An Osteopontin–NADPH Oxidase Signaling Cascade Promotes Pro–Matrix Metalloproteinase 9 Activation in Aortic Mesenchymal Cells
http://www.100md.com
Chung-Fang Lai, Venkat Seshadri, Kane Hu
参见附件。
the Department of Medicine (C.-F.L., V.S., K.H., J.-S.S., J.C., R.V., A.S., A.P.L., D.A.T.), Washington University School of Medicine, St Louis, Mo
the Department of Cell Biology and Neuroscience (D.T.D., S.R.R.), Rutgers University, Piscataway, NJ. Current address for S.R.R. is the Forsyth Institute, Boston, Mass.
Abstract
Osteopontin (OPN) is a cytokine upregulated in diabetic vascular disease. To better understand its role in vascular remodeling, we assessed how OPN controls metalloproteinase (MMP) activation in aortic adventitial myofibroblasts (AMFs) and A7r5 vascular smooth muscle cells (VSMCs). By zymography, OPN and tumor necrosis factor (TNF)- preferentially upregulate pro–matrix metalloproteinase 9 (pro-MMP9) activity. TNF- upregulated pro-MMP9 in AMFs isolated from wild-type (OPN+/+) mice, but pro-MMP9 induction was abrogated in AMFs from OPN–/– mice. OPN treatment of VSMCs enhanced pro-MMP9 activity, and TNF- induction of pro-MMP9 was inhibited by anti-OPN antibody and apocynin. Superoxide and the oxylipid product 8-isoprostaglandin F2 -isoprostane (8-IsoP) were increased by OPN treatment, and anti-OPN antibody suppressed 8-IsoP production. Like OPN and TNF-, 8-IsoP preferentially activated pro-MMP9. Superoxide, 8-IsoP, and NADPH oxidase 2 (Nox2) subunits were reduced in OPN–/– AMFs. Treatment of A7r5 VSMCs with OPN upregulated NADPH oxidase subunit accumulation. OPN structure/function studies mapped these activities to the SVVYGLR heptapeptide motif in the thrombin-liberated human OPN N-terminal domain (SLAYGLR in mouse OPN). Treatment of aortic VSMCs with SVVYGLR upregulated pro-MMP9 activity and restored TNF- activation of pro-MMP9 in OPN–/– AMFs. Injection of OPN-deficient OPN+/– mice with SVVYGLR peptide upregulated pro-MMP9 activity, 8-IsoP levels, and Nox2 protein levels in aorta and increased panmural superoxide production (dihydroethidium staining). At equivalent hyperglycemia and dyslipidemia, 8-IsoP levels and aortic pro-MMP9 were reduced with complete OPN deficiency in a model of diet-induced diabetes, achieved by comparing OPN–/–/LDLR–/– versus OPN+/–/LDLR–/– siblings. Thus, OPN provides a paracrine signal that augments vascular pro-MMP9 activity, mediated in part via superoxide generation and oxylipid formation.
Key Words: osteopontin superoxide diabetes metalloproteinase
Introduction
An epidemic of diabetes is assailing the world as Westernized lifestyles become prevalent.1 Diabetic macro- and microvascular diseases result in blindness, renal insufficiency, and cardiovascular mortality.1 For 2 decades, it has been appreciated that diabetes elicits genomic responses in the intima, media, and adventitia of muscular arteries2; hyperglycemia-induced changes in gene expression, mural oxidative stress, and matrix remodeling occur in all 3 vascular compartments.3 Because macrovascular disease progresses for several years before clinically overt diabetes,1 a better understanding of the mechanisms whereby glucose exerts changes in vascular structure and function is required.
The signals mediating low-grade medial and adventitial inflammation of diabetes are only beginning to be understood. Diabetes-induced vascular tumor necrosis factor (TNF)- expression and NADPH oxidase activate mural cell-mediated immunity as 1 early component.3 In addition to osmotic stress, both proximal (glucosamine, myoinositol) and distal (advanced glycosylation end products) glucose metabolites participate in vascular disease.3 We have identified that the cytokine osteopontin (OPN) is also upregulated in the aortic wall during diabetes.4,5 Intracellular glucose metabolism and transcriptional activation via upstream stimulatory factor 1 (USF1) and activator protein 1 (AP1) mediate OPN induction.5 OPN enhances adventitial myofibroblast (AMF) and vascular smooth muscle cell (VSMC) migration and proliferation,6 promotes vascular matrix metalloproteinase 9 (MMP9) activation,6,7 augments cell-mediated immunity,8 and inhibits matrix calcium accumulation.7–9 OPN is required for angiotensin II (AT-II)–induced aortic aneurysm formation.7 Because OPN transgenic mice exhibit increased arterial medial thickness and mural cell proliferation,6 hyperglycemia-induced vascular OPN expression may contribute to the intimal–medial thickening, enhanced remodeling, and reduced macrovascular compliance of dysglycemic patients.1 MMP9 is a key downstream mediator, because it promotes angiogenesis and vascular remodeling.10
To better understand the role of OPN in diabetic vascular disease, we assessed effects of OPN signaling on primary aortic AMF and aortic A7r5 VSMC physiology. We identify that OPN, via the N-terminal SVVYGLR motif (SLAYGLR in rodents), preferentially upregulates vascular pro-MMP9 activity by augmenting superoxide generation and oxylipid formation.
Materials and Methods
Reagents
Biochemicals were from Fisher or Sigma. Antibody to NADPH oxidase 2 (Nox2) was from Transduction Laboratories. Antibodies to other NADPH oxidase subunits, eIF2-, and normal goat or rabbit immunoglobulin were from Santa Cruz Biotechnology. Antibody against MMP9 was from Chemicon. Precast 10% Zymogram Gelatin Gel and 12% Zymogram Casein Gel zymography kits were purchased from Invitrogen. The human MMP9 Biotrak Activity Kit (Amersham technical bulletin RPN2614PL Rev-B, 2004) was purchased from GE Healthcare. Bovine OPN (bOPN), OPN neutralizing antibody, and TNF- were from R&D Systems. The 8-IsoP ELISA kit (no. 516351) and 8-IsoP (no. 16350) were purchased from Cayman. Dihydroethidium (DHE) was from Molecular Probes. The GRGDS and GRGES peptides were from Bachem. Custom synthetic heptapeptides SVVYGLR and SLAYGLR were from Invitrogen.
Quantitative Gelatin Zymography and Superoxide Detection in Cultured Cells
Before each analysis, 0.5 million aortic VSMCs were plated on a 15-cm tissue culture dish, grown until confluent, and then seeded at high cell densities, ie, 0.5 to 0.75x105 cells per well (96-well plates) or 1.0 to 1.5x105 cells per well (12-well plates). Treatment with 30.5 mmol/L glucose versus the 5.5 mmol/L glucose+25 mmol/L mannitol osmotic control was performed in basal medium, Eagle’s (BME). All other treatments were performed in serum-free DMEM (25 mmol/L glucose) for 18 hours. Twenty-microliter aliquots of serum-free conditioned media were analyzed by zymography (2 to 4 replicates) using precast 10% Zymogram Gelatin Gels (Invitrogen/Novex) per the instructions of the manufacturer (Novex technical bulletin IM-1002, Version B). With constant gentle agitation, gels were renatured for 30 minutes at room temperature, developed overnight at 37°C, fixed and stained with Colloidal Blue (Novex technical bulletin IM-6025), and extensively washed (>20 hours) to yield uniform background signal, and digital images of stained wet gels were captured using a Hewlett-Packard ScanJet 5370C bed scanner. Image analysis for was performed by importing scanned JPEG images into Kodak 1D image analysis software essentially as previously described11 but using net pixel intensity of the pro-MMP9 zymogram with uniform rectangular region of interest to quantify MMP activity. For 8-IsoP determination, 25- to 50-μL aliquots were used in immunoassays (n=4 per group). For superoxide detection, cells were plated as above, rinsed with DMEM, incubated with 5 μmol/L DHE in DMEM for 20 minutes, and rinsed, and light and fluorescence microscopy with digital image was capture performed as described.12,13 Image analysis for net pixel intensity was performed by importing JPEG images into Kodak 1D image analysis software using uniform region of interest area for pro-MMP9 f-bands.14 Data are expressed as net pixel intensity, normalized either to mass of extracted tissue protein or the experimental control as indicated. Statistical analyses were performed using Student’s unpaired t test (Microsoft Excel 2002), with graphical data presented as the mean±SE of independent replicates (n=4 to 7).
All other methods are detailed in the expanded Materials and Methods section in the online data supplement available at http://circres.ahajournals.org.
Results
TNF-– and Glucose-Enhanced Pro-MMP9 Activation Is Reduced in OPN–/– Primary Aortic VSMCs
To better understand the role of OPN in diabetic vascular disease, we examined the effects on MMP9 in primary aortic AMFs and A7r5 VSMCs. Previous data suggested that pro-MMP9 might be the preferentially activated by an OPN transgene.6 Therefore, we focused on enzyme activity via gelatin zymography, a robust MMP9 and pro-MMP9 assay.7 Like AT-II, OPN and TNF- increased cellular gelatinase activity; however, whereas AT-II activated both MMP9 and pro-MMP9, OPN and TNF- preferentially upregulated pro-MMP9 (Figure 1A). Image analysis of zymograms from multiple independent experiments (supplemental Figure IA) revealed that 50 nmol/L OPN treatment upregulated pro-MMP9 activity 5-fold (Figure 1B; P=0.004). Immunoprecipitation with anti-MMP9 antibody enriched this gelatinase activity, confirming that it was indeed pro-MMP9; minimal pro-MMP9 activity was immunoprecipitated from OPN–/– aortic AMFs (supplemental Figure IB). Basal pro-MMP9 levels were lower in OPN–/– AMFs (Figures 1C and 4D; see also supplemental Figure IB). Because AMF and VSMC OPN expression is entrained to ambient glucose concentration,4,5 we examined effects of glucose levels on pro-MMP9 activation. With OPN+/+ AMFs, basal pro-MMP9 activity was greater when cultured in 30.5 mmol/L glucose as compared with the isoosmotic control of 5.5 mmol/L glucose with 25 mmol/L mannitol (Figure 1C, top, lane 3 versus lane 1). MMP3 activity (casein zymogram, Figure 1C, bottom, lane 3 versus lane 1) was not affected by glucose. TNF- induction of pro-MMP9 was not enhanced by high glucose (Figure 1C, lane 4 versus lane 2). We next examined whether OPN was required for pro-MMP9 regulation using aortic AMFs isolated from OPN–/– mice. Under every condition, pro-MMP9 activity was reduced in OPN–/– AMFs versus OPN+/+ controls (Figure 1C, top, lanes 5 to 8 versus lanes 1 to 4; P=0.008). MMP3 activity was not diminished (Figure 1C, bottom, lanes 5 to 8 versus lanes 1 to 4). Moreover, neutralization of endogenous OPN in A7r5 cells with an anti-OPN antibody abrogated induction of pro-MMP9 activity by TNF- (Figure 1D) as compared with pro-MMP9 induction elicited in presence of control IgG. Thus, in aortic AMFs and A7r5 VSMCs, pro-MMP9 activation is regulated by OPN signaling.
Inhibitors of Superoxide Signaling Reduce Pro-MMP9 Activation by TNF- and OPN
Many inflammatory cytokines signal via nuclear factor (NF)-B, a well-characterized transcription factor that supports expression of MMP9.15 Thus, we considered that NF-B might mediate pro-MMP9 activation in aortic AMFs and VSMCs. The thiol reagent N-acetylcysteine (NAC)16 and IKK inhibitor BAY 11-7082 (BAY)17 both inhibit NF-B signaling but via distinct mechanisms (supplemental Figure IIA). Therefore, we examined the effects of these compounds on pro-MMP9 activation in A7r5 cells. As shown in Figure 2A, both compounds are active as inhibitors of TNF-–induced NF-B signaling, indicated by the transfected reporter NF-B–LUC. Surprisingly, whereas NAC inhibited pro-MMP9 activation at high cell densities, BAY did not (Figure 2B and supplemental Figure IIB and IIC). NAC also suppresses superoxide generation18; therefore, we assessed the effects of diphenyleneiodonium (DPI) and apocynin on pro-MMP9 activation. Apocynin (inhibits p47 and p67 cytosol-to-membrane translocation) and DPI (targets heme b in NADPH oxidase subunits as well as flavin centers) are mechanistically distinct inhibitors of cellular superoxide generation routinely used to functionally test contributions of NADPH oxidases to signal transduction.12,18 Like NAC, DPI reduced TNF- induction of pro-MMP9 by 50% (supplemental Figure IIB). Similar results were obtained with apocynin, which preferentially abolished TNF- induction of pro-MMP9 (Figure 2C and 2D). The enzyme superoxide dismutase (SOD) catalytically destroys superoxide to generate oxygen and hydrogen peroxide.18 As observed with NAC, apocynin, and DPI, treatment with SOD prevented pro-MMP9 activation by either TNF- or OPN (not shown). Thus, redox signaling participates in cytokine activation of pro-MMP9.
Superoxide and Superoxide-Generating NADPH Oxidase Subunits Are Regulated by OPN
Inhibition of pro-MMP9 activation by NAC, DPI, apocynin, and SOD strongly suggested that superoxide production was a component of OPN signaling. Therefore, we examined the effects of OPN on superoxide generation as revealed by fluorescence staining with DHE, a compound chemically oxidized to the fluorescent DNA intercalating agent ethidium in proportion to the amount of superoxide present.12,14 As compared with OPN+/+ aortic AMFs, OPN–/– aortic AMFs exhibited reduced DHE staining (Figure 3A). Because NADPH oxidases are embedded in the membrane bilayer, the short-lived superoxide radicals react with membrane lipids, generating oxylipids19 that demarcate and mediate oxidative signaling.18 One category of oxylipids, isoprostanes, arises from the chemical oxidation of arachidonic acid in phospholipids.20 Therefore, we assessed the relationship between OPN, isoprostane production (8-IsoP), and pro-MMP9 activation in AMFs. 8-IsoP is reduced by 50% in OPN–/– AMFs as compared with OPN+/+ controls (Figure 3B). Moreover, exogenous OPN protein upregulated DHE staining (Figure 3C) and 8-IsoP (Figure 3D) in A7r5 VSMCs. Furthermore, anti-OPN neutralizing antibody inhibited basal 8-IsoP production by 50%, presumably via inhibition of endogenous OPN,5 whereas an isotype-matched control antibody had no effect (Figure 3D).
NADPH oxidases are specific but heterogeneous complexes of p67, p47, p40, and p22 subunits with cell type–specific NADPH oxidase subunits.18 Other important interacting proteins (eg, the Rac family GTPases and cortactin) convey regulatory responses to vasculotropic hormones such as AT-II.21 In human and rodent aortic medial VSMCs, Nox1 is a predominant isoform,18,22 whereas Nox2 is highly expressed in aortic AMFs.23 Therefore, we assessed the levels of NADPH oxidase subunit in aortic AMFs from OPN+/+ versus OPN–/– mice. As shown in Figure 3E, Nox2, Nox1, p67, and p47 are reduced in cell extracts from OPN–/– AMFs as compared with OPN+/+ controls. No reduction of the housekeeping protein eIF-2 was observed in OPN–/– AMFs (Figure 3E). Because Pagano and colleagues had demonstrated a critical role for Nox2 in AMF physiology,23 we emphasized Nox2 and its associated subunits. Western blots analyzed from 3 independent sets of cultures demonstrated a >50% reduction in Nox2 accumulation in OPN–/– AMFs versus OPN-replete control cultures (Figure 3F). Of note, p67—a regulatory subunit required for electron transport between NADPH and the flavin center that generates superoxide18—is reduced along with Nox2 in OPN–/– AMFs (Figure 3E). Moreover, addition of exogenous OPN upregulated p67 in A7r5 VSMCs (supplemental Figure IIIA). Finally, a validated antisense oligonucleotide (ASO) directed to p47 significantly reduced 8-IsoP levels in A7r5 cells (supplemental Figure IIIB), confirming contributions of prototypic NADPH oxidases to A7r5 oxylipid formation. Thus, superoxide production, oxylipid formation, and NADPH oxidase subunit accumulation is regulated by OPN in aortic mesenchymal cells.
The OPN Peptide SVVYGLR Activates Vascular Pro-MMP9 and Superoxide Signaling In Vitro
OPN is a phosphorylated glycoprotein, processed by thrombin to generated N-terminal and C-terminal domains with unique bioactivities8 (supplement Figure IVA). Like full-length bOPN (Figure 1A and 1B above, and data not shown), the N-terminal recombinant murine OPN fragment dose-dependently activates pro-MMP9 (Figure 4A), whereas the C-terminal OPN fragment is inactive (not shown). The N-terminal domain of OPN contains v3 and 41 or 47 binding subdomains that mediate adhesion and signaling.8 Of note, a synthetic peptide corresponding to the SVVYGLR motif of human OPN upregulated pro-MMP9 activity (Figure 4B); GRGDS and GRGES peptides were insufficient. Similar results were observed with murine OPN SLAYGLR peptide (not shown). We confirmed these results using HA-VSMCs, and an independent immunoassay specific for human pro-MMP9; 50 μmol/L SVVYGLR upregulated pro-MMP9 approximately 6-fold (Figure 4C, P=0.029, 2-tailed t test). Induction of pro-MMP9 by TNF- is poor in AMFs from OPN–/– mice (Figure 4D; also Figure 1C). However, incubation with either 50 μmol/L SVVYGLR peptide (lanes 3 and 4) or 50 nmol/L bOPN (lanes 5 to 6) restored robust TNF- induction of pro-MMP9 (Figure 4D). Thus, OPN signaling via the SVVYGLR motif plays an important role in pro-MMP9 activation.
We next evaluated effects of OPN peptides on superoxide and oxylipid formation. Like full-length OPN, recombinant OPN N-terminal fragment increased A7r5 8-IsoP production; OPN C-terminal domain did not (Figure 5A). Moreover, the N-terminal heptapeptide SVVYGLR, but not GRGDS or GRGES, upregulated 8-IsoP (Figure 5B). SVVYGLR upregulated superoxide (DHE staining) in A7r5 VSMCs (2-fold; supplemental Figure VA) and OPN–/– AMFs (5-fold; not shown), quantified by digital image analysis. Thus, the OPN N-terminal domain encodes motifs necessary for pro-MMP9 activation and superoxide and oxylipid formation.
The OPN Peptide SVVYGLR Activates Vascular Pro-MMP9 and Superoxide Signaling In Vivo
We next injected SVVYGLR (n=5) or vehicle (n=5) into OPN+/– mice, an in vivo background partially deficient in endogenous OPN tone. Intraperitoneal SVVYGLR administration (8 μg/g) significantly upregulated pro-MMP9 activity in aortic extracts (P=0.03; Figure 6A). Mural actions of SVVYGLR were spatially resolved by DHE staining of aortic frozen sections; adventitial and medial tunics exhibited enhanced superoxide production (Figure 6B and 6C), indicating cellular responses in both compartments. Pretreatment of sections with PEG-SOD reduced nuclear DHE staining, but not elastin autofluorescence, confirming superoxide-generated signal (Figure 6B, e and f). SVVYGLR significantly upregulated aortic 8-IsoP in vivo (Figure 6D; P=0.019). Moreover, Western blot analysis of individual aortic extracts confirmed that, as observed with in vitro (supplemental Figure IIIA), aortic Nox2 protein levels were upregulated 2.3-fold by OPN SVVYGLR peptide in vivo (P=0.07 for trend approaching significance, supplemental Figure VB, n=5 per group). To examine the generality of signaling, we assessed pro-MMP9 activation by SVVYGLR in pulmonary tissues. As observed in aorta, SVVYGLR upregulates pro-MMP9 activity in lung (Figure 6E and 6F). Thus, OPN peptide SVVYGLR upregulates levels of vascular superoxide, aortic 8-IsoP formation, and tissue pro-MMP9 activation in vivo.
8-IsoP and Aortic Pro-MMP9 Are Reduced in OPN-Deficient LDLR–/– Mice Fed High-Fat Diabetogenic Diets
We wished to examine whether OPN signaling was important for vascular stress responses to diabetes in vivo. The male LDLR–/– mouse has proven useful for studies of diet-induced diabetes, dyslipidemia, and vascular disease.4,11 Therefore, we studied the effect of high-fat diabetogenic diets (Western diet) on OPN+/–/LDLR–/– versus OPN–/–/LDLR–/– mice. Reduced plasma OPN levels confirmed OPN deficiency in LDLR–/–/OPN–/– versus LDLR–/–/OPN+/– mice (Figure 7A). Seven weeks of a high-fat diet generated equivalent diabetes (mean±SEM: glucose of 514±31 versus 496±28 mg/dL, P=0.39; 1-tailed t test) and dyslipidemia (free fatty acids 2.1±0.05 versus 2.0±0.05 mmol/L, P=0.11; cholesterol 1804±196 versus 1807±129 mg/dL, P=0.49). By contrast, plasma 8-IsoP levels were lower in LDLR–/–/OPN–/– versus LDLR–/–/OPN+/– mice (Figure 7B; P=0.017). Moreover, aortic pro-MMP9 activity was reduced in diabetic LDLR–/–/OPN–/– versus LDLR–/–/OPN+/– (Figure 7C and 7D; P=0.04). Thus, OPN participates in isoprostane generation and aortic pro-MMP9 activation in the LDLR–/– mouse model of diet-induced diabetes.4
Oxylipids Preferentially Activate Pro-MMP9
Oxylipids not only reflect the presence of superoxide generation but also mediate inflammatory signals.20 Thus, we tested the ability of oxylipid to activate pro-MMP9, evaluating 8-IsoP, which is among the many isoprostanes generated from the heterogeneous oxylipid precursor oxidized palmitoyl arachidonoyl phosphatidylcholine (Ox-PAPC).19,20 Like OPN and TNF- (Figure 1A), 8-IsoP preferentially upregulated pro-MMP9 in HA-VSMCs and A7r5 aortic VSMCs (Figure 8A and 8B). Similar results were obtained in OPN–/– aortic AMFs treated with 8-IsoP (3.5-fold induction; P=0.02, not shown). However, in contrast to activation by TNF- and OPN (Figure 2B and supplemental Figure IIB), pro-MMP9 activation by 8-IsoP is insensitive to NAC (Figure 8B, right) and DPI (not shown), confirming the potential contributions of oxylipids as terminal effectors (supplemental Figure IIA). We also tested the ability of 8-IsoP to activate pro-MMP9 in the presence of anti-OPN antibody in A7r5 cells, a treatment that reduces pro-MMP9 activity (Figure 1D; and Figure 8C, lanes 1 and 2) and endogenous 8-IsoP production (Figure 3D). As shown in Figure 8C, whereas anti-OPN decreased pro-MMP9 activity (lanes 1 and 2), 8-IsoP treatment dose-dependently restored pro-MMP9 activity (lanes 3 to 5). In aqueous solutions, 3-morpholinosydnonimine-N-ethylcarbamide (SIN-1) decomposes to form superoxide and nitric oxide (and thus peroxynitrite).10 SIN-1 dose-dependently upregulated 8-IsoP accumulation in A7r5 cells (Figure 8D), confirming that superoxide generates oxylipids in aortic VSMCs. Moreover, like OPN and 8-IsoP, SIN-1 upregulated pro-MMP9 but also enhanced MMP9 activation (Figure 8E). Little effect on MMP2 was noted. Thus, OPN- and superoxide-regulated oxylipids can participate in the activation of pro-MMP9 in VSMCs.
Discussion
VSMCs and AMFs participate in every aspect of arterial pathobiology.23,24 Proliferative and synthetic responses contribute to mural thickness, neointima formation, fibrosis, calcification, and vascular stiffening.6 Contractile responses modulate arterial tone, wound repair, and hemostasis.3,23,24 Very few well-validated markers differentiate VSMCs of the media from adventitial AMFs24; however, AMFs exhibit higher levels of NADPH oxidase activity.23,24 Aortic AMFs elaborate robust metabolic responses in the setting of diabetes, dyslipidemia, and hypertension.23,24 Thus, both adventitial AMFs and mural VSMCs are important mediators of vascular disease progression.3,23,24
SVVYGLR treatment activated oxidative stress in both adventitial and medial tissue compartments, consistent with our observations made with aortic AMF and VSMC culture models. Other vascular beds are also responsive, because pulmonary pro-MMP9 activity was significantly upregulated by OPN SVVYGLR peptide. Under basal conditions, total NADPH oxidase activity is higher in adventitial versus medial arterial compartments.3,23 With diabetes, adventitial oxidase activity is further enhanced, and tunica media NADPH oxidase is dramatically upregulated.3 Because OPN is a hyperglycemia-induced cytokine,5 the macrovascular oxidative stress of diabetes3 is potentially augmented by paracrine OPN signaling. Consistent with this, both 8-IsoP levels and aortic pro-MMP9 activity are reduced in diabetic LDLR–/– mice lacking the OPN gene. Advanced glycosylation end products upregulate proinflammatory vascular cytokines such as TNF-.3 Thus, in our working model (supplemental Figure IIA), oxylipids19 such as 8-IsoP generated via OPN-dependent signals would augment mural pro-MMP9 in diabetes. The plasma 8-IsoP levels observed in vivo (Figure 7E) are 1000-fold lower than those necessary to upregulate pro-MMP9 in vitro. Although 8-IsoP is 1 bioactive oxylipid generated in response to OPN signaling, many other oxylipids are produced during vascular disease progression.19 Ox-PAPC, the oxylipid precursor of isoprostanes,19,20 also activates pro-MMP9 (not shown). Thus, although 8-IsoP is 1 oxylipid readily assayed and studied, the spectrum of vascular oxylipids that are upregulated in diabetes and participate in pro-MMP9 activation has yet to be determined. Of note, oxylipids such as isoprostanes or 4-hydroxynonenal could activate pro-MMP9 via covalent alkylation reactions involving pro-MMP9 propeptide.10
At high cell densities, we see no effect of BAY compound on OPN or TNF- pro-MMP9 activation in vitro. However, when cells are sparsely subconfluent, BAY compound weakly inhibits MMP9 induction, suggesting that NF-B signaling contributes at low cell densities; moreover, unlike TNF-, pro-MMP9 activation by TGF- is not altered by OPN deficiency (not shown). Thus, multiple pathways must participate in vascular MMP9 activation during development and disease.
Galis and Khatri first demonstrated that oxidative stress activates collagenases in phagocytes, including pro-MMP9 and MMP910; this resembles our results with SIN-1. Our data extend these observations by identifying that OPN signaling and oxylipids preferentially activate pro-MMP9 in aortic mesenchymal cells.7 However, the specific NADPH oxidase complex mediating this vascular response is as yet unknown and may differ between media and adventitia.18,23 Future experiments will identify the relative contributions of NADPH oxidase isoforms to OPN signaling in medial versus adventitial tunics.
With diabetes, periaortic AMFs express high levels of OPN at the earliest stages of disease.4 Both adventitial and medial oxidative stresses are enhanced in response to diabetes-induced NADPH oxidase activity.3,24 We propose that hyperglycemia-induced aortic OPN expression4,5 contributes to the vascular inflammation of diabetes, augmenting mural oxidative stress, pro-MMP9 activation, medial thickening, and matrix remodeling (supplemental Figure IIA). Thus, strategies that inhibit paracrine OPN–superoxide signaling may ameliorate the macrovascular injury of diabetes. Liaw, Giachelli, and colleagues first showed that neutralizing OPN reduced neointima formation25; whether this antibody concomitantly regulated vascular oxidative stress is currently unknown. Our data add to this evidence,25 highlighting a potential benefit in selective inhibition of OPN SVVYGLR signaling as 1 strategy to reduce inflammatory vascular remodeling in diabetes and, thus, preserve macrovascular structure and function.7,25
Acknowledgments
Sources of Funding
This work was supported by the American Diabetes Association and the National Heart, Lung, and Blood Institute.
Disclosures
None.
Footnotes
Original received August 29, 2005; resubmission received January 11, 2006; revised resubmission received April 6, 2006; accepted May 8, 2006.
References
Zimmet P, Alberti KG, Shaw J. Global and societal implications of the diabetes epidemic. Nature. 2001; 414: 782–787. [Order article via Infotrieve]
Lehmann RR, Nienhaus RH, Denes RM, Steinbach T. Changes of euchromatin/heterochromatin ratios in cell nuclei of the aortic adventitia in diabetic rats. Artery. 1987; 14: 66–75. [Order article via Infotrieve]
Zhang L, Zalewski A, Liu Y, Mazurek T, Cowan S, Martin JL, Hofmann SM, Vlassara H, Shi Y. Diabetes-induced oxidative stress and low-grade inflammation in porcine coronary arteries. Circulation. 2003; 108: 472–478.
Towler DA, Bidder M, Latifi T, Coleman T, Semenkovich CF. Diet-induced diabetes activates an osteogenic gene regulatory program in the aortas of low density lipoprotein receptor-deficient mice. J Biol Chem. 1998; 273: 30427–30434.
Bidder M, Shao JS, Charlton-Kachigian N, Loewy AP, Semenkovich CF, Towler DA. Osteopontin transcription in aortic vascular smooth muscle cells is controlled by glucose-regulated upstream stimulatory factor and activator protein-1 activities. J Biol Chem. 2002; 277: 44485–44496.
Isoda K, Nishikawa K, Kamezawa Y, Yoshida M, Kusuhara M, Moroi M, Tada N, Ohsuzu F. Osteopontin plays an important role in the development of medial thickening and neointimal formation. Circ Res. 2002; 91: 77–82.
Bruemmer D, Collins AR, Noh G, Wang W, Territo M, Arias-Magallona S, Fishbein MC, Blaschke F, Kintscher U, Graf K, Law RE, Hsueh WA. Angiotensin II-accelerated atherosclerosis and aneurysm formation is attenuated in osteopontin-deficient mice. J Clin Invest. 2003; 112: 1318–1331.
Denhardt DT, Noda M, O’Regan AW, Pavlin D, Berman JS. Osteopontin as a means to cope with environmental insults: regulation of inflammation, tissue remodeling, and cell survival. J Clin Invest. 2001; 107: 1055–1061.
Steitz SA, Speer MY, McKee MD, Liaw L, Almeida M, Yang H, Giachelli CM. Osteopontin inhibits mineral deposition and promotes regression of ectopic calcification. Am J Pathol. 2002; 161: 2035–2046.
Galis ZS, Khatri JJ. Matrix metalloproteinases in vascular remodeling and atherogenesis: the good, the bad, and the ugly. Circ Res. 2002; 90: 251–262.
Shao JS, Cheng SL, Charlton-Kachigian N, Loewy AP, Towler DA. Teriparatide (human parathyroid hormone (1–34)) inhibits osteogenic vascular calcification in diabetic low density lipoprotein receptor-deficient mice. J Biol Chem. 2003; 278: 50195–50202.
Bleeke T, Zhang H, Madamanchi N, Patterson C, Faber JE. Catecholamine-induced vascular wall growth is dependent on generation of reactive oxygen species. Circ Res. 2004; 94: 37–45.
Shao JS, Cheng SL, Pingsterhaus JM, Charlton-Kachigian N, Loewy AP, Towler DA. Msx2 promotes cardiovascular calcification by activating paracrine Wnt signals. J Clin Invest. 2005; 115: 1210–1220.
Dai X, Galligan JJ, Watts SW, Fink GD, Kreulen DL. Increased O2- production and upregulation of ETB receptors by sympathetic neurons in DOCA-salt hypertensive rats. Hypertension. 2004; 43: 1048–1054.
Rangaswami H, Bulbule A, Kundu GC. Nuclear factor-inducing kinase plays a crucial role in osteopontin-induced MAPK/IB kinase-dependent nuclear factor B-mediated promatrix metalloproteinase-9 activation. J Biol Chem. 2004; 279: 38921–38935.
Langen RC, Schols AM, Kelders MC, Van Der Velden JL, Wouters EF, Janssen-Heininger YM. Tumor necrosis factor- inhibits myogenesis through redox-dependent and -independent pathways. Am J Physiol Cell Physiol. 2002; 283: C714–C721.
Yang G, Abate A, George AG, Weng YH, Dennery PA. Maturational differences in lung NF-B activation and their role in tolerance to hyperoxia. J Clin Invest. 2004; 114: 669–678.
Cai H. NAD(P)H oxidase-dependent self-propagation of hydrogen peroxide and vascular disease. Circ Res. 2005; 96: 818–822.
Navab M, Ananthramaiah GM, Reddy ST, Van Lenten BJ, Ansell BJ, Fonarow GC, Vahabzadeh K, Hama S, Hough G, Kamranpour N, Berliner JA, Lusis AJ, Fogelman AM. The oxidation hypothesis of atherogenesis: the role of oxidized phospholipids and HDL. J Lipid Res. 2004; 45: 993–1007.
Morrow JD. Quantification of isoprostanes as indices of oxidant stress and the risk of atherosclerosis in humans. Arterioscler Thromb Vasc Biol. 2005; 25: 279–286.
Touyz RM, Yao G, Quinn MT, Pagano PJ, Schiffrin EL. p47phox associates with the cytoskeleton through cortactin in human vascular smooth muscle cells: role in NAD(P)H oxidase regulation by angiotensin II. Arterioscler Thromb Vasc Biol. 2005; 25: 512–518.
Sorescu GP, Song H, Tressel SL, Hwang J, Dikalov S, Smith DA, Boyd NL, Platt MO, Lassegue B, Griendling KK, Jo H. Bone morphogenic protein 4 produced in endothelial cells by oscillatory shear stress induces monocyte adhesion by stimulating reactive oxygen species production from a nox1-based NADPH oxidase. Circ Res. 2004; 95: 773–779.
Liu J, Ormsby A, Oja-Tebbe N, Pagano PJ. Gene transfer of NAD(P)H oxidase inhibitor to the vascular adventitia attenuates medial smooth muscle hypertrophy. Circ Res. 2004; 95: 587–594.
Sartore S, Chiavegato A, Faggin E, Franch R, Puato M, Ausoni S, Pauletto P. Contribution of adventitial fibroblasts to neointima formation and vascular remodeling: from innocent bystander to active participant. Circ Res. 2001; 89: 1111–1121.
Liaw L, Lombardi DM, Almeida MM, Schwartz SM, deBlois D, Giachelli CM. Neutralizing antibodies directed against osteopontin inhibit rat carotid neointimal thickening after endothelial denudation. Arterioscler Thromb Vasc Biol. 1997; 17: 188–193.
the Department of Medicine (C.-F.L., V.S., K.H., J.-S.S., J.C., R.V., A.S., A.P.L., D.A.T.), Washington University School of Medicine, St Louis, Mo
the Department of Cell Biology and Neuroscience (D.T.D., S.R.R.), Rutgers University, Piscataway, NJ. Current address for S.R.R. is the Forsyth Institute, Boston, Mass.
Abstract
Osteopontin (OPN) is a cytokine upregulated in diabetic vascular disease. To better understand its role in vascular remodeling, we assessed how OPN controls metalloproteinase (MMP) activation in aortic adventitial myofibroblasts (AMFs) and A7r5 vascular smooth muscle cells (VSMCs). By zymography, OPN and tumor necrosis factor (TNF)- preferentially upregulate pro–matrix metalloproteinase 9 (pro-MMP9) activity. TNF- upregulated pro-MMP9 in AMFs isolated from wild-type (OPN+/+) mice, but pro-MMP9 induction was abrogated in AMFs from OPN–/– mice. OPN treatment of VSMCs enhanced pro-MMP9 activity, and TNF- induction of pro-MMP9 was inhibited by anti-OPN antibody and apocynin. Superoxide and the oxylipid product 8-isoprostaglandin F2 -isoprostane (8-IsoP) were increased by OPN treatment, and anti-OPN antibody suppressed 8-IsoP production. Like OPN and TNF-, 8-IsoP preferentially activated pro-MMP9. Superoxide, 8-IsoP, and NADPH oxidase 2 (Nox2) subunits were reduced in OPN–/– AMFs. Treatment of A7r5 VSMCs with OPN upregulated NADPH oxidase subunit accumulation. OPN structure/function studies mapped these activities to the SVVYGLR heptapeptide motif in the thrombin-liberated human OPN N-terminal domain (SLAYGLR in mouse OPN). Treatment of aortic VSMCs with SVVYGLR upregulated pro-MMP9 activity and restored TNF- activation of pro-MMP9 in OPN–/– AMFs. Injection of OPN-deficient OPN+/– mice with SVVYGLR peptide upregulated pro-MMP9 activity, 8-IsoP levels, and Nox2 protein levels in aorta and increased panmural superoxide production (dihydroethidium staining). At equivalent hyperglycemia and dyslipidemia, 8-IsoP levels and aortic pro-MMP9 were reduced with complete OPN deficiency in a model of diet-induced diabetes, achieved by comparing OPN–/–/LDLR–/– versus OPN+/–/LDLR–/– siblings. Thus, OPN provides a paracrine signal that augments vascular pro-MMP9 activity, mediated in part via superoxide generation and oxylipid formation.
Key Words: osteopontin superoxide diabetes metalloproteinase
Introduction
An epidemic of diabetes is assailing the world as Westernized lifestyles become prevalent.1 Diabetic macro- and microvascular diseases result in blindness, renal insufficiency, and cardiovascular mortality.1 For 2 decades, it has been appreciated that diabetes elicits genomic responses in the intima, media, and adventitia of muscular arteries2; hyperglycemia-induced changes in gene expression, mural oxidative stress, and matrix remodeling occur in all 3 vascular compartments.3 Because macrovascular disease progresses for several years before clinically overt diabetes,1 a better understanding of the mechanisms whereby glucose exerts changes in vascular structure and function is required.
The signals mediating low-grade medial and adventitial inflammation of diabetes are only beginning to be understood. Diabetes-induced vascular tumor necrosis factor (TNF)- expression and NADPH oxidase activate mural cell-mediated immunity as 1 early component.3 In addition to osmotic stress, both proximal (glucosamine, myoinositol) and distal (advanced glycosylation end products) glucose metabolites participate in vascular disease.3 We have identified that the cytokine osteopontin (OPN) is also upregulated in the aortic wall during diabetes.4,5 Intracellular glucose metabolism and transcriptional activation via upstream stimulatory factor 1 (USF1) and activator protein 1 (AP1) mediate OPN induction.5 OPN enhances adventitial myofibroblast (AMF) and vascular smooth muscle cell (VSMC) migration and proliferation,6 promotes vascular matrix metalloproteinase 9 (MMP9) activation,6,7 augments cell-mediated immunity,8 and inhibits matrix calcium accumulation.7–9 OPN is required for angiotensin II (AT-II)–induced aortic aneurysm formation.7 Because OPN transgenic mice exhibit increased arterial medial thickness and mural cell proliferation,6 hyperglycemia-induced vascular OPN expression may contribute to the intimal–medial thickening, enhanced remodeling, and reduced macrovascular compliance of dysglycemic patients.1 MMP9 is a key downstream mediator, because it promotes angiogenesis and vascular remodeling.10
To better understand the role of OPN in diabetic vascular disease, we assessed effects of OPN signaling on primary aortic AMF and aortic A7r5 VSMC physiology. We identify that OPN, via the N-terminal SVVYGLR motif (SLAYGLR in rodents), preferentially upregulates vascular pro-MMP9 activity by augmenting superoxide generation and oxylipid formation.
Materials and Methods
Reagents
Biochemicals were from Fisher or Sigma. Antibody to NADPH oxidase 2 (Nox2) was from Transduction Laboratories. Antibodies to other NADPH oxidase subunits, eIF2-, and normal goat or rabbit immunoglobulin were from Santa Cruz Biotechnology. Antibody against MMP9 was from Chemicon. Precast 10% Zymogram Gelatin Gel and 12% Zymogram Casein Gel zymography kits were purchased from Invitrogen. The human MMP9 Biotrak Activity Kit (Amersham technical bulletin RPN2614PL Rev-B, 2004) was purchased from GE Healthcare. Bovine OPN (bOPN), OPN neutralizing antibody, and TNF- were from R&D Systems. The 8-IsoP ELISA kit (no. 516351) and 8-IsoP (no. 16350) were purchased from Cayman. Dihydroethidium (DHE) was from Molecular Probes. The GRGDS and GRGES peptides were from Bachem. Custom synthetic heptapeptides SVVYGLR and SLAYGLR were from Invitrogen.
Quantitative Gelatin Zymography and Superoxide Detection in Cultured Cells
Before each analysis, 0.5 million aortic VSMCs were plated on a 15-cm tissue culture dish, grown until confluent, and then seeded at high cell densities, ie, 0.5 to 0.75x105 cells per well (96-well plates) or 1.0 to 1.5x105 cells per well (12-well plates). Treatment with 30.5 mmol/L glucose versus the 5.5 mmol/L glucose+25 mmol/L mannitol osmotic control was performed in basal medium, Eagle’s (BME). All other treatments were performed in serum-free DMEM (25 mmol/L glucose) for 18 hours. Twenty-microliter aliquots of serum-free conditioned media were analyzed by zymography (2 to 4 replicates) using precast 10% Zymogram Gelatin Gels (Invitrogen/Novex) per the instructions of the manufacturer (Novex technical bulletin IM-1002, Version B). With constant gentle agitation, gels were renatured for 30 minutes at room temperature, developed overnight at 37°C, fixed and stained with Colloidal Blue (Novex technical bulletin IM-6025), and extensively washed (>20 hours) to yield uniform background signal, and digital images of stained wet gels were captured using a Hewlett-Packard ScanJet 5370C bed scanner. Image analysis for was performed by importing scanned JPEG images into Kodak 1D image analysis software essentially as previously described11 but using net pixel intensity of the pro-MMP9 zymogram with uniform rectangular region of interest to quantify MMP activity. For 8-IsoP determination, 25- to 50-μL aliquots were used in immunoassays (n=4 per group). For superoxide detection, cells were plated as above, rinsed with DMEM, incubated with 5 μmol/L DHE in DMEM for 20 minutes, and rinsed, and light and fluorescence microscopy with digital image was capture performed as described.12,13 Image analysis for net pixel intensity was performed by importing JPEG images into Kodak 1D image analysis software using uniform region of interest area for pro-MMP9 f-bands.14 Data are expressed as net pixel intensity, normalized either to mass of extracted tissue protein or the experimental control as indicated. Statistical analyses were performed using Student’s unpaired t test (Microsoft Excel 2002), with graphical data presented as the mean±SE of independent replicates (n=4 to 7).
All other methods are detailed in the expanded Materials and Methods section in the online data supplement available at http://circres.ahajournals.org.
Results
TNF-– and Glucose-Enhanced Pro-MMP9 Activation Is Reduced in OPN–/– Primary Aortic VSMCs
To better understand the role of OPN in diabetic vascular disease, we examined the effects on MMP9 in primary aortic AMFs and A7r5 VSMCs. Previous data suggested that pro-MMP9 might be the preferentially activated by an OPN transgene.6 Therefore, we focused on enzyme activity via gelatin zymography, a robust MMP9 and pro-MMP9 assay.7 Like AT-II, OPN and TNF- increased cellular gelatinase activity; however, whereas AT-II activated both MMP9 and pro-MMP9, OPN and TNF- preferentially upregulated pro-MMP9 (Figure 1A). Image analysis of zymograms from multiple independent experiments (supplemental Figure IA) revealed that 50 nmol/L OPN treatment upregulated pro-MMP9 activity 5-fold (Figure 1B; P=0.004). Immunoprecipitation with anti-MMP9 antibody enriched this gelatinase activity, confirming that it was indeed pro-MMP9; minimal pro-MMP9 activity was immunoprecipitated from OPN–/– aortic AMFs (supplemental Figure IB). Basal pro-MMP9 levels were lower in OPN–/– AMFs (Figures 1C and 4D; see also supplemental Figure IB). Because AMF and VSMC OPN expression is entrained to ambient glucose concentration,4,5 we examined effects of glucose levels on pro-MMP9 activation. With OPN+/+ AMFs, basal pro-MMP9 activity was greater when cultured in 30.5 mmol/L glucose as compared with the isoosmotic control of 5.5 mmol/L glucose with 25 mmol/L mannitol (Figure 1C, top, lane 3 versus lane 1). MMP3 activity (casein zymogram, Figure 1C, bottom, lane 3 versus lane 1) was not affected by glucose. TNF- induction of pro-MMP9 was not enhanced by high glucose (Figure 1C, lane 4 versus lane 2). We next examined whether OPN was required for pro-MMP9 regulation using aortic AMFs isolated from OPN–/– mice. Under every condition, pro-MMP9 activity was reduced in OPN–/– AMFs versus OPN+/+ controls (Figure 1C, top, lanes 5 to 8 versus lanes 1 to 4; P=0.008). MMP3 activity was not diminished (Figure 1C, bottom, lanes 5 to 8 versus lanes 1 to 4). Moreover, neutralization of endogenous OPN in A7r5 cells with an anti-OPN antibody abrogated induction of pro-MMP9 activity by TNF- (Figure 1D) as compared with pro-MMP9 induction elicited in presence of control IgG. Thus, in aortic AMFs and A7r5 VSMCs, pro-MMP9 activation is regulated by OPN signaling.
Inhibitors of Superoxide Signaling Reduce Pro-MMP9 Activation by TNF- and OPN
Many inflammatory cytokines signal via nuclear factor (NF)-B, a well-characterized transcription factor that supports expression of MMP9.15 Thus, we considered that NF-B might mediate pro-MMP9 activation in aortic AMFs and VSMCs. The thiol reagent N-acetylcysteine (NAC)16 and IKK inhibitor BAY 11-7082 (BAY)17 both inhibit NF-B signaling but via distinct mechanisms (supplemental Figure IIA). Therefore, we examined the effects of these compounds on pro-MMP9 activation in A7r5 cells. As shown in Figure 2A, both compounds are active as inhibitors of TNF-–induced NF-B signaling, indicated by the transfected reporter NF-B–LUC. Surprisingly, whereas NAC inhibited pro-MMP9 activation at high cell densities, BAY did not (Figure 2B and supplemental Figure IIB and IIC). NAC also suppresses superoxide generation18; therefore, we assessed the effects of diphenyleneiodonium (DPI) and apocynin on pro-MMP9 activation. Apocynin (inhibits p47 and p67 cytosol-to-membrane translocation) and DPI (targets heme b in NADPH oxidase subunits as well as flavin centers) are mechanistically distinct inhibitors of cellular superoxide generation routinely used to functionally test contributions of NADPH oxidases to signal transduction.12,18 Like NAC, DPI reduced TNF- induction of pro-MMP9 by 50% (supplemental Figure IIB). Similar results were obtained with apocynin, which preferentially abolished TNF- induction of pro-MMP9 (Figure 2C and 2D). The enzyme superoxide dismutase (SOD) catalytically destroys superoxide to generate oxygen and hydrogen peroxide.18 As observed with NAC, apocynin, and DPI, treatment with SOD prevented pro-MMP9 activation by either TNF- or OPN (not shown). Thus, redox signaling participates in cytokine activation of pro-MMP9.
Superoxide and Superoxide-Generating NADPH Oxidase Subunits Are Regulated by OPN
Inhibition of pro-MMP9 activation by NAC, DPI, apocynin, and SOD strongly suggested that superoxide production was a component of OPN signaling. Therefore, we examined the effects of OPN on superoxide generation as revealed by fluorescence staining with DHE, a compound chemically oxidized to the fluorescent DNA intercalating agent ethidium in proportion to the amount of superoxide present.12,14 As compared with OPN+/+ aortic AMFs, OPN–/– aortic AMFs exhibited reduced DHE staining (Figure 3A). Because NADPH oxidases are embedded in the membrane bilayer, the short-lived superoxide radicals react with membrane lipids, generating oxylipids19 that demarcate and mediate oxidative signaling.18 One category of oxylipids, isoprostanes, arises from the chemical oxidation of arachidonic acid in phospholipids.20 Therefore, we assessed the relationship between OPN, isoprostane production (8-IsoP), and pro-MMP9 activation in AMFs. 8-IsoP is reduced by 50% in OPN–/– AMFs as compared with OPN+/+ controls (Figure 3B). Moreover, exogenous OPN protein upregulated DHE staining (Figure 3C) and 8-IsoP (Figure 3D) in A7r5 VSMCs. Furthermore, anti-OPN neutralizing antibody inhibited basal 8-IsoP production by 50%, presumably via inhibition of endogenous OPN,5 whereas an isotype-matched control antibody had no effect (Figure 3D).
NADPH oxidases are specific but heterogeneous complexes of p67, p47, p40, and p22 subunits with cell type–specific NADPH oxidase subunits.18 Other important interacting proteins (eg, the Rac family GTPases and cortactin) convey regulatory responses to vasculotropic hormones such as AT-II.21 In human and rodent aortic medial VSMCs, Nox1 is a predominant isoform,18,22 whereas Nox2 is highly expressed in aortic AMFs.23 Therefore, we assessed the levels of NADPH oxidase subunit in aortic AMFs from OPN+/+ versus OPN–/– mice. As shown in Figure 3E, Nox2, Nox1, p67, and p47 are reduced in cell extracts from OPN–/– AMFs as compared with OPN+/+ controls. No reduction of the housekeeping protein eIF-2 was observed in OPN–/– AMFs (Figure 3E). Because Pagano and colleagues had demonstrated a critical role for Nox2 in AMF physiology,23 we emphasized Nox2 and its associated subunits. Western blots analyzed from 3 independent sets of cultures demonstrated a >50% reduction in Nox2 accumulation in OPN–/– AMFs versus OPN-replete control cultures (Figure 3F). Of note, p67—a regulatory subunit required for electron transport between NADPH and the flavin center that generates superoxide18—is reduced along with Nox2 in OPN–/– AMFs (Figure 3E). Moreover, addition of exogenous OPN upregulated p67 in A7r5 VSMCs (supplemental Figure IIIA). Finally, a validated antisense oligonucleotide (ASO) directed to p47 significantly reduced 8-IsoP levels in A7r5 cells (supplemental Figure IIIB), confirming contributions of prototypic NADPH oxidases to A7r5 oxylipid formation. Thus, superoxide production, oxylipid formation, and NADPH oxidase subunit accumulation is regulated by OPN in aortic mesenchymal cells.
The OPN Peptide SVVYGLR Activates Vascular Pro-MMP9 and Superoxide Signaling In Vitro
OPN is a phosphorylated glycoprotein, processed by thrombin to generated N-terminal and C-terminal domains with unique bioactivities8 (supplement Figure IVA). Like full-length bOPN (Figure 1A and 1B above, and data not shown), the N-terminal recombinant murine OPN fragment dose-dependently activates pro-MMP9 (Figure 4A), whereas the C-terminal OPN fragment is inactive (not shown). The N-terminal domain of OPN contains v3 and 41 or 47 binding subdomains that mediate adhesion and signaling.8 Of note, a synthetic peptide corresponding to the SVVYGLR motif of human OPN upregulated pro-MMP9 activity (Figure 4B); GRGDS and GRGES peptides were insufficient. Similar results were observed with murine OPN SLAYGLR peptide (not shown). We confirmed these results using HA-VSMCs, and an independent immunoassay specific for human pro-MMP9; 50 μmol/L SVVYGLR upregulated pro-MMP9 approximately 6-fold (Figure 4C, P=0.029, 2-tailed t test). Induction of pro-MMP9 by TNF- is poor in AMFs from OPN–/– mice (Figure 4D; also Figure 1C). However, incubation with either 50 μmol/L SVVYGLR peptide (lanes 3 and 4) or 50 nmol/L bOPN (lanes 5 to 6) restored robust TNF- induction of pro-MMP9 (Figure 4D). Thus, OPN signaling via the SVVYGLR motif plays an important role in pro-MMP9 activation.
We next evaluated effects of OPN peptides on superoxide and oxylipid formation. Like full-length OPN, recombinant OPN N-terminal fragment increased A7r5 8-IsoP production; OPN C-terminal domain did not (Figure 5A). Moreover, the N-terminal heptapeptide SVVYGLR, but not GRGDS or GRGES, upregulated 8-IsoP (Figure 5B). SVVYGLR upregulated superoxide (DHE staining) in A7r5 VSMCs (2-fold; supplemental Figure VA) and OPN–/– AMFs (5-fold; not shown), quantified by digital image analysis. Thus, the OPN N-terminal domain encodes motifs necessary for pro-MMP9 activation and superoxide and oxylipid formation.
The OPN Peptide SVVYGLR Activates Vascular Pro-MMP9 and Superoxide Signaling In Vivo
We next injected SVVYGLR (n=5) or vehicle (n=5) into OPN+/– mice, an in vivo background partially deficient in endogenous OPN tone. Intraperitoneal SVVYGLR administration (8 μg/g) significantly upregulated pro-MMP9 activity in aortic extracts (P=0.03; Figure 6A). Mural actions of SVVYGLR were spatially resolved by DHE staining of aortic frozen sections; adventitial and medial tunics exhibited enhanced superoxide production (Figure 6B and 6C), indicating cellular responses in both compartments. Pretreatment of sections with PEG-SOD reduced nuclear DHE staining, but not elastin autofluorescence, confirming superoxide-generated signal (Figure 6B, e and f). SVVYGLR significantly upregulated aortic 8-IsoP in vivo (Figure 6D; P=0.019). Moreover, Western blot analysis of individual aortic extracts confirmed that, as observed with in vitro (supplemental Figure IIIA), aortic Nox2 protein levels were upregulated 2.3-fold by OPN SVVYGLR peptide in vivo (P=0.07 for trend approaching significance, supplemental Figure VB, n=5 per group). To examine the generality of signaling, we assessed pro-MMP9 activation by SVVYGLR in pulmonary tissues. As observed in aorta, SVVYGLR upregulates pro-MMP9 activity in lung (Figure 6E and 6F). Thus, OPN peptide SVVYGLR upregulates levels of vascular superoxide, aortic 8-IsoP formation, and tissue pro-MMP9 activation in vivo.
8-IsoP and Aortic Pro-MMP9 Are Reduced in OPN-Deficient LDLR–/– Mice Fed High-Fat Diabetogenic Diets
We wished to examine whether OPN signaling was important for vascular stress responses to diabetes in vivo. The male LDLR–/– mouse has proven useful for studies of diet-induced diabetes, dyslipidemia, and vascular disease.4,11 Therefore, we studied the effect of high-fat diabetogenic diets (Western diet) on OPN+/–/LDLR–/– versus OPN–/–/LDLR–/– mice. Reduced plasma OPN levels confirmed OPN deficiency in LDLR–/–/OPN–/– versus LDLR–/–/OPN+/– mice (Figure 7A). Seven weeks of a high-fat diet generated equivalent diabetes (mean±SEM: glucose of 514±31 versus 496±28 mg/dL, P=0.39; 1-tailed t test) and dyslipidemia (free fatty acids 2.1±0.05 versus 2.0±0.05 mmol/L, P=0.11; cholesterol 1804±196 versus 1807±129 mg/dL, P=0.49). By contrast, plasma 8-IsoP levels were lower in LDLR–/–/OPN–/– versus LDLR–/–/OPN+/– mice (Figure 7B; P=0.017). Moreover, aortic pro-MMP9 activity was reduced in diabetic LDLR–/–/OPN–/– versus LDLR–/–/OPN+/– (Figure 7C and 7D; P=0.04). Thus, OPN participates in isoprostane generation and aortic pro-MMP9 activation in the LDLR–/– mouse model of diet-induced diabetes.4
Oxylipids Preferentially Activate Pro-MMP9
Oxylipids not only reflect the presence of superoxide generation but also mediate inflammatory signals.20 Thus, we tested the ability of oxylipid to activate pro-MMP9, evaluating 8-IsoP, which is among the many isoprostanes generated from the heterogeneous oxylipid precursor oxidized palmitoyl arachidonoyl phosphatidylcholine (Ox-PAPC).19,20 Like OPN and TNF- (Figure 1A), 8-IsoP preferentially upregulated pro-MMP9 in HA-VSMCs and A7r5 aortic VSMCs (Figure 8A and 8B). Similar results were obtained in OPN–/– aortic AMFs treated with 8-IsoP (3.5-fold induction; P=0.02, not shown). However, in contrast to activation by TNF- and OPN (Figure 2B and supplemental Figure IIB), pro-MMP9 activation by 8-IsoP is insensitive to NAC (Figure 8B, right) and DPI (not shown), confirming the potential contributions of oxylipids as terminal effectors (supplemental Figure IIA). We also tested the ability of 8-IsoP to activate pro-MMP9 in the presence of anti-OPN antibody in A7r5 cells, a treatment that reduces pro-MMP9 activity (Figure 1D; and Figure 8C, lanes 1 and 2) and endogenous 8-IsoP production (Figure 3D). As shown in Figure 8C, whereas anti-OPN decreased pro-MMP9 activity (lanes 1 and 2), 8-IsoP treatment dose-dependently restored pro-MMP9 activity (lanes 3 to 5). In aqueous solutions, 3-morpholinosydnonimine-N-ethylcarbamide (SIN-1) decomposes to form superoxide and nitric oxide (and thus peroxynitrite).10 SIN-1 dose-dependently upregulated 8-IsoP accumulation in A7r5 cells (Figure 8D), confirming that superoxide generates oxylipids in aortic VSMCs. Moreover, like OPN and 8-IsoP, SIN-1 upregulated pro-MMP9 but also enhanced MMP9 activation (Figure 8E). Little effect on MMP2 was noted. Thus, OPN- and superoxide-regulated oxylipids can participate in the activation of pro-MMP9 in VSMCs.
Discussion
VSMCs and AMFs participate in every aspect of arterial pathobiology.23,24 Proliferative and synthetic responses contribute to mural thickness, neointima formation, fibrosis, calcification, and vascular stiffening.6 Contractile responses modulate arterial tone, wound repair, and hemostasis.3,23,24 Very few well-validated markers differentiate VSMCs of the media from adventitial AMFs24; however, AMFs exhibit higher levels of NADPH oxidase activity.23,24 Aortic AMFs elaborate robust metabolic responses in the setting of diabetes, dyslipidemia, and hypertension.23,24 Thus, both adventitial AMFs and mural VSMCs are important mediators of vascular disease progression.3,23,24
SVVYGLR treatment activated oxidative stress in both adventitial and medial tissue compartments, consistent with our observations made with aortic AMF and VSMC culture models. Other vascular beds are also responsive, because pulmonary pro-MMP9 activity was significantly upregulated by OPN SVVYGLR peptide. Under basal conditions, total NADPH oxidase activity is higher in adventitial versus medial arterial compartments.3,23 With diabetes, adventitial oxidase activity is further enhanced, and tunica media NADPH oxidase is dramatically upregulated.3 Because OPN is a hyperglycemia-induced cytokine,5 the macrovascular oxidative stress of diabetes3 is potentially augmented by paracrine OPN signaling. Consistent with this, both 8-IsoP levels and aortic pro-MMP9 activity are reduced in diabetic LDLR–/– mice lacking the OPN gene. Advanced glycosylation end products upregulate proinflammatory vascular cytokines such as TNF-.3 Thus, in our working model (supplemental Figure IIA), oxylipids19 such as 8-IsoP generated via OPN-dependent signals would augment mural pro-MMP9 in diabetes. The plasma 8-IsoP levels observed in vivo (Figure 7E) are 1000-fold lower than those necessary to upregulate pro-MMP9 in vitro. Although 8-IsoP is 1 bioactive oxylipid generated in response to OPN signaling, many other oxylipids are produced during vascular disease progression.19 Ox-PAPC, the oxylipid precursor of isoprostanes,19,20 also activates pro-MMP9 (not shown). Thus, although 8-IsoP is 1 oxylipid readily assayed and studied, the spectrum of vascular oxylipids that are upregulated in diabetes and participate in pro-MMP9 activation has yet to be determined. Of note, oxylipids such as isoprostanes or 4-hydroxynonenal could activate pro-MMP9 via covalent alkylation reactions involving pro-MMP9 propeptide.10
At high cell densities, we see no effect of BAY compound on OPN or TNF- pro-MMP9 activation in vitro. However, when cells are sparsely subconfluent, BAY compound weakly inhibits MMP9 induction, suggesting that NF-B signaling contributes at low cell densities; moreover, unlike TNF-, pro-MMP9 activation by TGF- is not altered by OPN deficiency (not shown). Thus, multiple pathways must participate in vascular MMP9 activation during development and disease.
Galis and Khatri first demonstrated that oxidative stress activates collagenases in phagocytes, including pro-MMP9 and MMP910; this resembles our results with SIN-1. Our data extend these observations by identifying that OPN signaling and oxylipids preferentially activate pro-MMP9 in aortic mesenchymal cells.7 However, the specific NADPH oxidase complex mediating this vascular response is as yet unknown and may differ between media and adventitia.18,23 Future experiments will identify the relative contributions of NADPH oxidase isoforms to OPN signaling in medial versus adventitial tunics.
With diabetes, periaortic AMFs express high levels of OPN at the earliest stages of disease.4 Both adventitial and medial oxidative stresses are enhanced in response to diabetes-induced NADPH oxidase activity.3,24 We propose that hyperglycemia-induced aortic OPN expression4,5 contributes to the vascular inflammation of diabetes, augmenting mural oxidative stress, pro-MMP9 activation, medial thickening, and matrix remodeling (supplemental Figure IIA). Thus, strategies that inhibit paracrine OPN–superoxide signaling may ameliorate the macrovascular injury of diabetes. Liaw, Giachelli, and colleagues first showed that neutralizing OPN reduced neointima formation25; whether this antibody concomitantly regulated vascular oxidative stress is currently unknown. Our data add to this evidence,25 highlighting a potential benefit in selective inhibition of OPN SVVYGLR signaling as 1 strategy to reduce inflammatory vascular remodeling in diabetes and, thus, preserve macrovascular structure and function.7,25
Acknowledgments
Sources of Funding
This work was supported by the American Diabetes Association and the National Heart, Lung, and Blood Institute.
Disclosures
None.
Footnotes
Original received August 29, 2005; resubmission received January 11, 2006; revised resubmission received April 6, 2006; accepted May 8, 2006.
References
Zimmet P, Alberti KG, Shaw J. Global and societal implications of the diabetes epidemic. Nature. 2001; 414: 782–787. [Order article via Infotrieve]
Lehmann RR, Nienhaus RH, Denes RM, Steinbach T. Changes of euchromatin/heterochromatin ratios in cell nuclei of the aortic adventitia in diabetic rats. Artery. 1987; 14: 66–75. [Order article via Infotrieve]
Zhang L, Zalewski A, Liu Y, Mazurek T, Cowan S, Martin JL, Hofmann SM, Vlassara H, Shi Y. Diabetes-induced oxidative stress and low-grade inflammation in porcine coronary arteries. Circulation. 2003; 108: 472–478.
Towler DA, Bidder M, Latifi T, Coleman T, Semenkovich CF. Diet-induced diabetes activates an osteogenic gene regulatory program in the aortas of low density lipoprotein receptor-deficient mice. J Biol Chem. 1998; 273: 30427–30434.
Bidder M, Shao JS, Charlton-Kachigian N, Loewy AP, Semenkovich CF, Towler DA. Osteopontin transcription in aortic vascular smooth muscle cells is controlled by glucose-regulated upstream stimulatory factor and activator protein-1 activities. J Biol Chem. 2002; 277: 44485–44496.
Isoda K, Nishikawa K, Kamezawa Y, Yoshida M, Kusuhara M, Moroi M, Tada N, Ohsuzu F. Osteopontin plays an important role in the development of medial thickening and neointimal formation. Circ Res. 2002; 91: 77–82.
Bruemmer D, Collins AR, Noh G, Wang W, Territo M, Arias-Magallona S, Fishbein MC, Blaschke F, Kintscher U, Graf K, Law RE, Hsueh WA. Angiotensin II-accelerated atherosclerosis and aneurysm formation is attenuated in osteopontin-deficient mice. J Clin Invest. 2003; 112: 1318–1331.
Denhardt DT, Noda M, O’Regan AW, Pavlin D, Berman JS. Osteopontin as a means to cope with environmental insults: regulation of inflammation, tissue remodeling, and cell survival. J Clin Invest. 2001; 107: 1055–1061.
Steitz SA, Speer MY, McKee MD, Liaw L, Almeida M, Yang H, Giachelli CM. Osteopontin inhibits mineral deposition and promotes regression of ectopic calcification. Am J Pathol. 2002; 161: 2035–2046.
Galis ZS, Khatri JJ. Matrix metalloproteinases in vascular remodeling and atherogenesis: the good, the bad, and the ugly. Circ Res. 2002; 90: 251–262.
Shao JS, Cheng SL, Charlton-Kachigian N, Loewy AP, Towler DA. Teriparatide (human parathyroid hormone (1–34)) inhibits osteogenic vascular calcification in diabetic low density lipoprotein receptor-deficient mice. J Biol Chem. 2003; 278: 50195–50202.
Bleeke T, Zhang H, Madamanchi N, Patterson C, Faber JE. Catecholamine-induced vascular wall growth is dependent on generation of reactive oxygen species. Circ Res. 2004; 94: 37–45.
Shao JS, Cheng SL, Pingsterhaus JM, Charlton-Kachigian N, Loewy AP, Towler DA. Msx2 promotes cardiovascular calcification by activating paracrine Wnt signals. J Clin Invest. 2005; 115: 1210–1220.
Dai X, Galligan JJ, Watts SW, Fink GD, Kreulen DL. Increased O2- production and upregulation of ETB receptors by sympathetic neurons in DOCA-salt hypertensive rats. Hypertension. 2004; 43: 1048–1054.
Rangaswami H, Bulbule A, Kundu GC. Nuclear factor-inducing kinase plays a crucial role in osteopontin-induced MAPK/IB kinase-dependent nuclear factor B-mediated promatrix metalloproteinase-9 activation. J Biol Chem. 2004; 279: 38921–38935.
Langen RC, Schols AM, Kelders MC, Van Der Velden JL, Wouters EF, Janssen-Heininger YM. Tumor necrosis factor- inhibits myogenesis through redox-dependent and -independent pathways. Am J Physiol Cell Physiol. 2002; 283: C714–C721.
Yang G, Abate A, George AG, Weng YH, Dennery PA. Maturational differences in lung NF-B activation and their role in tolerance to hyperoxia. J Clin Invest. 2004; 114: 669–678.
Cai H. NAD(P)H oxidase-dependent self-propagation of hydrogen peroxide and vascular disease. Circ Res. 2005; 96: 818–822.
Navab M, Ananthramaiah GM, Reddy ST, Van Lenten BJ, Ansell BJ, Fonarow GC, Vahabzadeh K, Hama S, Hough G, Kamranpour N, Berliner JA, Lusis AJ, Fogelman AM. The oxidation hypothesis of atherogenesis: the role of oxidized phospholipids and HDL. J Lipid Res. 2004; 45: 993–1007.
Morrow JD. Quantification of isoprostanes as indices of oxidant stress and the risk of atherosclerosis in humans. Arterioscler Thromb Vasc Biol. 2005; 25: 279–286.
Touyz RM, Yao G, Quinn MT, Pagano PJ, Schiffrin EL. p47phox associates with the cytoskeleton through cortactin in human vascular smooth muscle cells: role in NAD(P)H oxidase regulation by angiotensin II. Arterioscler Thromb Vasc Biol. 2005; 25: 512–518.
Sorescu GP, Song H, Tressel SL, Hwang J, Dikalov S, Smith DA, Boyd NL, Platt MO, Lassegue B, Griendling KK, Jo H. Bone morphogenic protein 4 produced in endothelial cells by oscillatory shear stress induces monocyte adhesion by stimulating reactive oxygen species production from a nox1-based NADPH oxidase. Circ Res. 2004; 95: 773–779.
Liu J, Ormsby A, Oja-Tebbe N, Pagano PJ. Gene transfer of NAD(P)H oxidase inhibitor to the vascular adventitia attenuates medial smooth muscle hypertrophy. Circ Res. 2004; 95: 587–594.
Sartore S, Chiavegato A, Faggin E, Franch R, Puato M, Ausoni S, Pauletto P. Contribution of adventitial fibroblasts to neointima formation and vascular remodeling: from innocent bystander to active participant. Circ Res. 2001; 89: 1111–1121.
Liaw L, Lombardi DM, Almeida MM, Schwartz SM, deBlois D, Giachelli CM. Neutralizing antibodies directed against osteopontin inhibit rat carotid neointimal thickening after endothelial denudation. Arterioscler Thromb Vasc Biol. 1997; 17: 188–193.
您现在查看是摘要介绍页，详见ORG附件。