Induction of Endothelial-Leukocyte Interaction by Interferon-γ Requires Coactivation of Nuclear Factor-κB
Abstract—To determine whether nuclear factor (NF)-κB is necessary to confer endothelial cell responsiveness to interferon (INF)-γ in terms of vascular cell adhesion molecule (VCAM)-1 expression and leukocyte adhesion, human endothelial cells were treated with IFN-γ in the presence of low concentrations (LCs) of interleukin (IL)-1α (≤100 pg/mL), which activates NF-κB but does not induce VCAM-1 expression. Although IFN-γ induced major histocompatibility complex class II antigen expression and although a high concentration of IL-1α (10 ng/mL) induced leukocyte adhesion and VCAM-1 expression, neither IFN-γ nor LC IL-1α was able to induce VCAM-1 expression or leukocyte adhesion. However, the combination of IFN-γ and LC IL-1α induced VCAM-1 expression and increased leukocyte adhesion (67% and 49% of high-concentration IL-1α, respectively). Electrophoretic mobility shift assays and immunoblotting of nuclear extracts showed that IFN-γ activated signal transducers and activators of transcription (STAT)-1α and interferon regulatory factor (IRF)-1 but not NF-κB, whereas LC IL-1α activated NF-κB but not STAT-1α or IRF-1. Nuclear run-on studies showed that LC IL-1α is necessary but not sufficient for inducing VCAM-1 gene transcription and that the combination of IFN-γ and LC IL-1α is required for full VCAM-1 gene transcription. These findings suggest that factors that activate NF-κB can synergize with IFN-γ in promoting endothelial-leukocyte interaction.
- Received February 4, 2000.
- Accepted July 31, 2000.
The adhesion of leukocytes to vascular endothelium importantly contributes to many inflammatory and immunologic disorders, including early atherogenesis, in which monocyte but not granulocyte accumulation characterizes the formation of fatty streaks.1 Multiple protein families, including chemoattractants and cellular adhesion molecules, provide “traffic signals” for monocyte attachment to the vascular wall. These mediators include molecules overexpressed in atherosclerotic lesions, such as vascular cell adhesion molecule (VCAM)-1. The relatively monocyte-selective “inflammatory” state of the vascular endothelium during early atherogenesis is possibly explained by the fact that the VCAM-1 ligand, the integrin very late antigen-4, is expressed on monocytes but not on neutrophils.2
Cytokines such as interleukin (IL)-1α and tumor necrosis factor (TNF)-α, induce endothelial VCAM-1 expression and the adhesion of monocytes to the endothelial surface.3 The ability of these cytokines to promote VCAM-1 gene expression relates, in part, to their ability to activate the transcription factor nuclear factor (NF)-κB.4 However, recent analyses of the VCAM-1 promoter suggest that NF-κB activation alone may not suffice to induce VCAM-1 expression and that other transcription factors, such as interferon regulatory factor (IRF)-1, are required for the full induction of VCAM-1 gene transcription.5 Indeed, IRF-1 has been found to synergize with NF-κB in transactivating cytokine-inducible genes, such as inducible type II NO synthase6 and VCAM-1.7
A major product of activated T lymphocytes, interferon (IFN)-γ, induces functional changes in the vascular endothelium, including the expression of major histocompatibility complex (MHC) class II antigens.8 9 10 Activated T lymphocytes and MHC-II–positive vascular smooth muscle cells (SMCs) are colocalized within atherosclerotic lesions,11 12 and recent evidence suggests that T lymphocytes contribute to the development and rupture of atherosclerotic plaques.13 Interestingly, IFN-γ, which is an inducer of IRF-1, cannot alone induce endothelial VCAM-1 expression or substantial monocyte adhesion to endothelial cells. In contrast, IFN-γ potently stimulates VCAM-1 expression in SMCs.14 Because SMCs, but not endothelial cells, in serum-containing media display basal NF-κB activity,15 16 we hypothesized that NF-κB activation may permit IFN-γ–induced VCAM-1 expression in SMCs but not endothelial cells. Therefore, the aim of the present study was to determine whether minimal activation of NF-κB, which in itself is insufficient to induce VCAM-1 expression, is required for the induction of VCAM-1 expression by IFN-γ.
Human recombinant IL-1α was obtained from Hoffmann-La Roche. IL-1β, TNF-α, and IFN-γ were purchased from Genzyme. The monoclonal antibodies to VCAM-1 (Ab E1/6), E-selectin (Ab H18/7), ICAM-1 (Ab HU 5/3), and MHC class I (Ab W6/32) were kindly provided by M.A. Gimbrone (Brigham & Women’s Hospital, Boston, Mass). The monoclonal antibody to MHC class II (HLA-DRα, Ab I-2/IA antigen) was kindly provided by Arnold Freedman (Dana Farber Cancer Institute, Boston, Mass). Unless otherwise specified, all other reagents were obtained from Sigma Chemical Co.
Human saphenous vein endothelial cells were harvested enzymatically with type II collagenase at 0.1% as described17 18 and were maintained in medium 199 (GIBCO-BRL) containing HEPES (25 mmol/L), heparin (50 U/mL), endothelial cell growth factor (50 μg/mL), l-glutamine (5 mmol/L), antibiotics, and 5% FCS (Hyclone). Once they were grown to confluence, the cells were replated on low pyrogen fibronectin (1.5 μg/cm2) at 20 000 cells/cm2. Human saphenous vein endothelial cells isolated by these techniques form a confluent monolayer of polygonal cells and express von Willebrand factor as determined by their content of specific mRNA and immunoreactive protein.17 Cell number was assessed after trypsinization in a Neubauer hemocytometer (VWR Scientifics). Cellular viability was assessed by trypan blue exclusion. Bovine aortic endothelial cells (106 per 100-mm2 culture dish, grown in DME medium with 5% heat-inactivated FCS) were used within passage 3 for transfection experiments. Cell number was assessed by direct cell counting of adherent cells, after trypsin detachment, in a Neubauer hemocytometer and staining by trypan blue. The percentage of cells excluding trypan blue was taken as a measure of cell viability.
Cell-Surface Enzyme Immunoassay
The expression of cell-surface adhesion molecules was determined by enzyme immunoassays. The assays were performed by incubating endothelial monolayers with specific monoclonal antibodies (1:10 to 1:1000 dilutions), followed by the addition of biotinylated goat anti-mouse IgG (Vector Labs, Inc), and then with streptavidin–alkaline phosphatase (Zymed). Endothelial monolayers were washed 3 times with PBS between each incubation step, and the integrity of the monolayers was monitored by phase-contrast microscopy. The surface expression of each protein was determined spectrophotometrically at an absorbance of 450 nm after the addition of the chromogenic substrate 3,3′,5,5′-tetramethylbenzidine.
Leukocyte Adhesion Assay
Monocytoid U937 cells were obtained from American Tissue Culture Collection and grown in RPMI medium 1640 (GIBCO-BRL) containing 10% FCS. The U937 cells were concentrated by centrifugation to 1×106 cells/mL. For the adhesion assays, endothelial cells were grown to confluence in 6-well tissue culture plates, after which IFN-γ (10 to 1000 U/mL) or IL-1α (0.01 to 10 ng/mL) or both were added for an additional 24 or 48 hours to allow for the induction of VCAM-1 and MHC-II antigen, respectively. For control, some monolayers were pretreated with a monoclonal antibody against VCAM-1 (E1/6). The adhesion assay was performed by adding 1 mL of the concentrated U937 cell suspension to each monolayer under rotating conditions (63 rpm) at 21°C.18 After 10 minutes, nonadhering cells were removed by gentle washing with medium 199, and the monolayers were fixed with 1% paraformaldehyde. The number of adherent cells was determined by counting 6 different fields with use of an ocular grid and a ×20 objective (0.16 mm2 per field). Microscopic fields chosen for counting adherent leukocytes were randomly selected at half-radius distance from the center of the monolayers.
Nuclear Run-On Assay
Nuclei from 3 to 5×107 endothelial cells were prepared 16 hours after stimulation with IL-1α (low and high concentrations) and IFN-γ, either alone or in combination, and in vitro transcription with [α-32P]UTP (800 Ci/mmol) was performed as described.19 Linearized plasmids (1 μg) were immobilized on nylon membranes, hybridized to radiolabeled transcripts (≈5 to 8×107 cpm/mL) at 45°C for 48 hours in hybridization buffer containing 50% formamide, 5× SSC, 2.5× Denhardt’s solution, 25 mmol/L sodium phosphate buffer (pH 6.5), 0.1% SDS, and 250 μg/mL salmon sperm DNA, and washed in 1× SSC/0.1% SDS at 65°C before autoradiography. Densitometric analyses of autoradiographic bands for Northern hybridization were performed with the aid of Image software (National Institutes of Health).20
Electrophoretic Mobility Shift Assay
Endothelial cells were grown to confluence (≈5×105 cells) in 10-cm Petri dishes in serum-containing medium. After stimulation with appropriate cytokines for the indicated time intervals, the cells were scraped and collected into prechilled microfuge tubes. Nuclear and cytosolic extracts were prepared according to Dignam et al,21 with the additional step of washing nuclear pellets in low-salt buffer before high-salt extraction of nuclear proteins to remove any residual cytosolic contaminants. Aliquots were assayed for protein concentration by the BCA method (Pierce). Dithiothreitol was added to a final concentration of 1 mmol/L, and extracts were stored at −80°C.
The oligonucleotide 5′-AGTTGAGGGGACTTTCCCAGGC-3′, corresponding to the tandem κB binding site in the human VCAM-1 promoter, and a mutant oligonucleotide with a G→C substitution in the third nucleotide of the consensus motif were used to assess NF-κB activation. The oligonucleotide 5′-CATGTTATGCATATTCCTGTAAGTG-3′, containing the consensus binding site for signal transducers and activators of transcription (STAT)-1α (p91), and a mutant oligonucleotide with a CCT→GGA substitution were used to assess STAT-1α activation. The oligonucleotide 5′-GGAGTGAAATAGAAAGTCTG-3′, corresponding to the IFN-stimulatory response element (ISRE) site in the human VCAM-1 promoter was used to assess IRF-1 binding. All oligonucleotides were either synthesized or obtained from Santa Cruz Biotechnology. The oligonucleotides were end-labeled by T4 polynucleotide kinase (New England Biolabs) and [γ-32P]ATP (3000 Ci/mmol) and purified by Sephadex G-50 columns (Pharmacia).
Nuclear extract (10 μg) was added to 32P-labeled oligonucleotides (≈20 000 cpm) in a buffer containing 2 μg poly[dI·dC] (Boehringer-Mannheim), 10 μg BSA, 10 mmol/L Tris-HCl (pH 7.5), 50 mmol/L NaCl, 1 mmol/L dithiothreitol, 1 mmol/L EDTA, and 5% glycerol. After a 30-minute incubation, DNA-protein complexes were resolved on 4% nondenaturing polyacrylamide gels electrophoresed at 12 V/cm in 0.5× Tris-borate EDTA buffer. Specificity was determined by the addition of RelA (p65), p50, or STAT-1α antibodies (2 μg IgG per reaction, Santa Cruz Biotechnology) or excess unlabeled (cold) or mutant oligonucleotides (20 ng) to the nuclear extracts for 10 minutes before the addition of radiolabeled probe.
Whole-cell lysates were prepared in 2× SDS lysis buffer (250 mmol/L Tris-HCl [pH 6.8], 20% glycerol, 4% SDS, and 5% 2-mercaptoethanol), as described.22 Equivalent amounts of whole-cell lysates (30 to 40 μg of protein) or nuclear or cytosolic fractions (prepared as described below) were resolved on 10% or 12% SDS-polyacrylamide gels, followed by electrophoretic transfer to polyvinylidene difluoride membranes (Millipore).
Membranes were incubated in PBS containing 0.1% Tween 20 (PBS-T) with 5% nonfat dry milk for 1 hour at 37°C and then incubated for 1 hour with primary antibodies used at 0.4 μg/mL. Membranes were washed with PBS-T and incubated with horseradish peroxidase–conjugated donkey anti-rabbit IgG as a secondary antibody (Jackson Laboratories), which was diluted 1:15 000 in PBS-T and 5% dry milk. Protein bands were visualized with the use of Renaissance chemiluminescence reagents (DuPont NEN).
Multiple comparisons were performed by 1-way ANOVA, and individual differences were tested by the Fisher protected least significance difference test after the demonstration of significant intergroup differences by ANOVA.
Cytokine-Induced Expression of VCAM-1 and MHC-II
There was little or no basal VCAM-1 or MHC-II expression in vascular endothelial cells. After stimulation with IL-1α (10 ng/mL), VCAM-1 expression peaked at 14 hours and remained steady for up to 24 to 28 hours before declining to baseline at 48 hours. Similar VCAM-1 expression was observed with IL-1β (data not shown). The expression of MHC-II antigen after IFN-γ (1000 U/mL) stimulation was much slower, appearing after 24 hours and peaking at 48 hours. Stimulation with low concentrations of IL-1α (<0.2 ng/mL) did not induce endothelial VCAM-1 expression (Figure 1⇓). However, there was a steep concentration-dependent increase in VCAM-1 expression with higher concentrations of IL-1α (1 to 10 ng/mL), with maximal VCAM-1 expression occurring at an IL-1α concentration of 10 ng/mL. However, high concentrations of IL-1α (1 to 10 ng/mL) were unable to induce MHC-II expression. Treatment with increasing concentrations of IFN-γ (1 to 1000 U/mL) alone produced a gradual increase in MHC-II, but not VCAM-1, expression. However, the combination of IFN-γ (1000 U/mL) and a low concentration of IL-1α (0.1 ng/mL) significantly induced VCAM-1 expression (ie, 67% of high concentration of IL-1α). Similar synergistic effects of IFN-γ were obtained when low concentrations (ie, subthreshold for VCAM-1) of TNF-α (<0.2 ng/mL) were used instead of IL-1α.
Leukocyte Adhesion to Endothelial Cells
The effects of IL-1α (low and high concentrations) and IFN-γ, either alone or in combination, on monocytoid cell (U937) adhesion to endothelial cells were determined in a rotational adhesion assay (please see Figure I, which can be accessed online at http://atvb.ahajournals.org). Treatment of endothelial cells with IFN-γ (1000 U/mL) alone had no significant effect on U937 cell adhesion compared with control or unstimulated endothelial cells (140±30 versus 150±40 cells/mm2, P>0.05). Stimulation of endothelial cells with a low concentration of IL-1α (0.1 ng/mL) modestly increased U937 cell adhesion (420±50 cells/mm2), whereas stimulation of endothelial cells with a high concentration of IL-1α (10 ng/mL) substantially increased monocytoid cell adhesion (24 700±3700 cells/mm2). However, stimulation of endothelial cells with IFN-γ (1000 U/mL) and a low concentration of IL-1α (0.1 ng/mL) produced a 23-fold increase in U937 cell adhesion (9840±1020 cells/mm2, P<0.001). The induction of VCAM-1 expression by IFN-γ and a low concentration of IL-1α accounted for 70% of U937 cell attachment, inasmuch as pretreatment of endothelial cells with the blocking anti–VCAM-1 monoclonal antibody E1/6, at saturating concentrations, for 30 minutes before the adhesion assay decreased the adhesion of these cells by 70% (2950±320 cells/mm2, P<0.01).
Endothelial VCAM-1 and MHC-II Gene Transcription
To determine whether the regulation of VCAM-1 expression occurred at the level of gene transcription, we performed nuclear run-on assays with nuclei isolated from endothelial cells stimulated with IFN-γ and IL-1α (Figure 2⇓). Treatment with IFN-γ (1000 U/mL) induced MHC-II, but not VCAM-1, gene transcription, whereas the high concentration of IL-1α (10 ng/mL) induced VCAM-1, but not MHC-II, gene transcription. The low concentration of IL-1α (0.1 ng/mL) mildly induced VCAM-1 gene transcription, but in combination with IFN-γ, it substantially induced VCAM-1 gene transcription (ie, ≈80% of high concentration of IL-1α). Specificity was determined by the lack of hybridization to the prokaryotic vector DNA, pGEM. Equivalent β-tubulin gene transcription confirmed comparable in vitro loading conditions among the different treatment conditions.
Activation of NF-κB and IRF-1
Because IL-1α is known to activate NF-κB and because IFN-γ is known to activate STAT-1α and IRF-1, we performed electrophoretic mobility shift assays (EMSAs) and immunoblotting on nuclear extracts isolated from endothelial cells stimulated with IFN-γ and IL-1α to determine the potential roles of these transactivating factors in VCAM-1 gene transcription. Stimulation with IFN-γ (1000 U/mL) alone did not activate NF-κB (Figure 3⇓). However, the low concentration of IL-1α (0.1 ng/mL) and, to a greater extent, the high concentration of IL-1α (10 ng/mL) induced NF-κB activation. There was no further increase in NF-κB activation when endothelial cells were stimulated with the combination of IFN-γ (1000 U/mL) and a low concentration of IL-1α (0.1 ng/mL) compared with a low concentration of IL-1α alone. Specificity of the NF-κB band was evidenced by the decreased band intensity in the presence of excess unlabeled κB but not mutated κB oligonucleotides.
Stimulation of the type II IFN receptor by IFN-γ leads to the activation of Janus kinases and the subsequent tyrosine phosphorylation of STATs and the transcriptional induction of IRF-1.22 23 By EMSA, stimulation with IFN-γ (1000 U/mL), but not IL-1α (10 ng/mL), leads to the activation of STAT-1α (p91) (Figure 4⇓). Specificity of the band corresponding to STAT-1α decreased in the presence of excess unlabeled γ-activated sequence (GAS) probe but not mutated GAS oligonucleotides.
The IRF-1 gene contains functional GAS cis-acting elements for transcriptional induction. In a time-dependent manner, IFN-γ (1000 U/mL) induced IRF-1 activation after 2 hours (Figure 5⇓). Specificity of the IRF-1 band is evidenced by the obliteration of the IRF-1 band in the presence of excess cold or unlabeled ISRE oligonucleotide. Furthermore, immunoblotting of nuclear extracts with anti–IRF-1 antibody demonstrated that stimulation with IFN-γ (1000 U/mL), but not IL-1α (0.1 ng/mL), induced the expression of IRF-1 (Figure 6⇓). However, the combination of IFN-γ (1000 U/mL) and IL-1α (0.1 ng/mL) did not increase IRF-1 activation compared with that of IFN-γ alone. These findings indicate that IL-1α (0.1 ng/mL) is unable to activate STAT-1α or IRF-1.
Our findings indicate that IFN-γ can induce endothelial VCAM-1 expression in conjunction with “subthreshold” concentrations of IL-1α. The increase in VCAM-1 expression is due mainly to the increase in VCAM-1 gene transcription. Our results also indicate that there is no overlapping activation of transcription factors such as NF-κB, STAT-1α, and IRF-1 by these cytokines at the concentrations used, with NF-κB being activated selectively by low concentrations of IL-1α and with STAT-1α and IRF-1 being selectively activated by IFN-γ. Thus, our findings suggest that the mechanism by which IFN-γ and low concentrations of IL-1α induce VCAM-1 expression occurs via the combined action of NF-κB and IRF-1.
Although interaction between IL-1α and IFN-γ is a new finding in the present study, such interactions between TNF-α and IFN-γ have been previously documented.24 25 26 For example, the combination of TNF-α and IFN-γ activates the human class I MHC promoter synergistically, and this induction requires the binding of both NF-κB and IRF-1 to the class I MHC promoter.25 Furthermore, a synergistic interaction between transcription factors NF-κB and IRF-1 has been recently reported for the inducible type II NO synthase expression, whereby NF-κB and IRF-1 cooperate in inducing inducible type II NO synthase gene transcription.6 Our findings confirm this cooperative interaction in the VCAM-1 promoter, particularly at subthreshold concentrations of IL-1α. The potential clinical significance of these results is the demonstration that the lymphokine, IFN-γ, can modulate endothelial-leukocyte interaction, which is relevant in atherosclerosis and vascular inflammation.
The tandem NF-κB sites are necessary for the induction of VCAM-1 gene transcription by TNF-α or IL-1.27 However, recent analysis of the VCAM-1 promoter has identified an additional transcription factor, IRF-1, which is required for full induction of VCAM-1 gene transcription.5 IRF-1 is a transcription factor involved in the IFN-γ signal transduction pathway.23 24 For example, overexpression of IRF-1, although unable to transactivate the VCAM-1 promoter, synergized with overexpressed NF-κB in an ISRE-dependent manner.5 Furthermore, recombinant IRF-1 specifically interacted with the VCAM-1 ISRE, acting to increase the affinity of NF-κB to its cognate binding sites.5 During the preparation of the present article, a similar synergistic interaction has been reported for IFN-γ and TNF.7 However, the present study demonstrates that low concentrations of IL-1 are unable to induce VCAM-1 gene transcription without the added effects of IFN-γ, which alone also cannot induce VCAM-1 expression.
Most of the stimuli that are able to trigger adhesion molecule expression in endothelial cells, with the possible notable exception of IL-4,28 act by increasing the rate of gene transcription.4 The expression of VCAM-1 is dependent upon the activation of NF-κB.29 30 The induction of VCAM-1 by the combined treatment with IFN-γ and IL-1 is caused by an increased rate of VCAM-1 gene transcription, as demonstrated by nuclear run-on experiments. These experiments indicate that neither IFN-γ nor low concentrations of IL-1α could by themselves trigger VCAM-1 gene transcription. However, their combination leads to such an induction. The synergistic activation of VCAM-1 by IFN-γ and TNF/IL-1, albeit lower than that obtained with a maximal concentration of IL-1α or TNF-α, was substantial compared with the null effects of IFN-γ on the one hand and of barely detectable effects of the low IL-1α or TNF-α concentrations on the other. Augmented VCAM-1 protein corresponded to an increased adhesion of leukocytes to endothelial cells, largely attributable (on the basis of experiments with blocking antibodies) to VCAM-1 expression. On the basis of such a magnitude of gene activation, we propose a pathophysiological relevance for these observations in the setting of atherosclerosis and vascular inflammation.
These findings support a pathogenetic role of activated Th1 lymphocytes, the best known source of the production of IFN-γ, within the atherosclerotic plaque. Lymphocytes have been described in early as well as in advanced and especially “unstable” atherosclerotic plaques.11 12 Their production of IFN-γ could be one important pathogenetic link between their presence and monocyte recruitment from the circulation, in conjunction with local production of possibly low concentrations of IL-1α and TNF-α by activated macrophages and foam cells. Furthermore, inasmuch as other known triggers of atherogenesis, such as the advanced glycation end products31 and possibly minimally modified LDLs,32 also appear to act through endothelial NF-κB activation, likely overriding the tonic inhibition of such activation by NO, we anticipate a synergistic effect of IFN-γ on these other stimuli. The recent demonstration (in a murine cardiac transplant model of graft atherosclerosis) that arterial VCAM-1 expression in vascular endothelial and SMCs was markedly attenuated in IFN-γ–null mice supports the hypothesis that IFN-γ–induced factors, such as IRF-1, are required for maximal VCAM-1 gene expression.33 Therefore, the present results imply that factors or conditions such as hypercholesterolemia or diabetes mellitus, which minimally activate NF-κB, would make endothelial cells much more responsive to IFN-γ in terms of VCAM-1 expression.
This work was supported by National Institutes of Health grants HL-52233, HL-48743, and HL-09483, the Italian National Research Council, and the Deutsche Forschungsgemeinshaft. J.K.L. is an Established Investigator of the American Heart Association. We thank Michael A. Gimbrone, Jr, and Arnold Freedman for VCAM-1 and MHC-II antibodies, respectively.
Cybulsky MI, Gimbrone MA Jr. Endothelial-leukocyte adhesion molecules in acute inflammation and atherogenesis. In: Simionescu N, Simionescu M, eds. Endothelial Cell Dysfunctions. New York, NY: Plenum Press; 1992:129–140.
Gimbrone MA Jr. Vascular endothelium in health and disease. In: Haber E, ed. Molecular Cardiovascular Medicine. New York, NY: Scientific American Medicine; 1995:25–39.
Cybulsky MI, Gimbrone MA Jr. Endothelial expression of a mononuclear leukocyte adhesion molecule during atherogenesis. Science. 1991;251:788–791.
Neish A, Read M, Thanos D, Pine R, Maniatis T, Collins T. Endothelial interferon regulatory factor 1 cooperates with NF-κB as a transcriptional activator of vascular cell adhesion molecule-1. Mol Cell Biol. 1995;15:2558–2569.
Lechleitner S, Gille J, Johnson DR, Petzelbauer P. Interferon enhances tumor necrosis factor-induced vascular cell adhesion molecule 1 (CD106) expression in human endothelial cells by an interferon-related factor 1-dependent pathway. J Exp Med. 1998;187:2023–2030.
Pober JS, Gimbrone MA Jr. Expression of Ia-like antigens by human vascular endothelial cells is inducible in vitro: demonstration by monoclonal antibody binding and immunoprecipitation. Proc Natl Acad Sci U S A. 1982;79:6641–6645.
Pober JS, Gimbrone MA Jr, Cotran RS, Reiss CS, Burakoff SJ, Fiers W, Ault KA. Ia expression by vascular endothelium is inducible by activated T cells and by human gamma interferon. J Exp Med.. 1983;157:1339–1353.
Pober J, Gimbrone MA Jr, Lapierre L, Mendrick D, Fiers W, Rothlein R, Springer T. Overlapping patterns of activation of human endothelial cells by interleukin 1, tumor necrosis factor, and immune interferon. J Immunol. 1986;137:1893–1896.
van der Wal AC, Becker AE, van der Loos CM, Das PK. Site of intimal rupture or erosion of thrombosed coronary atherosclerotic plaques is characterized by an inflammatory process irrespective of the dominant plaque morphology. Circulation. 1994;89:36–44.
Davies MJ, Bland JM, Hangartner JRW, Angelini A, Thomas AC. Factors influencing the presence or absence of acute coronary artery thrombi in sudden ischemic death. Eur Heart J. 1989;10:203–208.
Lawrence R, Chang LJ, Siebenlist U, Bressler P, Sonenshein GE. Vascular smooth muscle cell express a constitutive NF-κB-like activity. J Biol Chem. 1994;269:28913–28918.
Bourcier T, Sukhova G, Libby P. The nuclear factor-κB signaling pathway participates in dysregulation of vascular smooth muscle cells in vitro and in human atherosclerosis. J Biol Chem. 1997;272:15817–15824.
De Caterina R, Libby P, Peng H-B, Thannickal V, Rajavashisth TB, Gimbrone MA Jr, Shin WS, Liao JK. Nitric oxide decreases cytokine-induced endothelial activation: nitric oxide selectively reduces endothelial expression of adhesion molecules and proinflammatory cytokines. J Clin Invest. 1995;96:60–68.
Liao JK, Shin WS, Lee WY, Clark SL. Oxidized low-density lipoprotein decreases the expression of endothelial nitric oxide synthase. J Biol Chem. 1995;270:319–324.
Rasband W. NIH Image Program, Version 1.49. Bethesda, Md: National Institutes of Health; 1993.
Dignam JD, Lebovitz RM, Roeder RG. Accurate transcription initiation by RNA polymerase II in a soluble extract from isolated mammalian nuclei. Nucleic Acids Res. 1983;11:1475–1489.
Bourcier T, Dockter M, Hassid A. Synergistic interaction of interleukin-1 beta and growth factors in primary cultures of rat aortic smooth muscle cells. J Cell Biol. 1995;164:644–657.
Shuai K, Schindler C, Prezioso VR, Darnell JE Jr. Activation of transcription by IFN-gamma: tyrosine phosphorylation of 91-kD DNA binding protein. Science. 1992;258:1808–1812.
Johnson DR, Pober JS. HLA class I heavy-chain promoter elements mediating synergy between tumor necrosis factor and interferons. Mol Cell Biol. 1994;14:1322–1332.
Johnson DR, Pober JS. Tumor necrosis factor and immune interferon synergistically increase transcription of HLA class I heavy- and light-chain genes in vascular endothelium. Proc Natl Acad Sci U S A. 1990;87:5183–5187.
Neish AS, Williams AJ, Palmer HJ, Whitley MZ, Collins T. Functional analysis of the human vascular cell adhesion molecule-1 promoter. J Exp Med.1992;176:1583–1593.
Iademarco MF, Barks JL, Dean DC. Regulation of vascular cell adhesion molecule-1 expression by IL-4 and TNF-α in cultured endothelial cells. J Clin Invest. 1995;95:264–271.
Peng HB, Libby P, Liao JK. Induction and stabilization of IκBα by nitric oxide mediates inhibition of NF-κB. J Biol Chem. 1995;270:14214–14219.
Khan BV, Harrison DG, Olbrych MT, Alexander RW, Medford RM. Nitric oxide regulates vascular cell adhesion molecule-1 gene expression and redox-sensitive transcriptional events in human vascular endothelial cells. Proc Natl Acad Sci U S A. 1996;93:9114–9119.
Schmidt AM, Hori O, Chen JX, Li JF, Crandall J, Zhang J, Cao R, Yan SD, Brett J, Stern D. Advanced glycation end products interacting with their endothelial receptor induce expression of vascular cell adhesion molecule-1 (VCAM-1) in cultured human endothelial cells in mice: a potential mechanism for the accelerated vasculopathy of diabetes. J Clin Invest.. 1995;96:1395–1403.
Berliner J, Territo M, Sevanian A, Ramin S, Kim JA, Bamshad B, Esterson M, Fogelman AM. Minimally modified low density lipoprotein stimulates monocyte endothelial interactions. J Clin Invest. 1990;85:1260–1266.