Vanadium Derivative Exposure Promotes Functional Alterations of VSMCs and Consequent Atherosclerosis via ROS/p38/NF-κB-Mediated IL-6 Production

Vanadium is a transition metal widely distributed in the Earth’s crust, and is a major contaminant in fossil fuels. Its pathological effect and regulation in atherosclerosis remain unclear. We found that intranasal administration of the vanadium derivative NaVO3 significantly increased plasma and urinary vanadium levels and induced arterial lipid accumulation and atherosclerotic lesions in apolipoprotein E-deficient knockout mice (ApoE−/−) murine aorta compared to those in vehicle-exposed mice. This was accompanied by an increase in plasma reactive oxygen species (ROS) and interleukin 6 (IL-6) levels and a decrease in the vascular smooth muscle cell (VSMC) differentiation marker protein SM22α in the atherosclerotic lesions. Furthermore, exposure to NaVO3 or VOSO4 induced cytosolic ROS generation and IL-6 production in VSMCs and promoted VSMC synthetic differentiation, migration, and proliferation. The anti-oxidant N-acetylcysteine (NAC) not only suppresses IL-6 production and VSMC pathological responses including migration and proliferation but also prevents atherosclerosis in ApoE−/− mice. Inhibition experiments with NAC and pharmacological inhibitors demonstrated that NaVO3-induced IL-6 production is signaled by ROS-triggered p38-mediated NF-κB-dependent pathways. Neutralizing anti-IL-6 antibodies impaired NaVO3-mediated VSMC migration and proliferation. We concluded that NaVO3 exposure activates the ROS-triggering p38 signaling to selectively induce NF-κB-mediated IL-6 production. These signaling pathways induce VSMC synthetic differentiation, migration, and proliferation, leading to lipid accumulation and atherosclerosis.


Introduction
Increasing evidence has indicated that exposure to environmental pollutants such as particulate matter (PM) and ultra-fine particles (UFPs), particularly residual oil fly ash (ROFA), lead to systemic pro-oxidant accumulation and inflammation, both of which play a critical role in the development of cardiopulmonary diseases [1,2], a major health issue worldwide with significant morbidity and mortality [3]. The main underlying cause of cardiovascular diseases is atherosclerosis, which is recognized as a chronic inflammatory disease [4]. Evidence has indicated that exposure to PM air pollution elicits numerous pathological and biological responses, including systemic inflammation and oxidants, which increases the relative risk of cardiovascular mortality [2,[5][6][7]. Although the toxic effects of PM on cardiovascular disease are usually assessed by its particle mass, there is increasing evidence that the chemical composition and size/morphology of PM may be a major contributor to its pathotoxicity toward the cardiovascular system [6,8]. Vanadium is one of the metal compounds found in atmospheric suspended matter and is classified as PM2.5 by its size. Considering the rich speciation chemistry of vanadium, the specific forms of vanadium oxides are likely a combination V 2 O 5 , VO 2 , V 2 O 3 , and some of their salts [9]. Suspended particles containing vanadium are inhaled into the respiratory system and reach the circulation and organs, adversely affecting the individual's physical health [10,11]. Inhalation of air containing vanadium can lead to pathological effects on human organ systems including the respiratory system, lungs, neurological system, eyes, hematopoietic tissue, immune system, and cardiac physiology, potentially through the induction of reactive oxygen species (ROS) and inflammatory responses [10][11][12]. These inhalations studies with mice have demonstrated that the form of vanadium is critical, with some being much more toxic than others [13,14].
Metallic vanadium, due to several oxidation states, forms numerous inorganic compounds including vanadyl sulfate, sodium metavanadate, sodium orthovanadate, and vanadium pentoxide. The pentavalent form (VO 3− ) and quadrivalent form (VO 2+ ) of vanadium exist in extracellular and intracellular body fluids, respectively, which are found to regulate the activity of various metabolic enzymes. Therefore, a low concentration (µg) of vanadium is involved in the development and growth of organisms and is considered essential for maintaining health. Furthermore, vanadium derivatives have been used as a common supplement to enhance athletes' weight training at doses up to 60 mg/d [15]. In vitro and in vivo studies have shown that vanadium compounds have anti-diabetic effects by increasing glucose transport activity and improving glucose metabolism [16]. Recently, vanadium compounds are being considered for the application in cancer treatment [17]. However, high air concentrations of vanadium occur in the occupational setting as a result of the presence of vanadium oxides in dust, leading to humans being exposed to an excessive amount of vanadium, with resulting toxicity. There is growing evidence that inhalation of vanadium-rich environmental pollutions, particularly vanadium pentoxide, can cause damage to human organ systems [10][11][12]. In addition, vanadium pentoxide is classified as possibly carcinogenic to humans by the IARC (International Agency for Research on Cancer). These observations indicate that vanadium derivatives with different physical and chemical properties exhibit different physicochemical properties, which confer beneficial properties or toxic effects. However, their adverse effects and susceptibility to vascular diseases such as atherosclerosis and the pathological mechanisms involved remain undetermined.
It is well-known that mice are most frequently used in various diseases such as atherosclerosis, a multifaceted disease which can involve a pathological progression. Therefore, the choice of the right transgenic mouse model is vital for atherosclerosis research. Since wild-type mice are not thought to develop atherosclerosis unless challenged for long periods with Western diets, transgenic mouse models of atherosclerosis have been developed to study experimental atherosclerosis. The first line of apolipoprotein E-deficient knockout mice (ApoE −/− ) was developed in 1992. Further, ApoE −/− mice are frequently used and are valuable tools in atherosclerosis research. Notably, ApoE −/− mice can develop extensive atherosclerotic lesions on a chow diet, and the resulting atherosclerotic lesions are comparable to human lesions. Therefore, ApoE −/− mice have been used to develop new drugs against atherosclerosis. Accordingly, ApoE −/− mice, but not wild-type mice, were used in this study to investigate the effect of vanadium derivatives, including NaVO 3 and VOSO 4 , on atherosclerosis.
Excessive and inappropriate activation of the innate immune system has been implicated in the development of chronic metabolic diseases including atherosclerosis and systemic inflammation [4,[18][19][20]. In general, the mechanisms underlying the induction of inflammatory discords leading to most pathological conditions remain to be determined. However, a disturbance in the reduction-oxidation (redox) equilibrium of cells and tissues may lead to an overwhelming proinflammatory state, which leads to cellular dysfunction and tissue injury. ROS such as superoxide, hydrogen peroxide, and hydroxyl radicals are highly reactive and thus harmful to health, causing chronic diseases such as atherosclerosis [21,22], a complex chronic vascular disease that progressively occludes the lumen of large and medium-sized arteries with plaques. This initiates with endothelial injury followed by immune and proinflammatory cell accumulation, lipid deposition, and progressive inflammatory responses [4,23,24]. Further, the inflammatory responses in the microenvironment of atherosclerotic lesions drive medial smooth muscle cell migration and proliferation into the intima, consequently causing plaque formation, which is a key event in the pathophysiology of atherosclerosis [3,[25][26][27]. Vascular smooth muscle cells (VSMCs) reside in the media of normal blood vessels, where they are quiescent and assume a contractile phenotype. Under pathological conditions, VSMCs transform into a highly proliferative synthetic phenotype with a loss of contractile markers (SM22α, α-SM actin, and E-cadherin) and induction of synthetic markers such as vimentin [3,27]. Furthermore, they contribute to vascular inflammation by producing proinflammatory cytokines such as interleukin 6 (IL-6) [28]. These VSMC pathophysiological alterations cause atherosclerotic lesions and plaque formation.
Given that ROS and inflammation play a pathogenic role in atherosclerosis and vanadium exposure can promote cellular ROS and inflammation, we hypothesized that excessive vanadium exposure may have pathological significance in VSMC survival, proliferation, and/or migration as well as atherosclerosis. Our results from in vivo and in vitro analyses reveal that vanadium derivatives, VOSO 4 and NaVO 3 , selectively induces IL-6-dependent VSMC pathological responses including phenotypic alternations, migration, and proliferation, with consequent atherosclerotic plaque formation which is mediated by NADPH oxidase-derived ROS generation, leading to p38-mediated NF-κB (nuclear factor kappa light chain enhancer of activated B cells) activation and NF-κB-dependent IL-6 production.

Intranasal Administration of NaVO 3 Induces Atherosclerosis in ApoE −/− Mice
This study showed that the plasma and urinary vanadium concentrations significantly increased in NaVO 3 -exposed mice (mean 407.5 ± 50.4 ng/mL in plasma, (n = 19) and 469 ± 147.4 µg/g creatinine in urine, (n = 10)) as compared to control mice exposed to endotoxin-free water (mean 30.41 ± 1.881 ng/mL in plasma, (n = 6) and 0.69 ± 0.18 µg/g creatinine in urine, (n = 5)) ( Figure 1A,B), suggesting that intranasal administered NaVO 3 can be absorbed into the circulation system. Furthermore, intranasal administration of NaVO 3 induced arterial lipid accumulation in the murine aorta but did not affect circulating lipid levels (cholesterol and triglyceride) compared to vehicle-exposed mice ( Figure 1C,D and Figure S1), and was accompanied by increasing plasma IL-6 levels ( Figure 1E). Notably, there were no damaging effects to the kidney, liver, or heart, but mild lung inflammation including inflammatory leukocyte infiltration was observed in the NaVO 3 -exposed mice ( Figure S2), suggesting that intranasal administration of NaVO 3 has a major effect on the arteries rather than the lungs and other organs. These results suggest that NaVO 3 may be an atherosclerosis initiator or inducer.
The decrease in VSMC differentiation marker proteins such as SM22α in atherosclerotic lesions is a common characteristic of atherosclerosis [3,29], we determined whether intranasal administration of NaVO 3 significantly decreased SM22α. A significant decrease in the immunopositive areas for SM22α in the atherosclerotic lesions ( Figure 1F) was found to be highly associated with enhanced neointimal formation in the atherosclerotic lesion of NaVO 3 -exposed mice ( Figure 1G). Quantitative analysis confirmed significant downregulation of SM22α in the atherosclerotic plaque of NaVO 3 -exposed mice as compared to vehicle-treated mice ( Figure 1H-I). In addition, NaVO 3 significantly modulated the expression of VSMC differentiation protein markers, decreased SM22α and E-cadherin, and increased vimentin in a dose-dependent manner ( Figure 1J). These results suggest that NaVO 3 induces atherosclerotic lesions by suppressing SM22α and E-cadherin and increasing vimentin expression to alter VSMC phenotype switching. The levels of vanadium in urine were normalized by creatinine (CRE). Each dot denotes an individual mouse. The solid black line denotes the mean value. (C) Lipid contents in aorta from mice treated with vehicle and NaVO3 were analyzed by Oil Red O staining. Scale bars represent 5 mm. (D) The area of positive staining for Oil Red O was quantified using ImageJ software as a percentage of total aortic area (n = 9 vehicle; n = 8 NaVO3). (E) Plasma interleukin 6 (IL-6) levels in vehicle treatment (n = 9) and NaVO3 treatment (n = 8) were measured by ELISA. (F) SM22α content in aorta tissues from mice was analyzed by immunohistochemistry. Immunopositive areas shown are shown at 400× magnification. An: adventitia; Lu: lumen; Neo: neointima; Med: media. Negative control represents staining with an irrelevant isotype control antibody. Scale bars represent 100 µm. (G) Paraffin-embedded tissues with atherosclerotic plaque (black arrow) were observed by hematoxylin and eosin (H&E) stain at 40× magnification. Scale bars represent 1 mm. (H) Immunopositive areas and atherosclerotic plaque in paraffin-embedded aorta tissues of mice treated with vehicle (n = 5) or NaVO3 (n = 10) were quantified using ImageJ software as a percentage of total aortic area in each section. Data in (A,B), (D,E), and (H,I) represent mean ± SEM. (J) Vascular smooth muscle cells (VSMCs) were treated with a different amount of NaVO3 for 48 h. Cell lysates were immunoblotted with antibodies for SM22α, E-cadherin, vimentin, or β-actin. Experiments were repeated three times with similar results. Data represent mean ± SEM of three experiments. * p < 0.05; ** p < 0.01; *** p < 0.001.

IL-6 Is Essential for VOSO 4 -and NaVO 3 -Induced VSMC Proliferation and Migration
Given that NaVO 3 induced atherosclerotic lesions and VSMC phenotype alternation accompanied by increasing plasma IL-6 levels ( Figure 1), we wondered if IL-6 secreted by VSMCs may be involved in vanadium salt-induced VSMC migration and proliferation. Our results revealed that VOSO 4 and NaVO 3 induced IL-6 production in VSMCs in a time-dependent and dose-dependent manner ( Figure 2E). Furthermore, IL-6 neutralizing antibodies but not control IgG significantly inhibited not only NaVO 3 -induced IL-6 production but also NaVO 3 -induced migration and proliferation of VSMCs ( Figure 2F-H), suggesting that NaVO 3 -induced VSMC proliferation and migration is mediated by the release of IL-6 into the extracellular milieu.

ROS-Mediated IL-6 Is Essential for NaVO 3 -Mediated VSMC Functions
Our results revealed that plasma ROS levels were increased in NaVO 3 -exposed mice with atherosclerosis compared with vehicle control mice ( Figure 3A). In addition, we wondered if NaVO 3 also induced VSMC ROS generation. Intracellular ROS level as assessed by the cell-permeable dye 6-carboxy-2,7-dichlorodihydrofluorescein diacetate (DCFDA) was increased in VSMCs exposed to NaVO 3 in a time-dependent manner ( Figure 3B). Notably, this ROS induction was suppressed by N-acetylcysteine (NAC) in a dose-dependent manner ( Figure 3C), which was accompanied by blocking VSMC proliferation and migration without affecting VSMC viability ( Figure 3D-E and Figure  S3A). In addition, NAC dose-dependently recused NaVO 3 -mediated reduction of SMα and SM22α and enhancement of vimentin ( Figure 3F). These results suggest that cytosolic ROS participates in NaVO 3 -mediated VSMC pathology. (n = 9) or NaVO 3 (n = 8) once weekly for 12 weeks were measured by thiobarbituric acid reactive substances (TBARS) assay. The solid black line denotes the mean value. (B) Quiescent VSMCs were treated with NaVO 3 for the indicated time. Intracellular ROS levels were measured by 6-carboxy-2,7-dichlorodihydrofluorescein diacetate (DCFDA). (C-E) After pretreating VSMCs with different concentrations of n-acetyl-l-cysteine (NAC) for 30 min, cells were stimulated with vehicle or NaVO 3 (1 µg/mL) for 24 h. (C) Intracellular ROS levels and mitochondrial ROS were measured by DCFDA. (D) VSMC proliferation was measured by BrdU incorporation assay. (E) VSMC migration was then measured by the transwell assays. (F) VSMCs were treated with NaVO 3 (1 µg/mL) with or without different concentrations of NAC for 48 h. Cell lysates were immunoblotted with antibodies for vimentin, smooth muscle α-actin (SMα), SM22α, or β-actin. Densitometry analysis of SMα, SM22α, and vimentin protein expression relative to β-actin. Data represent mean ± SEM of three experiments. * p < 0.05; ** p < 0.01; *** p < 0.001.
In addition to mitochondria, the NADPH oxidases family is thought to be responsible for the production of cytosolic ROS [30,31]. We thus used NADPH oxidases pharmacological inhibitor VAS2870 to investigate whether NADPH oxidases involve in NaVO 3 -induced cytosolic ROS production. The results revealed that VAS2870 dose-dependently reduced NaVO 3 -induced cytosolic ROS production in VSMCs and had no effect on VSMC viability. (Figure 4C and Figure S3B). This is accomplished by abolished NaVO 3 -induced VSMC IL-6 and VSMC migration ( Figure 4D,E). These results indicate that ROS derived by NADPH oxidases contributes to NaVO 3 -induced cytosolic ROS production and VSMC pathological responses.
Given that ROS and IL-6 are involved in NaVO 3 -induced VSMC proliferation and migration, we thus determined a pathological connection between ROS and IL-6. NAC significantly dose-dependently reduced NaVO 3 -induced IL-6 production ( Figure 4F). Conversely, anti-IL-6 neutralizing antibodies did not affect NaVO 3 -induced intracellular ROS production ( Figure 4G). These results indicate that NaVO 3 -induced IL-6 production is mediated through intracellular ROS and the ROS-mediated IL-6 is required for NaVO 3 -induced VSMC proliferation and migration.
2.6. Anti-Oxidant N-Acetylcysteine Prevents NaVO 3 -Induced Atherosclerosis in ApoE −/− Mice Induction of VSMC ROS generation by NaVO 3 could contribute to excessive plasma ROS and IL-6 and consequent atherosclerotic lesions. We thus determined whether exogenous anti-oxidant N-acetylcysteine administration reduces plasmas ROS and IL-6 levels induced by NaVO 3 and rescues mice from NaVO 3 -induced atherosclerosis. ApoE −/− mice exposed to NaVO 3 were injected with NAC (250 mg/kg) or vehicle three times a week for 12 weeks and atherosclerotic lipid accumulation and plasma ROS and IL-6 were measured. Mice treated with NAC suppressed NaVO 3 -induced vascular lipid accumulation and plasma IL-6 production ( Figure 6A-C), correlating with a significant decrease in plasma ROS ( Figure 6D). To further confirm the protective effect of NAC on atherosclerosis, the suppression of atherosclerotic pathological alternations was evaluated in the NaVO 3 -exposed mice administered with a different dose of NAC (150 and 250 mg/kg). The results suggest that although NAC had no effect on vanadium level of urine ( Figure S5A), it not only dose-dependently reduced NaVO 3 -induced plasma ROS but also inhibited NaVO 3 -induced atherosclerotic plaque and aortic lipid accumulation when compared with vehicle treatments ( Figure 6E-G and Figure S5B), which concurrently attenuated lung injury ( Figure S5C). In addition, NaVO 3 -mediated IL-6 enhancement and SM22α reduction in the atherosclerotic lesions were suppressed by NAC in a dose-dependent manner ( Figure 6H-J), indicating that NaVO 3 triggers ROS-meditated IL-6 production to induce atherosclerosis.  ApoE −/− mice administered NaVO 3 (4 mg/kg) once a week were injected intraperitoneally with saline (vehicle) or NAC (250 mg/kg) three times weekly for 12 weeks (n = 9 vehicle; n = 9 NaVO 3 ; n = 8 NaVO 3 + NAC). (A) Lipid contents in aorta were analyzed by Oil Red O staining. Scale bars represent 5 mm. (B) The area of positive staining for Oil Red O was quantified using ImageJ software as a percentage of total aortic area. (C) Plasma IL-6 levels were measured by ELISA. (D) Plasma ROS levels were measured by TBARS assay. (E-J) ApoE −/− mice administered NaVO 3 (4 mg/kg) once a week were injected intraperitoneally with different amounts of NAC (150 or 250 mg/kg) three times weekly for 12 weeks (n = 5 vehicle; n = 10 NaVO 3 ; n = 5 NaVO 3 + 150 mg/kg NAC; n = 8 NaVO 3 + 250 mg/kg NAC). (E) Plasma ROS levels were measured by TBARS assay. (F) Paraffin-embedded tissues with atherosclerotic plaque (black arrow) were observed by hematoxylin and eosin stain at 40× magnification. Scale bars represent 1 mm. (G) Atherosclerotic plaque in paraffin-embedded aorta tissues were quantified using ImageJ software as a percentage of total aortic area in each section. (H) IL-6 and SM22α contents in aorta tissues were analyzed by immunohistochemistry. Immunopositive areas are shown at 400× magnification. An: adventitia; Lu: lumen; Neo: neointima; Med: media. Negative control represents staining with an isotype control antibody. Scale bars represent 100 µm. (I) IL-6 and (J) SM22α immunopositive areas in paraffin-embedded aorta tissues were quantified using ImageJ software as a percentage of total aortic area in each section. Data represent mean ± SEM. The solid black line denotes the mean value. * p < 0.05; ** p < 0.01; *** p < 0.001.

Discussion
Although the causes of vascular diseases are complex, it is being recognized that environmental factors such as PM can play a significant role [6,33]. Metals in PM are considered to contribute to the particle's toxic effects. Numerous epidemiological studies link PM metal exposure to an increased risk of vascular diseases and cardiovascular mortality [5,6]. Vanadium is classified as a group 5 transition metal and is a redox-sensitive element. It exists in a large number of oxidation states, though the V(IV) and V(V) oxidation states are most common under physiological conditions. Vanadium pentoxide is considered an air toxicant with mutagenic effects, respiratory tract toxicity, and possible carcinogenic activity [11,13,34,35]. Despite much evidence indicating the pathological toxic effects of vanadium oxide on health, its pathological influence and mechanism on atherosclerosis remain unclear. In this study, we used ApoE −/− mice and found that exposure to NaVO 3 not only causes lung injury, consistent with previous reports, but also induces atherosclerotic plaque. The NaVO 3 -exposed mice developed downregulation of VSMC marker SM22α and upregulation of atherosclerotic-promoting cytokine IL-6 within atherosclerotic lesions. In addition, these atherosclerotic pathologies were accompanied by increasing plasma ROS and IL-6. Notably, these NaVO 3 -induced atherosclerotic pathologies in ApoE −/− mice were prevented by NAC treatment, indicating that ROS is required for NaVO 3 -driven atherosclerosis. Our results further show that VOSO 4 and NaVO 3 promote phenotypic transitions of VSMCs from the quiescent contractile state to the active synthetic state and VSMC proliferation and migration, which has been implicated in deleterious consequences of atherosclerosis [3,25,26,29]. The present study provides experimental evidence that vanadium salts, VOSO 4 and NaVO 3 , drive VSMC pathological responses by inducing ROS production and thereby activating p38/NF-κB signaling. Our results provide new insights into the pathological mechanism underlying VOSO 4 -and NaVO 3 -induced atherosclerosis and VSMC dysfunction.
Under physiological conditions, ROS function as signaling molecules, playing a vital role in maintaining vascular homeostasis and prevent cardiovascular inflammation and injury [31]. However, during pathological stress, aberrant concentrations of ROS in the vessel wall are produced and disturb the redox homeostasis and subsequently influence the function of vascular cells such as VSMCs. Given that transition metals with biological toxicity are due to the formation of oxidizing compounds that exert their toxic effects by the production of ROS [5,36], we propose that oxidizing vanadium compound induces ROS overproduction and activates ROS-mediated atherosclerotic signaling. This hypothesis is supported by anti-oxidant N-acetylcysteine (NAC) rescue of ApoE −/− mice from NaVO 3 -induced excessive plasma ROS and IL-6 and consequent atherosclerosis. In addition, NaVO 3 -driven VSMC pathological effects including ROS and IL-6 production, phenotypic alternation, migration, and proliferation are suppressed by NAC. These suggest that NaVO 3 increases vascular cell ROS production thereby increasing the plasma ROS level and mediating VSMC dysfunction and atherosclerosis.
Although ROS are generated from a number of sources including NADPH oxidases, xanthine oxidase, the mitochondrial respiratory chain, lipoxygenases, and nitric oxide synthases, the mitochondria and NADPH oxidases have been considered to be the major sources of ROS in VSMCs [21,37]. Here, we have provided experimental evidence that ROS derived from NADPH oxidases, but not mitochondria, are critical mediators of NaVO 3 -triggered VSMC pathological effects. Our results show that NaVO 3 significantly increases intracellular ROS but negligibly increases mitochondrial ROS production. However, NADPH oxidase inhibitor significantly reduces NaVO 3 -induced ROS generation and IL-6 production. These findings indicate that NADPH oxidases are a major source of ROS that contribute to NaVO 3 -mediated pathological signaling in VSMCs.
The pleiotropic cytokine IL-6 is a proatherogenic factor that is increased in atherosclerotic mouse plasma and in VSMCs in response to atherosclerotic stimulators. Accumulating evidence suggests IL-6 is implicated in the progression of atherosclerosis and plays a key role in inducing VSMC migration and proliferation [25,26,38,39]. Our findings indicate that NaVO 3 triggers VSMC migration and proliferation through ROS-mediated IL-6 production. This notion is supported by abrogation of NaVO 3 -induced VSMC IL-6 production, VSMC migration, and VSMC proliferation by NAC and NADPH oxidase inhibitor without affecting VSMC viability. We further show that anti-IL-6 antibodies suppressed NaVO 3 -induced VSMC migration and proliferation ( Figure 2G-H) but IL-6 antibodies did not affect VSMC ROS production ( Figure 4G). These results suggest that ROS-mediated IL-6 secretion induced by NaVO 3 acts in an autocrine or paracrine manner to regulate VSMC pathological function and thereby modulate atherosclerosis development. Although ROS-driven cellular signaling are reported to play an important role in controlling cell function and disease progression [40], little is known about the regulation of oxidizing vanadium derivative-modulated ROS-mediated IL-6 expression in VSMCs. In the present study, we found that NaVO 3 increased the phosphorylation level of p38, ERK, JNK, and NF-κB p65, but inhibiting ROS signaling by NAC only suppressed the levels of phosphor-p38 and phosphor-NF-κB p65. Furthermore, inhibition of p38 and NF-κB signaling by their pharmacological inhibitors significantly abolished NaVO 3 -induced VSMC migration and proliferation and IL-6 production without affecting ROS production in VSMCs. These results suggest that NaVO 3 promotes VSMC phenotypic switch, migration, and proliferation by increasing IL-6 production, which is mediated via ROS-activated p38/NF-κB-dependent signaling.
Vanadium is widely distributed throughout the Earth's crust at concentrations of 100-150 ppm (approximately 100-150 mg/kg) [41,42]; greater levels in urban locations have been reported and may reach higher values (up to 400 ppm) in areas polluted by fly ash due to a larger density of combustion sources capable of emitting PM-containing vanadium to the environment. Much evidence from epidemiological and toxicological studies suggests that vanadium exposure causes serious health problems such as lung disease, eye irritation, and hypertension in humans [43,44], but its toxic effects on vascular diseases remains unclear. In this study, we used an animal model of ApoE −/− mice to demonstrate that vanadium exposure has detrimental effects on increasing plasma ROS and atherosclerotic cytokine IL-6 and consequently promotes the synthetic phenotype in VSMCs, which leads to atherosclerosis. We provide evidence that the oxidizing vanadium salts, VOSO 4 and NaVO 3 , promote VSMC synthetic differentiation, migration, and proliferation and consequent atherosclerosis via ROS-mediated IL-6 induction. These findings provide mechanistic insights into the key vascular pathological role of VOSO 4 -and NaVO 3 -mediated ROS signaling in VSMC dysfunction and atherosclerosis as demonstrated by the anti-oxidant NAC, which not only suppresses VSMC pathological responses including migration and proliferation but also prevents atherosclerosis in ApoE −/− mice.
Our results suggest that ROS-mediated IL-6 signaling will be a valuable therapeutic target to prevent vanadium air pollution-mediated inappropriate smooth muscle cell function and atherosclerosis.
Our findings indicate that the oxidizing vanadium salts, VOSO 4 and NaVO 3, are capable of mediating VSMC migration and proliferation which lead to the development of atherosclerosis. We have provided evidence that these two vanadium salts promote VSMC migration and proliferation by activating NF-κB-mediated IL-6 production, which is signaled via ROS-dependent activation of the p38 MAPK. These findings provide mechanistic insights into the pathological effect of VOSO 4 and NaVO 3 in atherosclerosis and will be valuable for developing new therapeutic strategies against vascular diseases.

Preparation of NaVO3 and VOSO4 Solutions
NaVO3 (4 mg/mL) solution was freshly prepared by dissolving the required amounts of NaVO 3 in endotoxin-free water. Briefly, NaVO 3 was suspended in endotoxin-free water, heated at 55 • C for 20 min and vortex every 10 min until dissolved in endotoxin-free water. Before the experiment, the solution was allowed to stand at room temperature with a pH of about 6.8.
To prepare the VOSO 4 solution, the VOSO 4 was dissolved in the DMEM at a working concentration of 1 µg/mL (pH about 8.2~8.6).

Animals
Wild-type C57BL/6J mice (National Laboratory Animal Center, Taiwan) were used for primary vascular smooth muscle cell (VSMC) isolation, and ApoE −/− mice (C57BL/6 background, Jackson Laboratory, Bar Harbor, ME, USA) (8-12 weeks old) were randomly grouped to experiments and fed normal chow. Mice were housed in a specific pathogen-free animal facility at National Health Research Institutes, Taiwan. All experimental procedures were followed by the NIH guidelines for the laboratory animal's care and use and approved by the Institutional Animal Care and Use Committee of National Health Research Institutes, Taiwan (#NHRI-IACUC-106081-A, 02/01/2017).

Induction of Atherosclerosis in a Mouse Model
Eight-to-twelve-week-old ApoE −/− mice (C57BL/6 background, Jackson Laboratory) were randomly assigned and intranasally administered vehicle (endotoxin-free water) or NaVO 3 (4 mg/kg) once weekly with or without intraperitoneal injection of NAC (150 mg/kg or 250 mg/kg) three times a week for 12 weeks. This method followed the guidelines of experimental atherosclerosis studies described in the AHA (American Heart Association) Statement [45]. Following this, the mice were killed, and blood samples were collected for vanadium, ROS, and cytokine analysis. Mouse urine was collected on week 4 and 6 for urine vanadium level measurements. The aortas were harvested for analysis of atherosclerotic lesion. In addition, the lung, kidney, liver, and heart were collected for histological determination. To be the more specific in terms of the measurement of the atherosclerotic area in mouse lesion sites from the ascending aorta to the end of the arch, we made three 4-µm sections on each slide, with a total of 100 slides in a sequential manner. We picked one slide from every 10 slides and 10 slides for each group to perform H&E (hematoxylin and eosin) or immunohistochemistry (IHC) staining. The mouse experiments were approved by the Institutional Animal Care and Use Committee, National Health Research Institutes.

Cell Culture and Treatment
Primary VSMCs were isolated from 18.5-day-postconception embryonic mouse aortas of C57BL/6J mice (The National Laboratory Animal Center, Taipei, Taiwan) by enzymatic digestion of the aorta using collagenase and elastase as previously described [25]. In briefly, VSMCs were cultured in a growth medium containing Dulbecco's Modified Eagle medium (DMEM) supplemented with 10% FBS, penicillin (100 U/mL), and streptomycin (100 µg/mL) as previously described. Cells were passaged every 3 days, and experiments were performed on cells 5-10 passages after the primary culture. VSMCs were quiescent by starvation in 0.5% FBS in DMEM for 24 h before experiments.

Measurement of Vanadium Levels in Urine and Plasma
The vanadium concentrations in the urine and plasma were measured by inductively coupled plasma-mass spectrometry (ICP-MS, NexION 2000, PerkinElmer, Waltham, MA, USA). Before measurement, all the samples were diluted 1:10 with 70% nitric acid and microwave digestion to provide clean samples. All samples were quantified using linear calibration curve established from the analysis of a series of five standard solutions (0.1, 0.5, 1, 5, and 10 parts per billion (ppb)) with typical correlation coefficients (R 2 ) greater than 0.9999 for each analysis to ensure that the measurement process was reliable. The vanadium concentrations in the plasma samples were expressed as nanograms per milliliter (ng/mL); whereas the levels of vanadium in urine were normalized by creatinine (CRE) in the urine sample and expressed as micrograms per gram creatinine (µg/g CRE).

TBARS Assay
The plasma ROS was determined by measurement of lipid peroxidation via thiobarbituric acid reactive substances (TBARS) assay (Cayman Chemical, Ann Arbor, MI, USA) according to manufacturer's protocol. The concentration of TBARS in plasma were calculated from TBARS standard curve (0 to 400 µmol/L 1,1,3,3-tetramethoxypropane) and was expressed as µmol/L.

Oil Red O Staining
The aorta was dissected of connective tissue under a dissecting microscope and stained with 0.9% Oil Red O (in 66.6% isopropanol) solution for 15 min at room temperature. After destaining with 66.6% isopropanol for 10 min (three times), the aorta was fixed on dissection disk with an insect needle and inspected under a microscope.

Histology and Immunohistochemistry
For histological studies, mice were perfused with saline and subsequently with 10% formaldehyde. The tissues, including lung, liver, kidney, heart, and aorta, were immersed in formaldehyde for 24 h and then dehydrated with a graded series of ethanol and embedded in paraffin. Tissue blocks were cut into 4-µm-thick serial sections and stained with hematoxylin and eosin for examining the pathology, such as atherosclerotic plaque and lung inflammation and injury.
Prior to detection of tissue protein expression by immunohistochemistry, the aortic sections were deparaffinized with xylene and progressively rehydrated through graded alcohols. Antigen retrieval were performed by heating the sections on slides in an EDTA antigen retrieval buffer of pH 8 (Trilogy; Cell Marque Corporation) in an electric pressure cooker for 10 min. Sections were sequentially blocked by 3% H 2 O 2 for 20 min and blocking buffer (5% BSA in phosphate-buffered saline with 0.1% Tween 20 (PBST)) for an additional 30 min. Antibodies including SM22α (ab14106, Abcam) and IL-6 (ab6672, Abcam) antibodies were diluted in PBS. Sections were incubated at room temperature for 2 h or 4 • C for overnight with primary antibody and then washed in PBST. The sections were then incubated with HRP Labelled Polymer (Dako, Carpinteria, CA, USA) for 60 min and washed three times with PBST. The protein expression was visualized using the DAB Chromogen system (Dako). To validate the specificity of primary antibody, there was a negative control representing staining with an isotype control antibody. Slides were counterstained with hematoxylin. The immunopositive areas in the aorta tissues were quantified using ImageJ software. The signal values were expressed as the percent positive area out of the total tissue area.

Western Blot Analysis
Cellular proteins extracted from treated VSMCs were resolved with 4-12% SDS-PAGE and transferred to the polyvinylidene difluoride (PVDF) membranes [46].

Cell Viability and Cell Proliferation Assays
MTT assay was used to measure cell viability. Cell proliferation was determined by bromodeoxyuridine (BrdU) incorporation assays. Briefly, VSMCs were plated on microtiter plates in starvation medium for 24 h and treated with different amounts of VOSO 4 or NaVO 3 for the indicated times. For MTT assay, the treated VSMCs were incubated DMEM containing 0.5 mg/mL thiazolyl blue tetrazolium bromide (MTT) for 1 h, then lysed by DMSO. The formazan reduced by mitochondrial dehydrogenases of living cells were read by microplate reader at absorbance 540 nm. BrdU was added directly to cell cultures for 24 h starting at time of stimulation. The BrdU incorporation assays was performed according to the manufacturer's protocol (Millipore).

Migration Assays
To assess migration, quiesced VSMCs were treated with or without VOSO 4 or NaVO 3 (1 µg/mL) with or without inhibitor for 24 h. VSMCs were trypsinized, washed with PBS, re-suspended in medium with 0.5 % FBS, and then placed in the upper chamber of 24-well transwell plates (Millipore, 8-µm pore size) in triplicate (20,000 cells/well). The bottom chambers were filled with starvation medium containing PDGF-BB (Peprotech, 10 ng/mL) as a chemoattractant. After 4 h, the upper layer was scraped by cotton swab to remove non migratory cells, membrane fixed, and stained with crystal violet solution (0.1% crystal violet, 20% ethanol, and 1% formaldehyde in ddH 2 O). Cells that had migrated to the underside of the membrane were visualized by microscope. The crystal violet in the cells was extracted by crystal violet extraction buffer (50% ethanol and 0.1% acetic acid in ddH 2 O) and read by a microplate reader at absorbance 595 nm. The data were plotted as the fold change versus vehicle, arbitrarily set to 1.