LDL Induces Transcription Factor Activator Protein-1 in Human Endothelial Cells

From the Division of Biomedical Sciences, University of California, Riverside (Y.Z., H.L.L., O.F., L.V., N.W.M., D.S.S., M.B.S.), and the Department of Pathology, New York Medical College, Valhalla, NY (J.H-C.L.).

	Abstract

Top Abstract Introduction Methods Results Discussion References

 
Abstract—Low density lipoprotein (LDL) has been shown to perturb endothelial cells, with manifestations ranging from alterations in free radicals and arachidonate metabolism to stress fiber formation and monocyte recruitment. Some of these changes are regulated by LDL at the transcriptional level. Using mobility shift assays with consensus sequences for various transcription factors, we have detected an increase in activator protein 1 (AP-1), but not nuclear factor-B (NF-B), binding in human umbilical vein endothelial cells exposed to LDL. Following transfection, AP-1–driven chloramphenicol acetyltransferase and AP-1–driven-luciferase are upregulated by LDL. In contrast, there is no effect on NF-B–driven chloramphenicol acetyltransferase. AP-1 increases in a biphasic fashion, with the first peak occurring 6 hours after and the second 48 hours after exposure to LDL. This AP-1 binding increase involves c-Jun, but not c-Fos, as shown by gel supershift, Northern hybridization, and Western blotting analyses. c-Jun mRNA levels are elevated by 9 hours after and remain so until at least 24 hours after exposure to LDL. c-Jun protein levels increase at 12 hours and continue to rise for 24 hours after exposure to LDL. Moreover, this LDL-increased AP-1 binding is suppressed by several protein kinase (PK) inhibitors: the PKC inhibitor calphostin C, the cAMP-dependent PK inhibitor H89, and the tyrosine PK inhibitors genistein and lavendustin A. This study demonstrates that (1) LDL is an endothelial agonist distinct from other cell stimulators, such as cytokines, endotoxin, and phorbol 12-myristate 13-acetate, because LDL appears to activate human umbilical vein endothelial cells predominantly through the transcription factor AP-1 and not NF-B; and (2) LDL increases AP-1 via mechanisms involving multiple kinase activities and c-Jun transcription.

Key Words: LDL • activator protein 1 • c-Jun • c-Fos • human umbilical vein endothelial cells

	Introduction

Top Abstract Introduction Methods Results Discussion References

 
Low density lipoprotein is a well-established risk factor for atherosclerosis.^{1 2 3 4} When ECs are incubated with LDL in concentrations associated with premature coronary artery diseases, the cells become activated.^{5 6 7 8 9 10 11} Manifestation of this activation ranges from an alteration in free radicals⁵ and arachidonate metabolism^{6 7} to stress fiber formation⁸ and monocyte recruitment⁹ via ICAM-1¹⁰ and VCAM-1 induction.¹¹ Many inducible molecules in ECs, eg, VCAM-1, ICAM-1, and monocyte-specific chemoattractant protein-1 (MCP-1), contain a number of AP-1-like and NF-B-like binding motifs within their gene promoter regions.^{12 13 14 15} Both AP-1 and NF-B transcription factors have been reported to be involved in the regulation of these molecules.^{11 12 15 16 17} Recently, when the effects of LDL on VCAM-1 promoter activation were studied, we found LDL activation of AP-1 but not NF-B.¹¹ Thus, LDL-activated AP-1 may play an important role in the perturbation of ECs.

AP-1 is a key transcription factor that translates external stimuli into both short- and long-term changes of gene expression.^{18 19} AP-1 proteins consist of a variety of homodimers and heterodimers, including members of the Fos, Jun, and CREB/ATF families. These dimers are generated through the interactions between the leucine zipper motifs on each monomer.²⁰ Different members of the Fos, Jun, and CREB/ATF families exhibit different structural features. Such features lead to subtle differences in their DNA binding and transcriptional activation properties, and in turn, suggest specific functions in gene regulation for individual dimers.^{19 20} The activity of AP-1 is regulated by complex mechanisms. This regulation occurs through interactions with specific PKs and a variety of transcriptional coactivators. These mechanisms include posttranslational events acting on preexisting AP-1 proteins and transcriptional activation leading to increased amounts of AP-1 proteins.

In this study, the goal of which was to identify the transcription factor(s) activated by LDL in human ECs, LDL is shown to be an endothelial agonist distinct from other cell stimulators. Specifically, in cell culture medium containing 20% FBS, LDL activates HUVECs predominately through AP-1, whereas other cell stimulators activate these cells primarily through the transcription factor NF-B. On the basis of these results, which are the first to document such an effect of LDL in EC modulation, we propose a pathway for LDL activation of ECs that is distinct from other endothelial agonists.

	Methods

Top Abstract Introduction Methods Results Discussion References

 
Reagents
All PK inhibitors were from LC Laboratories. TNF- was a generous gift from Knoll Pharmaceutical, Whippany, NJ. Nonfrozen human plasma was obtained from the San Bernadino Blood Center, San Bernadino, Calif. Collagen was from Collaborative Biomedical Products. Collagenase was from Worthington Biochemical Corp. Recombinant human fibroblast growth factor was a generous gift from Dr J.A. Thompson, University of Alabama, Birmingham. [-³²P]dCTP (3000 Ci/mmol) and [-³²P]ATP (3000 Ci/mmol at 10 mCi/mL) were from ICN Biomedicals. The DECApriming II DNA labeling kit was from Ambion. Antibodies for the Western blotting and mobility supershift assays were from Santa Cruz Biotechnology Inc, and consensus sequences for NF-B, AP-1, and CREB were from Promega Corp. The CAT ELISA kit and Sephadex G-25 spin columns were from Boehringer Mannheim Biochemicals. Secondary antibodies for Western blotting and the ECL system were from Amersham Life Science Inc. BCA protein assay reagents were from Pierce Chemical Co. T4 DNA kinase was from New England Biolabs. Poly(dI-dC) · (dI-dC) was from Pharmacia Biotech. Reagents for polyacrylamide gel electrophoresis were from Bio-Rad Laboratories. All other reagents were of tissue culture or molecular biology grade and purchased from either Gibco BRL or Sigma Chemical Co.

Cell Cultures
HUVECs were extracted by the collagenase digestion method^{5 6 7 8} and cultured on plates coated with 50 µg/mL collagen. Cells were maintained in medium 199 (M199) supplemented with 20 mmol/L HEPES, pH 7.4, 20% FBS, 5 ng/mL of recombinant human fibroblast growth factor, antibiotics/antimycotics, and 90 µg/mL of heparin (EC medium).^{5 11} All experiments were performed with cells from passages two and three in EC-medium and cultured to confluence before the LDL incubation. COS-7 cells were obtained from American Type Culture Collection, Rockville, Md, and maintained in Iscove's modified Dulbecco's medium supplemented with 10% FBS.

LDL Isolation
LDL was isolated from nonfrozen human plasma as described.^{5 11} After two rounds of ultracentrifugation, LDL was dialyzed against two changes of saline solution containing 0.01% EDTA and 20 µmol/L BHT, followed by one change of M199 for HUVECs or of Iscove's modified Dulbecco's medium for COS-7 cells. The LDL preparations contained <0.0005 U endotoxin per milligram of cholesterol, as determined by the chromogenic Limulus test (BioWhittaker). For all studies unless indicated otherwise, LDL was used at a final cell culture concentration of 220 mg cholesterol per deciliter (5.72 mmol cholesterol per liter).¹¹

Nuclear Extract Preparation
HUVECs were cultured to confluence on 100-mm dishes and then exposed to LDL for 6 hours or for various times up to 2 days as described in the time-course study. Controls were parallel groups of cells without any treatment or cells stimulated for 2 hours with 10 ng/mL TNF- or 10 ng/mL PMA. After activation, the EC monolayers were extracted for nuclear proteins as previously described.¹¹ In brief, cells were allowed to swell on ice in a hypotonic buffer, their nuclei were pelleted, and nuclear proteins were extracted in a high-salt buffer. To minimize proteolysis, all buffers contained freshly added 0.2 mmol/L PMSF, 0.5 mmol/L DTT, 5 µg/mL aprotinin, 5 µg/mL leupeptin, and 5 µg/mL antipain. The nuclear extract was snap-frozen in liquid N₂ and stored at -80°C. Protein concentrations were determined by the BCA method of Pierce.

Electrophoretic Mobility Shift Assay
The consensus sequences for transcription factors AP-1 and NF-B were end-labeled with [-³²P]ATP using T4 DNA kinase. Unincorporated [³²P]ATP was removed via a Sephadex G-25 spin column. Binding of the labeled oligonucleotides to their corresponding factors was performed according to the method described¹¹ with modifications. In brief, a specific amount of nuclear protein (6 µg for NF-B and 3 µg for AP-1) was incubated with labeled oligonucleotide at room temperature for 20 minutes in a buffer containing 20 mmol/L HEPES (pH 7.9), 90 mmol/L KCl, 1 mmol/L MgCl₂, 1 mmol/L EDTA, 0.5 mmol/L DTT, 0.1 mg/mL poly(dI-dC) · (dI-dC), and 4% glycerol. Reaction mixtures were resolved on 5% nondenaturing polyacrylamide gels, and the binding patterns were detected by autoradiography. To test the specificity of binding, a 100-fold molar excess of unlabeled competing (or irrelevant) oligonucleotide was used for competition experiments. To identify the components involved in the LDL-increased AP-1 binding, a supershift assay was performed. The nuclear extracts were incubated with antibodies for 3 hours on ice before addition of the labeled probes.

Western Blotting Analysis
Protein analysis was performed according to the manufacture's protocol using the ECL system from Amersham Inc. In brief, 25 µg of nuclear extract mixed in sample buffer was boiled for 5 minutes. Proteins were separated on a 10% SDS polyacrylamide resolving gel with a 4% stacking gel and then transferred from the gel to a nitrocellulose membrane. After being blocked in 5% blocking reagent in Tris-buffered saline–Tween 20 (TBS-Tween), the membrane was incubated with primary antibodies against c-Jun or c-Fos for 1 hour at room temperature. After three washes with TBS-Tween, the membrane was incubated with a secondary antibody conjugated with horseradish peroxidase for another hour at room temperature, followed by three washes in TBS-Tween. The proteins were detected using ECL detecting reagents and autoradiography. Nuclear extracts from PMA-treated cells served as a positive control.

Northern Hybridization
Total RNA was isolated by the guanidinium isothiocyanate/CsCl ultracentrifugation method²¹ and subjected to Northern analysis²¹ for c-Jun and c-Fos expression. The c-Jun cDNA was from Dr L. Ransone, Salk Institute, La Jolla, Calif, and the c-Fos cDNA from Dr T. Curran, St Jude Children's Research Hospital, Memphis, Tenn. The probes were labeled with [-³²P]dCTP by the DECApriming II kit, and binding to the corresponding messages was detected by autoradiography.

Construction of pTATA-luc and p3xAP-1-TATA-luc Plasmids
A DNA fragment containing the TATA box of the herpes simplex virus thymidine kinase gene (HSV-TK) was cloned into pGL2-Basic (Promega) upstream from the luc gene to generate pTATA-luc. A doubled-stranded oligonucleotide containing a general AP-1 consensus binding site (Promega) was phosphorylated for ligation with T4 polynucleotide. It was ligated into the EcoRV site of pBluescript (SK^-) by using a 120:1 oligonucleotide-to-vector molar ratio. After transformation into bacteria, a plasmid containing the vector with an insert consisting of three copies of the oligonucleotide was identified by DNA sequencing. The trimeric AP-1 binding site was then removed from pBluescript (SK^-) by using KpnI and SacI and cloned upstream from the TATA box in TATA-luc to generate p(AP-1)₃-TATA-luc.

Transfection
The promoter reporter constructs for assessing AP-1 and NF-B activities, p3xAP1–37TKcat and p2xB-37TKcat, and their vector p-37TKcat were from Dr M. Meyer, European Molecular Biology Laboratory, Heidelberg, Germany.²² Plasmid pRSV-ß-gal, included in all transfection experiments for determining the relative transfection efficiency, was the generous gift of Dr T. Parks, Boehringer Ingelheim Pharmaceuticals Inc, Ridgefield, Conn. All plasmid DNAs were purified through two rounds of CsCl/ethidium bromide equilibrium centrifugation.²¹

HUVECs were seeded on 100-mm dishes and incubated overnight in EC medium. The next morning, the medium was switched to Dulbecco's modified Eagle's medium with 5% FBS for 2 hours before the transfection was performed. Supercoiled construct DNA was precipitated with calcium phosphate in the HEPES-buffered saline solution, pH 7.08 to 7.12, and then added dropwise to the dishes. After a 3 to 4-hour incubation with the calcium phosphate/plasmid mixture, the cells were washed with PBS and switched back to the regular EC medium. Each of the transfectants was cotransfected with pRSV-ß-gal to normalize the transfection efficiency and standardize the comparison. The posttransfected cells were then incubated for 48 hours with or without 220 mg/dL LDL. A parallel transfected group was treated with 100 ng/mL PMA for 16 hours before the samples were collected as a positive control. The promoter activities were measured by their reporter luc. Transfection results were expressed as luc activity normalized against ß-gal.

Transfection of COS-7 cells was performed in six-well plates by using the DEAE-dextran method with the ProFection kit from Promega. The cells were transfected with AP-1 (p3xAP1–37TKcat), NF-B (p2xB-37TKcat), and enhancer-less control (p-37TKcat) reporter gene constructs and pRSV–ß-gal at a 5:1 ratio. The posttransfected cells were then incubated for 48 hours with or without 220 mg/dL LDL. A parallel transfected group was treated with 100 ng/mL PMA as a positive control. The promoter activities were monitored by their reporter CAT, and the amount of CAT protein in the cell lysate was quantified by a CAT ELISA kit. Transfection results were expressed as CAT protein normalized against ß-gal.

	Results

Top Abstract Introduction Methods Results Discussion References

 
LDL Increases AP-1 Binding Activity
Confluent HUVECs were incubated in EC medium with LDL for 6 hours or with 10 ng/mL TNF- or 10 ng/mL PMA for 2 hours. Binding of AP-1 was increased by LDL, TNF-, or PMA as shown in Fig 1a. LDL had a greater effect than TNF- in elevating AP-1 binding. In contrast, NF-B, which was markedly activated by TNF- and PMA, was virtually absent in LDL-exposed HUVECs. LDL increased AP-1 binding in a dose-dependent fashion beginning at 160 mg/dL. This effect was maximal at 240 mg/dL (Fig 1b). The increase in AP-1 binding in HUVECs is biphasic as revealed in a time-course study: the first peak appeared at 6 hours and the second after 48 hours of exposure to LDL (Fig 2). Basal AP-1 binding was not changed as observed at all time points (data for only two time points shown). There was no AP-1 change relative to control on cells incubated with BHT (1 to 4 µmol/L) or EDTA (0.001% to 0.005%); these are the amounts of BHT or EDTA contained in the LDL incubation medium. Also, no effect was observed in cells exposed to M199 used in the last LDL dialysis. With a gel shift assay, neither AP-1 nor NF-B binding was increased in any of the above controls after a 6-hour incubation (data not shown).

View larger version (44K): [in this window] [in a new window]   Figure 1. Electrophoretic mobility shift assay for AP-1 and NF-B binding. Confluent HUVECs were incubated in (a) EC medium (Ctrl) with 220 mg/dL LDL for 6 hours (LDL), 10 ng/mL TNF- (TNF) for 2 hours, or 10 ng/mL PMA for 2 hours (PMA) or (b) EC medium with different concentrations of LDL as indicated for 6 hours. Nuclear extracts were prepared and quantified. 32 P-labeled consensus AP-1 or NF-B oligonucleotides were incubated for 30 minutes at room temperature with equal amounts of nuclear extract. Protein-DNA complexes were resolved by electrophoresis. Specific protein-DNA complexes are indicated by an arrow. The gel shift bands were quantified by densitometry and relative density is expressed as a percentage of control samples. Data presented are representative of >3 independent experiments. Within each experiment, 3 repetitive samples were run.

View larger version (44K): [in this window] [in a new window]   Figure 2. Temporal changes in AP-1 binding increased by LDL. Confluent HUVECs were incubated in EC medium with 220 mg/dL LDL at different periods of time as indicated. Nuclear extracts were prepared and quantified. Untreated samples were collected at 2 hours (lane 2, Ctrl) and 24 hours (lane 8, Ctrl) as basic controls. 32 P-labeled consensus AP-1 oligonucleotide was incubated with equal amounts of nuclear extract samples, and protein-DNA complexes were resolved by electrophoresis. Specific protein-DNA complexes are indicated by an arrow. The AP-1 bands were quantified by densitometry, and relative density is expressed as a percentage of the 2-hour basic control. Data presented are representative of 3 independent experiments. Within each experiment, 2 repetitive samples were run.

LDL Functionally Increases AP-1–Dependent Protein Expression
LDL increased AP-1 but not NF-B binding as demonstrated by gel shift assays. To further determine whether LDL could activate an AP-1–driven-promoter, we examined the effect of LDL by transfecting COS-7 cells with the following constructs: p3xAP1–37TKcat, p2xB-37TKcat, and the enhancer-less p-37TKcat. Fig 3a shows the averaged results of five independent experiments. p3xAP1–37TKcat was more efficient and produced a higher basic level of CAT protein than did p-37TKcat and p2xB-37TKcat. LDL doubled the promoter activities driven only by the AP-1 motif (P<.01) but not the activity of the promoter driven by the NF-B motif. PMA, on the other hand, significantly activated both AP-1–and NF-B–dependent reporter CAT activities (P<.001). Thus, LDL specifically regulates genes driven by the AP-1 motif and has no effect on the NF-B–dependent reporter, which is highly responsive to PMA. To confirm this finding in HUVECs, the TATA box of the HSV-TK promoter with or without the trimeric AP-1 binding sequences was constructed into a more sensitive reporter, the luc vector. These constructs, pTATA-luc and p3xAP-1-TATA-luc, were transfected into HUVECs. After exposure to LDL, the samples were collected and assayed for luc expression. As shown in Fig 3b, similar results were observed in HUVECs and COS-7 cells. LDL consistently induced AP-1–driven luc activity 2.13-fold compared with basal controls (25.60±8.12 versus 11.99±5.82, P<.05). Therefore, in both cell types, LDL was equally effective in activating AP-1–driven gene expression.

View larger version (23K): [in this window] [in a new window]   Figure 3. Effect of LDL on AP-1–driven reporter gene. a, COS-7 cells were cotransfected at a 5:1 ratio with pRSV-ß-gal and one of the reporter constructs: AP-1 (p3xAP1–37TKcat), NF-B (p2xB-37TKcat), or enhancer-less control (p37TKcat) reporter gene constructs. After recovery, transfected cells were incubated for 48 hours with or without 220 mg/dL LDL. A 100 ng/mL PMA treatment was used as a positive control. Promoter activities were monitored by their reporter CAT, and the CAT protein levels were normalized against ß-gal. Data presented are the mean±SD of 5 independent experiments, each performed in duplicate. b, HUVECs were cotransfected with pRSV-ß-gal and the luc reporter constructs p(AP-1)3-TATA-luc or pTATA-luc. Posttransfection cells were incubated for 48 hours with or without 220 mg/dL LDL. As a positive control, a parallel transfection group was treated with 100 ng/mL PMA. Promoter activities were measured by their reporter luc and normalized against ß-gal. Data are presented as the mean±SD of relative luc activities in 3 independent experiments, each performed in triplicate.

LDL-Induced AP-1 Binding Does Not Involve c-Fos
To further address the question of which members of the AP-1 family are present in nuclear extracts from LDL-stimulated ECs, supershift experiments were implemented. As shown in Fig 4, binding of AP-1 from LDL-treated cells' nuclear extracts could be supershifted by antibodies against c-Jun and JunD, but not by antibodies against c-Fos and p65, a component of NF-B. The results suggest that the LDL-increased AP-1 dimers in the nuclear extract largely contain c-Jun and JunD. Two different antibodies against c-Fos were tested in this study: one antibody (catalog No. sc-52, Santa Cruz Biotechnology Inc.) that reacts specifically with c-Fos and the other (catalog No sc-413, Santa Cruz Biotechnology Inc.) that reacts broadly with all Fos family members. Neither antibody could supershift the AP-1 band. The same antibodies could supershift the increased AP-1 band in a nuclear extract treated by PMA. Furthermore, since the AP-1 band could be completely blocked by unlabeled consensus AP-1 and partially decreased by unlabeled consensus CREB, but not by the NF-B sequence, the involvement of CREB members cannot be ruled out.

View larger version (77K): [in this window] [in a new window]   Figure 4. Components involved in AP-1 binding increased by LDL. HUVEC nuclear extracts were prepared from cells incubated with EC medium in the presence of LDL (220 mg/dL) for 6 hours or with PMA (10 ng/mL) for 2 hours. A gel shift assay was performed as mentioned above. For supershift, nuclear extracts were incubated with antibodies for 3 hours on ice before adding the labeled AP-1 probe. Nuclear extracts from untreated (lane 1), LDL-exposed (lanes 2–10), and PMA-treated (lanes 11–13) HUVECs were analyzed for binding activity. A 100-fold molar excess of unlabeled AP-1 probe (lane 3), NF-B probe (lane 4), or CREB probe (lane 5) was added as specific and nonspecific competitors. Antibodies against c-Jun, c-Fos, JunD, and NF-B p65 were included in the protein-DNA complex as labeled. Data presented are representative of 3 independent experiments. Within each experiment, 2 repetitive samples were run.

LDL Induces c-Jun but Not c-Fos in mRNA and Protein Levels
To determine the effects of LDL on the upregulation of subunits of AP-1 proteins and to further study the mechanism of AP-1 increases by LDL, a time-course study by Northern blotting and Western blotting analyses was performed. As shown in Fig 5, incubation of HUVECs with 220 mg/dL LDL for as long as 24 hours increased the steady-state levels of c-Jun mRNA. This increase started at 9 hours and rose for at least another 15 hours. In accordance with our findings in the supershift study, an increase of c-Fos was not observed by incubation with LDL at any time point (Fig 5a). As positive controls, these HUVECs could be induced to express both c-Jun and c-Fos by incubation with 50 ng/mL of PMA. In Northern analysis, PMA significantly increased both c-Jun and c-Fos mRNA within 15 minutes, reaching a peak at 30 minutes for c-Fos and at 45 minutes for c-Jun (Fig 5b). In Western blotting results, c-Jun protein, but not c-Fos, in nuclear extracts was also increased at 12 hours and remained at a high level until at least 24 hours after exposure to LDL (Fig 6).

View larger version (48K): [in this window] [in a new window]   Figure 5. Effects of LDL on c-Jun and c-Fos mRNA levels. a, Confluent HUVECs were incubated in EC medium with LDL (220 mg/dL) for different periods of time as indicated. An untreated sample was collected at 0.5 hour (lane 1) as a control. b, Confluent HUVECs were incubated in EC medium with 50 ng/mL PMA at different periods of time as indicated. RNA was isolated by the guanidinium isothiocyanate/CsCl ultracentrifugation method. Samples of 15 µg of total RNA were resolved by gel electrophoresis and then hybridized with -32P–labeled c-Jun and c-Fos cDNA, respectively. von Willebrand factor (vWF) mRNA levels from corresponding samples were also included to show equal loading. Resulting hybridization bands were quantified by densitometry and normalized against vWF. Relative density is expressed as a percentage of basic control (lane 1). Results shown are representative of 3 independent experiments.

View larger version (25K): [in this window] [in a new window]   Figure 6. Effects of LDL on c-Jun and c-Fos protein expression. Confluent HUVECs were incubated in EC medium with 220 mg/dL LDL for different periods of time as indicated, and nuclear proteins were extracted. Nuclear extract (25 µg) mixed in sample buffer was separated on a 10% SDS polyacrylamide resolving gel and then transferred to a nitrocellulose membrane. After being soaked in 5% blocking reagent, the membrane was incubated with primary antibodies against c-Jun or c-Fos, followed by incubation with a secondary antibody conjugated with horseradish peroxidase. The proteins were detected by autoradiography with ECL detecting reagents. Nuclear extracts from PMA-treated cells were used as a positive control. Results shown are representative of 3 independent experiments.

LDL-Induced AP-1 Binding Is Suppressed by Inhibitors of Various PKs
Since we observed that LDL increased AP-1 binding earlier than c-Jun upregulation at the mRNA and protein levels, we reasoned that posttranslational activation of AP-1 might be occurring. To investigate the possible role of PK activity in the induction of AP-1 by LDL, the effects of a group of PKA, PKC, and PTK inhibitors were examined. A 30-minute treatment with the PKA inhibitor H89, the PKC inhibitor calphostin C, or the PTK inhibitors genistein or lavendustin A was applied prior to a 6-hour incubation of LDL in HUVECs. As shown in Fig 7, all of these inhibitors appreciably reduced the LDL-induced AP-1 binding. These reductions of AP-1 binding by PK inhibitors are not due to general cell toxicity, since no increase in lactate dehydrogenase leakage into the medium was detected in cells with PK inhibitor pretreatments, as assessed by an in vitro toxicology assay kit (Sigma; data not shown). This result suggests that multiple kinase activities, especially those of PKC and PTK, may be involved in the process of AP-1 activation by LDL.

View larger version (40K): [in this window] [in a new window]   Figure 7. Effects of PK inhibitors on AP-1 binding induced by LDL. Before the 6-hour exposure to LDL, HUVECs were pretreated for 30 minutes with the PKA inhibitor H89 (5 µmol/L) the PKC inhibitor calphostin C (CPC, 100 nmol/L) or the PTK inhibitors genistein (GEN, 50 µmol/L) or lavendustin A (LAV, 1 µmol/L). Nuclear extracts were then prepared and assayed by gel shift. AP-1 bands were quantified by densitometry, and relative density was expressed as a percentage of basic control. Results shown are representative of 3 independent experiments. Within each experiment, 2 repetitive samples were run.

	Discussion

Top Abstract Introduction Methods Results Discussion References

 
The present report has examined the hypothesis that human native LDL activates human ECs by way of the transcription factor AP-1. This lipoprotein molecule regulates transcription factor AP-1–binding protein in ECs. Our findings demonstrate for the first time that LDL can increase AP-1 binding biphasically in human ECs. Biphasic AP-1 binding, albeit that induced by shear stress, has also been previously observed in ECs, but these were of bovine origin.²³ Analysis of this latter result, however, was limited by the use of gel shift assays only. In the present study, the LDL-increased c-Jun gene expression was seen at the mRNA and protein levels. c-Jun mRNA and protein were increased after the first peak of AP-1 binding. Thus, LDL appears to regulate AP-1 activity in two steps: (1) a posttranslational event acting on preexisting c-Jun molecules, leading to the first peak of increased AP-1 binding, and (2) induction of c-Jun gene expression, leading to an increase in the total amount of c-Jun, resulting in a second peak of AP-1 binding. These findings may be explained by positive autoregulation of AP-1 on the c-Jun promoter.²⁴ Angel et al²⁴ first demonstrated this autoregulatory loop in PMA-treated HeLa TK^- and HepG2 cells. The variant AP-1-like sequence 5'-GTGACATCAT-3' in the c-Jun promoter region seems to be the key element responsible for this response. Since this site is different from typical AP-1 sites, it may be recognized only by a subset of the complexes that bind to the consensus TPA response element (TRE) sequence.^{19 20} c-Jun homodimers²⁴ and c-Jun-ATF-2 heterodimers^{25 26} are the most likely candidates. Our present results may represent an example of this model in a paradigm simulating vascular injury.

The positive autoregulatory loop of AP-1 on the c-Jun promoter is an attractive model for signal amplification and conversion of transient early events into long-term effects on gene expression. Positive autoregulation, however, may occur in only some cells²⁷ under certain conditions. Fos proteins can associate with any of the Jun proteins to generate stable heterodimers with higher DNA-binding activity than that of Jun homodimers.^{19 20} In most tissues c-Fos expression is highly regulated, and its mRNA is expressed at only relatively low levels; however, it can be rapidly and transiently induced as early as 10 to 15 minutes after the addition of growth factors, PMA, cytokines, or serum. Following a brief peak of expression, c-Fos mRNA levels are substantially reduced and remain at basal levels in the absence of external stimuli.²⁸ Similar patterns of both c-Jun and c-Fos expression were observed in HUVECs after the addition of PMA. However, LDL did not induce c-Fos at either the mRNA or protein level in cell culture medium containing 20% FBS; this may explain why LDL-induced c-Jun proteins form homodimers and initiate a positive autoregulative loop. Since the anti-Fos antibody used in this study was broadly cross-reactive with other members of the Fos family, it is unlikely that a member of the Fos family is involved in LDL-induced AP-1 binding. Moreover, AP-1 may be only one of the transcription factors activated by LDL, or the c-Jun affected may only be within the subset of those AP-1 proteins specifically increased by LDL. In addition to increasing AP-1, LDL increases consensus CREB binding. Both AP-1 and CREB bands were blocked by the same unlabeled probe and partially cross-competed with each other by cold consensus oligonucleotides (data not shown). In vivo footprinting analysis has revealed that c-Jun and ATF-2 heterodimers bind to both the proximal and distal AP-1 binding sites in the c-Jun promoter.^{25 26} The possible involvement of other subsets of AP-1 complexes, such as c-Jun with other Jun dimers and c-Jun/ATF-2 heterodimers, cannot be ruled out. All components of the AP-1 complex are yet to be examined.

An objective in understanding the role of LDL in atherogenesis is to determine how it alters endothelial function.^{5 6 7 8 9 10 11 29} The present report extends previous studies by examining the mechanism of LDL induction of transcription factors. NF-B is believed to be a key oxidative-stress transcriptional regulation factor for cytokines,³⁰ PMA,³¹ lipopolysaccharide,³² and oxidatively modified LDL.^{33 34} All of these agonists induce adhesion molecules and other gene products through NF-B activation in ECs. NF-B activation and gene induction by mildly oxidized LDL appear to be due to the appearance of oxidized phospholipids.^{34 35} Recently, we demonstrated that lysophosphatidylcholine, a phospholipid product of LDL oxidation, can activate NF-B and induce ICAM-1 expression through a PTK-dependent pathway in HUVECs.³⁶ However, in the present study, activation of transcription factors other than NF-B appears to be the major intracellular signaling mechanism of EC activation by LDL.

AP-1 is a key regulator for converting numerous signals into genetic responses.^{19 27} It is, therefore, highly regulated by both the levels of synthesis of c-Jun and c-Fos proteins and by posttranslational modification, principally phosphorylation, of its components. The direct phosphorylation of c-Jun can be catalyzed by members of the MAPK family, JNKs,³⁷ also known as stress-activated protein kinases (SAPKs).^{38 39} The activities of JNKs/SAPKs, like those of other MAPK-related PKs, depend on the phosphorylation of specific threonine and tyrosine residues believed to be catalyzed by an upstream MAPKK-related PK.^{37 39} PKC activation appears to be a key regulator for AP-1 in PMA activation.^{19 26} On the other hand, TNF- activates JNK–AP-1 via a PTK pathway.⁴⁰ Both PKC and PTK may activate AP-1 through the MAPK-JNK pathway by phosphorylating and activating Ras, or c-Raf kinase.^{41 42 43} To further understand the pathway leading to activation of AP-1 by LDL in ECs, we have explored the potential role of PKs in AP-1 binding activity. All PKC, PKA, and PTK inhibitors examined in this study partially inhibited LDL-induced AP-1 binding in cultured HUVECs. Thus, multiple kinase activities are likely involved in the mechanism of LDL-induced AP-1 activation. These inhibitory effects by PK inhibitors on LDL-elevated AP-1 are not due to general cytotoxicity, as neither morphological changes nor an increase of lactate dehydrogenase in the medium was observed in the cells treated with any of the inhibitors at the indicated concentrations. HUVECs receiving the same concentration of these inhibitors were used in another experiment. Specifically, PKA and PKC inhibitors did not affect either NF-B activation or ICAM-1 induction by lysophosphatidylcholine, but genistein almost abolished them.³⁶ PKC inhibitors could not prevent the NF-B increase induced by TNF-, but they were effective in preventing the NF-B increase induced by PMA.³⁶ Thus, multiple PK activities may be involved in downstream signaling pathways in EC activation by LDL.

Exposure of mammalian cells to UV light or high-osmolarity solutions strongly induces clustering and internalization of cell surface receptors for epidermal growth factor, TNF-, and interleukin-1, which results in activation of the JNK cascade.⁴⁴ Likewise, exposure of ECs to LDL may perturb the cell surface or alter receptor conformation, thereby subverting signal pathways normally used by growth factors or cytokines. Prolonged incubation of ECs with LDL increased cellular cholesterol content and cholesterol phospholipid molar ratios of the EC membrane, resulting in a reduction in the relative EC plasma membrane fluidity.²⁹ LDL-pretreated ECs were hyperresponsive to interleukin-1 or TNF- induction of adhesion molecules (Zhu et al, unpublished observations 1997). Such enhancing effects may be explained by the results of synergistic effects between transcriptional factors induced by LDL or cytokines or by the results of LDL-changed membrane fluidity and altered cytokine receptors. Thus, LDL may initiate a variety of signal transduction pathways through affecting signaling at the level of the cell surface.

Collectively, the results of this and previous studies allow us to propose a pathway for EC activation by LDL, as shown in Fig 8. Exposure to LDL perturbs the cell membrane or alters receptor conformation and activates multiple kinase activities, such as the MAPK family and JNKs. Activation of JNK or other pathways activates preexisting c-Jun, which forms c-Jun homodimers to induce its own expression via the AP-1 sites on the promoter. This positive autoregulation of c-Jun can accumulate c-Jun by transcription and translation to increase total c-Jun protein. The accumulations of c-Jun, in the form of dimers with c-Jun or other subunits, further form more AP-1. The consistently elevated level of AP-1 further induces other specific cellular genes that are regulated by AP-1, such as ICAM-1, VCAM-1, and MCP-1, leading to EC activation.

View larger version (23K): [in this window] [in a new window]   Figure 8. Schematic representation of the signal pathway in EC activation by LDL.

In conclusion, LDL appears to be an endothelial agonist distinct from other cell stimulators such as cytokines, endotoxin, oxidized LDL, and phorbol esters since it activates HUVECs predominantly through transcription factor AP-1. LDL increases AP-1 via two steps: by activating preexisting c-Jun protein and stimulating c-Jun transcription. The increase in AP-1 may be due to the formation of c-Jun homodimers or heterodimers of Jun and transcription factors other than c-Fos. Finally, multiple kinase activities are involved in the mechanism of LDL-induced AP-1 activation in HUVECs. The c-Jun positive autoregulation initiated by LDL may be an important mechanism helping to explain LDL's chronic and long-lasting effects in EC activation and atherogenesis.

	Selected Abbreviations and Acronyms

 

AP-1 = activator protein 1 ATF = activating transcription factor CAT = chloramphenicol acetyltransferase CREB = cAMP-responsive element binding protein FBS = fetal bovine serum ß-gal = ß-galactosidase (HUV)EC = (human umbilical vein) endothelial cell ICAM-1 = intercellular adhesion molecule 1 JNK = Jun N-terminal kinase luc = luciferase MAPK(K) mitogen-activated protein kinase (kinase) NF-B = nuclear factor-B PK = protein kinase PMA = phorbol 12-myristate 13-acetate PTK = protein tyrosine kinase RSV = Rous sarcoma virus TNF- = tumor necrosis factor- VCAM-1 = vascular cell adhesion molecule 1

	Acknowledgments

 
This study was supported in part by the American Heart Association, New York State Affiliate Grants 94–213 (to Y.Z.) and 92–027GB (to J.H.L.) and National Institutes of Health grant HL43023 (to M.B.S.). We thank Drs T. Curran, L. Ransone, M. Meyer, and T. Parks for plasmids. We also thank Dr J.A. Thompson for recombinant human fibroblast growth factor. We also would like to thank Arshad Khan for editing this manuscript.

	Footnotes

 
Presented in part at the 69th American Heart Association Scientific Sessions, New Orleans, La, November 10–13, 1996, and published in abstract form in Circulation. 1996;94(suppl I):I-101.

Correspondence to Yi Zhu, MD, Division of Biomedical Sciences, University of California, Riverside, CA 92521.

Received February 4, 1997; accepted November 24, 1997.

	References

Top Abstract Introduction Methods Results Discussion References

 

1. Ross R. The pathogenesis of atherosclerosis: a perspective for the 1990s. Nature. 1993;362:801–809.[Medline] [Order article via Infotrieve]
2. Grundy SM. Role of low-density lipoproteins in atherogenesis and development of coronary heart disease. Clin Chem. 1995;41:139–146.[Abstract/Free Full Text]
3. Steinberg D, Witztum JL. Lipoprotein and atherogenesis. JAMA. 1990;264:3047–3052.[Abstract]
4. Luscher TF, Tanner FC, Tschudi MR, Noll G. Endothelial dysfunction in coronary artery disease. Annu Rev Med. 1993;44:395–418.[Medline] [Order article via Infotrieve]
5. Pritchard KA Jr, Groszek L, Smally DM, Sessa WC, Wu M, Villalon P, Wolin MS, Stemerman MB. Native low-density lipoprotein increases endothelial cell nitric oxide synthase generation of superoxide anion. Circ Res. 1995;77:510–518.[Abstract/Free Full Text]
6. Holland JA, Pritchard KA Jr, Rogers NJ, Stemerman MB. Perturbation of cultured human endothelial cells by atherogenic levels of low density lipoprotein. Am J Pathol. 1988;132:474–478.[Abstract]
7. Pritchard KA Jr, Wang PY-K, Stemerman MB. Atherogenic concentrations of low-density lipoprotein enhance endothelial cell generation of epoxyeicosatrienoic acid products. Am J Pathol. 1990;136:1383–1391.[Abstract]
8. Holland JA, Pritchard KA Jr, Rogers NJ, Stemerman MB. Atherogenic levels of low-density lipoprotein increase endocytotic activity in cultured human endothelial cells. Am J Pathol. 1992;140:551–558.[Abstract]
9. Pritchard KA Jr, Tota KR, Lin JH-C, Danishefsky KJ, Kurilla BA, Holland JA, Stemerman MB. Native LDL: EC recruitment of mononuclear cells. Arterioscler Thromb. 1991;11:1175–1181.[Abstract/Free Full Text]
10. Smalley DM, Lin JH, Italiano ML, Kobari Y, Stemerman MB, Pritchard KA Jr. Native low density lipoprotein increases endothelial cell adhesiveness by inducing ICAM-1. Arterioscler Thromb Vasc Biol. 1996;16:585–590.[Abstract/Free Full Text]
11. Lin JH-C, Zhu Y, Liao H-L, Groszek L, Stemerman M B. Native low density lipoprotein enhances the induction of vascular cell adhesion molecule-1 (VCAM-1) in human endothelial cells. Atherosclerosis. 1996;127:185–194.[Medline] [Order article via Infotrieve]
12. 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.[Abstract/Free Full Text]
13. Voraberger G, Schafer R, Stratowa C. Cloning of the human gene for intercellular adhesion molecule 1 and analysis of its 5' regulatory region. J Immunol. 1991;147:2777–2786.[Abstract/Free Full Text]
14. Degitz K, Lian-lie L, Caughman SW. Cloning and characterization of the 5' transcriptional regulatory region of the human intercellular adhesion molecule 1 gene. Biol Chem. 1991;266:14024–14030.[Abstract/Free Full Text]
15. Shyy JY, Lin MC, Han J ,Lu Y, Petrime M, Chien S. The cis-acting phorbol ester "12-O-tetradecanoylphorbol 13-acetate"-responsive element is involved in shear stress-induced monocyte chemotactic protein 1 gene expression. Proc Natl Acad Sci U S A. 1995;92:8069–8073.[Abstract/Free Full Text]
16. Roebuck AK, Rahman A, Lakshminarayanan V, Janakidevi K, Malik AB. H₂O₂ and tumor necrosis factor- activate intercellular adhesion molecule 1 gene transcription through distinct cis-regulatory elements within the ICAM-1 promoter. J Biol Chem. 1995;270:18966–18974.[Abstract/Free Full Text]
17. Ledebur HC, Parks TP. Transcriptional regulation of the ICAM-1 gene by inflammatory cytokines in human endothelial cells. Biol Chem. 1995;270:933–943.[Abstract/Free Full Text]
18. Lee W, Haslinger A, Karin M, Tjian R. Activation of transcription by two factors that bind promoter and enhancer sequences of the human metallothionein gene and SV40. Nature. 1987;325:368–372.[Medline] [Order article via Infotrieve]
19. Angel P, Karin M. The role of Jun, Fos and the AP-1 complex in cell-proliferation and transformation. Biochim Biophys Acta. 1991;1072:129–157.[Medline] [Order article via Infotrieve]
20. Angel PE, Herrlich PA. Characteristics of AP-1 subunits. In: Angel PE, Herrlich PA, eds. The Fos and Jun Families of Transcription Factors. Boca Raton, Fla: CRC Press Inc; 1994:3–14.
21. Ausubel FM, Brent G, Kingston RE, Moore DD, Seidman JG, Smith JA, Struhl K. Current Protocols in Molecular Biology. 1996;4:4.10.1–4.10.9.
22. Meyer M, Schreck R, Baeuerle PA. H₂O₂ and antioxidants have opposite effects on activation of NF- B and AP-1 in intact cells: AP-1 as secondary antioxidant-responsive factor. EMBO J. 1993;12:2005–2015.[Medline] [Order article via Infotrieve]
23. Lan Q, Mercurius KO, Davies PF. Stimulation of transcription factor NF-B and AP-1 in endothelial cells subjected to shear stress. Biochem Biophys Res Commun. 1994;201:950–956.[Medline] [Order article via Infotrieve]
24. Angel A, Hazue K, Smeal T, Karin M. The jun proto-oncogene is positively autoregulated by its product, Jun/AP-1. Cell. 1988;55:875–885.[Medline] [Order article via Infotrieve]
25. van Dam H, Duyndam M, Rotter R, Bosch A, de Vries-Smits L, Herrlich P, Zantema A, Angel P, van der Eb AJ. Heterodimer formation of c-Jun and ATF-2 is responsible for induction of c-Jun by 243 amino acid adenovirus E1A protein. EMBO J. 1993;12:479–487.[Medline] [Order article via Infotrieve]
26. Herr I, van Dam H, Angel P. Binding of promoter associated AP-1 is not altered during induction and subsequent repression of the c-Jun promoter by TPA and UV irradiation. Carcinogenesis. 1994;15:1105–1113.[Abstract/Free Full Text]
27. Mechta F, Yaniv M. Structure and regulation of the c-Jun promoter. In: Angel PE, Herrlich PA, eds. The Fos and Jun Families of Transcription Factors. Boca Raton, Fla: CRC Press Inc; 1994:115–129.
28. Stewart AF, Herrera RE, Nordheim A. Rapid induction of c-Fos transcription reveals quantitative linkage of RNA polymerase II and DNA topoisomerase I enzyme activities. Cell. 1990;60:141–149.[Medline] [Order article via Infotrieve]
29. Pritchard KA Jr, Schwarz SM, Medow MS, Stemerman MB. Effect of LDL on EC membrane fluidity and mononuclear cell attachment. Am J Physiol. 1991;260:C43–C49.[Abstract/Free Full Text]
30. Collins T, Read M, Neish A, Whitley DT, Maniatis T. Transcriptional regulation of endothelial cell adhesion molecules: NF-kB and cytokine-inducible enhancers. FASEB J. 1995;9:899–909.[Abstract]
31. Muller S, Kammerbauer C, Simmons U, Shibagaki N, Li L-J, Caughman SW, Degitz K. Transcriptional regulation of intercellular adhesion molecule-1: PMA-induction is mediated by NF-kB. J Invest Dermatol. 1995;104:970–975.[Medline] [Order article via Infotrieve]
32. Muller JM, Ziegler HW, Baeuerle PA. Nuclear factor kappa B, a mediator of lipopolysaccharide effects. Immunobiology. 1993;187:233–256.[Medline] [Order article via Infotrieve]
33. Liao F, Andalibi A, deBeer FC, Fogelman AM, Lusis AJ. Genetic control of inflammatory gene induction and NF-B-like transcription factor activation in response to an atherogenic diet in mice. J Clin Invest. 1993;91:2572–2579.
34. Berliner JA, Navab M, Fogelman AM, Frank JS, Demer LL, Edwards PA, Watson AD, Lusis AJ. Atherosclerosis: basic mechanisms. Oxidation, inflammation, and genetics. Circulation. 1995;91:2488–2496.[Abstract/Free Full Text]
35. Watson AD, Navab M, Hama SY, Sevanian A, Prescott SM, Stafforini DM, McIntyre TM, La Du BN, Fogelman AM, Berliner JA. Effect of platelet activating factor-acetylhydrolase on the formation and action of minimally oxidized-low density lipoprotein. J Clin Invest. 1995;95:774–782.
36. Zhu Y, Lin JH-C, Liao H-L, Verna L, Stemerman MB. Activation of ICAM-1 promoter by lysophosphatidylcholine: possible involvement of protein tyrosine kinases. Biochim Biophys Acta. 1997;1345:93–98.[Medline] [Order article via Infotrieve]
37. Derijard B, Hibi M, Wu I-H, Barrett T, Su B, Deng T, Karin M, Davis RJ. JNK1: A protein kinase stimulated by UV light and Ha-Ras that binds and phosphorylates the c-Jun activation domain. Cell. 1994;76:1025–1037.[Medline] [Order article via Infotrieve]
38. Kyriakis JM, Banerjee P, Nikolakaki E, Dai T, Rubie EA, Ahmad MF, Avruch J, Woodgett JR. The stress-activated protein kinase subfamily of c-Jun kinases. Nature. 1994;369:156–160.[Medline] [Order article via Infotrieve]
39. Sanchez I, Hughes RT, Mayer BJ, Yee K, Woodgett JR, Avruch J, Kyriakis JM, Zon LI. Role of SAPK/ERK kinase-1 in the stress-activated pathway regulating transcription factor c-Jun. Nature. 1994;372:794–798.[Medline] [Order article via Infotrieve]
40. Liu ZY, Ganju RK, Wang JF, Ona MA, Hatch WC, Zheng T, Avraham S, Gill P, Groopman JE. Cytokine signaling through the novel tyrosine kinase RAFTK in Kaposi's sarcoma cells. J Clin Invest. 1997;99:1798–1804.[Medline] [Order article via Infotrieve]
41. Kolch W, Heidecker G., Kochs G, Hummel R, Vahidi H, Mischak H, Finkenzeller G, Marme D, Rapp UR. Protein kinase C alpha activates RAF-1 by direct phosphorylation. Nature. 1993;364:249–252.[Medline] [Order article via Infotrieve]
42. Sozeri O, Vollmer K, Liyanage M, Frith D, Kour G, Mark GE III, Stabel S. Activation of the c-Raf protein kinase by protein kinase C phosphorylation. Oncogene. 1992;7:2259–2262.[Medline] [Order article via Infotrieve]
43. Liu ZG, Hsu H, Goeddel DV, Karin M. Dissection of TNF receptor 1 effector functions: JNK activation is not linked to apoptosis while NF-kappaB activation prevents cell death. Cell. 1996;87:565–576.[Medline] [Order article via Infotrieve]
44. Rosette C, Karin M. Ultraviolet light and osmotic stress: activation of the JNK cascade through multiple growth factor and cytokine receptors. Science. 1996;274:1194–1197.[Abstract/Free Full Text]