Preservation and structural adaptation of endothelium over experimental foam cell lesions. Quantitative ultrastructural study

Abstract

To assess the extent to which endothelial cell (EC) structure is modified by hyperlipidemia and by the formation of intimal plaques, we undertook a quantitative ultrastructural study of aortic EC of cynomolgus monkeys after 3 or 6 months on an atherogenic diet. We compared EC in lesion-free areas (LFA) with EC overlying focal discrete foam cell accumulations (FDA) or covering multilayered confluent plaques (MCP). There was a 15% increase in cross-sectional lumen surface profile length over FDA or MCP compared to LFA (p < 0.005) corresponding to the bulging contours of immediately underlying foam cells. There was, however, no increase in the number of EC per unit of surface area (26.2 ± 4.47 per 104 mm2 for LFA and 26.0 ± 4.22 for FDA) or, on cross-section, per 100 μm length of underlying internal elastic lamina (8.79 ± 2.42 for LFA, 8.26 ± 2.01 for MCP). Nor did the number of surrounding cells contacted by each cell over LFA or MCP differ from normolipemic controls (6.56 ± 0.85 for LFA and 5.58 ± 0.86 for MCP). Most ECs were markedly attenuated over lesions, and while the extent and complexity of lateral contact regions between adjacent EC was diminished, the number and complexity of basilar projections was greatly increased. These structures extended among the intimal foam cells to insert on the internal elastic lamina or on intimal matrix fibers, resulting in a 2.7-fold increase in the length of the abluminal portion of the EC profile. The perimeter of the transverse EC profiles was thereby increased from 41.4 ± 2.12 μm in LFA to 82.2 ± 5.21 μm over MCP (p < 0.0001). Polarization of EC in the direction of flow diminished as lesions developed. The ratio of length to width, as well as the standard deviation of the ratio, decreased from 3.51 ± 3.92 in LFA to 2.35 ± 0.25 over MCP, due mainly to increases in the proportion of the cell perimeter exposed to the lumen. Lesion localization bore no relationship to the orientation of EC in corresponding locations in the normolipemic controls or in LFA immediately adjacent to plaques. Organelles of EC in hyperlipidemic animals showed features suggestive of increased metabolic activity in all regions, and stress filaments were increased in the EC attenuated over lesions. There was no evidence of EC degeneration, necrosis, or sloughing regardless of lesion location, size, or complexity. We conclude that EC that cover FDA or MCP undergo alterations in thickness, configuration, extent of intercellular junctional overlap, and in the number and length of their abluminal attachment sites, although with no change in cell number. EC degeneration, disruption, or sloughing are not features of exposure to hyperlipidemia or of the initiation or early progression of diet-induced experimental atherosclerosis.