Impact of eNOS-Dependent Oxidative Stress on Endothelial Function and Neointima Formation

Abstract

Aims: Vascular oxidative stress generated by endothelial NO synthase (eNOS) was observed in experimental and clinical cardiovascular disease, but its relative importance for vascular pathologies is unclear. We investigated the impact of eNOS-dependent vascular oxidative stress on endothelial function and on neointimal hyperplasia.

Results: A dimer-destabilized mutant of bovine eNOS where cysteine 101 was replaced by alanine was cloned and introduced into an eNOS-deficient mouse strain (eNOS-KO) in an endothelial-specific manner. Destabilization of mutant eNOS in cells and eNOS-KO was confirmed by the reduced dimer/monomer ratio. Purified mutant eNOS and transfected cells generated less citrulline and NO, respectively, while superoxide generation was enhanced. In eNOS-KO, introduction of mutant eNOS caused a 2.3-3.7-fold increase in superoxide and peroxynitrite formation in the aorta and myocardium. This was completely blunted by an NOS inhibitor. Nevertheless, expression of mutant eNOS in eNOS-KO completely restored maximal aortic endothelium-dependent relaxation to acetylcholine. Neointimal hyperplasia induced by carotid binding was much larger in eNOS-KO than in mutant eNOS-KO and C57BL/6, while the latter strains showed comparable hyperplasia. Likewise, vascular remodeling was blunted in eNOS-KO only.

Innovation: Our results provide the first in vivo evidence that eNOS-dependent oxidative stress is unlikely to be an initial cause of impaired endothelium-dependent vasodilation and/or a pathologic factor promoting intimal hyperplasia. These findings highlight the importance of other sources of vascular oxidative stress in cardiovascular disease.

Conclusion: eNOS-dependent oxidative stress is unlikely to induce functional vascular damage as long as concomitant generation of NO is preserved. This underlines the importance of current and new therapeutic strategies in improving endothelial NO generation.

FIG. 1.

Characterization of C101A-eNOS in vitro. (A) Specific eNOS activity in purified WT- and C101A-eNOS preparations and (B) cell homogenates from stably transfected WT- and C101A-eNOS cells in dependence on BH₄ concentrations. Both enzymes were activated by BH₄, but the basal and the maximal specific eNOS activity in the C101A-eNOS cells was significantly lower than the activity of the WT cells (*p=0.006, n=3). (C) Western blot of a low-temperature SDS-gel representing a small amount of C101A-eNOS dimer in the purified mutant enzyme. Lane 1: denatured protein as monomer control, lane 2: 25 μg WT/MT-eNOS, lane 3: 50 μg WT/MT-eNOS. (D) Representative Western blot of pcDNA3-, WT-, and C101A-eNOS-transfected cells for detection of carbonyl groups as a marker for cellular protein oxidation. Twenty micrograms of total protein was loaded, and the band below 75 kDa was used for densitometric analysis. (E) Relative optical density (%) from six Oxyblots. The amount of carbonyl groups in the C101A-eNOS cells was significantly higher than in WT or pcDNA3- HEK 293 cells (*p=0.002, n=6). BH₄, (6R-)5,6,7,8-tetrahydro-L-biopterin; eNOS, endothelial NO synthase; HEK 293, human embryonic kidney 293.

FIG. 2.

Generation and basal characterization of C101A-Tg and C101A/eNOS-KO. (A) Construct used for insertion into fertilized eggs of C57BL/6 to generate C101A-eNOS-expressing mice (C101A-Tg) consisting of a 2.1 kb Tie-2 promoter, 4.1 kb bovine C101A-eNOS DNA, and a 10.6 kb Tie-2 enhancer. SpeI restriction sites were used for linearization of the plasmid before insertion. (B) Aortic and myocardial eNOS protein expression standardized to actin in C101A-Tg (mean±SEM *p<0.05 vs. controls, n=5–7). (C) No effect of C101A-eNOS overexpression on aortic endothelium-dependent relaxation induced by cumulative addition of acetylcholine (ACh, p=0.5548, n=6, two-way ANOVA). (D) Significant increase in nitrotyrosine residues as a marker for peroxynitrite formation (mean±SEM upper panel, *p<0.05, n=6) and representative Western blot of protein nitrotyrosine residues (middle panel) standardized to actin (lower panel) in myocardial tissue of C101A-Tg versus C57BL/6. (E) Protein expression of eNOS in the aorta, skeletal muscle, lung, and myocardium of C101A/eNOS-Tg (*p<0.05 vs. C57BL/6, n=6–12). The control values referring to eNOS protein expression in every tissue of C57BL/6 were set to 100% and are reflected by just one bar. (F) A representative experiment showing an increase of eNOS monomer level in native and denatured lung homogenates of C101A/eNOS-KO versus C57BL/6 (100 μg of total protein per lane). ANOVA, analysis of variance.

FIG. 3.

Role of C101A-eNOS as a source of increased vascular oxidative stress in C101A/eNOS-KO. Increased (A) aortic and (B) myocardial basal superoxide production in C101A/eNOS-KO, eNOS-KO, and C57BL/6. The cumulative counts/mg of dried tissue measured during incubation with 5 μM lucigenin are given (*p<0.05, two-way ANOVA, n=6–13). (C) Effect of NOS inhibition (L-NAME, 1 mM) on aortic and (D) myocardial superoxide production assessed by lucigenin chemiluminescence in C101A/eNOS-KO, eNOS-KO, and C57BL/6. Results are expressed as counts/tissue dry weight/min (mean±SEM) averaged over the last 8 min of measurement (*p<0.05 vs. all other conditions, one-way ANOVA, n=5–8). (E) Aortic and (F) myocardial nitrotyrosine levels before and after oral treatment with the NO synthase inhibitor, L-NA, in eNOS-KO and C101A/eNOS-KO. Results are given as mean±SEM percentage relative to eNOS-KO (upper panel, *p<0.05 vs. eNOS-KO, one-way ANOVA, n=5–11). The middle panel shows a representative Western blot for protein nitrotyrosine residues, the lower panel shows β-actin expression as loading control. L-NA, N^ω-nitro-L-arginine; L-NAME, NG-nitro-L-arginine methyl ester.

FIG. 4.

Preserved endothelial function in C101A/eNOS-KO. (A) Relaxation response of aortic rings of C101A/eNOS-KO, eNOS-KO, and C57BL/6 induced by cumulative addition of acetylcholine (*p<0.05 vs. C57BL/6, two-way ANOVA, n=6–14). (B) Aortic acetylcholine-induced relaxation in C101A/eNOS-KO in the presence of 100 U/ml of pegylated SOD (*p<0.05, n=4). (C) Effect of the NOS inhibitor, L-NAME, (D) the sGC inhibitor, ODQ, and (E) the NO scavenger, Fe(DETC)₂, on relaxation response to acetylcholine in aortic rings of C101A/eNOS-KO and C57BL/6 (*p<0.05 vs. C57BL/6, two-way ANOVA, n=4–9). (F) Aortic relaxation response to cumulative addition of NO donor, SNAP, in C101A/eNOS-KO, eNOS-KO, and C57BL/6 (*p<0.05 for pD₂ and two-way ANOVA, n=5–6). ODQ, 1H-[1,2,4] oxadiazolo[4,3-a]quinoxalin-1-one; sGC, soluble guanylyl cyclase; SNAP, S-nitroso-N-acetyl-D,L-penicillamine; SOD, superoxide dismutase.

FIG. 5.

Negligible effects of C101A-eNOS on neointimal hyperplasia. (A) Representative sections stained by hematoxylin and eosin demonstrate pronounced neointimal hyperplasia, media thickening, and luminal narrowing in the ligated A. carotis sinistra of all studied genotypes (450 μm from the ligation). Scale bar indicates 100 μm. (B) Neointima formation is shown for different distances from the ligation site in C57BL/6, eNOS-KO, and C101A/eNOS-KO. Values are mean±SEM, *p<0.05 vs. C57BL/6, two-way ANOVA, n=4–6). (C) Statistical evaluation of the neointimal area (*p<0.05 vs. C57BL/6, ^#p<0.05 vs. eNOS-KO, n=4–6). (D) Intima/media ratio for different distances from ligation in C57BL/6, eNOS-KO, and C101A/eNOS-KO (*p<0.05 vs. C57BL/6, two-way ANOVA, n=4–6) and (E) statistical analysis after summarizing the ratios for each animal at each distance point (*p<0.05 vs. C57BL/6, ^#p<0.05 vs. eNOS-KO, one-way ANOVA, n=4–6). To see this illustration in color, the reader is referred to the web version of this article at www.liebertpub.com/ars

FIG. 6.

Effects of C101A-eNOS on vascular remodeling. (A) Representative photomicrographs of hematoxylin and eosin-stained transverse sections of the right (unligated side, A. carotis dextra) and left (ligated side, A. carotis sinistra) common carotid arteries (850 μm from the ligation). Scale bar indicates 200 μm. (B) Effects of A. carotis sinistra ligation on the vessel area in C57BL/6, (C) eNOS-KO, and (D) C101A/eNOS-KO (*p<0.05 vs. A. carotis dextra, two-way ANOVA, n=4–5). (E) Vessel area ratio of left common carotid artery from the ligated animal over the intact right common carotid artery from the same mouse (*p<0.05 vs. C57BL/6, two-way ANOVA, n=4–5). To see this illustration in color, the reader is referred to the web version of this article at www.liebertpub.com/ars