Identifying key factors in the development of atherosclerosis
A conspicuous feature of atherosclerosis is its highly non-uniform distribution within arteries. This characteristic implies the existence of powerful local risk factors.
The current consensus is that these “localising factors” include:
- Low, oscillatory levels of haemodynamic shear stress at the blood:wall interface (the frictional force exerted on each area of the artery wall by the flow of blood);
- Elevated permeability of the endothelium to circulating macromolecules such as low density lipoprotein (LDL, the major cholesterol carrier in blood).
However, the evidence for these views is mutually contradictory. The low shear stress hypothesis of Caro et al is based on the observation that areas downstream of the flow divider of aortic side branches are spared of disease; these regions are expected to experience high wall shear stress. The insudation theory (or lipid hypothesis) of Anitschkow is based on the observation that experimental atherosclerosis in rabbits develops preferentially in such areas, and that these regions show elevated uptake of circulating macromolecules. There is thus a fundamental contradiction at the heart of our understanding of atherogenesis.
We propose that this apparent contradiction can be resolved by taking age into account. Essentially we argue that the pattern of disease changes with age, and that confusion has arisen from comparing young animals with mature people. The argument is presented briefly below and reviewed in greater detail here.
Arterial lesions change location with age
The distribution of lesions in human arteries changes with age (see publication): lipid deposition occurs downstream of side branches in the aorta of human fetuses, neonates and infants, in young adults it occurs at the lateral margins of the same ostia, and at later ages it occurs upstream of the ostia.
Elevated uptake of circulating macromolecules by the arterial wall changes location with age
Net uptake of albumin is greater downstream than upstream of side branches in aortas of immature animals but shows the opposite pattern in mature animals (see publication). Furthermore, short-term uptake of albumin, reflecting the rate of influx into the wall (and most likely depending on endothelial permeability), also follows these patterns (see publication). When mapped around the branch, rather than just upstream and downstream of it, influx shows a remarkable spatial correlation with the pattern of lesions at each age (see publication):
Thus when age is taken into account, the patterns of lesions in rabbits and people are similar and appear to be explained by the insudation theory:
Why do transport patterns change with age?
Studies in perfused vessels and in vivo have shown that the mature but not the immature pattern of transport can be reversed by inhibiting nitric oxide synthesis (see publication). Hence the switch is related to a change in the way in which NO is synthesised or the way in which it is coupled to transport pathways.
Curiously, however, the pattern of lesions is not modified by chronic administration of L-arginine, the substrate for NO production (see publication) or by inhibitors of NO synthesis (Staughton, T.J, and Weinberg, P.D. (2004) In: “Trends in Atherosclerosis Research” (L. V. Clark, Ed.), Nova Biomedical Books, NY).
The mature pattern can also be reversed by modifying shear stress around the branch mouth link (see publication). This is consistent with the NO-dependence because flow is the major stimulus of baseline NO synthesis.
We are currently developing non-invasive methods for assessing arterial nitric oxide bioactivity – for more information click here.
We are also investigating the effects of shear on endothelial permeability and morphology by culturing endothelial cell monolayers – for more information click here.
Additionally, we are looking at the effects of shear on transport using modelling techniques, since several key properties – local inhomogeneities in plasma macromolecule concentration due to concentration polarisation, different rates of convection towards the well between regions near and away from inter-endothelial cell junctions, and the modifying effects of the endothelial glycocalyx layer – cannot currently be measured (see publication)(see publication).
Do flow patterns near arterial branches change with age?
An unanswered question is whether the change in transport and lesion patterns with age reflects a change in the wall response to flow (as suggested by the changing role of nitric oxide) or whether the flow pattern itself changes. Unfortunately, it is not currently possible to measure flow near the wall around branches under physiological conditions to answer this directly. Surrogate indices of shear stress patterns or numerical models have to be used. One surrogate index is the morphology of the endothelial cells lining the wall. These cells, and their nuclei, elongate with increasing shear stress and align with the flow. Patterns of endothelial nuclear shape around rabbit aortic branches do change with age, suggesting that patterns of shear depend on age (see publication); the patterns suggest that lesions correlate with high shear (see publication).
Numerical simulations of flow are challenging because arterial flow is so complex (see videos courtesy of Dr. P. Vincent – Aeronautics, Imperial College). Also, the geometrical definitions have to be very precise (see publication), and a number of boundary conditions that significantly affect patterns of wall shear stress – such as the proportion of aortic flow entering the branch (see publication) or the phasing of the flow between the aorta and the side branch (see publication) – are incompletely characterised.
Thus it is currently unclear whether shear stress patterns around branches do change with age (as indicated by endothelial cell morphology) or do not (as suggested by computational studies).
A new technique that might help resolve this uncertainty is ultrasound particle image velocimetry (“echo PIV”). The scattering patterns produced by blood-borne particles are compared between two consecutive B-mode images in order to estimate local velocities. Preliminary data have been obtained in the rabbit aorta (see publication).