Cigarette smoking, endothelial injury and cardiovascular disease

Abstract

Despite the fact that the epidemiological evidence linking cigarette smoking with cardiovascular disease is overwhelming, the precise components of cigarette smoke responsible for this relationship and the mechanisms by which they exert their effect have not yet been elucidated. There are however, some promising pointers as a result of recent developments and this review concentrates on new evidence since earlier reviews of this topic. It is now known that the endothelium has a vastly more important role than was ever thought to be the case a decade ago. Its role in health and disease is increasingly understood, as is the relationship between endothelial injury and the development of atherosclerosis. There is considerable evidence that cigarette smoking can result in both morphological and biochemical disturbances to the endothelium both in vivo and in cell culture systems. Cigarette smoke is a complex mixture and only a few components have been extensively studied. Nicotine and carbon monoxide are much less damaging than is whole smoke. However the free radical components of cigarette smoke have been shown to cause damage in model systems. Further work will be necessary to consolidate the evidence base but the data reported in this review suggest that the free radical components of cigarette smoke may be responsible for the morphological and functional damage to endothelium that has been observed in model systems.

Epidemiological evidence

Since the first cohort studies indicating that death from coronary heart disease is more common in smokers than nonsmokers were published in the 1950s (Doll & Hill 1956; Hammond & Horn 1958), there have been many reports which convincingly demonstrate that cardiovascular disease is the most important cause of smoking-related premature death (Fowler 1993). The prospective investigation of British Doctors commenced in 1951 showed that half the excess mortality from smoking results from cardiovascular disease with approximately one third due solely to coronary heart disease (Doll & Peto 1976). Prospective studies on smokers have shown that the risk of death from coronary heart disease is roughly doubled for both men and women (Fowler 1993) although there is evidence that the risk is greater for women (Vriz et al. 1997). Doll (1983) estimated that cigarette smoking accounted for 25% of the overall mortality from ischaemic heart disease. More recently in the United States, Manley (1997) has estimated that one fifth of all heart disease-related deaths are due to cigarette smoking and that smoking, as a single factor, doubles the risk of heart failure. Cigarette smokers are also one and a half times more at risk of stroke than nonsmokers (Fowler 1993). Cigarette smoking is an independent risk factor in the development of atherosclerotic lesions in the internal pudendal and common penile arteries of young impotent men (Rosen et al. 1991). It is a significant predictor of the presence of intracranial internal carotid artery atherosclerosis (Ingall et al. 1991). In the Edinburgh Artery Study, cigarette smoking was shown to have a direct effect on the risk of aortic aneurysm which was independent of atherosclerosis (Lee et al. 1997). This study also showed that smoking increases the risk of peripheral artery disease more than heart disease (Fowkes et al. 1992). Perhaps not surprisingly, cigarette smokers who regularly undertake physical activity have a reduced cardiovascular mortality rate than those who have a sedentary existence (Hedblad et al. 1997). Increased total exposure to tobacco in humans and a smoking pattern which maximizes nicotine yield are also associated with an increased risk of peripheral arterial disease (Powell et al. 1997).

Endothelial injury

Despite the strong epidemiological evidence that exists linking cigarette smoking and cardiovascular disease, the mechanisms by which cigarette smoking causes disease and the components of smoke responsible remain poorly understood. Endothelial injury is considered to be a key initiating event in the pathogenesis of atherosclerosis (Ross 1986,1993) and it has therefore seemed reasonable to hypothesize that cigarette smoke, or some components of it, may exert its effects by damaging the endothelium. Components of smoke that gain access to the circulation will come in contact with blood and with the vascular endothelial cells that form a monolayer lining the vessels. These cells are now known to have a crucial role in controlling the blood circulation and to be highly active metabolically (Gryglewski et al. 1988; Vane & Botting 1995). Even minor disturbances to their normal functioning could have significant implications for the initiation and development of atherosclerosis. Endothelial functions, including an increase in permeability and decreased nitric oxide (NO) production along with increased expression of adhesion molecules and adherence of leukocytes to the vessel wall, have been shown to be impaired by risk factors for cardiovascular disease such as hypertension, hyperlipidaemia and hyperglycaemia (Haller 1997). This review examines the evidence that cigarette smoking can bring about vessel wall damage with particular reference to the vascular endothelium.

Damaging effects of cigarette smoking on human arteries

Direct evidence that cigarette smoking can result in atherosclerosis continues to be obtained from human investigations. Using serial quantitative coronary arteriography, Waters et al. (1996) have shown that cigarette smoking accelerates coronary progression and new lesion formation. Interestingly, Lovastatin was shown to slow the progression of coronary atherosclerosis and prevent the development of new coronary lesions in smokers (Waters et al. 1996). The relative thickness of the intima and media of an artery as measured by ultrasound can be used to provide an index of atherosclerosis. Carotid artery atherosclerosis determined by ultrasound determination of plaque thickness, a measure of atherosclerosis, is accelerated depending on the level of cigarette use and this is independent of age, hypertension and diabetes (Dempsey & Moore 1992). Cigarette smoking has an adverse effect on the intimal medial ratio and stenosis of the carotid artery as measured by ultrasound (Howard et al. 1994; Tell et al. 1994; Diezroux et al. 1995). Exposure to environmental tobacco smoke (passive smoking), and cigarette smoking have been shown using the intimal/medial ratio as an index to cause progression of atherosclerosis in the carotid artery (Howard et al. 1998). Cigarette smoking was shown to be an independent indicator of the severity of coronary and thoracic aortic atherosclerosis in men receiving coronary angiography and transoesophageal echocardiography (Inoue et al. 1995). In uraemic patients on chronic dialysis cigarette smoking is associated with carotid atherosclerosis as revealed using echo-colour-Doppler evaluation (Malatino et al. 1999). Studies on the coronary arteries of trauma victims aged between 15 and 34 years of age suggest that intermediate atherosclerotic lesions progress rapidly into advanced lesions in smokers and that intima formerly having early lesions is replaced by intima with raised lesions (Zieske et al. 1999). This study also demonstrated that smoking was associated with more extensive fatty streaks and raised lesions in the abdominal aorta (McGill et al. 1997). Radiological observations on women aged between 45 and 64 years showed a direct association between the development of calcified deposits, which have been shown to represent atherosclerotic change, in the abdominal aorta and the number of cigarettes smoked per day over a nine year period (Witteman et al. 1993). This study suggests that the rate of atherosclerotic change may be reduced by the cessation of smoking, but a residual effect appears to be present for a decade (Witteman et al. 1993). Using ultrasonography to look at carotid intima and media thickness, the evidence indicates that atherosclerosis is mediated by components of tobacco smoke other than nicotine (Bolinder et al. 1997).

Cigarette smoking and endothelial modification

Direct evidence that cigarette smoking could result in endothelial injury was first obtained from morphological observations on the umbilical arteries from smoking mothers. These reports, along with others on the effects of cigarette smoking on endothelial morphology, have been previously reviewed by Pittilo (1988,1990). The morphological alterations, which included the endothelium having an irregular appearance along with membrane disturbances evidenced by the formation of blebs or microvillous-like projections, were also associated with functional changes to the endothelium. In laboratory rats exposed to cigarette smoke Pittilo et al. (1982) were able to demonstrate that the morphological alterations to the endothelium of the thoracic aorta were accompanied by reduced endothelial prostacyclin production as well as the adhesion of platelets to apparently intact cells. Figures 1 – 7 illustrate the morphological changes associated with cigarette smoking. The ability of cigarette smoking to reduce endothelial prostacyclin production is well established and the data have been previously reviewed (Pittilo 1990; Woolf et al. 1993).

Figure 1.

Scanning electron micrograph of control upper thoracic rat aorta. The endothelial cells have a flattened appearance and the borders can just be discerned. The cell layer is intact with no exposure of subendothelial collagen —× 1200. Note: Figures 1 – 7 are scanning electron micrographs of rat aorta that have been perfusion fixed. This type of fixation preserves the vasculature from artefacts such as shrinkage and contraction.

Figure 7.

Scanning electron micrograph of part of the rat abdominal aorta seen in Figure 6. At this magnification the endothelial cells appear extremely irregular and a very large number of white blood cells can be seen adhering to the endothelium —× 1200.

Cigarette smoking and the components of smoke responsible for endothelial injury

Cigarette smoke is a complex mixture and only a few components have been examined in isolation to assess their effects on endothelial morphology and function. The effects of nicotine on both the morphology and prostacyclin production of rabbit and cultured human endothelial cells has been reviewed by Bull (1988),Pittilo (1990) and Woolf et al. (1993), and although deleterious effects result from exposure, these are much more limited than those observed with whole smoke (Bull et al. 1985,1988; Pittilo et al. 1990). Clinical studies support these findings in that observations on pipe smokers and users of transdermal nicotine devices suggest that components of smoke other than nicotine are the most important causes of acute cardiovascular events (Benowitz 1997). The effects of carbon monoxide exposure have also been reviewed (Pittilo 1988,1990) and are conflicting, with some reports providing evidence of morphological or functional alterations and others reporting no effects.

Rats exposed to nicotine show an increased frequency of endothelial cell death which results in enhanced transendothelial leakage of macromolecules including low density lipoproteins (Lin et al. 1992). Nicotine has been shown to enhance the release of platelet-derived growth factor by bovine aortic smooth muscle cells and is associated with alterations in the cytoskeleton (Cucina et al. 2000). Nicotine significantly stimulates DNA synthesis and endothelial cell proliferation in bovine pulmonary artery endothelial cells at concentrations lower than those obtained in the blood after smoking (Villablanca 1998). In rabbits, administration of nicotine is associated with an acceleration of intimal hyperplasia following endothelial removal (Hamasaki et al. 1997).

Data derived from epidemiological and animal studies suggest that carbon monoxide is not atherogenic (Smith & Steichen 1993). However, exposure of cockerels to carbon monoxide does not have a discernible effect upon arteriosclerotic plaque development (Penn et al. 1992). Bovine pulmonary artery endothelial cells exposed to carbon monoxide release higher concentrations of nitric oxide and it has been suggested that carbon monoxide causes oxidative stress through competition for intracellular binding sites which increase steady state levels of nitric oxide and the generation of peroxynitrite by the endothelium (Thom et al. 1997). Carbon monoxide suppresses the production of endothelin-1 and platelet-derived growth factor B by endothelial cells (Kourembanas et al. 1997).

There is experimental evidence with cockerels to indicate that exposure to environmental tobacco smoke at levels equivalent to those routinely encountered by people in smoke-filled environments is sufficient to promote arteriosclerotic plaque development (Penn & Snyder 1993; Penn et al. 1994). In cholesterol-fed rabbits, passive smoking increases aortic and pulmonary artery atherosclerosis in a dose-responsive way and is independent of changes in serum lipids (Zhu et al. 1993; Sun et al. 1994). One review of the literature concluded that passive smoking increases platelet activity, accelerates atherosclerotic lesion formation, and increases tissue damage following ischaemia or myocardial infarction (GLantz & Parmley 1995). However, it has been argued that the available experimental and epidemiological evidence to date does not support an association between environmental tobacco smoke exposure and an increased risk of heart disease (Denson 1999; Lee & Roe 1999).

Woolf et al. (1993) and Pittilo & Woolf (1993,1994) focused on the possible importance of free radicals as the mediator of cigarette smoke-induced endothelial damage for a number of reasons. Cigarette smoke has been known for many years to be a rich source of free radicals (Lyons et al. 1958). The scanning electron microscope changes seen in laboratory rats following smoke exposure (Pittilo et al. 1982) (Figs 1,2,3,4,5,6,7), or cultured cells following exposure to plasma from humans after smoking (Pittilo et al. 1984), were similar to the changes seen in cardiac myocytes following free radical-induced lipid peroxidation (Noronha-Dutra & Steen 1982,1984). Finally, there is a strong association between lipid peroxidation and endothelial cell injury (Hennig & Chow 1988). Woolf et al. (1993) and Pittilo & Woolf (1994) reviewed some of the supportive evidence and considered that free radical-induced lipid peroxidation might eventually be shown to be the mechanism responsible for the epidemiological relationship between cigarette smoking and atherosclerosis (Pittilo & Woolf 1994). It has been suggested that the high levels of lipid peroxidation and increased formation of reactive oxygen species within the vascular wall in atherosclerosis can overwhelm cellular antioxidant defence mechanisms and human oxidatively modified LDL-induced expression of antioxidant stress proteins in vascular cells (Siow et al. 1999).

Figure 2.

Scanning electron micrograph from control thoracic aorta from an animal that has been sham-smoked. The endothelium is regular and a white blood cell can be seen in contact with the endothelial surface. The movement of white blood cells across the endothelium is a normal feature but they are typically only seen in small numbers —× 1200.

Figure 3.

Scanning electron micrograph of the thoracic aorta from a rat exposed to cigarette smoke. The endothelium is irregular and some minor projections can be seen protruding from the cells. These morphological alterations are accompanied by biochemical disturbances —× 1200.

Figure 4.

Scanning electron micrograph of the abdominal aorta from a rat exposed to cigarette smoke. There is some minor irregularity of the cells but platelets can be seen adhering to what may be functionally abnormal, but nevertheless morphologically intact, endothelium —× 1200.

Figure 5.

Scanning electron micrograph of rat thoracic aorta following exposure to cigarette smoke. Both a white cell and platelets can be seen adhering to morphologically abnormal endothelium —× 1400.

Figure 6.

Scanning electron micrograph of the upper abdominal aorta from a rat exposed to cigarette smoke. At lower magnification, the endothelium can be seen to be extremely irregular with a suggestion of apertures between some raised endothelial cells. Adhering to the surface of the blood vessel are large numbers of leucocytes. Adhesion of leucocytes to the endothelium is increased following cigarette smoking —× 500

On the contrary, studies showing no differences in free radical markers between smokers and nonsmokers suggest that any free radical activity generated from smoking is adequately scavenged. No significant differences in free radical markers were found between young adult smokers and nonsmoking volunteers (Leonard et al. 1995a). Free radical activity due to cigarette smoking also appears to be adequately scavenged in young adults with diabetes who are free of significant macrovascular disease (Leonard et al. 1995b). Furthermore, there is evidence to indicate that chemical modification of glutathione is a major damage mechanism of filtered cigarette smoke with free radical oxidations being less significant (Maranzana & Mehlhorn 1998).

Despite these findings, there remains strong evidence to implicate free radicals as an important cause of cigarette smoke-induced endothelial injury. The morphological alterations seen with cultured cells exposed to plasma obtained from volunteers after smoking (Pittilo et al. 1984) are associated with functional changes to the endothelium including activation of the hexose-monophosphate shunt, a sharp increase in the total glutathione content of the culture medium, release of angiotensin-converting enzyme from the cells and a decrease in the ability of the cells to produce ATP, all of which indicate both oxidative stress and cell injury (Noronha-Dutra et al. 1993a). Smoking is also a major determinant of increased plasma-free radical activity in dyspeptic subjects (Phull et al. 1998).

Free radicals and cigarette smoke

Cigarette smoke consists of two distinct populations of free radicals; the principal free radical in the tar phase is a quinone/hydroquinone (Q/QH₂) complex which is an active redox system that can reduce molecular oxygen to produce superoxide, and in the gas phase small oxygen-centred and carbon-centred radicals produced in a steady state by the oxidation of nitric oxide (NO) to nitrogen dioxide (NO₂) react with reactive species in smoke such as isoprene (Church & Pryor 1985; Pryor et al. 1990). The radicals associated with the particulate or tar phase of cigarette smoke are long lived whereas the radicals associated with the gas or vapour phase have structures that predict a very short lifespan, although spin trap experiments demonstrate that these radicals may exist for periods in excess of five minutes (Church & Pryor 1985; Pryor et al. 1990). Aqueous extracts of cigarette tar auto-oxidize to produce semiquinone, hydroxyl, and superoxide radicals in air-saturated buffered aqueous solutions (Zang et al. 1995).

Direct-electron resonance measurements suggest that superoxide and hydroxl radicals are produced during the auto-oxidation of hydroquinone and catechol-related species in aqueous extracts of cigarette tar (Zang et al. 1995). Fresh cigarette smoke contains from 300 to 500 parts per million of nitric oxide (Cueto & Pryor 1994). The nitric oxide radical is produced slowly from nitric oxide radical donors such as amine complexes, peroxinitrite and other reactants including nitrogen oxides (Shinagawa et al. 1998). Nitric oxide can react with hydrogen peroxide to produce singlet oxygen which may also be important in mediating cigarette smoke-induced endothelial damage (Noronha-Dutra et al. 1993b). Certainly there is evidence to suggest that nitric oxide in cigarette smoke is important as a mediator of oxidative damage (Epperlein et al. 1996) and that DNA damage by cigarette smoke may invoke reactive nitrogen species (Spencer et al. 1995) as well as reactive oxygen species (Liu et al. 1999). Analysis of cigarette smoke demonstrates that the main source of oxygen and hydrogen peroxide results from polyphenols in the particulate phase, and a synergistic effect is observed between these polyphenols and nicotine (Kodama et al. 1997). The vapour phase contains a factor which produces reactive oxygen metabolites from hydrogen peroxide (Kodama et al. 1997). Free radicals in the gas phase of cigarette smoke, identified using electron spin resonance spectrometry, were identified as mainly alkoxyl and alkyl free radicals with the latter making up approximately two thirds of the total spectral components (Zhao et al. 1991).

Haemoglobin-impregnated conventional cigarette filters are capable of withholding nitric oxide, nitrogen oxides, hydrogen peroxide, carbon monoxide, aldehydes, trace elements and carcinogenic nitrosocompounds from cigarette smoke (Deliconstantinos et al. 1994). This is interesting in view of the fact that the red blood cells of cigarette smokers contain more glutathione than those of nonsmokers and are more protective to endothelial cells in culture from hydrogen peroxide-mediated damage than are the red cells from nonsmokers (Toth et al. 1986).

Cigarette smoking and endothelial dysfunction

More recent studies that have not been the subject of previous reviews continue to provide evidence that cigarette smoke can alter vascular endothelium not only morphologically but functionally. In recent years, the discovery of nitric oxide has radically altered our understanding of vascular control (Vallance 1997). Both nitric oxide and carbon monoxide generated within the blood vessel wall are important cellular messengers involved in the regulation of vascular smooth muscle tone. (Siow et al. 1999). Carbon monoxide generated through haeme oxygenase inhibits mitogen-induced proliferation of vascular smooth muscle cells and there is evidence that endogenous carbon monoxide serves as a protective factor limiting the excessive vascular smooth muscle cell proliferation associated with vascular disease (Togane et al. 2000).

Cigarette smoking is associated with dose-related and potentially reversible impairment of endothelium-dependent arterial dilation in asymptomatic young adults, consistent with endothelial dysfunction (Celermajer et al. 1993). It augments endothelial-derived vaso relaxation but has no effect on endothelium-independent vaso relaxation (Nene et al. 1997). It leads to a significant decrease in endothelium-dependent dilatation of the brachial artery (Lekakis et al. 1997). It causes immediate constriction of proxymal and distal epicardial coronary arteries and an increase in coronary vessel tone despite an increase in myocardial oxygen demand (Quillen et al. 1993), and cigarette smokers show an impairment in basal but not stimulated nitric oxide mediated vaso dilation (McVeigh et al. 1996).

Cigarette smoking has been shown to affect endothelial nitric oxide synthase activity and protein levels (Wang et al. 2000). It results in a decrease in exhaled nitric oxide in humans (Verleden et al. 1999) suggesting that it may inhibit the enzyme nitric oxide synthase (Kharitonov et al. 1995). Cigarette smoke extracts cause an irreversible inhibition of nitric oxide synthase activity in pulmonary artery endothelial cells (Su et al. 1998) and a reduction in constitutive nitric oxide synthase activity in the rat in a dose-dependent manner (Ma et al. 1999a). Conversely, other studies in the rat have shown that cigarette smoking results in an increase in nitric oxide synthase gene expression and protein production (Wright et al. 1999). However, lower respiratory tract nitric oxide concentrations are increased following cigarette smoking (Chambers et al. 1998) although, in humans, plasma and urinary levels of nitrate, a metabolite of inhaled nitric oxide, have been shown to be unchanged suggesting that nitric oxide is not absorbed from the inhaled smoke (Rangemark & Wennmalm 1996).

Inhibition of nitric oxide synthase activity may explain the impaired endothelium-independent vasodilation associated with cigarette smoking (Su et al. 1998). The decrease in endothelial nitric oxide synthase activity may also partly explain the high risk of pulmonary and vascular disease in cigarette smokers. Cigarette smoking affects nitric oxide-mediated coagulation of coronary artery tone which is associated with a decrease in the bioactivity of nitric oxide (Kugiyama et al. 1996). It has been shown to increase isoprostane levels and reduce the generation of prostacyclin, l-arginine and l-citrulline in umbilical arteries and veins, and correlates with a direct vasoconstrictive effect (Obwegeser et al. 1999). Chronic smoking in the rat leads to age-independent moderate hypertension and a decrease in penile nitric oxide synthase activity (Xie et al. 1997).

It has been reported that exposure of cultured bovine pulmonary artery endothelial cells to carbon monoxide results in increased release of endothelial nitric oxide (Thom et al. 1997). The acute cytotoxicity from carbon monoxide was owing to nitric oxide-derived oxidants (Thom et al. 1997). Incubation of isolated rabbit aortas with cigarette smoke extract inhibits endothelium-dependent relaxation in a dose-dependent manner (Ota et al. 1997). If this incubation is carried out with free radical scavengers, attenuation of the inhibition occurs. This suggests that free radicals in cigarette smoke extract induce the impairment of endothelium-dependent relaxation and this may be partly because of the suppression of nitric oxide production (Ota et al. 1997). The same author has demonstrated, using cultured human endothelial cells, that cigarette smoke extract suppressed endothelial release of stable metabolites of nitric oxide and this was attenuated by free radical scavengers (Ota et al. 1997). The superficial femoral veins from rabbits exposed to cigarette smoke have a significant decrease in endothelium-dependent relaxation in response to acetylcholine without smooth muscle injury (Freischlag et al. 1999). Ascorbic acid protects rabbit arteries from cigarette smoke-induced endothelial injury by reducing the impairment of endothelium-dependent acetylcholine relaxation caused by smoking, probably as a result of oxygen-free radicals (Mays et al. 1999).

It has been suggested that the presence of hypoxia and exogenous nitric oxide, which lead to endothelial-dependent and independent vaso-relaxation secondary to cigarette smoking, may serve to explain the apparent augmentation of endothelial-derived relaxation in the rat (Nene et al. 1997). Cigarette smoking has been shown in the rat to increase aortic endothelial regeneration and serum levels of nitric oxide following balloon injury of the thoracic aorta (Sarkar et al. 1999). Acute hypoxia causes pulmonary vessel constriction, and chronic hypoxia causes smooth muscle cell replication and extracellular matrix accumulation, resulting in vessel wall remodelling (Kourembanas et al. 1997). Oxidized low-density lipoprotein, hypoxia and pro-inflammatory cytokines induce haem oxygenase expression and activity in vascular endothelial and smooth muscle cells (Siow et al. 1999).

Cultured human umbilical endothelial cells from smokers convert significantly more LDL into an atherogenic form than do cells derived from nonsmokers (PechAmsellem et al. 1996). The LDL modifications were strongly thiol-dependent and the enhanced superoxide production seen in the cells derived from smokers was dependent on the presence of cysteine in the medium (PechAmsellem et al. 1996). There is also evidence that cigarette smoking results in endothelial dysfunction in hypercholesterolemic patients by enhancing the oxidation of LDL (Heitzer et al. 1996). Cigarette smoke-modified low-density lipoprotein has been reported to impair endothelium-dependent relaxation in isolated rabbit arteries (Kagota et al. 1996).

Cells derived from smokers have higher levels of intracellular glutathione than those from nonsmokers and it may be that stimulation of cysteine uptake by the cells reflecting the enhanced total glutathione content could account for the enhanced superoxide production, all of which may be relevant to the pathophysiology of smoking-related cardiovascular disease (PechAmsellem et al. 1996). It has already been noted that there is evidence to indicate that chemical modification of glutathione is a major damage mechanism of filtered cigarette smoke with free radical oxidations being less significant (Maranzama & Mehlhorn 1998). Endothelial cell cGMP production is decreased in a dose-dependent manner following cigarette smoking and endothelial cell detachment is increased following smoking (Nagy et al. 1997). Externally added thiols protect endothelial cells from damage and it has been suggested that they may bind an unknown component of smoke to bring about this protection (Nagy et al. 1997). In humans, the endothelial dysfunction observed following cigarette smoking lasts for about one hour and is not attenuated with repeat exposure (Lekakis et al. 1998).

Cigarette smoking is associated with increased monocyte-endothelial cell adhesion in humans (Dovgan et al. 1994; Adams et al. 1997). Studies on hamsters using intravital microscopy and scanning electron microscopy demonstrate that cigarette smoking results in leucocytes adhering in clusters to the aortic endothelium and that aggregates of leukocytes and platelets are formed (Lehr et al. 1993,1994). There is evidence that phagocytes may employ myeloperoxidase-generated reactive nitrogen intermediates as a physiological pathway for initiating lipid peroxidation as well as forming biologically active lipid and sterol oxidation productions in vivo (Schmitt et al. 1999). Increased generation of oxygen free radicals by polymorphonuclear leucocytes may be responsible for the enhanced risk of certain smoking-related diseases (Kalra et al. 1991). The promotion of neutrophil infiltration and free radical production in rats exposed to cigarette smoke has been shown to contribute significantly to the development of experimental inflammatory bowel disease (Guo et al. 1999).

Cigarette smoking has been shown to increase re-endothelialization in the rat following large vessel injury and this is associated with an increase in serum nitric oxide levels (Sarkar et al. 1999). There is evidence that cigarette smoke-induced cell proliferation in the pulmonary arterial vessels is partly mediated through stimulation of endothelin-A receptors (Dadmanesh & Wright 1997). Conversely, cigarette smoke represses angiogenesis in the rat (Ma et al. 1999b). Cigarette smoking, but not transdermal nicotine delivery, is associated with borderline increases in plasma endothelin-1 levels but these are restricted to the first 10 minutes after the onset of smoking (Goerre et al. 1995).

In a study of monozygotic twins discordant for smoking Calori et al. (1996) found that cigarette smoking was associated with an atherogenic lipid profile along with changes in platelets and white cells which might reflect endothelial cell damage. Cigarette smoking increases total serum cholesterol levels (Jorge et al. 1995). Smokers have mean serum levels that are higher than nonsmokers for total and low-density lipoprotein cholesterol and triglyceride whereas high-density lipoprotein cholesterol levels are lower (Tang et al. 1997). There is evidence that cigarette smoking results in increased platelet consumption in human atherosclerotic vessels as well as the production of larger platelets which are more active (Kario et al. 1992). Cigarette smoking causes marked inhibition of substance P-induced tissue plasminogen activator released in vivo in humans (Newby et al. 1999). Serum from cigarette smokers contains higher levels of von Willebrand factor and was more cytotoxic to endothelial cells in vitro than serum from nonsmokers (BLann & McCollum 1993).

Conclusions

Since the publication of previous reviews on cigarette smoking, endothelial injury and cardiovascular disease (Pittilo 1988,1990; Pittilo & Woolf 1993,1994; Woolf et al. 1993), there have been new reports that demonstrate that cigarette smoking is associated with blood vessel wall damage including endothelial injury. Cigarette smoke is a complex mixture containing a range of individual components and only a small number of these have been examined in isolation. Whilst components such as nicotine and carbon monoxide have been the subject of a number of investigations, their damaging effects in model systems are less than those seen with whole smoke. Free radicals are an important component of cigarette smoke and our understanding of their chemistry and effects on biological systems has greatly increased. It seems likely that they will be further implicated as critical to the link between cigarette smoking and cardiovascular disease.