Reactive Oxygen Species Generation and Atherosclerosis
Witold N. Nowak,Jiacheng Deng,Xiong Z. Ruan,Qingbo Xu
DOI: https://doi.org/10.1161/atvbaha.117.309228
2017-01-01
Abstract:HomeArteriosclerosis, Thrombosis, and Vascular BiologyVol. 37, No. 5Reactive Oxygen Species Generation and Atherosclerosis Free AccessReview ArticlePDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissions ShareShare onFacebookTwitterLinked InMendeleyReddit Jump toFree AccessReview ArticlePDF/EPUBReactive Oxygen Species Generation and Atherosclerosis Witold N. Nowak, Jiacheng Deng, Xiong Z. Ruan and Qingbo Xu Witold N. NowakWitold N. Nowak From the Cardiovascular Division, King’s BHF Centre, King’s College London, United Kingdom (W.N.N., J.D., Q.X.); Centre for Nephrology and Urology, Health Science Centre, Shenzhen University, China (X.Z.R.); and Centre for Nephrology, University College London, United Kingdom (X.Z.R.). Search for more papers by this author , Jiacheng DengJiacheng Deng From the Cardiovascular Division, King’s BHF Centre, King’s College London, United Kingdom (W.N.N., J.D., Q.X.); Centre for Nephrology and Urology, Health Science Centre, Shenzhen University, China (X.Z.R.); and Centre for Nephrology, University College London, United Kingdom (X.Z.R.). Search for more papers by this author , Xiong Z. RuanXiong Z. Ruan From the Cardiovascular Division, King’s BHF Centre, King’s College London, United Kingdom (W.N.N., J.D., Q.X.); Centre for Nephrology and Urology, Health Science Centre, Shenzhen University, China (X.Z.R.); and Centre for Nephrology, University College London, United Kingdom (X.Z.R.). Search for more papers by this author and Qingbo XuQingbo Xu From the Cardiovascular Division, King’s BHF Centre, King’s College London, United Kingdom (W.N.N., J.D., Q.X.); Centre for Nephrology and Urology, Health Science Centre, Shenzhen University, China (X.Z.R.); and Centre for Nephrology, University College London, United Kingdom (X.Z.R.). Search for more papers by this author Originally published1 May 2017https://doi.org/10.1161/ATVBAHA.117.309228Arteriosclerosis, Thrombosis, and Vascular Biology. 2017;37:e41–e52Atherosclerosis is the key component of most cardiovascular diseases, including stroke and myocardial infarction.1–3 An inflamed endothelium recruits inflammatory cells, such as monocytes, via the expression of various mediators and chemokines.4–6 This, in addition to the accumulation of cholesterol and smooth muscle cells (SMCs) in the intima, leads to the transformation of monocytes into foam cells, which consume dead cells and lipids.7–9 The atherogenic process involves multiple cell types, that is, endothelial cells (ECs),10 SMCs,11 immune cells,12 and stem/progenitor cells,13 in which levels of both intracellular and extracellular reactive oxygen and nitrogen species play a fundamental role in vascular cell homoeostasis and eventually affects the development of atherosclerosis.14 Fine-tuning of cellular redox status is a prerequisite for the well-being of vascular system. Although too much oxidative stress can be detrimental, some basal levels are crucial for proper cell signaling. Recently, a number of publications in ATVB and other journals have demonstrated substantial progress in research into oxidative stress vascular disease, especially atherosclerosis.14–17 In the present article, we highlight these updated publications, providing insights into the mechanisms of reactive oxygen species (ROS) generation in pathophysiological conditions of the vessel wall, and the contribution of redox imbalance to lesion formation via influencing vascular cell (dys)functions.Oxidative stress is defined as a cellular condition where the damaging effect of oxidant is greater than the beneficial effect of antioxidants. Major oxidants are based on O2 molecules, which are taken in during respiration, with higher reactivity than molecular O2, and are known as ROS.14 ROS are, thus, broadly defined as oxygen-containing chemical species with higher reactive properties. Major ROS include superoxide (O2·−) and hydroxyl (HO·) free radicals, as well as nonradical molecules, such as hydrogen peroxide (H2O2). Primary sources of oxidative stress in vessel wall are mitochondria,18 uncoupled nitric oxide synthase,19 lipoxygenase,20 myeloperoxidase,21 xanthine oxidase (XO),22 and importantly NAD(P)H oxidases23 (Figure 1). However, the influence of ROS-producing enzymes, especially NADPH oxidases, in the development of atherosclerosis is ambiguous.23 Nox4-derived mitochondrial ROS were detrimental in older mice with atherosclerosis,24 and dominant negative mutant form of Nox4 decreased atherosclerosis formation.25 On the other hand, Nox4 knockout aggravated atherosclerosis,26 especially in diabetic mice.27Download figureDownload PowerPointFigure 1. Sources of ROS in the vascular wall. LOX indicates lipoxygenase; MPO, myeloperoxidase; NOS, nitric oxide synthase; NOX, NAD(P)H oxidase; SOD, superoxide dismutase; and XO, xanthine oxidase.High levels of oxidative stress can be counteracted by complex antioxidant cell systems that are crucial for the maintenance of redox balance. The key player that plays a primary role in the regulation of antioxidant gene response is nuclear factor erythroid 2–related factor 2 (Nrf2),28 encoded by the Nfe2l2 gene. However, depending on the mouse model used and cell type analyzed or even sex of animals, Nrf2 showed both pro- and antiatherogenic properties.29 Wire injury induces higher neointima formation in Nfe2l2−/− mice than in control animals.30,31 Global knockout of Nfe2l2 in Apoe−/− mice decreased the formation of atherosclerotic lesions.32–35 Transplantation of Nfe2l2−/− bone marrow to Apoe−/−35 or Ldlr−/−36 recipients attenuated atherosclerosis, what underlines proatherogenic activity of Nrf2 in myeloid cells. However, increased formation of plaque in Ldr−/− mice transplanted with Nfe2l2−/− bone marrow was also reported.37 On the other hand, activation of Nrf2 in SMCs38,39 or ECs40 was protective against atherosclerosis. Furthermore, knockouts of potent antioxidant enzymes, for example, Gpx1,41Prdx2,42Hmox1,43 can also aggravate plaque formation. Importantly, basal levels of ROS were crucial for the activation of endoplasmic reticulum (ER) stress response,44 maintenance of SMC contractile phenotype,45 or differentiation of SMC from stem cells.46,47 Thus, the regulation of oxidative stress is complex, and investigation of its role in the pathogenesis of atherosclerosis remains an important subject of many studies. Therefore, the aim of this article is to summarize the latest advances in the research on the role of oxidative stress in the modulation of cells that can affect the development of atherosclerosis.Endothelial CellsThe healthy endothelium is key for the functional maintenance of vascular system.48,49 Sustained ROS levels can contribute to the endothelial dysfunction, and further to its senescence and activation of an inflammatory response, and in turn lead to the development of atherosclerosis. NADPH oxidases, especially Nox4, play an ambiguous role in the development of atherosclerosis. Craige et al50 reported recently that Apoe−/− mice with endothelial-specific Nox4 overexpression (Apoe−/− Nox4EC) showed significantly smaller lesions than control Apoe−/− animals. Apoe−/− Nox4EC aortas contained, however, similar numbers of macrophages and did not differ from Apoe−/− mice in expressions of macrophage or inflammatory markers, that is, E- and P-selectin, VCAM1 (vascular cell adhesion molecule 1), and ICAM1 (intercellular adhesion molecule 1). On the contrary, Apoe−/− Nox4EC aortas contained higher Treg numbers, while numbers of effector T cells and plasma CXCL9 (chemokine [C-X-C motif] ligand 9) concentration were lower than in Apoe−/− controls. Both mouse lung ECs and human aortic ECs overexpressing Nox4 showed attenuated CXCL9 induction after interferon-γ.50In another study, downregulation of Nox4 in human aortic ECs, cultured under high glucose conditions, resulted in the increase of profibrotic CTGF (connective tissue growth factor), while endothelial H2O2 decreased.51 On the other hand, Nox4 overexpression caused a significant reduction in p-SMAD3 (phosphorylated mothers against decapentaplegic homolog 3). In diabetic Apoe−/− mice, aortic expression of both Nox1 and Nox4 increases after 10 weeks of diabetes mellitus. Further development of atherosclerosis in diabetic Apoe−/− mice led to the decline in Nox4 expression, while levels of Nox1 remained high. Importantly, Nox4 expression in carotid plaques is lower in patients with cardiovascular events or diabetes mellitus. Nox4 deletion in diabetic Apoe−/− mice increased, while deletion of Nox1 in diabetic Apoe−/− attenuated plaque formation in comparison to control diabetic Apoe−/− mice. Nox4-deficient Apoe−/− diabetic mice showed lower levels of aortic H2O2 while superoxide levels were increased. Nox1 deletion resulted in a decrease in both superoxide and tyrosine nitration. Furthermore, diabetic Nox4−/−Apoe−/− showed elevated serum and aortic MCP-1 (monocyte chemotactic protein 1) concentrations, as well as macrophage accumulation in the vessel wall. Nox4 knockdown also led to the increase of fibrillar collagens I and III in plaques, which was associated with elevated transforming growth factor-β expression and p-SMAD3 levels.51Nox4 and H2O2 regulate the response to ER stress, which is crucial for the induction of unfolded protein response.44 Importantly, both ER stress and response to it are activated during the development of atherosclerosis.52 Wu et al44 showed that tunicamycin-induced ER stress elevated H2O2 in ER in the Nox4-dependent way. Increased H2O2 concentrations led in turn to the oxidation of sarco/endoplasmic reticulum Ca2+-ATPase and an increase in cytosolic calcium concentration. High levels of Ca2+ caused next the activation of RasGRF (Ras-specific guanine nucleotide releasing factor), which then induced via Ras UPR (unfolded protein response), that is, BiP (binding immunoglobulin protein), CHOP (CCAAT enhancer–binding protein homologous protein), and phosphorylation of eIF2α (eukaryotic initiation factor 2α)44 (Figure 2A). Another important player in the ER stress response is XBP1 (X-box-binding protein 1).53 However, while spliced XBP1 (XPB1s), which positively regulates UPR, contributes to EC apoptosis and atherosclerosis formation, the unspliced XBP1 (XBP1u) can induce the antioxidant response.54 Interestingly, disturbed flow elevated expression of antioxidant genes in ECs in XBP1u- and HDAC3 (histone deacetylase 3)-dependent manner. Increased expression of XBP1u and HDAC3 in HUVECs has decreased with KDR (kinase insert domain receptor) or PI3K (phosphatidylinositol-4,5-bisphosphate 3-kinase) inhibition. What is more, spliced XBP1 decreased HDAC3 levels, while overexpression of XBP1u protected it. Furthermore, high expression of XBP1u protected ECs from H2O2-induced oxidative stress. The latter effect was mediated by the increased stability of Nrf2 transcription factor and upregulation of its target—Hmox154 (Figure 2B).Download figureDownload PowerPointFigure 2. Endoplasmic reticulum (ER) stress leads to the activation of Nox4 on ER surface but not cell membrane and increases H2O2 levels. Increased oxidative stress inactivates sarco/endoplasmic reticulum Ca2+-ATPase (SERCA), what causes the increase of cytoplasmatic Ca2+, activation of RasGRF, and Ras. Ras induces then unfolded protein response, namely BiP (binding immunoglobulin protein), CHOP (CCAAT enhancer–binding protein homologous protein), and phosphorylation of eIF2α (A). Disturbed flow causes ligand-independent activation of KDR, which induces formation of mTORC2-Akt1-XBP1u-HDAC3 complex and increases Akt1 phosphorylation, what contributes to the enhances stability of nuclear factor erythroid 2–related factor 2 (Nrf2). Nrf2, the regulator of antioxidant response, induces HMOX1 expression, which protects endothelial cells from oxidative stress (B).Laminar shear stress increases expression of transcription factor EB (TFEB) in cultured human ECs.55 TFEB is a basic helix–loop–helix transcription factor, which regulates lysosomal biogenesis.56 Laminar shear stress induced TFEB translocation to the nucleus and decreased mTOR (mechanistic target of rapamycin) activity in ECs. Interestingly, atherosclerosis-resistant regions of mouse aorta had higher TFEB levels than atherosclerosis-prone ones. TFEB overexpression reduced levels of H2O2 and superoxide, as well as expression of SELE, MCP1, VCAM1, IL1B, IL6, and IL8, in ECs and human coronary artery ECs. ECs overexpressing TFEB also had a higher expression of HMOX1, SOD2, and TXN1. Importantly, TFEB can bind to its putative binding sites in the HMOX1 promoter and SOD2 intron 2. However, knockdown only of HMOX1 in TFEB-overexpressing EC potently attenuated the anti-inflammatory effect of TFEB. Furthermore, although TFEB is known to regulate autophagy,56ATG5 silencing or pharmacological inhibition of autophagy did not block TFEB anti-inflammatory activity. mTie2-TFEB mice overexpressing TFEB in ECs showed less leukocyte rolling after lipopolysaccharide treatment than wild-type (WT) animals. Finally, Apoe−/−mTie2-TFEB mice on high cholesterol diet had less atherosclerotic lesions than Apoe−/− control animals.55Nrf2 in steady state conditions is sequestered by Keap1 and directed to the proteasomal degradation.57 Therefore, the regulation of Nrf2 activity is of potential clinical interest. Recent work by Xie et al58 showed that hydrogen sulfide could enhance Nrf2 activity in diabetic atherosclerotic mice and ECs treated with high glucose and oxidized low-density lipoproteins (LDLs). Streptozotocin-treated Ldlr−/− mice on a high-fat diet had less atherosclerotic lesions when treated with GYY4137—H2S donor. H2S treatment reduced ROS levels and VCAM1 and ICAM1 expressions and enhanced Nrf2 nuclear translocation in both diabetic atherosclerotic mice and ECs treated with high glucose and oxidized LDLs. H2S did not show the protective effect in Ldlr−/−Nfe2l2−/− mice or in ECs with silenced Nfe2l2. Interestingly, H2S enhanced Nrf2 activity by S-sulphydrylation of Keap1 at Cys151.58Interestingly, lysophosphatidylcholines (LPC)—proinflammatory lipids—can contribute to the ROS formation in atherosclerosis.59 The concentration of LPC, as well as expression of Pla2g7 and Pla2g4c, were higher in Apoe−/− aortas than in control ones. LPC quickly induced both mitochondrial ROS and to lesser levels cytoplasmatic ROS. Inhibition of NADPH oxidases further increased ROS levels. LPC increased cytoplasmatic and mitochondrial Ca2+, what was crucial for the induction of mitochondrial ROS production. LPC elevated mitochondrial proton leak but not ATP production and did not affect levels of mitochondrial superoxide dismutase or uncoupling protein 3. Furthermore, LPC increased expression of proatherogenic genes, such as ICAM1, IL6, and MMP2, what was attenuated by mitoTEMPO—mitochondrial ROS scavenger. Finally, mitoTEMPO inhibited LPC-induced monocyte adhesion to ECs in vitro, and leukocyte rolling and numbers of total and inflammatory monocytes in Apoe−/− high-fat diet mice.59Mitochondrial oxidative stress can be decreased by thioredoxin 2 (Txn2) that acts together with thioredoxin reductase 2 (Txnrd2) and peroxiredoxin 3 (Prdx3).60 Importantly, the endothelial-specific knockout of Txnrd2 impaired angiogenesis and arteriogenesis after femoral artery ligation and vascular function.61 Lack of Txnrd2 in ECs rendered them more proinflammatory and prothrombotic. Embryonic endothelial progenitor cells deficient in Txnrd2 showed higher intracellular ROS levels and more positive mitochondrial potential. Such cells had also decreased angiogenic properties in vitro, what could be reversed by Txnrd2 re-expression.61 Therefore, the imbalance of redox-mediated signaling and the consequence results in endothelial dysfunction that is a key event for the development of atherosclerosis.Smooth Muscle CellsFine-tuning of SMC proliferation and differentiation status is crucial for the development of atherosclerosis62 because they can contribute to the formation of neointima but, on the other hand, can stabilize the plaque and prevent its rupture.11,63 Oxidative stress in atherosclerosis results primarily from the activity of NAD(P)H oxidases.64 However, their role in SMCs, especially in the context of atherosclerosis, remains ambiguous. Mice lacking Nox1 developed less neointima in femoral arteries after the wire injury than control animals.65 On the other hand, SMC-specific overexpression of Nox1 did not enhance neoplasia formation. Nox1-deficient SMCs were further characterized by lower proliferation rate and migration than WT SMCs. Overexpression of Nox1 slightly enhanced both of the latter characteristics. Lower migration rate in Nox1-deficient cells was associated with changes in the regulation of actin cytoskeleton, that is, cofilin phosphorylation and mDia1 expression increased while PAK1 levels were lower.65 Furthermore, expression of Nox1 was higher in neointimal SMCs than in normal medial cells.66 Elevated levels of Nox1 were associated with ERK1/2 (extracellular signal-regulated kinases 1/2) activation and enhanced MMP-9 (matrix metallopeptidase 9).66Importantly, Nox1 forms canonical or hybrid systems with p47phox in place of NoxO1 in the hybrid system. Production of superoxide by hybrid Nox1 is regulated by ezrin–radixin–moesin-binding phosphoprotein 50 (EBP50).67 Expression of EBP50 is elevated in SMC after the angioplasty.68 EBP50 increased SMC proliferation with inhibiting antimitogenic parathyroid hormone type 1 receptor69 and regulation of p21cip1 and Skp2.70 Al Ghouleh et al67 recently showed that EBPO50−/− SMCs produce less superoxide than WT cells when treated with H2O2 or angiotensin II. In line with that, aortic rings from both EBPO50−/− or Nox1−/− mice show no increase in superoxide production after stimulation with angiotensin II. What is more, femoral arteries from EBPO50−/− mice had lower lipid peroxidation after lipopolysaccharide treatment than control animals. EBPO50 binds to the p47phox subunit of the hybrid Nox1 system and promotes Nox1 activity.67Expression of NADPH oxidases changes dynamically during the development of atherosclerosis.71 Nox1 levels were increased in aortas Apoe−/−Ldr−/− prior to the lesion formation and then normalized in later stages. Nox4 expression was elevated on later stages of atherosclerosis. SMCs isolated from mice with advanced atherosclerosis had lower Nox1 and higher Nox4 and p22phox levels and WT SMCs produced more superoxide or H2O2, but also expressed more catalase, SOD1 (superoxide dismuate 1), and SOD2. Furthermore, elevated Nox4 expression in SMCs increased their apoptosis in basal conditions and after stimulation with H2O2.71 Vendrov et al24 showed that while knockout of Ncf1 gene, encoding for p47phox, an essential regulatory subunit of both Nox1 and Nox2, is protective in young Apoe−/− mice, the effect is lost in aged animals.24 Areas of aortic atherosclerotic lesions, staining with DHE, levels of DNA oxidative damage, or macrophage infiltration were similar in aged Apoe−/− and Apoe−/−Ncf1−/− animals. What is more, SMCs from both types of aged mice showed similar upregulation of superoxide and H2O2 when treated with thrombin. Importantly, neither xanthine oxidation inhibition, nor lipoxygenase inhibition but Nox1/4 inhibitor attenuated ROS induction after stimulation with thrombin. The mitochondrial source of elevated ROS in aged SMCs was confirmed with the analysis of H2O2 levels in isolated mitochondria. Both SMC in aortas of aged animals or mitochondria in SMC from aged mice showed high staining against Nox4. When Nox4 was inhibited, levels of mitochondrial ROS in aged SMCs decreased. Interestingly, elevated levels of Nox4 were also detected in samples from older human donors. Last but not least, treatment with mitoTEMPO reduced thrombin-induced VCAM1 upregulation in SMCs in vitro, but also aortic wall stress and lesion area, and prevented oxidative damage in Apoe−/− mice.24SMC-specific expression of a human-dominant negative form of NOX4 (NOX4 P437H) resulted in decreased neointima formation after carotid artery denudation in FVB/N mice25 or lower atherosclerosis in Apoe−/− mice.72 Protective effect of dominant negative NOX4 in Apoe−/− mice was also associated with lower numbers of infiltrating macrophages. Overexpression of NOX4 P437H resulted in inhibition of soluble epoxide hydroxylase-2 expression and decreased proinflammatory signaling and levels of VCAM1, MCP1, and ICAM1, what was mirrored by lower macrophage adhesion.72 Importantly, soluble epoxide hydroxylase-2 downregulation reduced the expression of thrombospondin-1,72 which was crucial for the observed effects of NOX4 P437H on proliferation and migration in Apoe+/+ mice.25 On the other hand, Nox4 in Apoe−/− mice had a protective effect both in mice with spontaneous atherosclerosis development or after partial carotid artery ligation.26Nox4 deletion led to the decrease in H2O2 in mouse aortas, associated with the increase in superoxide levels. Microarray analysis of gene expression in WT or Nox4-deficient animals showed increased proinflammatory signaling in aortas from the Nox4-deficient mice.26Nox4 was protective also in diabetic Apoe−/− mice with streptozotocin-induced diabetes mellitus. Such mice had lower aortic expression of smooth muscle contractile markers, while levels of PDGF (platelet-derived growth factor), vimentin, and osteopontin increased.27Nox4 deletion further increased aortic expression of PDGF, collagen I, and Ki-67. SMCs isolated from Nox4-deficient animals had lower expression of smooth muscle contractile markers, produced less H2O2 but more superoxide and upregulated Nox1. Silencing of the latter gene resulted in the attenuation of PDGF-BB and osteopontin production and Ki-67 expression levels. Abnormal expression of PDGF-BB in Nox4−/− SMCs could also be reversed by PDGF inhibition or treatment with H2O2.27 What is more, Nox4 together with Nrf2 play significant roles in the CD38 signaling pathway, which is crucial for the maintenance of SMC contractile phenotype.45 Coronary arterial myocytes isolated from Cd38−/− mice had lower calponin and SM22α expression, accompanied by an increase in vimentin and PCNA (proliferating cell nuclear antigen). 7-Ketocholesterol, known atherogenic stimulus, further exacerbated the effect of Cd38 knockout. Lack of CD38 caused a decrease in Nrf2 expression and activity. Nfe2l2 silencing in coronary arterial myocytes enhanced reduction of contractile markers caused by 7-ketocholesterol. CD38 activity is crucial for the induction of Nox4 and superoxide production, which is then necessary for the Nrf2 translocation to the nucleus and maintenance of arterial myocytes contractile phenotype.45Interestingly, Nfe2l2 overexpression in human aortic SMCs resulted in an increase of GCLC, GCLM, and HMOX1 expression and decreased SMC proliferation.38 The latter effect could be partially reversed with tin protoporphyrin IX, which is an HMOX1 (heme oxygenase 1) inhibitor. Furthermore, Ad Nfe2l2 reduced oxidative stress, macrophage infiltration, and MCP-1 levels in balloon-injured aortas in rabbits. Finally, Nfe2l2 gene transfer in vivo decreased both SMC proliferation and apoptosis, and therefore, it had no effect on neointimal neoplasia.38 What is more, PDGF induces Nrf2 nuclear translocation and increases expression of its target genes: Nqo1, Hmox1, and Txnrd1.31Nfe2l2 silencing enhances PDGF-induced migration of SMCs. Furthermore, it leads to prolonged activation of Rac1, which can increase NADPH oxidase activity and ROS production. Nfe2l2 knockout also caused enhanced PDGF signaling, namely ERK1/2 phosphorylation, which can be inhibited with antioxidants, such as N-acetylcysteine. Femoral injury in Nfe2l2−/− mice resulted in decreased lumen area and higher neointima formation, with no significant changes in media and with greater ERK1/2 activation in neointima.31 Sulforaphane, which is Nrf2 activity inductor, reduces neointima formation in injured femoral arteries of mice fed a Western diet.39 In vitro, sulforaphane reduced leptin-induced SMC proliferation, cyclin D1 expression, and phosphorylation of both p70S6kinase and ribosomal S6 protein.39 Ashino and et al30 showed that injured femoral arteries contained regions of TUNEL+ (terminal deoxynucleotidyl transferase dUTP nick-end labeling) apoptotic cells that overlapped with regions rich in Nrf2high cells. Silencing of Keap1 in rat SMC resulted in higher Nrf2 levels and activity, resulted in higher caspase-3/7 activity, and increased apoptosis. Nfe2l2 silencing rescued the proapoptotic phenotype of Keap1-deficient cells. Finally, wire injury of femoral arteries in Nfe2l2−/− mice led to the enhanced formation of neointima, which contained less apoptotic cells. Therefore, authors proposed that Nrf2 protect from neointima also because of the enhancement of SMC apoptosis.30Interestingly, although Atg7−/− SMCs had higher Nrf2 activity, expression of its target genes, and resistance to oxidative stress, they showed accelerated senescence.73 The latter effect was related to the defective autophagy and accumulation of Sqstm1/p62, which is a Nrf2 target, but also contributes to its activation.74 In vivo, deletion of Atg7 in SMCs caused enhanced senescence and promoted neointima formation.73 On the other hand, carotid artery ligation in Sqsmt1−/− mice caused more neointima formation.75 SMCs isolated from Sqstm1−/− aortas proliferated and migrated faster.75 Overexpression of HMOX1, enzyme degrading toxic heme to biliverdin, carbon monoxide and ferrous iron, and target of Nrf2 activity in SMCs decreased their migration in response to PDGF-BB.76 Inhibitory effect of heme oxygenase-1 depended on its enzymatic activity and was mimicked by carbon monoxide. Increased activity of heme oxygenase-1 led to the upregulation of both VEGFA (vascular endothelial growth factor A) and VEGFR1 (vascular endothelial growth factor receptor 1). The latter one formed a complex with PDGFRβ, which attenuated PDGFRβ signaling.76 Activation of another Nrf2 target—NAD(P)H: quinone oxidoreductase-1—by β-lapochone reduced neointima formation after balloon injury in rat carotid arteries and inhibited PDGF-induced SMC proliferation. Increased NQO1 (4-nitroquinoline 1-oxide) activity caused AMPK activation mediated by LKB1 but not CaMKKβ (Ca2+/calmodulin-dependent protein kinase kinase beta).77 Taken together, Nrf2 is a crucial player in maintaining the hemostasis of the vessel wall via influencing inflammatory response, SMC proliferation, and neointimal formation.Increased SMC apoptosis may lead to the plaque instability and rupture.78 Wnt/β-catenin/WISP-1 (WNT1-inducible-signaling pathway protein 1) represents example of the pathway that can protect SMCs from apoptosis79 and oxidative stress but, on the other hand, enhance SMC migration and contribute to intimal thickening.80 Mill et al79 showed recently that macrophage-derived Wnt5a could protect SMC from H2O2-induced apoptosis. Exogenous Wnt5a increased amounts of active and nuclear β-catenin in SMC and induced TCF (transcription factor) signaling. However, TCF pathway was attenuated when cells were costimulated with H2O2, but this effect was dependent on the active β-catenin. Protective activity of Wnt5a against H2O2-induced activity was lost after LRP5/6 silencing, which suggests that it is regulated by Wnt/Frizzled/LRP5/6 pathway. Importantly, Wnt5a induced cytoprotective WISP-1 in CREB-dependent manner. Interestingly, SMC in stable plaques in human coronary arteries showed higher staining for WISP-1 than those in unstable plaques.79 Wnt2 also induced WISP-1 expression in SMC in β-catenin-dependent manner.80 Levels of Wnt2 and WISP-1 were elevated in ligated carotid arteries. Furthermore, WISP-1 overexpression promoted intimal thickening, which was reduced in Wnt2+/− animals.80Finally, although high glutathione levels and, that is, high ratio of reduced to oxidized glutathione are considered protective, too much of it can also lead to the enhanced vascular remodeling. Network analysis of gene expression in human carotid neointima showed that the most significantly downregulated network was the one with GPX1 in its hub.81 Furthermore, Apoe−/− mice with deleted Gpx1 were characterized with increased atherosclerotic plaque, which contained more SMCs but not macrophages. However, increased macrophage infiltration in Gpx1−/−Apoe−/− was previously reported by Torzewski et al.41 Furthermore, balloon angioplasty and stenting in Gpx1+/+Apoe−/− mice decreased Gpx1 expression. Global knockout of Gpx1 in Apoe−/− mice caused an increase in superoxide levels in media but not adventitia or endothelium. Effect of lack of Gpx1 on SMCs was mediated by the high activity of ROS1 receptor tyrosine kinase. Inhibition of tyrosine kinase activity or ROS1 silencing could decrease enhanced proliferation and migration of Gpx1−/−Apoe−/− SMCs. Gpx1−/− SMCs contained more glutathione, which resulted from its higher synthesis. Elevated levels of glutathione caused reductive stress that contributed to S-glutathionylation and inactivation of SHP-2 (tyrosine-protein phosphatase nonreceptor type 11) phosphatase, what then inhibited ROS1 inactivation81 (Figure 3). Interestingly, Izawa et al82 reported that buthionine sulfoximine, which is glutathione inhibitor, elevated vessel wall superoxide levels, but on the other hand decreased angiotensin II–induced vessel remodeling.Download figureDownload PowerPointFigure 3. Reductive stress contributes to ROS1 activation. Vascular injury leads to the downregulation of Gpx1, what is accompanied with elevated GSH synthesis. The increase of cellular GSH leads to S-glutathionylation of SHP-2 phosphatase, what in turn inhibits its activity and leads to prolonged activation of ROS1 receptor tyrosine kinase. ROS1 activity contributes t