Diet-Induced Hyperhomocysteinemia Exacerbates Neointima Formation in Rat Carotid Arteries After Balloon Injury
Background—Increasing evidence indicates that elevated plasma homocysteine levels are associated with an increased risk of atherosclerosis and endothelial dysfunction, although little specific information on the mechanisms responsible for the atherogenic effects of homocysteine or on the in vivo contribution made by hyperhomocysteinemia to atherosclerosis is currently available. Because homocysteine is known to exert a direct inhibitory effect on endothelial cell growth in vitro, we hypothesized that this effect contributes to the progression of atherosclerotic lesions initiated by endothelial damage caused by mechanical injury.
Methods and Results—We prepared diet-induced hyperhomocysteinemic rats in which neointima formation after balloon injury to the common carotid artery was assessed. Moderate hyperhomocysteinemia (plasma homocysteine levels 3- to 4-fold higher than control) significantly exacerbated neointima formation. Oral administration of folate, which had a homocysteine-lowering effect, diminished neointima formation induced by moderate hyperhomocysteinemia. Furthermore, the attenuation of reendothelialization was shown in diet-induced hyperhomocysteinemic rats with Evans blue staining.
Conclusions—Diet-induced hyperhomocysteinemia, even mild to moderate, exacerbates neointima formation after denuding injury, making hyperhomocysteinemia a likely risk factor for postangioplasty restenosis. It may be mediated through an inhibitory effect of homocysteine on reendothelialization. Homocysteine lowering with folate supplementation can effectively ameliorate the detrimental effects of moderate hyperhomocysteinemia. Clinical trials would seem to be warranted.
Homocysteine is a sulfur-containing amino acid whose concentrations are regulated by nutritional factors, mainly methionine and vitamin cofactors such as folate and vitamins B6 and B12.1 The prevalence of mild-to-moderate hyperhomocysteinemia in the general population is well known, and an increasing number of clinical studies have shown that elevated plasma homocysteine levels are associated with an increased risk of atherosclerosis, independent of other known risk factors.2 In addition, recent progress in human genetics has revealed that ≥1 polymorphisms among the enzymes involved in metabolizing homocysteine may be responsible for the elevated plasma homocysteine levels and may thus constitute a genetic risk factor for atherothrombotic disease. For example, we and others have found an association between atherosclerotic disease and a common, functionally important mutation (C677T) of the key metabolic enzyme, 5,10-methylenetetrahydrofolate reductase (MTHFR), particularly when folate levels are low.3 4 5 6 7 The genetic polymorphism of MTHFR is a genetic marker for atherosclerotic diseases as well as a determinant of susceptibility to folate insufficiency. Given the potential clinical significance of homocysteinemia, this means that folate supplementation may serve as an effective antiatherosclerotic therapy.
Homocysteine induces both atherogenic and thrombogenic mediators in cultured vascular cells8 9 10 and causes endothelial dysfunction in humans and monkeys.11 12 However, little is known about the mechanisms underlying the atherogenic effects of homocysteine or about the in vivo contribution made by mild-to-moderate hyperhomocysteinemia to atherosclerosis. The latter is important in clinical terms, because even though mild-to-moderate hyperhomocysteinemia itself may not be atherogenic, it may lead to progression of atherosclerosis in concert with other stimuli such as mechanical injury.
The aim of the present study was to test this hypothesis using the rat balloon injury model, which is the most extensively studied model of the neointima formation typically encountered in human atherosclerotic coronary arteries after percutaneous transluminal coronary angioplasty.13 14
Hyperhomocysteinemia was induced by dietary modification, and the extent of neointima formation and reendothelialization after balloon injury was estimated. In addition, the therapeutic effect of oral folate administration was tested and the clinical merit of folate supplementation was discussed.
The experimental protocol used in this study was designed in accordance with the Guide for Animal Experimentation, Faculty of Medicine, University of Tokyo.
Eight-week-old male Sprague-Dawley rats were divided into subgroups and maintained for 4 weeks before experimentation either on control chow containing 0.36% methionine, 0.00019% folate, 0.0% homocysteine, and 0.27% cysteine or on 1 of the following modified diets containing either 1.2% methionine, 2.0% methionine, 1.2% methionine+0.0019% folate, 2.0% methionine+0.0019% folate, 1.0% homocysteine, or 1.0% cysteine. Food and water were provided ad libitum.
Measurement of Homocysteine and Folate
Blood samples were drawn from the tail vein, promptly centrifuged, and stored at −20°C. Plasma homocysteine was measured as the total homocysteine by high-performance liquid chromatography (HPLC) with fluorescence detection, as previously described.15 Plasma concentrations of methionine and cystine were measured by HPLC using ninhydrin. Serum folate was measured with a commercially available radioimmunoassay kit.
Rat Arterial Injury Model
After 4 weeks on their respective diets, rats (weighing 400 to 450 g) were anesthetized with chloral hydrate (370 mg/kg IP). Balloon denudation of the left common carotid artery was performed as previously described.13 Mortality attributable to anesthesia or postsurgical complications was <10%. The right common carotid artery served as a control.
Evaluation of Neointimal Hyperplasia
Rats were euthanized with a lethal dose of anesthetic 14 days after balloon injury, after which the carotid arteries were perfused with 4% paraformaldehyde and PBS. Each injured left carotid artery was excised from the proximal edge of the omohyoid muscle to the carotid bifurcation. The middle third of the segment was then fixed in 4% paraformaldehyde for 12 hours and embedded in paraffin. Cross sections (6 μm) cut from each sample at intervals of 500 μm were stained with hematoxylin/eosin and photographed. The photomicrographs were then scanned and analyzed with NIH Image 1.58 Software (National Institutes of Health). The sections with the largest neointima/media ratios were subjected to statistical analysis. An average for each parameter was calculated on the basis of the animals in the respective group.
Analysis of Reendothelialization of Arterial Segments
Reendothelialization was assessed in rats maintained on control chow, the 2.0% methionine diet, the 1.2% methionine diet, or the 1.2% methionine+0.0019% folate diet with Evans blue dye, which stains areas of nonendothelialized artery blue.16 17 18 Thirty minutes before euthanasia, rats received an intravenous injection via the tail vein of 0.5 mL 0.5% Evans blue. After fixation in situ in 100% methanol, the initially endothelium-denuded segment of left common carotid artery (defined as the total surface area of the harvested arterial segment, from just under the proximal edge of the omohyoid muscle to the carotid bifurcation) was dissected free, incised longitudinally, and photographed under a dissecting microscope. A single observer blinded to the experimental regimen carried out the planimetric analysis of the photographs using a computerized sketching program on a digitizing board (NIH Image 1.58 Software). Reendothelialized areas were defined macroscopically as those areas not stained by the Evans blue dye.
Functional Study of Carotid Rings
To test whether endothelial function is impaired in diet-induced hyperhomocysteinemic rats, we examined endothelium-dependent vasodilator, acetylcholine-induced relaxation of arterial rings isolated from rats fed control chow (n=10) and 2.0% methionine (n=7). The right (uninjured) and left (injured) common carotid arteries were carefully removed to avoid damaging endothelium 14 days after balloon injury. After blood and connective tissue had been removed, the common carotid artery was divided into 3 cylindrical segments (proximal, middle, and distal portions). Each ring segment (2 mm long) was mounted between 2 stainless steel wires in 10 mL of organ bath containing Krebs bicarbonate solution bubbled with a mixture of 95% O2 and 5% CO2 to obtain rapid mixing of drugs. One wire was attached to a fixed support, and the other was connected to a force-displacement transducer. The rings were suspended under 1.0 g of tension. The preparation was allowed to equilibrate for 90 minutes and preconstricted by phenylephrine (10−9 mol/L). To obtain a dose-response curve for acetylcholine (10−8 to 10−5 mol/L) and sodium nitroprusside (10−8 to 10−5 mol/L), a stock solution of drugs was added cumulatively to the organ bath. To evaluate the endogenous NO production, we further analyzed acetylcholine-induced vasodilatation after blockade of NO production with N-nitro-l-arginine methyl ester (L-NAME; 10−5 mol/L). Data were expressed as percentage relaxation of phenylephrine-induced preconstriction.
All data are presented as mean±SEM. Differences between groups were evaluated by ANOVA and Scheffé’s F test. Values of P<0.05 were considered significant.
Effect of Diet on Plasma Homocysteine, Methionine, Cystine, and Serum Folate Levels
Plasma homocysteine concentrations in rats fed 1.2% methionine were significantly higher than in rats fed control chow (27.2±5.0 versus 7.6±0.5 μmol/L) (Figure 1A⇓). Still higher plasma homocysteine levels were observed in rats fed diets containing either 2.0% methionine or 1.0% homocysteine (83.9±21.7 and 86.2±11.6 μmol/L, respectively). Below-control homocysteine levels were observed in rats fed 1.0% cysteine (5.4±0.3 μmol/L). In addition, we observed a significant homocysteine-lowering effect of folate in rats fed 1.2% methionine+0.0019% folate compared with 1.2% methionine alone (10.1±1.6 versus 27.2±5.0 μmol/L). However, folate had no apparent effect on plasma homocysteine in rats fed 2.0% methionine (83.9±21.7 versus 88.7±15.2 μmol/L).
Mildly elevated serum folate was seen in rats maintained on 1.2% or 2.0% methionine, on 1.0% homocysteine, or on 1.0% cysteine, although there were no significant differences among the 4 groups (Figure 1B⇑). As expected, folate levels were significantly elevated in rats receiving the high-folate diet. Also as expected, plasma methionine was significantly higher in the high-methionine groups; however, methionine levels were not elevated in the 1.0% homocysteine group, and there was a significant decrease in the 1.0% cysteine group (Figure 1C⇑). High dietary folate did not affect plasma methionine concentrations in our study. Furthermore, there were no significant differences in plasma cystine levels among rats fed 2.0% methionine (28.0±4.1), 1.0% homocysteine (23.5±3.8), or 1.0% cysteine (27.0±3.6), all of which were significantly higher than control (3.5±0.4 μmol/L). Cystine concentrations were also unaffected by dietary folate (1.2% methionine, 18.8±3.9 versus 1.2% methionine+0.0019% folate, 20.9±3.0 μmol/L).
Analysis of Neointima Formation
Neointimal thickening was significantly greater in rats fed diets containing 1.2% or 2.0% methionine or 1.0% homocysteine than in the controls (Figure 2⇓).
Rats fed control chow had an average neointimal area of 0.21±0.01 mm2, compared with 0.34±0.02 mm2 in the 1.2% methionine group, 0.35±0.02 mm2 in the 2.0% methionine group, and 0.35±0.03 mm2 in the 1.0% homocysteine group (Figure 2B⇑). In contrast, no significant changes in medial area were observed (Figure 2A⇑).
Calculation of neointima/media ratios yielded analogous results (Figure 2C⇑). The average neointima/media ratio in rats maintained on control chow was 1.41±0.10, compared with 2.15±0.13 in the 1.2% methionine group, 2.21±0.09 in the 2.0% methionine group, and 2.03±0.12 in the 1.0% homocysteine group. Luminal areas were significantly smaller in rats fed 1.2% methionine (0.11±0.02 mm2), 2.0% methionine (0.09±0.01 mm2), or 1.0% homocysteine (0.09±0.01 mm2) than in rats fed control diet (0.19±0.02 mm2) (Figure 2D⇑). Rats fed a diet containing 1.0% cysteine exhibited an average neointimal area of 0.21±0.01 mm2, a neointima/media ratio of 1.34±0.07, and a luminal area of 0.21±0.02 mm2, values similar to those seen in rats maintained on control chow.
Homocysteine lowering with high dietary folate diminished neointima formation such that both neointimal area and neointima/media ratios were significantly lower in rats fed 1.2% methionine+0.0019% folate than in rats fed 1.2% methionine alone (neointimal area, 0.24±0.02 versus 0.34±0.02 mm2; neointima/media ratio, 1.56±0.11 versus 2.15±0.13). In addition, rats on the high-folate diet exhibited a concomitant increase in luminal area (0.17±0.03 versus 0.11±0.02 mm2).
Conversely, rats fed 2.0% methionine+0.0019% folate (n=8) exhibited an average neointimal area of 0.33±0.03 mm2, a neointima/media ratio of 2.24±0.11, and a luminal area of 0.10±0.01 mm2, values similar to those seen in rats maintained on a 2.0% methionine diet.
No structural or histological differences were seen in the uninjured right common carotid arteries of any group (data not shown).
We summarized the changes of plasma amino acids and neointima/media ratio after balloon injury in a demonstrative chart (Table⇓).
Analysis of Reendothelialization
To estimate the extent of areas that remained to be deendothelialized at 14 days after balloon injury, we used Evans blue dye. Reendothelialization was observed to spread from the marginal portion of the initially deendothelialized area to the central portion. The attenuation of reendothelialization was observed in the 1.2% high-methionine-diet (n=11) and 2.0% high-methionine-diet groups (n=13) versus the normal-diet group (n=13). Although there was no difference in sizes of the initially deendothelialized areas in the 3 groups, rats receiving a normal diet had a reendothelialized area of 5.9±0.6 versus 3.0±0.5 mm2 in the 1.2% high-methionine-diet group (P=0.0021 versus normal-diet group) and 2.2±0.4 mm2 in the 2.0% high-methionine-diet group (P<0.0001 versus normal-diet group) (Figure 3A⇓). Expressed as a percentage of the total area that was initially deendothelialized, the reendothelialized area was 37.4±2.9% in the normal-diet group, compared with 18.8±2.5% in the 1.2% high-methionine-diet group (P<0.0001 versus normal-diet group) and 13.7±2.4% in the 2.0% high-methionine-diet group (P<0.0001 versus normal-diet group) (Figure 3B⇓).
As shown in Figure 3⇑, the attenuation of reendothelialization was ameliorated in the 1.2% methionine+0.0019% folate–diet group (n=9) compared with the 1.2% high-methionine-diet group (reendothelialized area, 5.4±0.4 versus 3.0±0.5 mm2 [P=0.0030]; percent reendothelialization, 33.9±2.5% versus 18.8±2.5% [P=0.0061]).
Functional Study of Carotid Rings
In a dose-dependent manner, acetylcholine (10−8 to 10−5 mol/L) elicited significantly greater peak dilatations of uninjured common carotid arteries from rats fed control chow than those fed 2.0% high-methionine diet with Rmax (maximal response) of 98.4±1.0% and 66.2±9.4%, respectively (P=0.0005) (Figure 4A⇓).
As shown in Figure 4B⇑ through 4D, there was a significant difference in acetylcholine-induced relaxation of proximal and distal portions of injured common carotid arteries between the normal-diet group and the 2.0% high-methionine-diet group (proximal portion: Rmax 34.9±8.2% versus 10.1±4.2% [P=0.025]; distal portion: R(10−6 mol/L Ach) 32.6±7.0% versus 12.3±5.3% [P=0.039]; Rmax 37.0±8.1% versus 17.2±6.8% [P=0.081]), whereas no difference in the relaxation responses to acetylcholine was shown in the middle portion (Rmax 15.8±4.9% versus 13.5±8.2%).
Moreover, we compared the endothelial function according to the portion of injured common carotid arteries in each diet group. In the normal-diet group, the middle portion elicited a weaker relaxation response to acetylcholine than the proximal and distal portions (Rmax 34.9±8.2% [proximal] versus 15.8±4.9% [middle] [P=0.046]; 37.0±8.1% [distal] versus 15.8±4.9% [middle] [P=0.030]). In the 2.0% high-methionine-diet group, any portion had poor endothelial function, and there was no difference according to the portion (Figure 5⇓).
Relaxation to acetylcholine was abolished in all vessels after pretreatment with the NO synthesis inhibitor L-NAME 10−5 mol/L. There was no significant difference in relaxation responses to sodium nitroprusside (10−8 to 10−5 mol/L) among the carotid ring segments described above (data not shown).
The present analysis yielded 3 major findings: first, diet-induced hyperhomocysteinemia, even mild to moderate, exacerbated neointima formation in the rat balloon injury model; second, hyperhomocysteinemia attenuated reendothelialization of balloon-injured arteries; and third, folate supplementation, which had a homocysteine-lowering effect, attenuated the hyperhomocysteinemia-induced exacerbation of neointima formation.
Numerous clinical studies have shown that hyperhomocysteinemia is a major independent risk factor for vascular diseases.2 Through genetic analysis of the enzymes mediating homocysteine metabolism, we and others have shown that prevalent mild-to-moderate hyperhomocysteinemia contributes to the onset of atherosclerotic disease and that folate administration may be an effective antiatherosclerotic strategy.3 4 5 6 7
In addition, a variety of in vitro studies have shown that homocysteine may contribute to the atherosclerotic and thrombotic processes by modulating vascular cell proliferation and by promoting prothrombotic activities.8 9 10
The in vivo contribution of hyperhomocysteinemia to atherosclerosis has been less extensively investigated, however. Harker et al19 20 showed that infusion of homocystine into baboons elicited patchy desquamation of vascular endothelium in the short term and led to formation of a neointima composed of proliferating smooth muscle cells in the long term. Because homocysteine concentrations are regulated by dietary intake of mainly methionine and vitamin cofactors, analysis of diet-induced hyperhomocysteinemia in animals should provide useful information on the in vivo effect of homocysteine and on potential therapeutic strategies. In the present study, therefore, we used a dietary model of hyperhomocysteinemia to examine the effect of homocysteine on reendothelialization and neointima formation after balloon injury as well as the therapeutic effect of oral folate supplementation.
Neointima Formation and Reendothelialization
Balloon catheter injury to the rat carotid artery triggers a sequence of events leading to formation of a thickened neointima, and many mediators and compounds have been discussed16 17 18 21 ; in particular, the endothelium plays an essential role in neointima formation after balloon injury. Asahara et al16 showed that application of vascular endothelial growth factor, which promotes reendothelialization, attenuates neointimal formation. Similarly, Krasinski et al17 showed that increased endothelial recovery is correlated with diminished neointimal hyperplasia in a dose-response manner. Because homocysteine is known to exert a direct inhibitory effect on endothelial growth, hyperhomocysteinemia would be expected to attenuate reendothelialization and exacerbate neointima formation after balloon injury.
The present findings show that augmented neointimal formation occurred at plasma homocysteine concentrations ≈3- to 4-fold higher than control, which are similar to the concentrations associated with increased risk of vascular disease in humans. Comparing neointima formation in the high-methionine group, exhibiting hypermethioninemia, and the homocysteine group, exhibiting the same plasma methionine levels as the control group, we concluded that it was highly unlikely that neointima formation observed in the groups on high-methionine diets was due to a direct effect of methionine, because the same degree of neointima formation was observed in the homocysteine-diet group with normomethioninemia.
To distinguish whether this effect is due to increased plasma levels of cystine, we examined the neointima formation in rats fed a high-cysteine diet. The high-cysteine diet had no effect on neointima formation, which is in agreement with an earlier report that homocysteine, but not cysteine, inhibits DNA synthesis in vascular endothelial cells via reduction of p21ras methylation.10
Furthermore, reendothelialization was significantly attenuated in diet-induced hyperhomocysteinemic rats, suggesting that the augmented neointima formation might be secondary to impaired reendothelialization. However, the precise mechanism by which homocysteine attenuates reendothelialization remains to be investigated, and the other mechanisms—for example, in vivo effects of homocysteine on proliferating vascular smooth muscle cells—might be related to a predisposition to neointima formation in this model.
Ross22 suggested that factors including oxidized LDL, mechanical injury, and homocysteine act in concert to cause endothelial injury that predisposes arteries to progression of neointimal thickening. Consistent with this theory, we observed that in the presence of a balloon injury, elevated plasma homocysteine exacerbated the progression of neointimal thickening, even when homocysteine levels remaining within the expected physiological range, making hyperhomocysteinemia a likely risk factor for postangioplasty restenosis.
Therapeutic Application of Folate
The beneficial effects of vitamins on plasma homocysteine levels and on the progression of atherosclerosis have yet to be verified. Nonetheless, we observed that oral folate administration ameliorated moderate hyperhomocysteinemia and diminished neointima formation, possibly mediated through the increased reendothelialization, whereas in severe hyperhomocysteinemia (2.0% high-methionine-diet group), the plasma homocysteine concentrations and the neointima formation after balloon injury could not be improved by additional high-folate diet. This is a useful control group to separate the independent influence of folate from its homocysteine-lowering effect.
It is suggested that control of hyperhomocysteinemia by dietary modification or folate supplementation, if not severe, could decrease the consequent vascular events caused by hyperhomocysteinemia.
Consequently, additional studies and/or clinical trials aimed at determining whether homocysteine-lowering regimens might effectively prevent postangioplasty restenosis in humans would seem to be warranted.
Analysis of Endothelial Function in This Model
The functional analysis of endothelium in homocysteinemia could give us a great deal of information on the atherogenic effect of homocysteinemia. We examined the endothelial function in both the injured and the noninjured arteries using ex vivo models of carotid rings. In the analysis of uninjured common carotid arteries, we confirmed the endothelial dysfunction in hyperhomocysteinemia, which is consistent with other recent reports.23 24 And through the functional analysis of injured arteries, we demonstrated 2 findings: (1) reendothelialization initiated in the marginal part of the deendothelialized area spreads to the center part, and (2) reendothelialization is impaired in hyperhomocysteinemia, which is in concordance with the analysis using Evans blue dye. This is the first report on the functional analysis of injured endothelium under hyperhomocysteinemia.
In summary, diet-induced mild-to-moderate hyperhomocysteinemia, the extent of which is comparable to the levels that are associated with a predisposition to common atherosclerotic diseases in humans, is shown to promote neointima formation in a rat balloon injury model. It may be mediated through an inhibitory effect of homocysteine on reendothelialization. Moreover, folate administration might be a promising therapy for vascular events caused by hyperhomocysteinemia.
This work was supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science, and Culture, Japan; the Japan Cardiovascular Research Foundation; TMFC; and the Ryoichi Naito Foundation for Medical Research (to Dr H. Kurihara); and by the Japan Heart Foundation; a Pfizer Pharmaceuticals Grant for Research on Coronary Artery Disease; a Research Grant of the Tokyo Hypertension Conference; the Research Foundation for Community Medicine Research Meeting on Hypertension and Arteriosclerosis; and a Research Grant from the Sankyo Foundation of Life Science (to Dr Morita).
- Received May 5, 2000.
- Revision received July 18, 2000.
- Accepted July 20, 2000.
- Copyright © 2001 by American Heart Association
1. Mudd SH, Levy HL, Skovby F. Disorders of transsulfuration. In: Scriver CR, Beaudet AL, Sly WS, et al, eds. The Metabolic and Molecular Bases of Inherited Disease. New York, NY: McGraw-Hill; 1995:1279–1327.
Morita H, Taguchi J, Kurihara H, et al. Genetic polymorphism of 5,10-methylenetetrahydrofolate reductase (MTHFR) as a risk factor for coronary artery disease. Circulation. 1997;95:2032–2036.
Morita H, Kurihara H, Tsubaki S, et al. Methylenetetrahydrofolate reductase gene polymorphism and ischemic stroke in Japanese. Arterioscler Thromb Vasc Biol. 1998;18:1465–1469.
Morita H, Kurihara H, Sugiyama T, et al. Polymorphism of the methionine synthase gene: association with homocysteine metabolism and late-onset vascular diseases in the Japanese population. Arterioscler Thromb Vasc Biol. 1999;19:298–302.
Gallagher PM, Meleady R, Shields DC, et al. Homocysteine and risk of premature coronary heart disease: evidence for a common gene mutation. Circulation. 1996;94:2154–2158.
Harmon DL, Doyle RM, Meleady R, et al. Genetic analysis of the thermolabile variant of 5,10-methylenetetrahydrofolate reductase as a risk factor for ischemic stroke. Arterioscler Thromb Vasc Biol. 1999;19:208–211.
Tsai J-C, Perrella MA, Yoshizumi M, et al. Promotion of vascular smooth muscle cell growth by homocysteine: a link to atherosclerosis. Proc Natl Acad Sci U S A. 1994;91:6369–6373.
Wang H, Yoshizumi M, Lai K, et al. Inhibition of growth and p21ras methylation in vascular endothelial cells by homocysteine but not cysteine. J Biol Chem. 1997;272:25380–25385.
Chambers JC, McGregor A, Jean-Marie J, et al. Demonstration of rapid onset vascular endothelial dysfunction after hyperhomocysteinemia: an effect reversible with vitamin C therapy. Circulation. 1999;99:1156–1160.
Schwartz SM, deBlois D, O’Brien ERM. The intima: soil for atherosclerosis and restenosis. Circ Res. 1995;77:445–465.
Asahara T, Bauters C, Pastore C, et al. Local delivery of vascular endothelial growth factor accelerates reendothelialization and attenuates intimal hyperplasia in balloon-injured rat carotid artery. Circulation. 1995;91:2793–2801.
Krasinski K, Spyridopoulos I, Asahara T, et al. Estradiol accelerates functional endothelial recovery after arterial injury. Circulation. 1997;95:1768–1772.
Spyridopoulos I, Principe N, Krasinski KL, et al. Restoration of E2F expression rescues vascular endothelial cells from tumor necrosis factor–induced apoptosis. Circulation. 1998;98:2883–2890.
Harker LA, Slichter SJ, Scott CR, et al. Homocystinemia: Vascular injury and arterial thrombosis. N Engl J Med. 1974;291:537–543.
Harker LA, Ross R, Slichter SJ, et al. Homocystine-induced arteriosclerosis: the role of endothelial cell injury and platelet response in its genesis. J Clin Invest. 1976;58:731–741.
Rakugi H, Kim D-K, Krieger JE, et al. Induction of angiotensinogen converting enzyme in the neointima after vascular injury: possible role of restenosis. J Clin Invest. 1994;93:339–346.
Ungvari Z, Pacher P, Rischak K, et al. Dysfunction of nitric oxide mediation in isolated rat arterioles with methionine diet-induced hyperhomocysteinemia. Arterioscler Thromb Vasc Biol. 1999;19:1899–1904.
Lang D, Kredan MB, Moat SJ, et al. Homocysteine-induced inhibition of endothelium-dependent relaxation in rabbit aorta: role for superoxide anions. Arterioscler Thromb Vasc Biol. 2000;20:422–427.