Induction of Tissue Factor Expression in Human Endothelial Cells by CD40 Ligand Is Mediated via Activator Protein 1, Nuclear Factor B, and Egr-1*

Induction of tissue factor expression in endothelial cells via ligation of CD40 probably figures prominently in the pathogenesis of prevalent inflammatory diseases, including atherosclerosis. However, the molecular mechanisms of tissue factor gene expression triggered by CD40 ligand (CD40L) in this cell type remain unknown. We demonstrate here that the tissue factor promoter region 278 bp to 121 bp contains the CD40L-responsive elements, consisting of activator protein 1 (AP-1) , nuclear factor (NF) B-, and Egr-1binding sites. Mutations of either the AP-1or NFBbinding sites markedly reduced the CD40L-dependent promoter activation. The AP-1 and NFB sites displayed constitutive and CD40L-enhanceable DNA binding activity, respectively. Of note, mutation of the Egr-1-binding sites, previously not associated with CD40 signaling, impaired activation of the tissue factor promoter. Accordingly, CD40L strongly induced Egr-1 protein expression and DNA binding activity to all three bindings sites. In contrast to CD40L, other established inducers of tissue factor in endothelial cells, interleukin-1 or tumor necrosis factor , did not increase the expression of Egr-1. In conclusion, induction of tissue factor gene expression in human endothelial cells by CD40L involves AP-1 and NFB as well as Egr-1, a pathway previously not implicated in CD40 signaling and distinct from that employed by certain other proinflammatory cytokines.

Thrombosis, resulting in complete or partial vessel occlusion, plays a crucial role in the pathogenesis of cardiovascular diseases such as atherosclerosis and thromboembolism (1,2). Intravascular coagulation also complicates conditions such as septic shock. For well over a century pathologists have recognized that altered endothelial coagulant potential contributes to thrombotic diathesis. Understanding the molecular details of endothelial coagulability has emerged recently. Expression of procoagulants on the surface of endothelial cells (EC), 1 the cell type at the blood interface, has attracted particular attention in this regard. Tissue factor (TF), a 47-kDa membrane-bound glycoprotein, serves as a key initiator of blood coagulation. Of note, EC express little or no TF constitutively in vitro and in vivo. However, levels of this procoagulant increase markedly in the endothelium of the diseased vasculature (3).
Despite its potential pathophysiologic role, knowledge regarding the regulation of TF gene activity in EC remains incomplete. Endotoxin (lipopolysaccharide (LPS)), tumor necrosis factor ␣ (TNF␣), interleukin 1␤ (IL-1␤), vascular endothelial growth factor, and phorbol 12-myristate 13-acetate (PMA) induce the expression of this procoagulant in EC (4,5). Interestingly, the transcriptional regulation of TF expression in different cell types appears to involve distinct signaling pathways (6). Functional studies of the human TF promoter in various cell types, including EC, identified two promoter regions implicated in TF gene expression. The first region, termed the LPSresponse element (LRE), contains two activator protein-1 (AP-1)-and one nuclear factor B (NF-B)-like binding sites. The LRE mediates inducibility of the TF promoter in response to proinflammatory stimuli (e.g. IL-1␤ and TNF␣), LPS, and PMA (4,7,8). The second region, the serum-response region (SRR), contains three Egr-1-binding sites. This region mediates the induction of TF gene expression via Egr-1 in response to serum, PMA, shear stress, low density lipoprotein, hypoxia, vascular endothelial growth factor, and LPS (5, 9 -14). Thus, the molecular pathways employed to activate the TF gene appear to depend on the stimulus applied, e.g. proinflammatory cytokines appear not to signal via the Egr-1-binding sites located in the SRR.
Recently, we and others have identified the immunomodulatory dyad CD40/CD40L as a novel regulator of TF protein and activity in various cell types, including EC, in vitro and in vivo (15)(16)(17)(18)(19). Further supporting the potential role of CD40L in the modulation of the thrombotic balance of EC, CD40 ligation diminishes the expression of thrombomodulin by this cell type, the "anticoagulant" receptor for thrombin (16,17). Moreover, in human and experimental atheroma, the expression of CD40L correlates and co-localizes with that of TF in vivo (19,20).
Although many cell types can express TF, EC occupy a central position in the homeostasis of blood coagulation as they provide the contact surface of the tissue with coagulation factors containing blood. The CD40 signaling mechanisms mediating TF expression in this cell type, however, remain unknown. Therefore, the present study characterized the cisacting regulatory elements, which mediate the CD40L-induced TF promoter activity in EC.

EXPERIMENTAL PROCEDURES
Reagents-Human recombinant IL-1␤ and TNF␣ were obtained from Endogen (Cambridge, MA); PMA and polymyxin B were purchased from Sigma. Human recombinant CD40L was obtained from Leinco Technologies (St. Louis, MO).
Cell Isolation and Culture-Human vascular EC were isolated from saphenous veins and cultured in dishes coated with gelatin as described elsewhere (21). Cells were maintained in growth medium containing medium 199 (M199; BioWhittaker, Walkersville, MD), supplemented with 1% penicillin/streptomycin (BioWhittaker), 5% fetal bovine serum (Atlanta Biologicals, Norcross, GA), 100 g/ml heparin (Sigma), and 50 g/ml endothelial cell growth factor (Pel-Freez Biologicals, Rogers, AK), and used throughout passages 2-4. Culture media and fetal bovine serum contained less than 40 pg of endotoxin/ml as determined by chromogenic limulus amoebocyte assay analysis (QLC-1000; BioWhittaker). Purity of EC cultures was Ն99% as characterized by immunostaining with anti-von Willebrand factor monoclonal antibody (Dako, Carpinteria, CA) or by flow cytometry with anti-CD31 monoclonal antibody (BD PharMingen). EC were cultured 16 h before the experiments in M199 supplemented with 0.1% human serum albumin (Immuno-US, Rochester, MI). All experiments employing recombinant CD40L were performed in the presence of 1 g/ml polymyxin B.
Flow Cytometry-Human vascular EC were harvested by trypsinization, fixed (PBS, 4% paraformaldehyde, 15 min), washed once with ice-cold PBS, 0.2% bovine serum albumin (Ca 2ϩ /Mg 2ϩ -free), and subsequently incubated (30 min, 4°C) with fluorescein isothiocyanate-conjugated IgG (Ancell, Bayport, MN) or TF antibody (American Diagnostica). After immunofluorescence labeling, cells were washed twice with PBS, 0.2% bovine serum albumin and analyzed in a BD PharMingen FACScan TM flow cytometer. Data were analyzed using CellQuest TM software (BD PharMingen). At least 20,000 viable cells per condition were analyzed for the determination of percent positive cells.
Transfections-Human vascular EC were subcultured the day before transfection (1:2) into 150-cm 2 flasks. For transfection, cells were harvested by trypsinization, washed once in electroporation medium (Opti-MEM I, Invitrogen; supplemented with 1% fetal bovine serum), and 3 ϫ 10 5 cells/well (6-well plate) were incubated (30 min, 4°C) with lucifer- ase plasmid and pSV-␤-Gal (5 g each) in electroporation medium. Electroporation was performed in electroporation cuvettes with a gap of 0.4 cm in a Gene-Pulser (Bio-Rad) at a time constant of 32 ms. After electroporation, cells were incubated for 30 min at room temperature, resuspended in growth medium, and equally distributed to 6-well plates. Following a 20-h recovery, EC were cultured (16 h) in serum-free medium as described above. Following stimulation with 10 g/ml recombinant CD40L (6 h), cell lysates were assayed for luciferase (BD PharMingen) and ␤-galactosidase activity (Tropix, Galacto-Light Plus, Bedford, MA) in a luminometer (Berthold, Bad Wildbad, Germany). Luciferase activity of each sample was normalized to ␤-galactosidase activity to correct for transfection efficiency (approximately 10 -15%). Values were determined in triplicate and are presented as mean Ϯ S.D.
Preparation of Nuclear Extracts-Confluent EC monolayers were washed twice with ice-cold PBS, harvested by scraping into 1 ml of ice-cold PBS, and pelleted in a 1.5-ml microcentrifuge tube (300 ϫ g, 5 min). The cells were lysed in buffer A (10 mM HEPES, pH 7.9, 10 mM KCl, 1.5 mM MgCl 2 , 0.5% Nonidet P-40, 1 mM DTT, 1 mM PMSF, and 1 g/ml aprotinin, leupeptin, and pepstatin A) on ice for 10 min. Nuclei were pelleted by centrifugation (3,500 ϫ g, 5 min). The supernatant was collected, precleared (14,000 ϫ g, 10 min), and stored as the cytosolic fraction at Ϫ80°C until use. Nuclei were washed once with buffer A before extraction (30 min, 4°C) of nuclear proteins with buffer B (20 mM HEPES, pH 7.9, 1.5 mM MgCl 2 , 420 mM NaCl, 0.2 mM EDTA, 1 mM DTT, 1 mM PMSF, 1 g/ml aprotinin, leupeptin, and pepstatin A). The nuclear extracts were collected after centrifugation (14,000 ϫ g, 10 min) and combined with an equal volume of buffer C (20 mM HEPES, pH 7.9, 100 mM KCl, 0.2 mM EDTA, 20% glycerol, 1 mM DTT, 1 mM PMSF, and 1 g/ml aprotinin, leupeptin, and pepstatin A). Nuclear extracts used for detection of TF-specific NF-B DNA binding activity were prepared employing a modified protocol as described (23). Finally, extracts were frozen in liquid nitrogen and stored at Ϫ80°C until use. Protein concentrations in extracts were determined by the BCA method (Pierce).

FIG. 3. CD40L-induced TF promoter activity in human endothelial cells involves AP-1-and NF-B-binding sites.
A, human EC cultures were transiently transfected with plasmids containing the wild-type TF promoter, pTF(Ϫ227), or those with either mutation in the AP-1 sites (AP-1 m ) or NF-B site (NF-B m ). Induction of TF promoter activity by CD40L (10 g/ml, 6 h) is expressed as percent of the wild-type promoter. Six experiments performed with EC from different donors yielded similar results. B-D, confluent human EC were cultured (16 h) in serum-free medium and then in the absence or presence (1 h) of recombinant CD40L (10 g/ml). Protein-DNA binding was performed with nuclear extracts (AP-1, 2 g; NF-B, 10 g) and oligonucleotides containing binding sites for the proximal (AP-1 P ) and distal (AP-1 D ) AP-1 as well as NF-B sites of the TF promoter. To confirm the specificity of the protein-DNA complexes, a 25-fold molar excess of non-labeled competitor oligonucleotides containing AP-1, NF-B, or Sp1 consensus binding sites or specific antibodies (4 g) was applied (arrowhead indicates supershift, n.s. indicates nonspecific bands). Protein-DNA complexes were separated using 6% nondenaturing acrylamide gels. Three experiments performed with EC from different donors yielded similar results.
Statistical Analysis-Data are presented as mean Ϯ S.D. and groups were compared using Student's t test. A value of p Ͻ 0.05 was considered significant.

CD40L Responsive Elements of the Human TF Promoter Localize Downstream of
Ϫ278 bp-Ligation of CD40 on human vascular EC induced de novo synthesis and cell surface expression of immunoreactive TF protein as determined by Western blot, radioimmunoprecipitation, and FACS analysis (Fig. 1), in accordance with previous reports (16 -18). Notably, stimulation with CD40L further induced the expression of TF mRNA, suggesting transcriptional modulation via this cytokine (Fig. 1). To identify putative CD40L response elements in the TF promoter in human EC, we performed promoter reporter gene analysis utilizing 5Ј-deletion constructs of the wild-type TF promoter (Ϫ278 to ϩ121 bp) ( Fig. 2A). Ligation of CD40 on human EC transiently transfected with wild-type TF promoter (pTF(Ϫ278)) markedly increased the promoter activity compared with non-treated cells (fold induction: 2.70 Ϯ 0.78; n ϭ 5, Fig. 2B). However, 5Ј-deletion of the wild-type promoter between Ϫ278 and Ϫ194 bp diminished the constitutive and CD40L-inducible promoter activity significantly compared with the wild-type promoter (53 Ϯ 11 and 16 Ϯ 23%, respectively; n ϭ 5, p Ͻ 0.05). Deletion of the promoter region to Ϫ111 bp further decreased the basal promoter activity (21 Ϯ 7%; n ϭ 5) but continued to show diminished inducibility of the promoter by CD40L compared with wild-type (28 Ϯ 24%; n ϭ 5). Deletion of the promoter to Ϫ67 or Ϫ21 bp eventually resulted in the loss of detectable promoter activity with and without exposure to CD40L (Fig. 2B). These data indicated that a region of the human TF promoter downstream of Ϫ278 bp contains cis-acting elements required for induction of the promoter by CD40L in human EC.

The Induction of Human Tissue Factor Promoter Activity by Recombinant CD40L in Human Endothelial Cells Requires Functional AP-1-and NF-B-binding Sites-
The significantly impaired inducibility of the TF promoter activity by CD40L associated with 5Ј-deletions indicated the involvement of the LRE region. This region contains two AP-1 and one NF-B binding sites. To determine whether these sites indeed participate in the CD40L-mediated induction of TF, we transfected EC with plasmids containing mutations in the proximal and distal AP-1-(pTF(Ϫ227)AP-1 m ) or NF-B (pTF(Ϫ227)NF-B m )binding sites. Mutations in either binding site significantly and independently diminished the CD40L-mediated induction of TF promoter activity compared with wild-type (AP-1, 33 Ϯ 18%; NF-B, 18 Ϯ 17%, n ϭ 6, p Ͻ 0.05; Fig. 3A). These findings demonstrated an obligatory role for both AP-1-and NF-Bbinding sites in the induction of TF gene expression following CD40 ligation.
More detailed analysis of AP-1 and NF-B binding activity in EC following CD40 ligation used EMSA employing oligonucleotides containing the TF promoter-specific proximal (AP-1 P ) and distal (AP-1 D ) AP-1 as well as NF-B DNA-binding sites. Nuclear extracts from unstimulated as well as CD40L-stimulated EC formed a prominent protein-DNA complex with the distal and proximal AP-1 sites (Fig. 3, B and C). Specificity of this complex for AP-1 was demonstrated by competition with a 25-fold molar excess of unlabeled AP-1 but not of Sp1 consensus oligonucleotides (Fig. 3, B and C). To determine the composition of the AP-1 complexes formed at the proximal and distal AP-1 sites, specific antibodies to members of the AP-1 transcription factor family were added to the DNA-protein binding assay. Supershifted protein-DNA complexes occurred after addition of antisera against c-Jun and JunD, but not JunB, c-Fos, and FosB to the DNA binding reaction performed with either the proximal or distal AP-1 sites (Fig. 3, B and C). The observation that nuclear extracts from unstimulated and CD40Lstimulated EC yielded similar results suggested constitutive binding of c-Jun and JunD to the AP-1 sites of the TF promoter in human EC.
Similarly, we performed EMSA employing oligonucleotides containing the TF promoter-specific NF-B site. Incubation of nuclear proteins from unstimulated EC with the TF promoterspecific NF-B site resulted in constitutive complex formation with low abundance. Ligation of CD40L, however, enhanced

FIG. 4. CD40L-dependent activation of the TF promoter in human endothelial cells involves Egr-1-binding sites.
A, human EC were transiently transfected with plasmids containing the wild-type TF promoter and derivatives containing mutations of the Egr-1-binding (Egr-1 m ) sites in regions I, II, or III. Induction of TF promoter activity by CD40L (10 g/ml, 6 h) is expressed as percent of the wild-type promoter. Five experiments performed with EC from different donors yielded similar results. B, confluent human EC were cultured (1 h) in the absence or presence of recombinant CD40L (10 g/ml). Protein-DNA binding was performed with nuclear extracts (2 g) and oligonucleotides containing Egr-1-binding sites located in regions I, II, or III (RI, RII, and RIII) of the SRR in the human TF promoter. To confirm the specificity of protein-DNA complexes, a 25-fold molar excess of non-labeled competitor oligonucleotides containing either Egr-1 or AP-1 consensus binding sites, respectively, was applied. Protein-DNA complexes were separated using 6% nondenaturing acrylamide gels. Three experiments performed with EC from different donors yielded similar results. complex formation, which ceased in the presence of an excess of an unlabeled oligonucleotide containing a NF-B but not an AP-1 consensus binding site. Furthermore, addition of antibodies against c-Rel and p65, but not of p50, RelB, and p52, diminished formation of the CD40L-enhanced complex (Fig.  3D), indicating that CD40 ligation enhanced binding of a c-Rel/ p65 heterodimer to the TF-specific NF-B site in EC.

Egr-1 Modulates Induction of the Human Tissue Factor Promoter by Recombinant CD40L in Human Endothelial Cells-
The analysis of TF promoter activity employing the 5Ј-deletion constructs described above did not permit evaluation of the involvement of binding sites located in the SRR because of the loss of inducibility in these markedly shortened constructs. Although the three Egr-1-binding sites may participate in the induction of TF promoter activity in EC by shear stress and vascular endothelial growth factor (5,11), previous work has not associated these sites with TF induction by proinflammatory cytokines (25).
Exploration of Egr-1 involvement employed TF promoter constructs mutated in all three Egr-1-binding sites. Surprisingly, mutations in the Egr-1 sites of the SRR significantly reduced the induction of the promoter activity by CD40L compared with wild-type (46 Ϯ 31%, n ϭ 5; p Ͻ 0.05, Fig. 4A).
These results indicated that Egr-1 indeed contributes to the CD40L-dependent induction of TF gene expression in human EC.
To test CD40L-inducible protein-DNA binding activity to the Egr-1 sites, we performed EMSA employing oligonucleotides containing overlapping regions (RI, RII, and RIII) of the SRR, each containing a single Egr-1-binding site. Incubation of nuclear extracts from CD40L-stimulated EC with labeled oligonucleotides containing RI, RII, or RIII of the SRR all demonstrated formation of a protein-DNA complex, which was competed with oligonucleotides containing a consensus Egr-1binding site but not with those containing an AP-1 consensus binding site (Fig. 4B). Notably, comparison of nuclear extracts obtained from CD40L-stimulated EC with those from cultures treated with PMA, an established and potent inducer of Egr-1 protein expression and Egr-1-specific DNA binding activity, showed formation of protein-DNA complexes of the same electrophoretic mobility, further supporting the CD40L-inducible binding of Egr-1 to the TF promoter SRR (Fig. 5A). In addition, CD40L-inducible complexes in RI, RII, and RIII supershifted with an antibody specific for Egr-1 but not with irrelevant antibodies specific for Sp1 or c-Rel (Fig. 5B). Altogether these FIG. 5. CD40L induces binding of Egr-1 to three Egr-1-binding sites of the TF promoter in human endothelial cells. Confluent human EC were cultured (1 h) in the absence or presence of recombinant CD40L (10 g/ml) or PMA (50 ng/ml). Protein-DNA binding was performed with nuclear extracts (2 g) and oligonucleotides containing Egr-1-binding sites located at regions I, II, or III (RI, RII, and RIII) in the human TF promoter. To confirm the specificity of protein-DNA complexes, a 25-fold molar excess of nonlabeled competitor oligonucleotides containing Egr-1, Sp1, or AP-1 consensus binding sites (A) or specific antibodies (B) (4 g) were applied (arrowhead indicates supershift). Protein-DNA complexes were separated using 6% nondenaturing acrylamide gels. Three experiments performed with EC from different donors yielded similar results. studies demonstrated that the ligation of CD40 induced the formation of protein-DNA complexes specific for Egr-1.
CD40L Induces the Expression of Egr-1 Protein in Human Vascular Endothelial Cells-To verify our surprising observation that ligation of CD40 on EC promotes TF expression via Egr-1, we further analyzed whether CD40L induced the expression of Egr-1 protein in human EC. Incubation of EC with CD40L indeed triggered accumulation of the Egr-1 protein in nuclear extracts, a previously unsuspected function. Notably, this increase in Egr-1 protein resembled that induced by PMA (Fig. 6A). Interestingly, IL-1␤ and TNF␣ did not elevate either expression of Egr-1 protein (Fig. 6A) or Egr-1-specific protein-DNA complexes (Fig. 6B), as demonstrated by Western blot and EMSA, respectively.

DISCUSSION
Various lines of evidence support a role for CD40 signaling in regulating the thrombotic balance in blood vessels (15)(16)(17)(18)(19)(20). However, the molecular pathways of CD40L-induced TF expression have remained unknown. We demonstrate here that optimal induction of TF expression in human EC requires not only binding activity of AP-1 and NF-B family members but also of Egr-1, a transcription factor previously not associated with CD40 signaling and not employed by other established mediators of TF expression, namely IL-1␤ and TNF␣ (Fig. 7).
Promoter-reporter gene studies employing 5Ј-deletion constructs localized CD40L-sensitive elements to a region downstream of Ϫ278 bp in the human TF promoter. This region contains the recognized LRE as well as the SRR containing the main regulatory cis-elements for TF gene expression (AP-1 and NF-B as well as Egr-1-binding sites, respectively) (6). Notably, induction of TF promoter activity by proinflammatory cytokines, such as IL-1␤ and TNF␣, utilizes binding sites located in the LRE (4), whereas those in the SRR are employed by mediators, such as serum, shear stress, hypoxia, and vascular endothelial growth factor (5,9,12,13). Of further interest, heretofore only PMA and LPS have been reported to utilize both elements, LRE and SRR, to induce TF gene expression (4,9,14,24).
Promoter reporter gene studies demonstrated the involvement of AP-1-and NF-B-binding sites in CD40L-mediated TF expression, in common with other proinflammatory cytokines, such as IL-1␤ and TNF␣ (6). Notably, the constitutive binding of c-Jun and JunD to the AP-1 sites of the TF promoter was not affected by CD40L, whereas CD40 ligation markedly enhanced the binding of c-Rel/p65 proteins to the NF-B site (Fig. 3). The constitutive binding of c-Jun and JunD proteins to the AP-1 sites agrees with the basal TF promoter activity. Nevertheless, CD40L-induced phosphorylation of bound c-Jun by the stressactivated protein kinase c-Jun NH 2 -terminal kinase (26) might contribute to increased transcriptional activity of c-Jun (27). The CD40L-dependent activation of NF-B binding to the TF promoter most likely depends on the inactivation of IB␣ via IKK as demonstrated for other cytokines in various cell types, including EC (28,29). The requirement for members of both the AP-1 and NF-B/Rel transcription factor family to maximally induce the TF promoter by CD40L might indicate the necessity of cooperative binding of c-Jun/JunD to the AP-1 sites and c-Rel/p65 to the NF-B site, both located closely in the LRE of the TF promoter. This hypothesis gains support from recent studies that demonstrated interaction of c-Jun as well as JunD with p65 exhibiting enhanced functional DNA binding activity (30,31). In addition, modifying the spacing between the AP-1 sites and the NF-B site in the LRE by insertion of additional base pairs abolished induction of the TF promoter by LPS in monocytes (32).
We further tested whether CD40L, besides PMA and LPS, might utilize binding sites in both elements, LRE and SRR, for optimal induction of TF gene expression, indicating that CD40 signaling differs from that of previously studied proinflammatory mediators, e.g. IL-1␤ and TNF␣. In contrast to the proinflammatory cytokines IL-1␤ or TNF␣, CD40L suprisingly induced Egr-1 protein expression and DNA binding activity in EC, notably to a degree similar to that achieved by PMA. Although a recent report suggested the activation of Egr-1 by FIG. 6. Egr-1 protein expression in human endothelial cells is induced by CD40L but not by IL-1␤ or TNF␣. Confluent human EC were cultured (1 h) in the absence and presence of IL-1␤ (10 ng/ml), TNF␣ (50 ng/ml), recombinant CD40L (10 g/ml), or PMA (50 ng/ml). A, nuclear extracts (20 g) were separated by SDS-PAGE, and expression of Egr-1 protein was determined by Western blotting using anti-Egr-1 rabbit polyclonal antibody (1:500). B, protein-DNA binding was performed with nuclear extracts (2 g) employing oligonucleotides containing consensus DNA-binding sites for Egr-1. To confirm specificity of protein-DNA complexes a 25-fold molar excess of non-labeled competitor oligonucleotides containing either a Egr-1 or AP-1 consensus binding site or specific antibodies (4 g) was applied (arrowhead indicates supershift). Protein-DNA complexes were separated using 6% nondenaturing acrylamide gels. Three experiments performed with EC from different donors yielded similar results. CD40L Induces TF Expression via AP-1, NF-B, and Egr-1 in EC TNF␣ in EC (33), lack of Egr-1 inducibility via TNF␣ in this cell type reported here is in accordance with the previous report of Pendurthi et al. (25). Notably, all three Egr-1-binding sites located downstream of the LRE in the human TF promoter appear to function in CD40L-mediated TF expression.
Therefore, optimal induction of the TF promoter by CD40L in EC probably requires not only interaction with the transcription factors of the AP-1 and c-Rel/NF-B families but also Egr-1. The finding that members of all three transcription factor families participate in the induction of TF expression agrees with recent observations in LPS-stimulated monocytic cells (14).
The present findings indicate that CD40 ligation on EC results in broader activation of intracellular signaling pathways targeting the TF gene than the "classical" cytokines IL-1␤ and TNF␣. This finding may further help to understand why CD40L activates certain genes (e.g. caspase-1 (ICE) and stromelysin-3) (21, 34) not affected by IL-1␤ or TNF␣.
The present results provide new insight into the transcriptional regulation of endothelial TF expression following CD40 ligation and illustrate a role for Egr-1 activation in the regulation of the thrombogenic potential of endothelial cells by CD40L. Thus, pharmacologic inhibition of CD40L signaling pathways might provide new therapeutic options to prevent thrombotic complications in inflammatory processes such as atherosclerosis.