to steady flow (Re = 233) for varying time periods. The resulting cell morphology was analyzed in terms of shape index and angle of orientation relative to the flow direction. In both models a progressive elongation and alignment of the endothelium in the flow direction was observed following 8, 12, and 24 hours. This change, however, was significantly less pronounced in the anatomical model (as observed from morphological variations indicative of localized flow features). Differences were also observed between the inner and outer walls at the disease-prone proximal region. Since morphological adaptation is a visual indication of endothelial shear stress activation, the use of anatomical models in endothelial genetic and biochemical studies may offer better insight into the disease process.

Figure 1

(A) Simplified straight tubular model (B) anatomically accurate model of a 57 year male with no significant coronary artery disease who died of complications arising from colorectal surgery.

Figure 2

(A) Schematic representation of the experimental flow system. The closed-loop flow system used in this study consisted of the cell culture model(s), individual vented reservoirs, and a low-pulsatility 8-roller peristaltic pump (Ismatec, A-78002-34) coupled using biocompatible peroxide cured silicone tubing. All fittings and tubing were sterilized by autoclaving. (B) Illustration of the morphometric parameters calculated for each endothelial cell.

Figure 3

The approximate location of the proximal RCA regions under study (Region 1 and Region 2) indicated on the anatomical cast; (A) ventral view (C) dorsal view.

Figure 4

Light microscope images of EC morphological changes in the tubular Sylgard™ model. HAAECs were subjected to a steady laminar shear stress of magnitude 22 dynes/cm² for 8, 12, and 24 hours, (B-D) respectively. (A) Represents the no flow control. (Bar = 100 μm, Magnification = 100×). The arrow points in the direction of net flow.

Figure 5

Cell shape index time history for tubular model experiments, static, 8 hrs, 12 hrs and 24 hrs; SI = 1 corresponds to a perfect circle, while SI = 0 corresponds to a line; all values are expressed as mean ± standard deviation (n = 791, 562, 508, 276 respectively).

Figure 6

Histograms illustrating the distribution of cell angles of orientation for the static (no flow) control, and the 8 hour, 12 hour, and 24 hour tubular model flow experiments (n = 791, 562, 508, 276 respectively).

Figure 7

Bar graph illustrating cell shape index time history for both the tubular and anatomical models. The symbol (*) denotes a significant difference in the means; all values are expressed as mean ± standard deviation (n = 566, 562, 298, 344 respectively).

Figure 8

Light microscope images of ECs in anatomical Sylgard™ models following 8, 12, and 24 hours, (B-D) respectively. (A) represents the static (no flow) control. (Bar = 100 μm, Magnification = 100×). The arrow points in the direction of net flow. Certain images locations are blurred due to the local curvature.

Figure 9

Histograms illustrating the distribution of cell angles of orientation for the static, 8 hour, 12 hour, and 24 hour anatomical model flow experiments (n = 566, 562, 298, 344 respectively).

Figure 10

Bar graph illustrating cell shape index variation between the inner and outer walls of the anatomical model in Region 1 and 2. The symbol (*) denotes a significant difference in the means (n = 325, 266, 112, 86 respectively).

Figure 11

Histograms illustrating the distribution of cell angles of orientation for the inner and outer walls of the anatomical model in Region 1 and 2 at the 24 hour time point (n = 325, 266, 112, 86 respectively).