Jan 1, 2016 - ceous cholesteryl ester rich core, or âatherosâ of the atherosclerotic plaque. It is important to recognize that the death and disruption of the ...
Clin. Cardiol. 14, 1-1-16 (1991)
The Pathogenesis of Atherosclerosis: An Overview c. J.
SCHWARTZ, M . D . , A.
J. VALENTE, Ph.D., E. A. SPRAGUE. Ph.D., J . L. KELLEY,Ph.D., *R. M.NEREM,Ph.D
Department of Pathology, University of Texas Health ScienceCenter, San Antonio, Texas; *GeorgiaInstitute of Technology, Atlanta, Georgia, USA
Summary: In this unifying hypothesis directed to the etiology and pathogenesis of atherosclerosis, the importance of focal arterial lesion-prone sites has been emphasized. Key initial participants in these sites include the focal intimal influx and accumulation of low-density lipoprotein (LDL) and a preferential recruitment of blood monocytes. Both are further enhanced in the presence of hyperlipidemia, when the quantity of intimal LDL and the oxidative potential of the intima exceed the capacity of macrophages to remove, via the non-down-regulatingscavenger receptor, cytotoxic anionic (Ox-LDL) macromolecules. Foam cells, pathognomonic of the fatty streak, form during the receptor-mediateduptake of Ox-LDL by the macrophages. Interstitial free radicals and the excess of Ox-LDL particles injure and kill cells, including the foam cells, with the formation of the necrotic extracellular lipid core, a key transitional step in lesion progression. Monocytemacrophage recruitment to the intima is likely to be regulated not only by a multiplicity of endothelial adhesive cytokines, integrins, and selectins, but also by the monocyte-specificchemoattractant, MCP-1, constitutively synthesized and secreted by intimal smooth muscle and endothelial cells. Its synthesis and secretion is augmented by mildly oxidized LDL. Free radicals, pivotal in the oxidation of LDL, and derived from activated macrophages, and also endothelial and smooth muscle cells. Smooth muscle cells migrate from the media through the intimal endothelial layer (IEL) and proliferate under the regulation of a number of mitogens, including plateletderived gmwth factor (PDGF). Collagen synthesis by
smooth muscle cells is substantial. Lymphocytes, as a source of interferons, invade the plaque and are present in the adventitia in substantial numbers, likely representing an autoimmune response in the later stages of plaque development. Platelets and mural thrombosis directly contribute to subsequent plaque growth, particularly after plaque rupture or fissure and disruption of the thromboresistant endothelial cells (EC). Plaque regression in all likelihood involves the conversion of the inert pool of extracellular lipid to a metabolically active intracellular pool and subsequent clearance by the high-density lipoprotein mediated reverse cholesterol transport system. The atherogenic cascades so described conceptually represent arterial inflammatory and healing processes occurring in a hyperlipidemic environment. Many components of pathogenesis are the targets for modulation by genetic, hemodynamic and selected risk factors. The prevention and treatment of the disease should logically target reduction in plasma LDL levels, the inhibition of the oxidative modification of lipoproteins, including LDL, by free radical scavengers, and augmentation of the reverse cholesterol transport system.
Key words: lesion-prone sites, endothelium, macrophages, chemoattractants, oxidized LDL, thrombosis, antioxidants, smooth muscle, growth factors, necrosis, regression
Introduction
Supported in part by NIH Grants HL-26980, HL-07446 and HL-41175. Address for reprints: Colin J . Schwartz, M.D. Department of Pathology The University of Texas Health Science Center 7703 Floyd Curl Drive San Antonio, TX 78284-7750
Atherosclerosis is a disease by no means unique to twentieth century man. Advanced calcific lesions were found in ancient Egyptian mummies, but progress in our comprehending the etiology and pathogenesis of atherosclerosis has been slow and recent. In part a significant problem has been and continues to be the long time frame (years) for plaque development in humans, be it continuous or episodic, the fact that most human data are crosssectional rather than longitudinal, and that many animal models have a greatly accelerated rate of lesion progres-
1-2
Clin. Cardiol. Vol. 14, February 1991
sion (weeks or months). The latter is illustrated dramatically in the cholesterol-fat-fed rabbit, where the plasma and tissue changes are striking in the course of but several weeks. Smaller and more subtle changes occurring over many years could ultimately prove to be biologically of great significance in the development of human atherosclerosis. Yet another constraint to progress is the difficulty inherent in the interpretation of clinical trials or epidemiologic studies which address the influence of risk factors or risk factor modification (e.g., cholesterol lowering) on complex clinical end-points such as myocardial infarction or sudden unexpected cardiac death. A statistical reduction in coronary heart disease risk cannot correctly nor directly be extrapolated to an influence on atherosclerosis alone, as many additional factors such as occlusive thrombosis, thromboembolism, ventricular fibrillation, or pump failure can combine to influence cardiovascular morbidity or mortality. Hence the need for disease-specific studies. From the landmark observations of Anitschkow I and Anitschkow and Chalatow,* implicating cholesterol as the putative etiologic agent in rabbit atherosclerosis, we have been deluged with a plethora of “causative” hypotheses, most of which have focused on but one aspect of pathogenesis to the exclusion of all others. Such unifocal hypotheses have variously emphasized the platelet, thrombosis, injury, cholesterol, lipoprotein filtration, hemodynamics, the smooth muscle cell, and senescence. The need for a hypothesis which integrates all the known sets of information is of paramount importance. We shall take a short step in this direction, at the same time noting some areas where the information base appears inadequate or needs clarification. Another aspect of this complex problem deserves cornment. Neufeld and Goldbourt3 examined the “known” risk factors for coronary artery disease, and concluded that, after all are considered, the residual variance exceeds 50%. Even if the input were to be refined by more recent trial and epidemiologic data, the implications of this observation cannot be ignored. Clearly there may be other risk factors which are as yet unidentified and/or there are additional powerful genetic targets for atherogenesis, determining either individual susceptibility or resistance. How indeed can one quantitate genetic risk, or the roles of mural thrombosis or focal hemodynamic stresses in atherogenesis’? We shall explore the concept of atherosclerosis as an inflammatory disease, holistically the result of interactive cascades among injurious stimuli and the healing responses of the arterial wall occurring in a hyperlipidemic and dyslipoproteinemic environment, in which a multiplicity of inflammatory cytokines, mitogens, chemoattractants, and other humoral factors have modulating roles. The location and the nature and extent of the ensuing lesions are determined in part by focal hemodynamic and genetic influences, respectively.
Initial Events in Atherogenesis It is not surprising that attempts to reconstruct a dynamic sequence in the evolution of an atherosclemtic plaque from the morphology or composition of an advanced lesion have been frustrating. Identification of the earliest events in lesion development is important if we are to understand critical etiologic and pathogenic mechanisms which become obscured as the lesions progress. In experimental animals the focal sites of predilection for either spontaneous or dietary-induced atherosclerosis can be consistently demarcated before plaques become visible macroscopically or microscopically. Such lesionprone or prelesion areas are delineated by their in vivo uptake of the protein-binding azo dye, Evans Blue. The Evans Blue model has been studied extensively and characterized both morphologically and biochemically. Salient features of the lesion-prone (blue) areas include an increased endothelial permeability to and intirnal accumulation of plasma proteins including a l b ~ m i n , ~ . ~ fibrinogen,6 and low-density lipoprotein (LDL-ApoB).’.* Lesion-prone sites when viewed enface (Fig. 1) also exhibit a greater endothelial cell (EC) t u r n ~ v e rIn . ~ addition, the prelesion area EC are relatively more polyhedral or cobblestone relative to nonlesion-prone-areas, and their surface glycocalyx is some 2-5-fold thinner. l o . l l It is of particular interest that blood monocyte recruitment to the intima occurs preferentially in these prelesion (blue) areas and that accumulation of these cells in the subendothelial space of the prelesion sites is preferentially and strikingly accelerated in the presence of dietaryinduced hyperlipidemia. That these areas of in vivo Evans Blue dye uptake are in fact prelesion sites where atherosclerotic plaques subsequently develop either spontaneously or as a result of dietary supplementation has been established by several investigators. McGill et a l l 3 noted that areas of Evans Blue accumulation in the canine aorta delineate sites destined to develop atherosclerotic lesions. Fry14 confirmed these observations in the pig, noting that the focal pattern of dye uptake is “virtually identical to the pattern of sudanophilia in early experimental atherosclerosis. These conclusions are consistent with the observed distribution of spontaneously-occurring aortic atherosclerosis in the aging pig, together with the enhanced accumulation of cholesterol and cholesteryl esters in lesion prone (blue) areas relative to nonlesion-prone (white) sites in the cholesterol-fat-fed pig. Table I summarizes the salient features of the prelesion sites, and in Figure 1, the initial events in the development of atherosclerosis are illustrated schematically. On the basis of all the available data it appears that at least two processes are pivotal to the initiation of the atherogenic cascades: (1) an enhanced focal endothelial transcytosis of plasma proteins including LDL, which accumulate in the widened “edematous” proteoglycan-rich subendothelial space (SES); and (2) the preferential ”
C. J . Schwartz ef af. : Atherogenesis
1-3
FIG. I Schematic illustrating key features of arterial lesion-prone and nonlesion-prone areas as demarcated by the Evans Blue model. Both areas are exposed to similar circulating plasma concentrations of proteins including albumin (yellow), fibrinogen (green), and low-density lipoprotein (LDL) (brown). But as shown, in lesion-prone sites there is a greater influx and intimal accumulation of each. Blood monocytes (blue) are likewise preferentially recruited through the endothelium to the subendothelial space (SES). In lesion-prone areas the EC exhibit an en face cobble-stone morphology relative to nonlesion-prone areas, where the surface glycocalyx is some threefold thicker. Additional features of lesion-prone areas are described in Table I. Lesion-prone areas exist before lesions are visible macroscopically or microscopically. With hvpercholesterolemia and hyperlipidemia these areas preferentially accumulate plasma proteins including LDL and blood monocytes.
recruitment of blood monocytes to the intima, a process that is markedly augmented by even short periods of hyperlipidemia (Fig. 1). Thus the lesion-prone SES has two key participants in atherogenesis, namely the monocyte/macrophage and low-density lipoprotein (LDL). Monocyte recruitment to the intima of lesion-prone areas is presumably due to an enhanced generation of chemoattractants, of which SMC-CF (MCP-I), a 14 kD monomeric cationic peptide, synthesized and secreted by both arterial smooth muscle cells (SMC) and endothelial cells (EC), is of particular interest. Furthermore, the production of MCP- 1 by both SMC and EC is stimulated by minimally modified (oxidized) LDL,*O while oxida-
tively modified LDL (Ox-LDL) is itself chemotactic.21 Cells participating in the generation of reactive oxygen species through which the oxidative modification of the LDL may be mediated include the macrophage, EC, and SMC * How do the LDL particles enter the arterial SES? While a number of mechanisms may contribute to the endothelial transcytosis of LDL, on the basis of earlier particulate probe studies with native ferritin (diameter 11 nm), the in vivo transendothelial movement of the ferritin particles in lesion-prone sites was observed to be via pinocytotic vesicles,22a mechanism that in all likelihood accounts substantially for the influx of LDL (diameter 22 nm). These
TABLEI Salient structural and functional characteristics of arterial lesion-prone sites
Lesion-prone areas Nonlesion-prone areas
[13lI]-
(13111-
Albumin (re1 %)
Fibrinogen (re1 %)
LDL (Apo-B)
166.4
164.3 100
++ +
100
Intimal cholesterol accumulation (re1 %)
200 100
EC glycocalyx thickness
Intimal monocyte recruitment
EC turnover (re1 %)
SES thickness
+ +++
++
137 100
++++ +
+
Clin. Cardiol. Vol. 14, February 1991
1-4
studies established that the enhanced endothelial transcytosis in lesion-prone sites is not attributable to a greater number of vesicles, nor the ferritin-carrying capacity of each vesicle, but rather to the number of vesicles actively involved in the transport process (Table 11). It is possible, therefore, that the activation of pinocytotic vesicles for transport is subject to regulatory control. Whether the LDL (B/E) receptor plays any role in the transcytosis of LDL has yet to be clarified. It should also be noted that the enzymatic (neuraminidase) modifications to the cell surface can markedly enhance the uptake of LDL by endothelial cells,23a finding consistent with the observation of a thinner sialo-protein containing glycocalyx in lesion prone areas. I I Yet another route possibly accounting for the greater intimal influx of LDL in lesion-prone sites has been proposed by Lin et ul. 24 who have demonstrated an association between EC LDL leakage and EC turnover. Lesion-prone (blue) sites exhibit a greater EC turnover (Table 1). It is important to note that focal increases in arterial LDL concentrations precede fatty streak development in cholestrol-fed rabbits.2sand that a selective retention of LDL may contribute to the accumulation of LDL in the SES.26 It is also of interest that noradrenaline at physiologic concentrations can increase the arterial flux of LDL by a receptor-independent pathway.
Foam Cell Formation Thus far we have set the pathobiologic stage for lesion progression, emphasizing that in lesion-prone sites the combination of an enhanced blood monocyte recruitment to the SES and the spontaneously greater endothelial transcytosis of LDL to the arterial intima are pivotal initiating events. Both are markedly augmented in the presence of hypercholesterolemia and hyperlipidemia. A schematic of arterial foam cell formation is presented in Figure 2 , where the basic pathobiologic processes of the lesion-prone area are accelerated. Blood monocytes attach to the morphologically intact endothelium (EC), which is presumably activated, producing a spectrum of
TABLE I1 Endothelial transport of native fenitin in lesion-prone and nonlesion-prone areas of the pig aorta % Vesicles Number of containing Number of ferritin vesicles/mm* ferritin particleslvesicle
Lesion-prone areas Nonleaionprone areas
26.6 (NS)
18.6"
1.8 (NS)
27.8
9.0
I .4
"p
