Nitric Oxide Attenuates Vascular Smooth Muscle Cell Activation by Interferon- (cid:103) THE ROLE OF CONSTITUTIVE NF- (cid:107) B ACTIVITY*

Atherogenesis involves cellular immune responses and altered vascular smooth muscle cell (SMC) function. Cytokines such as interleukin (IL)-1 (cid:97) and interferon- (cid:103) (IFN- (cid:103) ) may contribute to this process by activating SMC. To determine whether the anti-atherogenic medi- ator, nitric oxide ( (cid:122) NO), can modulate cytokine-induced SMC activation, we investigated the effects of various (cid:122) NO-generating compounds on the expression of intercellular and vascular cell adhesion molecules (ICAM-1 and VCAM-1). Induction of ICAM-1 expression by IL-1 (cid:97) and VCAM-1 expression by IFN- (cid:103) was attenuated by (cid:122) NO donors but not by cGMP analogues. Nuclear run-on as- says and transfection studies using various VCAM-1 promoter constructs linked to the chloramphenicol acetyl- transferase reporter gene showed that (cid:122) NO repressed IFN- (cid:103) -induced VCAM-1 gene transcription, in part, through inhibition of nuclear factor- (cid:107) B (NF- (cid:107) B). Electrophoretic mobility shift assay revealed that SMC pos- sess basal constitutive NF- (cid:107) B activity, which was augmented by treatment with IL-1 (cid:97) . In contrast, IFN- (cid:103) induced and activated interferon regulatory factor (IRF)-1 but had little effect on basal constitutive NF- NO had no inhibitory effect on IRF-1 activation but did inhibit basal and IL-1 (cid:97) -stimulated NF- (cid:107) B activation. These findings suggest that the induction of ICAM-1 and VCAM-1 expression requires NF- (cid:107) B activation and that (cid:122) NO attenuates IFN- (cid:103) -induced VCAM-1 expression primarily by inhibiting basal consti- tutive NF- (cid:107) B activity in SMC. specificity The human VCAM-1 promoter constructs containing the CAT reporter were previously described by Neish et al. (28). Human SMC were transfected with each reporter plasmid (25 (cid:109) g) using the calcium phos- phate precipitation method (10). As an internal control for transfection efficiency, pRSV. (cid:98) GAL plasmid (10 (cid:109) g) was co-transfected in all exper- iments. Preliminary results using (cid:98) -galactosidase staining indicate that cellular transfection efficiency was approximately 15%. Cells (60– 70% confluent) were stimulated 48 h after transfection with IFN- (cid:103) (1000 units/ml) in the presence and absence of GSNO (0.2 m M ), and cellular extracts were prepared 12 h later using lysis buffer (100 (cid:109) g/ml leupeptin, 50 (cid:109) g/ml aprotinin, 0.1 m M phenylmethylsulfonyl fluoride, 5 m M EDTA, 5 m M EGTA, 100 m M NaCl, 5 m M Tris-HCl, pH 7.4) and one freeze-thaw cycle. The cellular extracts were centrifuged at 12,000 (cid:51) g for 10 min, and the supernatant was subjected to CAT and (cid:98) -galacto-sidase assay as described previously (10, 17). The relative CAT activity was calculated as the ratio of CAT to (cid:98) -galactosidase activity. Each experiment was performed three times in duplicate, and all experi- ments included both positive (highly expressed pSV40.CAT)

Atherosclerotic lesions contain proliferating intimal smooth muscle cells (SMC) 1 and cytokines such as tumor necrosis factor (TNF)-␣ and interleukin (IL)-1 (1)(2)(3). Although the involvement of cytokines in atherogenesis is well established, their signaling events leading to SMC activation and proliferation are still poorly understood. Recent studies have suggested that many cytokines activate the oxidant-sensitive transcription factor, nuclear factor-B (NF-B) (4,5), which may be important in mediating SMC activation and proliferation (6,7). Activated SMC express proinflammatory genes such as intercellular and vascular cell adhesion molecules (ICAM-1 and VCAM-1) (8,9). Indeed, we have shown that cytokines such as IL-1␣ and TNF-␣ can activate NF-B and induce the expression of VCAM-1 in human vascular endothelial cells (10). It is not known, however, whether ⅐ NO can similarly modulate cytokine-induced NF-B activity in SMC.
SMC responds to endothelium-derived nitric oxide ( ⅐ NO), which has emerged as an important modulator of vascular tone via stimulation of soluble guanylyl cyclase (11,12). However, ⅐ NO may have other important effects on SMC such as inhibition of SMC activation and proliferation (13,14). Supplementation of L-arginine, the precursor of ⅐ NO, lessens the extent of atherosclerosis in diet-induced hypercholesterolemic rabbits (15). In vivo transfer of the type III ⅐ NO synthase gene into balloon-injured vessels decreases intimal SMC proliferation in rat carotid arteries (16). These studies demonstrate that ⅐ NO can antagonize the effects of cytokines and growth factors, in part, by attenuating SMC activation and proliferation. Although the mechanism(s) by which ⅐ NO exerts its inhibitory effect(s) on SMC is not presently known, recent studies from our laboratory have indicated that ⅐ NO can modulate endothelial activation via cGMP-independent inhibition of cytokineinduced NF-B activation (17,18). Thus, ⅐ NO production in the vessel wall may influence SMC not only in their vasomotor functions, but also perhaps in their more prolonged transcriptional responses to NF-B activation by cytokines.
The cellular immune response in atherosclerotic lesions is evidenced by the marked infiltration of T-lymphocytes (19,20). Although the precise role of T-lymphocytes in the vessel wall has not been established, recent findings suggest that T-lymphocytes can modulate SMC activation, in part, through the lymphokine, interferon-gamma (IFN-␥) (21). In contrast to cytokines such as TNF-␣ and IL-1␣, IFN-␥ is not known to activate NF-B or induce VCAM-1 expression in endothelial cells (22). IFN-␥, however, can potently induce the expression of VCAM-1 and major histocompatability complex class II anti-* This work was supported by National Institutes of Health Grants HL-05280 and HL-52233 (to J. K. L.) and Grant HL-34636 (to P. L.) and by an American Heart Association grant-in-aid award (to J. K. L.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Since SMC, but not endothelial cells, possess basal constitutive NF-B/Rel-like activity (6,7,31), we hypothesized that the presence of this basal constitutive NF-B activity may contribute to the differential responses of vascular wall cells to IFN-␥. The purpose of this study, therefore, was to determine the role of basal constitutive NF-B activity in mediating the effects of IFN-␥ and ⅐ NO on VCAM-1 expression in SMC. We found that ⅐ NO can modulate IFN-␥-induced SMC activation through its effects on basal constitutive NF-B activity.
Cellular confluence was maintained for all treatment conditions. Cellular viability was assessed by morphology, cell number, DNA content, and trypan blue exclusion. The amount of DNA was measured by a Microfluor reader (Dynatech Laboratories, Inc., Chantilly, VA) using a fluorescent dye (Hoechst 33258) that binds specifically to DNA (Calbiochem).
Cell Surface Enzyme Immunoassay-Cytokine-stimulated SMC were cultured on 96-well Falcon plates (Lincoln Park, NJ), rinsed with phosphate-buffered saline and 2% fetal calf serum, and incubated with the indicated murine monoclonal antibody to human ICAM-1 and VCAM-1 for 2 h. After rinsing three times with phosphate-buffered saline, cells were incubated with biotinylated secondary antibody (horse anti-mouse IgG, Vector Labs, Inc., Burlingame, CA, 1:1000) for 1 h before incubation with streptavidin-alkaline phosphatase (Zymed, South San Francisco, CA) for 30 min. Cells were then treated with p-nitrophenylphosphate (1 g/ml) for 30 min at room temperature. Light absorbance was measured in a plate reader (Dynatech Laboratories) at 410 nm, using cells without primary antibody as a blank.
Electrophoretic Mobility Shift Assay-Nuclear extracts were prepared as described (34). Oigonucleotides corresponding to the B (5Ј-TGCCCTGGGTTTCCCCTTGAAGGGATTTCCCTCC-3Ј) and ISRE (5Ј-GGAGTGAAATAGAAAGTCTGTG-3Ј) sites in the VCAM-1 promoter were radiolabeled with [␥-32 P]ATP and T 4 polynucleotide kinase (New England Biolabs) and purified by G-50 Sephadex columns (Pharmacia). Nuclear extracts (10 g) were added to 32 P-labeled oligonucleotides (ϳ20,000 cpm, 0.2 ng) in a buffer containing 4 g of poly(dI⅐dC) (Boehringer Mannheim), 10 g of bovine serum albumin, 10 mM Tris-HCl (pH 7.5), 50 mM NaCl, 1 mM dithiothreitol, 1 mM EDTA, and 5% glycerol (total volume of 20 l). DNA-protein complexes were resolved on 4% nondenaturing polyacrylamide gel electrophoresed at 12 V/cm for 3 h in low ionic strength buffer (0.5 ϫ TBE) at 4°C. For supershift assays, the indicated antibody (15 g/ml) was added to the nuclear extracts for 10 min before the addition of radiolabeled probe. To determine the specificity of shifted bands, excess unlabeled oligonucleotide (20 ng) was added directly to the nuclear extracts for 10 min before the addition of corresponding radiolabeled probe.
Transfection and Chloramphenicol Acetyltransferase (CAT) Assay-The human VCAM-1 promoter constructs containing the CAT reporter were previously described by Neish et al. (28). Human SMC were transfected with each reporter plasmid (25 g) using the calcium phosphate precipitation method (10). As an internal control for transfection efficiency, pRSV.␤GAL plasmid (10 g) was co-transfected in all experiments. Preliminary results using ␤-galactosidase staining indicate that cellular transfection efficiency was approximately 15%. Cells (60 -70% confluent) were stimulated 48 h after transfection with IFN-␥ (1000 units/ml) in the presence and absence of GSNO (0.2 mM), and cellular extracts were prepared 12 h later using lysis buffer (100 g/ml leupeptin, 50 g/ml aprotinin, 0.1 mM phenylmethylsulfonyl fluoride, 5 mM EDTA, 5 mM EGTA, 100 mM NaCl, 5 mM Tris-HCl, pH 7.4) and one freeze-thaw cycle. The cellular extracts were centrifuged at 12,000 ϫ g for 10 min, and the supernatant was subjected to CAT and ␤-galactosidase assay as described previously (10,17). The relative CAT activity was calculated as the ratio of CAT to ␤-galactosidase activity. Each experiment was performed three times in duplicate, and all experiments included both positive (highly expressed pSV40.CAT) and negative (promoterless p.CAT) controls.
Data Analysis-Band intensities from Northern blots, nuclear run-on assays, and electrophoretic mobility shift assay (EMSA) blots were analyzed densitometrically by the NIH Image program (35). All values are expressed as mean Ϯ S.E. compared with controls and among separate experiments. Paired and unpaired Student's t tests were employed to determine the significance of changes in absorbance values and densitometric measurements. p values of less than 0.05 were considered significant.
To exclude possible cellular toxicity produced by the ⅐ NO donors, we examined their effects on cell number, DNA content, and trypan blue exclusion. We found that neither SNP nor SIN-1, at concentrations used in our study, significantly affected cellular viability with respect to cell number, DNA content, and trypan blue exclusion (Table I). This result agrees with immunohistochemical analyses showing that treatment with ⅐ NO donors did not appreciably affect SMC morphology (Fig. 2).
Stimulation of SMC with either IL-1␣ or IFN-␥ did not induce type II ⅐ NO synthase expression by Northern analyses or result in increased ⅐ NO production by SMC as measured by nitrite production (data not shown). In addition, activation of soluble guanylyl cyclase by exogenous ⅐ NO did not contribute to the observed decrease in cytokine-induced ICAM-1 and VCAM-1 expression, since neither 8-bromo-cGMP (1 mM) nor dibutyryl-cGMP (1 mM) inhibited ICAM-1 or VCAM-1 surface expression (Table II). In fact, there was a slight increase in ICAM-1 and VCAM-1 expression with higher concentrations of 8-bromo-cGMP (0.1 mM to 1.0 mM). 8-bromo-cGMP (1 mM), however, did stimulate cGMP-and probably cAMP-dependent protein kinase activity (Fig. 3).

Inhibition of Constitutive NF-B Activity by Nitric Oxide
imental conditions, hybridization was linear and nonsaturable. The density of each VCAM-1 band was standardized to the density of its corresponding ␤-tubulin band. The specificity of each band was determined by the lack of hybridization to the nonspecific pGEM cDNA vector. In unstimulated SMC (control), there was little basal VCAM-1 transcriptional activity. IFN-␥ augmented VCAM-1 gene transcription by 20-fold. Cotreatment with GSNO (0.2 mM) resulted in only a 3-fold induction, indicating repression of VCAM-1 gene transcription by ⅐ NO. ⅐ NO Inhibits Activation of NF-B, but Not IRF-1-EMSA showed that under our basal culture conditions, there were two constitutive bands corresponding to NF-B that were both "supershifted" in the presence of antibody to p65, whereas only the lower band was supershifted in the presence of antibody to p50 (Fig. 7A). These findings suggest that the composition of NF-B binding to the tandem B sites of the VCAM-1 promoter probably consists of the p65 homodimer (top band) and p50-p65 heterodimer (lower band). The anti-c-Rel antibody neither obliterated nor supershifted these basally active NF-B bands (data not shown). IFN-␥ (1000 units/ml) slightly augmented, while IL-1␣ (10 pg/ml) caused an increase in, basal constitutive NF-B activation. Higher concentrations of IL-1␣ (0.1-10 ng/ ml) produced an even greater activation of NF-B (data not shown). Treatment with ⅐ NO donors inhibited both basal constitutive and IL-1␣-(10 pg/ml) stimulated NF-B activation.
Using the VCAM-1 ISRE oligonucleotide, several different antibodies to p91 (STAT-1␣) failed to supershift any bands induced by IFN-␥ (data not shown), suggesting that interferon-stimulated gene factor-3 (ISGF-3) does not bind to the ISRE of VCAM-1 promoter and, therefore, may play only a limited role in the transcriptional activation of the VCAM-1 promoter by IFN-␥. However, IFN-␥ did induce IRF-1 in a cycloheximidesensitive, time-dependent manner (data not shown). The induction and activation of IRF-1 appeared no sooner than 2 h after stimulation with IFN-␥ and was not inhibited by treatment with ⅐ NO (Fig. 7B).

Induction of VCAM-1 Gene Transcription by IFN-␥ Requires
B Enhancer Element-Transient transfection studies using various VCAM-1 promoter constructs (F 0 , F 3 , and F 4 ) linked to the CAT reporter gene demonstrated that the two tandem B enhancer elements in the VCAM-1 promoter are required for transcriptional induction and repression by IFN-␥ and ⅐ NO (Fig. 8). IRF-1 activity (Fig. 7B), ⅐ NO attenuates IFN-␥-induced VCAM-1 expression via inhibition of basal constitutive NF-B activity. DISCUSSION We have shown that ⅐ NO can attenuate the surface expression of ICAM-1 and VCAM-1 on SMC in response to stimulation with IL-1␣ and IFN-␥, respectively. The mechanism for ⅐ NO's effect is independent of cGMP production, occurs at the transcriptional level, and involves inhibition of both basal constitutive and IL-1␣-stimulated NF-B activity. These findings agree with our earlier findings that ⅐ NO decreases cytokineinduced endothelial expression of VCAM-1 and ICAM-1 via inhibition of NF-B activation (10). However, SMC differ from endothelial cells in exhibiting basal constitutive NF-B activity (6,18). Indeed, we observed a small amount of SMC activation under basal culture conditions as exhibited by low levels of VCAM-1 mRNA expression, gene transcription, and promoter activity. The presence of basal constitutive NF-B activity has also been shown to be important in mediating SMC proliferation (7).
Previous studies have shown that ⅐ NO inhibits SMC proliferation via a cGMP-dependent mechanism (13,14). However, the expression of ICAM-1 and VCAM-1 were not affected by increasing concentrations of two different cGMP analogues that are able to stimulate protein kinase activity. Indeed, several groups have shown that ⅐ NO can exert non-cGMP-dependent effects on other cell types such as platelets (36), fibroblasts (37), and macrophages (38). Interestingly, the inhibitory effects of ⅐ NO on basal and stimulated NF-B activation resemble those of antioxidants such as N-acetylcysteine and pyrrolidine dithiocarbamate (39,40). Antioxidants have been shown to inhibit SMC proliferation, and at higher concentrations they appear to induce SMC apoptosis (41). SMC did not exhibit any signs of cellular toxicity with the concentrations of ⅐ NO donors used. Furthermore, the actual amount of ⅐ NO released was probably comparable with the levels achieved by the continuous release of ⅐ NO from cytokine-induced type II ⅐ NO synthase (42). Such localized high concentrations of ⅐ NO are readily achieved within the vicinity of cytokine-activated SMC, endothelial cells, or macrophages in atherosclerotic lesions.
Atherosclerotic plaques contain a variety of cell types including SMC, macrophages, and lymphocytes (1,2,20). Immunohistochemical analyses of cellular subtypes in plaques have revealed that most of the lymphocytes are T-cells (19,20). IFN-␥, a major product of activated T-cells, exerts a variety of paracrine effects on neighboring cells and, thus, may modulate the evolution of atherosclerotic lesions. For example, IFN-␥ can inhibit collagen production by SMC (43), augment the expression of major histocompatability complex class I, and induce the expression of major histocompatability complex class II antigens on endothelial cells and SMC (24,44), and in combination with other proinflammatory cytokines, it can induce apoptotic death of SMC (45,46). Consequently, SMC within human and experimental atheroma can express increased levels of ICAM-1 and VCAM-1, indicating a state of activation compared with those in normal vessels (47). However, the expression of these adhesion molecules on SMC in atheroma is quite heterogeneous (48). This may be attributed to the locally produced effects of cytokines and endogenously released ⅐ NO or to a heterogeneous population of intimal SMC that responds differently to external stimuli. In any case, factors such as cytokines, ⅐ NO, and antioxidants that can regulate the expression of ICAM-1 and VCAM-1 may modulate the course of atherogenesis.
IFN-␥ activates at least two transcription factors, ISGF-3 and IRF-1, which are capable of binding to the ISRE (25,27). ISGF-3 is a multicomplex DNA binding protein that contains the Janus kinase substrates, STATs (p91/84, p113) (25). Upon phosphorylation, ISGF-3 translocates into the nucleus, where it can bind to the ISRE of target genes. However, phosphorylation of p91 or GAF, but not p113, allows GAF to migrate to the nucleus by itself and participate in DNA-binding complexes that recognize a different DNA binding motif, the ␥-activated sequence (26). The IRF-1 gene contains ␥-activated sequence elements in its promoter, and the expression of IRF-1 is induced by activated GAF in response to IFN-␥ or TNF-␣ (27,30). IRF-1 binds to ISRE sites in the promoters of IFN-␣/␤, inducible type II ⅐ NO synthase, and IFN-inducible genes such as VCAM-1 (27,30). The induction and activation of IRF-1 is linked to tumor-suppressive properties and, in some instances, to the induction of apoptosis following DNA damage or in response to serum-depriving conditions (49). In our study, we do not find evidence of ISGF-3 binding to ISRE of the VCAM-1 promoter. However, the induction and binding of IRF-1 to ISRE, although not sufficient by itself, was necessary for the induction of VCAM-1 in response to IFN-␥.
The induction of VCAM-1 expression by IFN-␥ also required the two tandem B motifs in the VCAM-1 promoter constructs, F 0 and F 3 , and ⅐ NO's inhibitory effect on IFN-␥-induced VCAM-1 expression in SMC depends not on inhibition of IRF-1 induction or activity but on inhibition of basal constitutive NF-B activity. These results indicate that basal constitutive NF-B is necessary, but by itself is only modestly sufficient to transactivate the VCAM-1 gene in SMC. A more robust transcriptional induction of the VCAM-1 gene by IFN-␥ is mediated by the synergistic effects of basal constitutive NF-B and IFN-␥-stimulated IRF-1. These results are in agreement with a previous study showing that cooperativity between IRF-1 and NF-B is necessary and sufficient in transactivating the VCAM-1 gene in vascular endothelial cells (30). Consequently, the inability of IFN-␥ to stimulate VCAM-1 expression in endothelial cells may result from the lack of basal constitutive NF-B activity in endothelial cells (18,31). Interestingly, endothelial cells, but not SMC, have basal constitutive ⅐ NO production that may render NF-B inactive under basal conditions. Indeed, treatment with the type III ⅐ NO synthase inhibitor, N -arginine methyl ester, inhibits basal ⅐ NO production in endothelial cells and leads to the activation of NF-B (10,17).
In summary, we have identified an important mechanism by which ⅐ NO inhibits IFN-␥-induced VCAM-1 expression in SMC. Our findings add to the evidence that ⅐ NO may be anti-atherogenic through its inhibitory effects on not only cytokine-stimulated NF-B activation, but also on basal NF-B activity. These results provide new insights into how ⅐ NO may modulate SMC inflammatory activation in a manner highly relevant to the evolution of human atheroma.
FIG. 8. VCAM-1 promoter constructs, F 0 , F 3 , and F 4 , showing putative cis-acting elements. VCAM-1 promoter activity was assessed by CAT assays in human SMC transfected with plasmid vectors containing no promoter (p.CAT), the SV40 promoter (pSV2.CAT), and the indicated VCAM-1 promoter constructs. Cells were then stimulated with IFN-␥ (1000 units/ml) in the absence (control) or presence of GSNO (0.2 mM). The promoter activity for each condition was standardized to ␤-galactosidase activity (relative CAT activity). The asterisk represented a significant change in promoter activity between IFN-␥ alone and IFN-␥ with ⅐ NO (p Ͻ 0.05).