Quest for Fire
Seeking the Source of Pathogenic Oxygen Radicals in Atherosclerosis
Abstract—There is an accumulating body evidence that atherosclerosis is either caused by or accompanied by oxidative events in the vessel wall. These oxidative events have been implicated in proatherogenic modification of proteins, alteration of gene expression, promotion of inflammation, remodeling of vessels, and perturbations of vascular tone. This body of literature has led to a dogma that oxidation is a prerequisite for the atherosclerotic process. In particular, oxidation of lipoproteins by activated macrophages in the subintimal space has been postulated to be an important early step in the atherosclerotic process.
In macrophages, the predominant source of reactive oxygen species (ROS) is the NADPH oxidase. This enzyme system was first characterized in the neutrophil and is composed of at least 5 subunits: 2 cytosolic components, p47phox and p67phox; 2 membrane-bound components, a small, nonglycosylated p22phox and a larger, glycosylated gp91phox (which together compose the cytochrome b558); and a low-molecular-weight G protein, rac-1.1 Numerous mutations in these subunits have been described in humans with chronic granulomatous disease (CGD).2 Mice lacking gp91phox have been created and have a phenotype characteristic of CGD. Because this enzyme system is such a potent source of ROS, it has been strongly suspected to be involved in lipid oxidation and the initiation of atherosclerosis.
Kirk et al,3 in the current issue of Arteriosclerosis, Thrombosis, and Vascular Biology, have provided us with an extraordinarily important study addressing the relative importance of the NADPH oxidase and the role of macrophage-derived superoxide in the atherosclerotic process. These investigators examined the development of atherosclerotic lesions in 2 different experimental settings. In one, both control mice and mice lacking gp91phox were fed atherogenic diets. In these animals, the high-fat diet resulted in similar increases in plasma lipids, and the development of atherosclerotic lesions was identical between the control and the gp91phox-deficient mice. A genetic cross between the gp91phox knockouts and apoE knockouts produced offspring with a small but nonsignificant decrease in plasma lipid levels versus gp91phox–apoE(+/+) controls but again, showed no change in atherosclerotic lesion size.
The article by Kirk et al3 represents a significant advance in our understanding of atherosclerosis. The study clearly shows that the macrophage NADPH oxidase is not critical in the development of atherosclerosis. Does this mean that ROS do not have a role in atherosclerosis? We cannot push our conclusions that far. The data prove that superoxide generated from a gp91phox-containing oxidase is not important in atherosclerosis, but they do not exclude a role for other sources of radicals. Importantly, the authors did not show that vessels from the gp91phox-deficient animals indeed had reduced superoxide production. Quantification of vascular superoxide production by using techniques such as lucigenin-enhanced chemiluminescence or electron spin resonance would have provided invaluable information on whether the gp91phox knockout or the double knockout altered vascular superoxide production. Measurements of parameters of lipid peroxidation, although difficult in these small animals, would have also been helpful. Such parameters include measurements of conjugated dienes, thiobarbituric acid–reactive substances, antibodies against oxidized lipids, or oxidized protein epitopes in either the blood or the vessel wall. In the absence of these measurements, it is impossible to conclude that oxidative processes can be excluded as having a role in atherosclerosis.
Furthermore, the study by Kirk et al3 does not definitively exclude an NADPH-like enzyme as having a role in atherosclerosis. Recently, it has become evident that nonphagocytic cells in the vessel wall, including endothelial cells, vascular smooth muscle cells, and cells of the adventitia, can produce ROS and seem to contain enzyme systems similar to the NADPH oxidase of neutrophils.4 These “vascular oxidases” may utilize different subunits than do the neutrophil oxidases. In isolated endothelial cells, reverse transcriptase–polymerase chain reaction has demonstrated the expression of p22phox, gp91phox, p47phox, and p67phox, but Northern blot analysis was able to detect significant amounts of only p22phox mRNA. Immunoperoxidase staining revealed the presence of p47phox and p67phox protein. Although no antibodies were available to detect the 2 membrane-bound components, heme spectroscopy was unable to detect a cytochrome b558 signal.5 In vascular smooth muscle, p22phox has been cloned and is quite abundant.6 However, there is no evidence to suggest that gp91phox is present. Recently, a protein from smooth muscle cells has been cloned that shares some homology with the neutrophil gp91phox but is nonglycosylated. This protein, termed MOX-1, functions in both ROS generation and cell growth.7 It is suspected that MOX-1 may play the role of gp91phox in vascular smooth muscle cells. Importantly, MOX-1 would be present in the gp91phox-deficient mice, and their vascular smooth muscle cells could continue to produce large amounts of ROS. In this regard, Miller et al8 have shown that medial cells are an important source of superoxide in atherosclerosis.
Related to the aforementioned considerations, there is no reason to assume that only 1 cell type or 1 enzyme source of radicals is important in the oxidation of lipids. Several different reactive oxygen intermediates can be shown in vitro to initiate lipid oxidation and peroxidation. These include superoxide, hydrogen peroxide, peroxynitrite, hypchlorous acid, and the hydroxyl radical. In the case of severe hypercholesterolemia, as produced in the cholesterol-fed mice or the apoE-deficient mice studied by Kirk et al,3 it is quite likely that many different ROS, derived from many different sources, can react with the abundant lipoproteins that accumulate in the vessel wall. Alternatively, lipoxygenase enzymes react with esterified fatty acids to directly form the lipid alkoxyl radical (LO·). This lipid radical can react with molecular oxygen to form a lipid alkylperoxyl radical (LOO·). In this manner, lipoxygenase may initiate lipid peroxidation without forming superoxide or any of superoxide’s reactive products. Of substantial importance, a 12-lipoxygenase–deficient mouse has been created and crossed with the apoE-deficient mouse. Mice lacking both the apoE gene and the lipoxygenase gene had a dramatic reduction in atherosclerotic lesions compared with mice lacking the apoE gene only.9 Thus, it is possible that in the normal situation, superoxide (and other radicals) derived from the phagocytic NADPH oxidase contributes to lipid oxidation, but when this enzyme is absent, other sources are perfectly capable of filling the bill.
In the previous several paragraphs, we have offered several alternative explanations for the lack of an effect of the gp91phox knockout on atherosclerosis development. As discussed, 1 viable alternative is that lipoxygenase enzymes or other enzymes are far more important in lipid peroxidation than is the NADPH oxidase. Another lesson that has repeatedly been hammered home from studies of knockout mice is that substantial redundancy exists in biology, and deletion of 1 enzymatic pathway often leads to compensation by 1 or more alternative pathways. Lipoprotein oxidation is critically important in allowing macrophages and other cells to remove lipids from the interstitial space, as modified lipoproteins are more readily cleared by scavenger receptors (which recognize the modified forms) than by the native LDL receptor. Thus, lipoprotein oxidation is a normal adaptive process, and it would not be surprising to find that this occurs via several pathways.
On a lighter note, the burning of carbon-based fuel is an oxidation reaction. In 1982, an excellent movie was released, Quest for Fire, directed by Jean-Jacques Annaud. This film depicted a prehistoric odyssey of early man, searching far and wide for fire, which was clearly important for survival. Today, we are continuing to look for little fires, in the form of oxidation reactions occurring in the vessel wall, with the ultimate goal of stamping them out. This quest is made difficult by the elusive nature of their sources and by the likelihood that several sources exist. Hopefully, we will be as successful as the protohumans in Quest for Fire, who eventually found their fire, learned to control it, and evolved to our present state.
Rotrosen D, Yeung CL, Leto TL, Malech HL, Kwong CH. Cytochrome b558: the flavin-binding component of the phagocyte NADPH oxidase. Science. 1992;256:1459–1462.
Dinauer MC, Pierce EA, Bruns GAP, Curnutte JT, Orkin SH. Human neutrophil cytochrome b light chain (p22-phox): gene structure, chromosomal location, and mutations in cytochrome-negative autosomal recessive chronic granulomatous disease. J Clin Invest. 1990;86:1729–1737.
Kirk EA, Dinauer MC, Rosen H, Chait A, Heinecke JW, LeBoeuf RC. Impaired superoxide production due to a deficiency in NADPH oxidase fails to inhibit atherosclerosis in mice. Arterioscler Thromb Vasc Biol. 2000;20:1529–1535.
Rajagopalan S, Kurz S, Munzel T, Tarpey M, Freeman BA, Griendling KK, Harrison DG. Angiotensin II-mediated hypertension in the rat increases vascular superoxide production via membrane NADH/NADPH oxidase activation: contribution to alterations of vasomotor tone. J Clin Invest. 1996;97:1916–1923.
Jones SA, O’Donnell VB, Wood JD, Broughton JP. Hughes EJ, Jones OTG. Expression of phagocyte NADPH oxidase components in human endothelial cells. Am J Physiol. 1996;271(Heart Circ Physiol. 40):H1626–H1634.
Miller FJ Jr, Gutterman DD, Rios CD, Heistad DD, Davidson BL. Superoxide production in vascular smooth muscle contributes to oxidative stress and impaired relaxation in atherosclerosis. Circ Res. 1998;82:1298–1305.