Role of Activating Protein-1 in the Regulation of the Vascular Cell Adhesion Molecule-1 Gene Expression by Tumor Necrosis Factor- a *

Endothelial cell surface expression of VCAM-1 is one of the initial steps in the pathogenesis of atherosclero-sis. The inflammatory response transcription factor nuclear factor (NF)- k B plays an important role in the regulation of VCAM-1 expression by various stimuli in-cluding tumor necrosis factor (TNF)- a . Other transcription factors may modulate this response through interaction with NF- k B factors. Since c-Fos/c-Jun (activating protein-1 (AP-1)) are expressed in vascular endothelium during proinflammatory conditions, we investigated the role of AP-1 proteins in the expression of VCAM-1 by TNF- a in SV40 immortalized human microvascular endothelial cells (HMEC). TNF- a induced expression of both early protooncogenes, c- fos and c- jun . The ability of TNF- a to activate the k B-motif ( k L- k R)-dependent VCAM-1 promoter-chloramphenicol acetyltransferase (CAT) reporter gene lacking a consensus AP-1 element was markedly inhibited by co-transfection of the expression vector encoding c- fos ribozyme, which decreases the level of c- fos by degrading c- fos mRNA, or c- fos or c- jun oligonucleotides. Conversely, co-transfection of c-Fos and c-Jun encoding expression vectors potentiated the p65/NF- k B-mediated transactivation of the VCAM-1 promoter-CAT reporter gene. Furthermore the c-Fos encoding expression vector potentiated by 2-fold the transactivation done by the calcium phosphate co-precipitation technique. Overexpres- sion studies were performed by co-transfection of the expression vectors encoding the p65 subunit of NF- k B and/or c-Fos or c-Jun with the reporter gene p85VCAMCAT (5–10 m g) into HMEC cells as described previously (23). The promoterless plasmid, poLUC (24), was used for adjusting the amount of transfected DNA. The cells were harvested and cell extracts were prepared by three cycles of rapid freeze-thaw in 0.25 M Tris, pH 8.0. Protein content was determined using the Bradford (25) technique. The same amounts of proteins were assayed for chloram- phenicol acetyltransferase (CAT) activity according to standard proto-cols (26). The CAT activity was expressed as percent of chloramphenicol converted to acetyl chloramphenicol. Acetylated and unacetylated forms of chloramphenicol were separated on thin layer chromatogra- phy, and their amounts were quantified by phosphorimaging. Each assay was performed in duplicate or triplicate, and the results reported are representative of at least two separate experiments. Expression Vectors and CAT Reporter Genes— The eucaryotic expres- sion vector, cytomegalovirus-p65, contains the p65 cDNA cloned between a cytomegalovirus promoter, b -globin intron, and simian virus 40 poly(A) signal (27). Gal/p65 contains the DNA binding domain of the yeast transcription vector Gal-4 and the transactivation domain of p65. pGal-CAT is a reporter gene containing four tandem elements of Gal-4 attached to the reporter gene CAT. These vectors generous Dr. C. The reporter gene, p85VCAMCAT, contains coordinates 2 85 to 1 12 of the human VCAM-1 promoter These reporter genes generous gifts The expression vec- tors encoding c-Fos and c-Jun protein were kindly provided fos Nuclear Extract Preparation— Confluent HMEC cells were exposed to TNF- a (100U/ml) for 1–2 h. Nuclear extracts were prepared by a modification of the method of Dignam et al. (29). Briefly, after washing with phosphate buffered saline, cells were centrifuged and the cell pellet suspended in 500 m l buffer A (10 m M HEPES, pH 7.9, 1.5 m M MgCl 2 , 10 m M KCl, and 1.0 m M DTT). After recentrifugation, the cells were resuspended by gentle pipetting up and down in 80 m l of buffer A containing 0.1% Triton X-100. After incubating for 10 min at 4 °C, the homogenate was centrifuged, and the nuclear pellet was washed once with buffer A and resuspended in 70 m l of buffer C (20 m M HEPES, pH 7.9, 25% (v/v) glycerol, 0.42 M NaCl, 1.5 m M MgCl 2 , 0.2 m M EDTA, 1 m M DTT). This suspension was incubated for 30 min at 4 °C followed by centrifugation at 20,000 3 g for 10 min. The resulting supernatant (nuclear extract) was stored at 2 70 °C. Protein concentrations were determined by the Bradford (25) method. To minimize proteolysis, all buffers contained 1.0 m M phenylmethylsulfonyl fluoride, aprotinin (10 mg/ml), leupeptin (10 mg/ml), and antipain (10 mg/ml). Facility of Emory University. These oligonucleotides inhibit the translation of their corresponding mRNAs. Northern Blot Analysis— Total cellular RNA was isolated by a single extraction using an acid guanidium thiocyanate-phenol-chloroform (32). Total cellular RNA (10 m g) was size fractionated using 1% agarose formaldehyde gels in the presence of 1 m g/ml ethidium bromide. The RNA was transferred to a nitrocellulose filter and covalently linked by ultraviolet irradiation using a Stratlinker UV cross-linker (Stratagene, La Jolla, CA). Hybridizations were performed at 68 °C for 1 h in 5 3 SSC (1 3 5 150 m M NaCl, 15 m M sodium citrate), 1% sodium dodecyl sulfate, 5 3 Denhardt’s solution, 50% formamide, 10% dextran sulfate, and 100 m g/ml sheared denatured salmon sperm DNA. Approximately 1–2 3 10 6 cpm/ml of labeled probe (specific activity, . 10 8 cpm/ m g of DNA) was used per hybridization. Following hybridization, filters were washed with a final stringency of 0.2 3 SSC at 65 °C. The nitrocellulose membrane was soaked for stripping the probe with boiled water prior to rehybridization. mRNA bands were quantified using PhosphorImager. 32 P-Labeled DNA probes were prepared using the random primer oligonucleotide kit from Stratagene. The VCAM-1 probe was a Hin dIII- Xho I fragment of the cDNA consisting of nucleotide 132–1814 (33). The ICAM-1 probe was an Eco RI fragment of human cDNA

Several proinflammatory agents (e.g. tumor necrosis factor-␣ (TNF-␣), 1 interleukin-1␤, and lipopolysaccharide) induce VCAM-1 gene expression in endothelial cells through activating the redox-sensitive transcription factor nuclear factor-B (NF-B) (1)(2)(3). Promoter studies suggest that induction of the VCAM-1 gene by NF-B factors requires two NF-B motifs (present at Ϫ63 and Ϫ77) (1,2). Since activated NF-B is present in atherosclerotic lesions (4) and is involved in the regulation of several proinflammatory responsive genes in endothelial cells (5), it is considered to be one of the major vascular inflammatory transcriptional factors.
Activation protein-1 (AP-1, c-Fos/c-Jun) designates another class of transcription factors that mediate inflammatory responses by various cytokines and growth factors in different cell types (6,7). Inhibition of c-fos and/or c-jun by various agents (e.g. nordihydroguatiaretic acid (NDGA), curcumin, and N-acetylcysteine) has been shown to inhibit inflammatory responses in different tissues (6,7). AP-1 proteins are dimers of the Fos (e.g. c-Fos, FosB, and Fra1) and the Jun (e.g. c-Jun, JunD, and JunB) families (8 -15) generated through interaction between leucine zipper motifs (16 -21). Members of the Fos family cannot make stable homodimers and therefore cannot act as transcriptional activators by themselves. AP-1 proteins are regulated by several growth factors, cytokines, and by various agents that induce oxidative stress (6,7). Since several of these agents and the conditions that activate AP-1 proteins are present in atherogenic conditions, it is likely that AP-1 proteins play a role in the regulation of this inflammatory disease of the vasculature.
This study investigated the role of AP-1 proteins in the regulation of the inflammatory response gene VCAM-1. We found that AP-1 proteins mediate some of the effects of TNF-␣ in the regulation of endothelial VCAM-1 expression. At least part of these effects of AP-1 is independent of direct binding of AP-1 to its cognate enhancer element and is likely effected through direct interactions with NF-B factors.

MATERIALS AND METHODS
Cell Culture and Reagents-Human dermal microvascular endothelial cells (HMEC) were immortalized with simian virus 40 product, large T antigen (22). HMEC cells were generously provided by E. W. Ades (Centers for Disease Control and Prevention, Atlanta, GA) and maintained in EBM medium obtained from Clonetics (San Diego, CA) supplemented with 15% fetal bovine serum (Atlanta Biological, Norcross, GA), 10 ng/ml epidermal growth factor (Intergen, Purchase, NY), and 1 g/ml hydrocortisone (Sigma). Human recombinant TNF-␣ was obtained from Boehringer Mannheim. Curcumin and NDGA were purchased from Sigma. All other reagents were of reagent grade.
CAT Assay-One day prior to transfection, HMEC cells were split at the ratio that would give 30 -40% confluence. The transfection was * This work was supported in part by an American Heart Association national grant-in-aid, an American Heart Association Georgia Affiliate grant-in-aid, and an Emory University research grant (to M. A.), and National Institutes of Health Grant PO1-HL48667 (to R. M. M. and R. Wayne Alexander). 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  done by the calcium phosphate co-precipitation technique. Overexpression studies were performed by co-transfection of the expression vectors encoding the p65 subunit of NF-B and/or c-Fos or c-Jun with the reporter gene p85VCAMCAT (5-10 g) into HMEC cells as described previously (23). The promoterless plasmid, poLUC (24), was used for adjusting the amount of transfected DNA. The cells were harvested and cell extracts were prepared by three cycles of rapid freeze-thaw in 0.25 M Tris, pH 8.0. Protein content was determined using the Bradford (25) technique. The same amounts of proteins were assayed for chloramphenicol acetyltransferase (CAT) activity according to standard protocols (26). The CAT activity was expressed as percent of chloramphenicol converted to acetyl chloramphenicol. Acetylated and unacetylated forms of chloramphenicol were separated on thin layer chromatography, and their amounts were quantified by phosphorimaging. Each assay was performed in duplicate or triplicate, and the results reported are representative of at least two separate experiments.
Expression Vectors and CAT Reporter Genes-The eucaryotic expression vector, cytomegalovirus-p65, contains the p65 cDNA cloned between a cytomegalovirus promoter, ␤-globin intron, and simian virus 40 poly(A) signal (27). Nuclear Extract Preparation-Confluent HMEC cells were exposed to TNF-␣ (100U/ml) for 1-2 h. Nuclear extracts were prepared by a modification of the method of Dignam et al. (29). Briefly, after washing with phosphate buffered saline, cells were centrifuged and the cell pellet suspended in 500 l buffer A (10 mM HEPES, pH 7.9, 1.5 mM MgCl 2 , 10 mM KCl, and 1.0 mM DTT). After recentrifugation, the cells were resuspended by gentle pipetting up and down in 80 l of buffer A containing 0.1% Triton X-100. After incubating for 10 min at 4°C, the homogenate was centrifuged, and the nuclear pellet was washed once with buffer A and resuspended in 70 l of buffer C (20 mM HEPES, pH 7.9, 25% (v/v) glycerol, 0.42 M NaCl, 1.5 mM MgCl 2 , 0.2 mM EDTA, 1 mM DTT). This suspension was incubated for 30 min at 4°C followed by centrifugation at 20,000 ϫ g for 10 min. The resulting supernatant (nuclear extract) was stored at Ϫ70°C. Protein concentrations were determined by the Bradford (25) method. To minimize proteolysis, all buffers contained 1.0 mM phenylmethylsulfonyl fluoride, aprotinin (10 mg/ml), leupeptin (10 mg/ml), and antipain (10 mg/ml).
Gel Shift Assays-The oligonucleotide containing L-R of the VCAM-1 promoter (VCAM-1 wild type oligo) was synthesized. Its sequence is as follows: 5Ј-CTGCCCTGGGTTTCCCCTTGAAGGGATTTCC-CTCCGCCTCTGCAACAAGCTCGAGATCCTATG-3Ј. The sequences of L and R are double underlined. The single underlined sequences represent an unrelated tail sequence added to serve as a template for synthesis of the double-stranded DNA. To prepare double-stranded DNA, first an oligonucleotide 5Ј-CATAGGATCTCGAGC-3Ј (complementary to the 3Ј unrelated tail, single underlined sequence) was annealed to VCAM-1 wild type oligo. The second strand was extended with DNA polymerase (Klenow fragment) in a reaction mixture containing 50 Ci of [ 32 P]dCTP and 0.5 mM of cold dATP, dGTP, and dTTP. The reaction was followed by the addition of 0.5 mM cold dCTP to ensure completion of the second strand. Unincorporated nucleotides were removed by column chromatography over a Sephadex G-50 column. The DNA binding reaction was performed at 30°C for 15 min in a volume of 20 ml, which contained 225 g/ml bovine serum albumin, 1.0 ϫ 10 5 cpm 32 P-labeled probe, 0.1 mg/ml poly(dI-dC), and 15 ml of binding buffer (12 mM HEPES, pH 7.9, 4 mM Tris, 60 mM KCl, 1 mM EDTA, 12% glycerol, 1 mM DTT, and 1 mM phenylmethylsulfonyl fluoride). After the binding reaction, the samples were subjected to electrophoresis in 1 ϫ Trisglycine buffer using 4% native polyacrylamide gels.
Northern Blot Analysis-Total cellular RNA was isolated by a single extraction using an acid guanidium thiocyanate-phenol-chloroform (32). Total cellular RNA (10 g) was size fractionated using 1% agarose formaldehyde gels in the presence of 1 g/ml ethidium bromide. The RNA was transferred to a nitrocellulose filter and covalently linked by ultraviolet irradiation using a Stratlinker UV cross-linker (Stratagene, La Jolla, CA). Hybridizations were performed at 68°C for 1 h in 5 ϫ SSC (1 ϫ ϭ 150 mM NaCl, 15 mM sodium citrate), 1% sodium dodecyl sulfate, 5 ϫ Denhardt's solution, 50% formamide, 10% dextran sulfate, and 100 g/ml sheared denatured salmon sperm DNA. Approximately 1-2 ϫ 10 6 cpm/ml of labeled probe (specific activity, Ͼ10 8 cpm/g of DNA) was used per hybridization. Following hybridization, filters were washed with a final stringency of 0.2 ϫ SSC at 65°C. The nitrocellulose membrane was soaked for stripping the probe with boiled water prior to rehybridization. mRNA bands were quantified using PhosphorImager. 32 P-Labeled DNA probes were prepared using the random primer oligonucleotide kit from Stratagene. The VCAM-1 probe was a HindIII-XhoI fragment of the cDNA consisting of nucleotide 132-1814 (33). The ICAM-1 probe was an EcoRI fragment of human cDNA (34).

TNF-␣ Induces c-fos and c-jun in HMEC Cells-Because
TNF-␣ is known to induce c-fos and c-jun in various cell types (7,35,36), we tested whether similar responses occur in HMEC. Confluent HMEC cells were treated with TNF-␣ (100 units/ml) for 15-120 min and RNA was prepared for analysis of c-fos and c-jun expression. As shown in Fig. 1, TNF-␣ treatment produced a severalfold induction of c-fos and c-jun mRNA. c-fos mRNA was elevated 15 min after TNF-␣ treatment, peaked at nearly 30 min, and return to basal level at 2 h, kinetics similar to c-fos induction by serum and growth factors (37)(38)(39). The kinetics of induction of c-jun by TNF-␣ was similar to that observed for c-fos.
Inhibition of c-fos or c-jun blocks TNF-␣ Activation of the VCAM-1 Promoter-TNF-␣ induces p85VCAMCAT (a minimal VCAM-1 promoter-coordinates Ϫ85 to ϩ12, containing two NF-B motifs, L-R, but lacking a consensus AP-1 binding site, attached to the reporter gene CAT) through activating NF-B proteins (1)(2)(3)23). We investigated whether endogenous c-Fos contributes to the TNF-␣-mediated transactivation of p85VCAMCAT. The endogenous level of c-fos was inhibited by co-transfection of an expression vector encoding anti c-fos ribozyme that specifically degrades c-fos mRNA (28). As expected, TNF-␣ (100 units/ml) activated p85VCAMCAT severalfold above the basal levels (Fig. 2, lanes 3 and 4) in transient transfection assays in HMEC cells compared with untreated cells (lanes 1 and 2). In cells co-transfected with the expression vector encoding c-fos ribozyme, TNF-␣ was unable to transactivate p85VCAMCAT (lanes 5 and 6). Under similar conditions the expression vector expressing c-fos ribozyme had no effect on the constitutive activity of the SV40 promoter-driven reporter gene CAT, pSV2CAT (compare lanes 7 and 8 with lanes 9 and 10). Under similar conditions antisense oligonucleotides c-fos or c-jun (30, 31) inhibited the ability of TNF-␣ to activate p85VCAMCAT nearly 80% (data not shown). These results suggest that endogenous c-Fos and c-Jun contribute to TNF-␣-

FIG. 1. Effect of TNF-␣ on the regulation of c-fos and c-jun gene expression in HMEC cells.
A, confluent HMEC cells were treated with 100 units/ml TNF-␣ for 15-120 min. Total RNA was prepared and analyzed for expression of c-fos and c-jun mRNAs as described under "Materials and Methods." Shown are autoradiograms of sequential probings of same blot with cDNA probes specific for c-fos and c-jun. To show equal loading, the blots were stained with ethidium bromide. mediated transactivation of p85VCAMCAT.
Co-expression of the Expression Vectors Encoding c-Fos or c-Jun Potentiates the p65-mediated Transactivation of p85VCAMCAT-p65 is a member of the rel family and a component of the NF-B heterodimers. p65 potently transactivates p85VCAMCAT (23,40). We investigated whether overexpression of c-Fos and c-Jun potentiates the p65-mediated transactivation of p85VCAMCAT. Eucaryotic expression vectors encoding c-Fos or c-Jun and/or p65 were co-transfected with p85VCAMCAT into HMEC cells and the promoter activity assessed. Consistent with previous observations, co-expressed p65 transactivated p85VCAMCAT (Fig. 3) (23, 40). Co-expression of the expression vectors encoding c-Fos and c-Jun dosedependently potentiated the p65-mediated transactivation of p85VCAMCAT with maximum potentiation above 30-fold at 5 g of the each expression vector (Fig. 3A). When co-transfected individually, both c-fos and c-jun expression vectors also potentiated the p65-mediated transactivation of p85VCAMCAT (Fig.  3B). These results suggest that both c-Fos and c-Jun functionally interact with the p65 subunit of NF-B to potentiate p65mediated transactivation of p85VCAMCAT.
Co-expression of the Expression Vector Encoding c-Fos Potentiated the Gal/p65-dependent Transactivation of pGal-CAT-p65 has a DNA binding domain and two transactivation domains (41). The transactivation domains of p65 require co-factor proteins for transactivation activity (42). We investigated whether c-Fos could interact with the transactivation domains of p65. We utilized an expression vector encoding chimeric transcriptional factor Gal/p65 and a Gal/p65 driven reporter gene pGal-CAT. Gal/p65 contains a DNA binding domain of the yeast transcriptional factor Gal-4 and transactivation domains of p65 subunit of NF-B. pGal-CAT contains four multimerized DNA elements of the transcriptional factor Gal-4 attached to a reporter gene, CAT. Co-transfected Gal/p65 potently transactivated pGal-CAT in transient transfection assays (42) (Fig. 4).  (Fig. 5, lanes 2 and 6) compared with the control (lanes 1 and 5). Preincubation of nuclear extracts with antibodies to either c-Fos or c-Jun inhibited the binding of NF-B-like proteins to the L-R (lanes 3 and 7). To test the specificity of this interaction, the nuclear extracts were incubated with antibodies to either c-Fos or c-Jun in the presence of the peptides to which the antibodies were raised. In the presence of the peptide to which the antibodies c-Fos were raised, the antibody c-Fos was unable to inhibit the binding of NF- endogenous VCAM-1 expression, we investigated the effect of curcumin, an inhibitor of c-jun/c-fos expression (43). We investigated the dose-dependent effect of curcumin on the expression of the VCAM-1 gene by TNF-␣. Pretreatment of HMEC cells with 10 M curcumin inhibited TNF-␣-activation of VCAM-1 but not ICAM-1 expression at the mRNA levels ( Fig. 6A). At the same concentration, curcumin did not inhibit the TNF-␣-activation of NF-B (Fig. 6B). These results suggest that curcumin at 10 M inhibits the TNF-␣ activation of VCAM-1 through a mechanism independent of NF-B activation.
NDGA Inhibits the Expression of VCAM-1 by TNF-␣ without Inhibiting the Activation of NF-B-NDGA inhibits several of the effects of TNF-␣ by inhibiting c-fos expression (7). We investigated the effect of NDGA on the TNF-␣-activation of VCAM-1 expression in HMEC cells. Pretreatment of the cells with NDGA at all concentrations tested (1-10 M) inhibited the TNF-␣-activation of VCAM-1 expression (Fig. 7A). Pretreatment of the cells with NDGA at the same concentrations did not inhibit the TNF-␣ activation of NF-B (Fig. 7B). These results suggest that activation of the lipoxygenase pathway and c-Fos activation may mediate, at least in part, the effect of TNF-␣ in regulating the expression of VCAM-1 in HMEC cells. DISCUSSION We demonstrate that AP-1 proteins c-Fos/c-Jun are involved in TNF-␣-activation of VCAM-1 expression. At least a part of the effect of c-Fos/c-Jun is mediated through interaction with NF-B, and is independent of the AP-1 element present on the VCAM-1 promoter. Furthermore, we demonstrate that antiinflammatory agents and inhibitors of c-Fos/cJun expression inhibit the TNF-␣-activation of VCAM-1 in HMEC cells through a pathway that is independent of NF-B activation (Fig. 8). These studies suggest that in addition to NF-B, c-Fos/ c-Jun also mediate the effect of TNF-␣ in the regulation of VCAM-1 expression.
Inhibition of endogenous c-Fos/c-Jun by the expression vector encoding ribozyme c-fos (28) or by phosphorothioate antisense oligonucleotides c-fos or c-jun (30,31) inhibited the TNFactivation of 85VCAMCAT, a minimal NF-B-driven VCAM-1 promoter-reporter gene lacking the AP-1 element. These studies suggest that the effect of c-fos and c-jun is mediated, at least in part, through interaction with NF-B factors. While it is tempting to speculate that antisense c-fos and c-jun, and ri- Co-transfection of the expression vector encoding the p65 subunit of NF-B completely inhibited the Gal/p65-driven activity of pGal-CAT in HMEC cells as observed earlier in COS cells (42). These studies suggested that certain cofactors such as bridging proteins, mediators, or intermediary proteins that associate or interact with the transactivation domain of p65 are necessary for transactivation by p65. We investigated whether c-Fos or c-Jun could act as a co-factor in the transactivation by p65. Consistent with the 85VCAMCAT studies, co-transfection of the expression vector encoding c-fos potentiated the Gal-p65 driven transactivation of pGal-CAT. Under similar conditions, co-transfection of the expression vector encoding c-Jun did not potentiate the transactivation activity of Gal/p65 suggesting that the interaction of c-Jun with p65 may require its DNA binding domain. These studies at least suggest that c-Fos could act as a cofactor through interaction with the transactivation domain of p65 in HMEC cells.
AP-1 proteins can regulate the expression of several genes in various ways. 1) Through binding to the AP-1 element, AP-1 proteins can activate the transcription of various genes containing the AP-1 element; 2) after binding to their DNA element, AP-1 proteins can synergize with other transcriptional factors and modulate the transactivation mediated by them; and 3) independent of their DNA binding element, AP-1 proteins can interact with other transcriptional factors and modulate the transactivation mediated by them (44 -46). Since deletion or mutation of the AP-1 element present in the VCAM-1 promoter had no effect on activation of the VCAM-1 promoter by TNF-␣ (1, 2), the AP-1 element does not appear to play a role in TNF-␣-activation of the VCAM-1 promoter. This would suggest that the regulation by AP-1 proteins of VCAM-1 is mainly through interaction with NF-B factors and is independent of the AP-1 element.
Curcumin inhibits transcriptional activation of the c-jun gene by PMA and TNF-␣ and blocks the increase in TRE binding activity of c-Jun/AP-1 protein (43,47). Curcumin decreases TPA-induced nuclear abundance of c-Fos protein through the induction of a hyperphosphorylated unstable form of c-Fos (48). Curcumin has also been shown to inhibit the activation of NF-B (49). In our experimental conditions, curcumin (10 M) blocked the TNF-␣-activation of VCAM-1 expression in HMEC cells without inhibiting NF-B activation. These studies suggest that transcriptional factors other than NF-B, such as c-Fos/c-Jun/AP-1, may also play an important role in the regulation of VCAM-1 gene expression by TNF-␣.
NDGA, an inhibitor of fos expression (7), inhibits the TNF-␣ activation of VCAM-1 expression without blocking NF-B activation. These results suggest that NDGA inhibited the transactivation activity of NF-B either by directly changing the phosphorylation of nuclear NF-B or by inhibiting the activity of other factors that interact with NF-B. In recent studies, the phosphorylation of nuclear NF-B was analyzed under conditions that inhibited TNF-␣ activation of VCAM-1 expression but did not inhibit nuclear translocation of NF-B (50). These studies suggest that factors other than phosphorylation of NF-B are involved in inhibiting the transactivation activity of NF-B. Consistent with these observation, our studies suggest that AP-1 proteins may play an important role in regulating the activity of NF-B.
The antibodies to c-Fos and c-Jun inhibited the binding of NF-B to the L-R motif, as observed earlier with Ig/HIV B motif (44). Assuming that c-Fos and c-Jun are part of the L-R-bound NF-B complex, the antibodies to c-Fos or c-Jun could inhibit the binding of NF-B to L-R either by changing the conformation of the DNA-binding domain of NF-B or directly interfering with the binding of NF-B to L-R. The other possibility is that c-Fos and c-Jun are not part of the NF-B complex but that their presence in the nucleus facilitates the binding of NF-B to the NF-B motif. These results suggest that the TNF-␣-activated signal transduction pathways leading to activation of NF-B and AP-1 may interact in the nucleus to regulate VCAM-1 gene expression.
In conclusion, our studies have demonstrated that TNF-␣ induces 85VCAMCAT (a minimal VCAM-1 promoter-CAT reporter gene) and expression of VCAM-1 gene through interaction of c-Fos and c-Jun with NF-B proteins, suggesting that the AP-1/NF-B protein complex may be one of the transactivating complexes induced by TNF-␣ in HMEC cells. Furthermore, c-Fos/c-Jun may play a role in the regulation of VCAM-1 gene expression under conditions such as growth state of the cell and shear stress that affect the levels of c-Fos/c-Jun in endothelial cells. FIG. 8. Factors affecting the nuclear activity of NF-B. Agonists, such as TNF, activate nuclear translocation of NF-B. In the nucleus several other factors may regulate the transactivation activity of NF-B. For example, AP-1 factors can interact with NF-B to modulate its transactivation activity. Independent of NF-B activation signaling pathway, TNF appears to also activate other signaling pathway(s) that regulate the nuclear activity of NF-B. These studies suggest that NDGA and curcumin can inhibit the TNF-activation of VCAM-1 expression through inhibiting the pathways that regulate the nuclear activity of NF-B.