doi: /01.ATV

(1)

Martin Braddock

John H. McVey, Nigel Mackman, Jason M. Reese, Daniel G. Gorman, Callum Campbell and

Parul Houston, Marion C. Dickson, Valerie Ludbrook, Brian White, Jean-Luc Schwachtgen,

Mediated by Egr-1

Fluid Shear Stress Induction of the Tissue Factor Promoter In Vitro and In Vivo Is

Print ISSN: 1079-5642. Online ISSN: 1524-4636

is published by the American Heart Association, 7272 Arteriosclerosis, Thrombosis, and Vascular Biology

doi: 10.1161/01.ATV.19.2.281

1999;19:281-289

Arterioscler Thromb Vasc Biol.

http://atvb.ahajournals.org/content/19/2/281

World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://atvb.ahajournals.org//subscriptions/ at:

is online Arteriosclerosis, Thrombosis, and Vascular Biology

Information about subscribing to

Subscriptions:

http://www.lww.com/reprints

Information about reprints can be found online at:

Reprints:

document. Answer

Permissions and Rights Question and under Services. Further information about this process is available in the

permission is being requested is located, click Request Permissions in the middle column of the Web page which Copyright Clearance Center, not the Editorial Office. Once the online version of the published article for

can be obtained via RightsLink, a service of the Arteriosclerosis, Thrombosis, and Vascular Biology

in

Requests for permissions to reproduce figures, tables, or portions of articles originally published

Permissions:

by guest on February 27, 2014

http://atvb.ahajournals.org/

Downloaded from http://atvb.ahajournals.org/ by guest on February 27, 2014 Downloaded from

(2)

In Vitro and In Vivo Is Mediated by Egr-1

Parul Houston, Marion C. Dickson, Valerie Ludbrook, Brian White, Jean-Luc Schwachtgen,

John H. McVey, Nigel Mackman, Jason M. Reese, Daniel G. Gorman,

Callum Campbell, Martin Braddock

Abstract—Hemodynamic forces such as fluid shear stress have been shown to modulate the activity of an expanding family

of genes involved in vessel wall homeostasis and the pathogenesis of vascular disease. We have investigated the effect of shear stress on tissue factor (TF) gene expression in human endothelial cells (ECs) and in a rat arterial model of occlusion. As measured by reverse transcriptase polymerase chain reaction, exposure of ECs to 1.5 N/m2

shear stress resulted in a time-dependent induction of endogenous TF transcripts of over 5-fold. Transient transfection of TF promoter mutants into cultured ECs suggests the involvement of the transcription factor Egr-1 in mediating the response of the TF promoter to shear stress. To address the importance of flow induction of Egr-1 in vivo, we have established a flow-restricted rat arterial model and determined the level of expressed Egr-1 and TF at the site of restricted flow using immunohistochemistry. We report an increase in the level of Egr-1 and TF protein in ECs expressed at the site of restricted flow. Elevated expression of Egr-1 and TF is restricted to a highly localized area, as evidenced by the fact that no significant increase in level can be detected at arterial sites distal to the site of occlusion. These findings suggest a direct role for Egr-1 in flow-mediated induction of TF and further substantiate the importance of shear stress as a modulator of vascular endothelial gene function in vivo. (Arterioscler Thromb Vasc Biol. 1999;19:281-289.)

Key Words: fluid shear stressntissue factor promoternEgr-1 transcription factornvascular endothelial cells

E

ndothelial cells (ECs) lining the inner surface of blood vessels are constantly exposed to blood flow. Locally disturbed flow at arterial curvatures or bifurcations is character-ized by both low and high oscillatory wall shear stresses that are conveyed to the cell as both a change in pressure and a change in the stretch capacity of the EC lining.1–3_{As a consequence of}

the local perturbation in laminar shear stress, a number of genes have been identified whose promoter activity is positively or negatively regulated by shear stress (reviewed in References 4 to 8). Several cis-acting elements have been implicated in mediat-ing shear modulation of gene expression, and the term shear stress responsive element (SSRE) was first proposed after the identification of the sequence GAGACC.9 _{This SSRE was}

demonstrated to confer shear stress responsiveness on the platelet-derived growth factor (PDGF) B promoter,9_a

heterolo-gous SV40 promoter,10_{and has since been located in a number}

of other promoters that have been shown to be shear stress responsive in ECs.11 _{The GAGACC sequence binds nuclear}

factorkB (NF-kB) p50 –p65 heterodimers, and consistent with this observation, the wild-type HIV-1 LTR, but not an LTR lacking the NF-kB binding site, was also shown to be responsive

to shear stress.5_{Other transcription factors that have been shown}

to mediate shear stress activation of a promoter are AP1,12,13

Sp1,14_{and Egr-1.}15

Recently, the transcription factor Egr-1 has been suggested to play a role as a common factor in vascular injury16 trans-activating the expression of a number of genes in the

vascular endothelium.17_{Fluid shear stress has been shown to}

stimulate the production of Egr-1 RNA15_{and protein,}18_and

upregulation of the PDGF A proximal promoter by shear stress involves displacement of prebound Sp1 by Egr-1.15

Previous studies have reported increased activity of tissue factor (TF) under flow conditions, as measured by the conversion of factor X to factor Xa.19 –21_{We now report the}

identification of a minimal TF promoter that is shear stress responsive and relies on the cellular transcription factor Egr-1 for shear stress–mediated trans-activation. We show that Egr-1 and TF are both upregulated at the site of arterial stenosis in vivo. Thus, TF is a further gene to be modified in this manner by biomechanical forces, adding more evidence for the importance of Egr-1 as a fluid shear stress–responsive transcription factor.

Received March 26, 1998; revision accepted June 23, 1998.

From the Endothelial Cell Gene Expression Group, Vascular Disease Unit, Glaxo-Wellcome Medicines Research Centre, Stevenage (P.H., M.C.D., V.L., B.W., J.-L.S., C.C., M.B.), and the Haemostasis Research Group, MRC Clinical Sciences Centre, London (J.H.M.), UK; the Department of Immunology, The Scripps Research Institute, La Jolla, California (N.M.); and the Department of Engineering, University of Aberdeen, Scotland (J.M.R., D.G.G.).

Correspondence to Martin Braddock, Endothelial Cell Gene Expression Group, Vascular Disease Unit, Glaxo-Wellcome Medicines Research Centre, Gunnels Wood Road, Stevenage, Herts SG1 2NY, England. E-mail [email protected]

Arterioscler Thromb Vasc Biol. is available at http://www.atvbaha.org

281 by guest on February 27, 2014

(3)

Methods

EC Culture

Human aortic ECs (HAECs) and human umbilical vein ECs (HUVECs) were purchased from Clonetics, frozen in liquid nitrogen. Cells were thawed and maintained in endothelial growth medium (EGM-2) supplemented with 2% fetal calf serum, vascular endothe-lial growth factor (2 ng/mL), recombinant insulin-like growth factor-1 (5 ng/mL), ascorbic acid (75mg/mL), human epidermal growth factor (10 ng/mL), gentamycin (50mg/mL), amphotericin B (50mg/mL), heparin (0.5 ng/mL), human fibroblast growth factor (2 ng/mL), and hydrocortisone (500 ng/mL). The cells were passaged using trypsin/EDTA (Life Sciences) and maintained in a humidified atmosphere of 5% CO2at 37°C.

Cell Transfection and Shear Stress Conditions

Cells were transfected by electroporation using an EasyJect electro-porator (Flowgen). Cells were used up to a maximum of passage 5 for all transfection experiments. Briefly, cells were grown to con-fluence, passaged, and electroporated using 23106_{cells/mL per data} point. DNA was prepared by standard alkaline lysis and column purification (Qiagen). Reporter DNA (10 mg) was transfected, together with 1mg of human cytomegalovirus (CMV) promoter-b (Promega), which functioned as a normalizing plasmid for transfec-tion efficiency. Protein extracts containing an equal number of

b-galactosidase units were assayed for chloramphenicol acetyltrans-ferase (CAT) or luciacetyltrans-ferase activity. For all transfections, 30mg of pBluescript (Stratagene) was used as an inert carrier DNA. In

trans-activation experiments, the total amount of DNA for all

experiments was constant and made up to a total of 30mg DNA with pBluescript. Cells were electroporated at 350 V as described,22_with the modification that the cells were not synchronized. After electro-poration, 5 mL of complete EGM-2 medium was added and cells were plated out into slide flaskettes (Nunc) in duplicate. Laminar shear stress was applied to transfected cells at 1.5 N/m2_{for 1 or 3} hours using parallel-plate flow chambers as described,4_{set in series} in a closed circulating system gassed with 5% CO2and maintained at 37°C. Parallel-plate flow chambers were custom made at the Glaxo-Wellcome Department of Bioengineering. The wall shear stress Tw, expressed in newtons per square meter was calculated using the standard equation Tw53yQ/2a2b,23where y is the viscosity at 37°C in newton seconds per square meter, Q is the volumetric flow rate in milliliters per second, a is the half-channel height in centimeters, and b is the channel width in centimeters. Laminar shear stress was generated by the use of a Masterflex 7518 – 60 peristaltic pump and was shown to be linear within the range of 0.4 to 12.0 N/m2_.

CAT, Luciferase, andb-Galactosidase Assays

Qualitatively identical results were obtained in both HAECs and HUVECs. After shear stress stimulation, cells were washed in PBS and lysed in reporter lysis buffer (Promega) for assay of luciferase andb-galactosidase activity. For CAT assay, cells were disrupted in TEN buffer (10 mmol/L Tris-HCl, pH8; 1 mmol/L EDTA; 100 mmol/L NaCl) using a sonicator (MSE). CAT activity was measured using the Quan-T-CAT system as described24 (Amer-sham). Luciferase activity was measured using a luminometer (Wallac) and b-galactosidase activity assayed spectrophotometri-cally in the presence of the appropriate substrate (Promega). CAT and luciferase values were normalized tob-galactosidase levels, and each data point represents the average of at least 3 separate experiments.

Semiquantitative Polymerase Chain Reaction of Endogenous TF

RNA was extracted, using a total RNeasy extraction kit (Qiagen), from 23105_{cells, which were either static controls or subjected to 1} or 3 hours of laminar flow at 1.5 N/m2_{. Semiquantitative reverse} transcriptase polymerase chain reacton (RT-PCR) was performed as described.25 _{cDNA was synthesized from 500 ng total RNA by} reverse transcription. The reaction was carried out using MgCl2 (5 mmol/L), 53RT buffer (Promega), dNTPs (1 mmol/L; Promega), RNase inhibitor (20 U; Promega), oligo dT (500 ng; Boehringer

Mannheim), and 12 U AMV reverse transcriptase (Promega) in a total volume of 20 mL. The reaction was incubated at 42°C for 30 minutes and the enzymes were denatured for 10 minutes at 95°C. The cDNA reaction product (3 mL) was added to the following: MgCl2(1.5 mmol/L), 103buffer (Perkin-Elmer), dNTPs (200mmol/L), Taq polymerase (2.5 U; Perkin-Elmer) with 20 pmol of TF primers (forward 59TTA CCT GGA GAC AAA CCT CGG 39; from nucleotide (nt) 501 to 521 [relative to the TF ATG at11] and reverse 59ATG ACC ACA AAT ACC ACA GCT 39; from nt 890 to 870) and GAPDH primers (Clontech; forward 59TGA GTA CGT CGT GGA GTC CAC 39; from nt 71 to 96 [relative to the RNA start site at11] and reverse 59ACC AGG AAA TGA GCT TGA CA 39; from nt 1030 to 1053). The reaction products were denatured for 5 minutes at 95°C followed by 26 cycles of 95°C for 1 minute, 55°C for 1 minute, and 72°C for 1 minute and a final elongation at 72°C for 10 minutes. The PCR products were detected by staining a 1.8% agarose gel with SYBR green (Molecular Probes). Quantitation was carried out using a Storm system and ImageQuant software (Molec-ular Dynamics).

Plasmid Constructions

TF promoter constructs have been described previously.26,27_Results are expressed as shear stress activation above the basal activity of the promoter construct. Plasmids expressing Egr-1 as wild type (p930) or mutant (1.5 zinc fingers deleted, p858) driven by the CMV promoter were a gift from Dr Vikas Sukhatme, Harvard Medical School, Boston, Mass. Plasmid pSP13, harboring an Egr-1 binding site and a TATA box, was constructed using an oligonucleotide with the sequence 59 GATCTGAATTCGCCCTAGCTCGCGGGGG-CGTCGCACTCTCTATAAAGGCCGGAACAGCTGAAAGA 39, which was cloned into BglII/HindIII– digested pGL3 basic luciferase reporter vector (Promega).

Electrophoretic Mobility Shift Assay and Immunoblotting

Electrophoretic mobility shift assays (EMSAs) were carried out using purified proteins commercially available (Promega and Santa Cruz) with an EMSA gel-shift kit (Promega). The oligonucleotide used for EMSA experiments represents positions252 nt to289 nt in the TF promoter and has the sequence 59 GGAGCG-GCGGGGGCGGGCGCCGGGGGCGGGCAGAGG 39. The Egr-1 site extends from274 nt to266 nt and the Sp1 site from270 nt to

261 nt.32_{P-labeled DNA probes were generated using T4} polynu-cleotide kinase, and unincorporated radiolabeled nupolynu-cleotide was removed as recommended by the manufacturer (Qiagen). Binding reactions were carried out at 20°C for 10 minutes using labeled probe (105_{cpm per reaction, 10}8_cpm/_m_{g) in the presence or absence of} Egr-1 and/or Sp1 binding activity (30 ng per reaction) at the ratios described. The products of the binding reactions were separated on 6% acrylamide retardation gels (Novex) and run in 0.53 Tris-borate–EDTA buffer at 100 V for 90 minutes. Dried-down gels were exposed to autoradiography using Hyperfilm (Amersham).

The specificity of the TF antibody was determined using cell extracts of peripheral blood mononuclear cells (106 _{cells), which} were either unstimulated or stimulated with lipopolysaccharide (LPS; 100 ng/mL) for 6 hours (a kind gift from K. Buchan). Cells were disrupted by sonication, and protein extracts were run out on an 8% SDS-PAGE gel against MultiMark protein standards (Novex) and transferred to nitrocellulose by immunoblotting. Protein bands were detected using a horseradish peroxidase– conjugated second antibody and enhanced chemiluminescense detection (ECL; Amersham).

Rat Arterial Occlusion and Determination of Flow

Male Han Wistar rats (300 to 400 g) were anesthetized with inactin (120 mg/kg IP). After anesthesia, the trachea was cannulated and the animals were spontaneously allowed to respire room air. Body temperature was monitored by a rectal probe and was maintained at 37°C by the homeothermic blanket system (Harvard). The left jugular vein and left carotid artery were exposed by blunt dissection, and the left jugular vein was cannulated for intravenous administra-tion of supplementary anesthetic. The left carotid artery was

instru-282 Fluid Shear Stress Activation of TF by Egr-1

(4)

mented from proximal to distal with an ultrasonic flow probe (Transonic Systems Inc; 0.5 mm) and an adjustable mechanical occluder. The mechanical occluder was constructed by the Glaxo-Wellcome Department of Bioengineering following the description of a wire occluder.28_{The flow probe was connected to a flowmeter} (Transonic model T206) for continuous display of phasic carotid artery blood flow. Blood flow was displayed on a thermal array recorder (Gould TA4000) linked to an MI2_{data acquisition system} (Modular Instruments Inc.) After an equilibration period of at least 15 minutes to establish stable carotid artery blood flow, a stenosis was applied by gradually adjusting the mechanical occluder. The degree of occlusion was regulated, and the final stenosis resulted in approximately a 25%, 50%, or.75% reduction in mean carotid blood flow. Data were collected from occlusion experiments that generated approximately a 50% stenosis in 7 animals. The stenosis was maintained for 60 minutes, and the carotid artery was excised and placed in formalin ready for immunohistochemical analysis. The diameter of the left carotid artery was determined before and after the application of stenosis and expressed as percentage stenosis. The degree of stenosis in terms of decreased carotid blood flow was determined from the mean blood flow data and expressed as percentage decrease from the preocclusion mean value, together with the shear stress value over the stenosis for each of 7 separate experiments. Confirmation that flow regimes administered in vitro and in vivo represented laminar and not turbulent shear stress was obtained using standard computational fluid dynamics equations.29 The Reynolds number ReD, was calculated by the equation

ReD54pQ/pmD, where p is the fluid density (in kilograms per cubic

meter), Q is the volumetric flow rate (in cubic meters per second),m is the fluid dynamic viscosity (in newton seconds per square meter), and D is the hydraulic diameter of the vessel29_{(in meters). For the} arterial experiments, D is simply the diameter of the artery. The flow chamber used for in vitro flow experiments has a rectangular cross-section, so the appropriate hydraulic diameter D is 2ab/(a1b), where a and b are the width and height of the channel, respectively (in meters). Applying these equations to the two cases under consideration, the minimum Reynolds number obtained was 16 for the arterial occlusion experiments and 1240 for the in vitro flow experiments. Both values for ReD are within the critical Reynolds

number for flow in vessels of 2300, which indicates the onset of turbulent flow conditions, assuming that the maximal wall shear stress occurs in the throat of the stenosis. The shear stress was calculated at the stenosis, assuming that as the blood flow passes over the stenosis, the momentum boundary does not separate from the wall, a reasonable assumption for a smooth stenosis such as generated by our occluder. In addition, we assume that the flow is fully developed within the occluded artery. Given these assumptions, for laminar flow in an artery, the average wall shear stress Twin the artery is Tw50.5pU

2_{f, where U is the mean velocity in the arterial} occlusion, f is the friction factor, and p is the fluid density (in kilograms/cubic meter).29

Immunohistochemistry

Arterial sections were embedded in paraffin and dewaxed as de-scribed.30_{Sections were blocked for 20 minutes in 1% hydrogen} peroxide and then rinsed in PBS. Anti–Egr-1 antibody (100 mL; Santa Cruz), rabbit anti-human TF143antibody (J. McVey; 143 refers to amino acids 1 through 143), or rabbit anti-human von Willebrand factor (vWF) antibody (Dako) was diluted 1:100 in 0.5% BSA/PBS, applied to the slides, and left for 45 minutes. Slides were washed in 0.5% Tween 80/PBS and then in PBS. Biotinylated anti-rabbit secondary antibody (Vector) diluted in 0.5% BSA, 0.8% NaCl/PBS was applied to slides for 30 minutes. Avidin-biotin complex (ABC; Vector laboratories) was diluted in 0.5% BSA with a 1:100 dilution of avidin and biotin and 0.8% NaCl. This mix was left to stand for 30 minutes, and 100mL was then added to each slide and incubated for 30 minutes. Slides were washed in 0.5% Tween 80/PBS and then rinsed with PBS. Detection of antigen-antibody complexes was carried out with the diaminobenzidine detection system (Sigma) using a 5-minute detection time. Slides were then counterstained in Meyer’s hematoxylin for 10 seconds, dehydrated, cleared, and mounted for microscopy.

Results

Endogenous TF Expression is Upregulated by Shear Stress

Laminar shear stress has been shown to activate the expres-sion of a number of genes involved in vessel wall homeosta-sis and vascular injury.4 – 8_{We reasoned that a change in shear}

stress may activate the expression of TF. To address this, we subjected HAECs to shear stress of 1.5 N/m2

for 1 and 3 hours and compared the level of TF mRNA transcripts produced in comparison to ECs that had not received shear stress stimu-lation. The level of TF mRNA transcripts was measured using RT-PCR23_{and compared with the level of GAPDH mRNA,}

which served as an internal control. As shown in Figure 1, endogenous TF mRNA levels increased after only 1 hour to 3.6-fold that of the static control and further increased to 5.2-fold that of the static control after 3 hours of shear stress. The basal level of TF mRNA was barely detectable under zero flow conditions, in agreement with previous observations.19 –21

Identification of a Novel SSRE in the TF Promoter

TF promoter deletion constructs are described in Figure 2A. For transfection experiments, 10 mg of promoter reporter construct was transfected into HAECs by electroporation and cells were seeded into slide flaskettes and subjected to zero flow or flow for 3 hours at 1.5 N/m2

at 18 hours posttrans-fection (Figure 2B). Previous studies have shown that NF-kB and AP1 are required for response of the TF promoter to LPS stimulation in monocytic cells and ECs.31 _{Plasmid p526,}

which contains 1.373 kb of promoter sequence, including NF-kB and AP1 binding sites, was transfected in HAECs and reporter assays were carried out on cell lysates from cultures harvested in the absence or presence of flow. We observed a stimulation of promoter activity by flow of approximately 8-fold (Figure 2B). Plasmid p529, which contains 383 nt of promoter sequence, showed similar activation in response to Figure 1.Shear stress activation of transcription of the endoge-nous TF gene in HAECs. RT-PCR showing shear stress upregu-lation of TF transcription with time at 1.5 N/m.2_GAPDH stan-dards were used as an internal control. Data shown are representative of at least 3 independent experiments. M repre-sents molecular weight markers shown in base pairs.

(5)

flow. Plasmid construct p543 contains 111 nt of promoter sequence and harbors binding sites for Sp1 and Egr-1 transcription factors but does not contain a functional NF-kB or AP1 binding site. This promoter fragment was also shear responsive to the same extent as plasmids p526 and p529. This result suggests that the response of the TF promoter resides within a 111-bp fragment and may be functioning via the Sp1 and Egr-1 transcription factors.

Previous studies have identified several classes of shear stress–responsive genes whose response to shear is either rapid and transient, rapid and longer lived, or slow and sustained.4 – 8 _{To characterize the response of the TF}

pro-moter, we sought to determine the minimum time of exposure to shear that would result in the largest increase in promoter activity. To do this, we transfected HAECs with plasmid p526 and 18 hours after transfection applied shear stress for time intervals of up to 24 hours, removing samples at the time points indicated. The results shown in Figure 3 show a peak

of promoter activity at 4 hours after the application of stimulus, with activity gradually declining to basal levels at 24 hours, in good agreement with a previous study.14_{We then}

asked whether this activity was correlated with the level of Egr-1 protein produced under shear. We have recently shown a rapid induction of Egr-1 protein after only 15 minutes’ exposure to shear, which was sustained over a 7-hour period and rapidly declined to a low but significant level after 18 hours under flow conditions.18 _{These data suggest that the}

response of the TF promoter to shear stress is rapid and transient and parallels the shear induction of Egr-1, which is likewise rapidly induced by shear stress.

The sequence of the minimal shear-responsive element in the TF promoter contains overlapping Sp1/Egr-1 transcrip-tion factor binding sites distinct from 2 upstream Sp1 binding sites. We reasoned that shear stress activation of this TF promoter construct may be mediated via this Sp1/Egr-1 interaction and carried out EMSAs and mutational analysis of this region to determine the role of these factors. Deletion of the two distal Sp1 sites (compare p529 with p543; Figure 2B.) had little effect on the ability of p543 to be activated by shear stress, presumably because the basal level of transcription could be maintained by binding of Sp1 to its proximal cognate binding sites, as has been reported.32_{Using EMSA,}

we asked whether Sp1 and Egr-1 could bind together or whether the binding of one transcription factor could displace the binding of the other factor at an overlapping Sp1/Egr-1 binding site. We prepared a32

P-labeled oligonucleotide probe spanning the overlapping Sp1/Egr-1 binding site and incu-bated the probe with either Sp1 or Egr-1. Next, we attempted to compete off Sp1 with Egr-1 and Egr-1 with Sp1 over a range excess of proteins (Figure 4). Prebound Sp1 was readily displaced by an equimolar quantity of Egr-1, whereas pre-bound Egr-1 required a 30-fold excess of Sp1 for complete displacement. Under no conditions did we detect binding of both transcription factors to the same site, arguing that there may be circumstances in ECs that favor the binding of one factor over the other. These results are consistent with previous findings16,26_{and suggest that TF trans-activation by}

shear stress involves displacement of Sp1 by Egr-1.

Figure 2.A, Diagram of 59deletion mutants of the TF promoter. Promoter coordinates are indicated relative to the transcription start site at11. Transcription factor binding sites within the TF promoter are indicated. Constructs p526, p529, p533, and p543 have been described.27_{B, Response of 5}9_{deletion mutants of}

the TF promoter to shear stress. HAECs were transfected with promoter reporter plasmids, as indicated in Methods, and cells flowed at 1.5 N/m2_{for 3 hours. The basal level of promoter}

activities is shown, together with the activated levels. The shear activation ratio is derived after normalization of reporter gene activity by cotransfection with a CMVb-galactosidase plasmid. Results shown are the average of at least 3 separate

experiments.

Figure 3.Time course of the activation of the TF promoter by shear stress. Plasmid p526 was transfected into HAECs as described in Figure 4. Reporter gene activity was determined at the time points indicated in the figure.

284 Fluid Shear Stress Activation of TF by Egr-1

(6)

A Mutation in the Egr-1 Binding Site in the TF Promoter Abolishes the Response to Shear Stress

To define more precisely the role of Sp1 and Egr-1 in shear stress trans-activation of the TF promoter, we analyzed the effect of mutations in the overlapping Sp1/Egr-1 binding site that were known to abolish binding of Sp1 and Egr-1 in vitro.26_{Mutations were in either Egr-1: pTF (Egr-1}

m) or Sp1:

pTF (Sp1m) or in both Egr-1 and Sp1 binding sites: pTF

(Egr-1m/Sp1m) and are shown in Figure 5A. HUVECs were

transfected and subjected to shear stress. Extracts were prepared and reporter gene assays carried out. As shown in Figure 5B, a mutation in the Egr-1 binding site abolished the response of the TF promoter to shear stress, with a mutation in the Sp1 binding site having no significant effect. As expected, mutations in both Egr-1 and Sp1 binding sites abolished the response of this promoter to shear stress.

Superactivation of the TF Promoter in HUVECs Transfected With Egr-1 and Subjected to Flow

We reasoned that it should be possible to superactivate the TF promoter by a combination of flow induction of endogenous Egr-1 and supplementation of Egr-1 added exogenously via transfection. It has recently been reported that in HeLa cells, a truncated TF promoter identical to that used in this study can be trans-activated by Egr-1 transfection26_{and that}

trans-activation mediated via Egr-1 proceeds via its cognate bind-ing site. The construct pTF(EGR-1m), which lacks a

func-tional Egr-1 binding site, was not responsive to Egr-1 stimulation.26 _{We titrated the pTFWT promoter such that a}

low but significant level of shear stress–mediated trans-activation could be detected but could be further increased by transfection with the pCMV–Egr-1 (p930) construct but not the CMV–Egr-1 (p838) construct harboring a 1.5–zinc finger deletion in the Egr-1 protein (data not shown). This procedure corresponded to 5mg of promoter reporter construct and 5 ng of trans-activator plasmid per transfection experiment. Using this quantity of DNA, we transfected HUVECs and subjected

them to shear stress at 1.5 N/m2

for 3 hours. After this time, cells were harvested and reporter gene assays carried out (Figure 6A). The results show a clear superactivation of the TF minimal promoter under both transfection and flow conditions with 5 ng of p930 DNA. In the static control, 5 ng of p930 DNA failed to trans-activate pTFWT. Interestingly, transfection with 10 ng or 50 ng of p930 DNA reduced the level of superactivation. These data further suggest that

trans-activation of the TF promoter may be mediated by

Egr-1 and that under limiting concentrations of Egr-1, shear induction of endogenous Egr-1 may augment the response of the TF promoter to this transcription factor.

An Egr-1 Binding Site is a Novel SSRE

To further substantiate our findings that Egr-1 is implicated in the shear stress response of the TF promoter, we constructed a minimal synthetic promoter with an Egr-1 binding site driving expression of a luciferase reporter gene. Plasmid pSP13 was transfected into HUVECs, which were then exposed to a laminar shear stress of 1.5 N/m2

for 3 hours. Promoter activity was compared with cells not exposed to laminar shear stress, and the results are shown in Figure 6B. A promoter comprising an Egr-1 binding site placed 20 nt upstream of a TATA box could be activated 3.3-fold by shear stress. Thus, the nonameric sequence 59GCGGGGGCG 39 may act as an SSRE.

Flow Induces Egr-1 and TF Proteins in a Rat Arterial Model of Stenosis

Taken together, our experiments in vitro suggest a model whereby TF upregulation by shear stress is mediated by Egr-1 displacement of prebound Sp1. For these data to have greater physiological relevance, we wanted to know whether a transient increase in blood velocity and hence wall shear stress could upregulate Egr-1 and TF in vivo. To answer this question, we established a model of arterial stenosis in the rat in which we induced a graded stenosis by means of a Figure 4.EMSA demonstrating binding of Sp1 and Egr-1 to an oligonucleotide spanning positions289 to252 of the TF promoter. Prebound Egr-1 is competed off fully by a 30-fold molar excess of Sp1, whereas prebound Sp1 is competed off by an equimolar amount of Egr-1.

(7)

mechanical occluder.28_{This experiment was carried out on 7}

different animals, achieving a mean degree of stenosis of 4762.6% and a mean shear stress of 5.3461.65 N/m2

compared with a mean shear stress of 0.8660.13 N/m2

in unoccluded arteries. A typical experiment is shown in Figure 7. After the animals were killed, both the test and control carotid arteries were fixed, sectioned, and probed with Egr-1, TF, and vWF antibodies. (Figure 7A through 7C). There were no detectable levels of Egr-1 and TF proteins in the endothe-lial lining of the control artery, but they were substantially increased in the test artery under conditions of 47% stenosis, as indicated by the brown staining. The antibody was specific for TF as assayed by immunoblotting of TF in LPS-induced peripheral blood mononuclear cellular extracts (Figure 7D, and an intact endothelium in the control and test arteries was confirmed by immunodetection of vWF, as shown in Figure 7C.) Under conditions in which either a 22% or an 85% stenosis was delivered to the vessel, sections stained for both Egr-1 and TF proteins showed a similar increase in the production of Egr-1 (data not shown).

Discussion

This study identifies a novel SSRE in the promoter of the TF gene. This SSRE constitutes an overlapping Sp1/Egr-1 site that is capable of binding either Sp1 or Egr-1 protein in

vitro.14_{The TF promoter is a complex promoter, capable of}

responding to a number of stimuli that act through signaling pathways, some of which are well characterized and some poorly understood (reviewed in Reference 31). Chemical stimuli; for example, 12-O-decanoyl-13-phorbol acetate, interleukin-1b, tumor necrosis factor-a, and bacterial endo-toxin LPS, all induce transcription of TF mRNA via trans-acting factors.31–37 _{Regulation by the dietary pigment}

cur-cumin is more complex, exerting an effect through both NF-kB and AP1 sites.38_{However, with respect to shear stress}

induction of TF expression, the situation appears to be a little more clear. In this report, deletion analysis has shown that shear stress induction of the TF promoter requires a promoter element comprising nt2111 to114. This range excludes the binding sites for AP1 and NF-kB in the distal enhancer (at nt 2227 to 2172). Thus, shear stress induction of the TF promoter requires only functional Sp1 and Egr-1 binding sites and a TATA box. Evidence for the involvement of Egr-1 is derived from a number of additional experiments.

Figure 5.A, TF promoter mutants. Plasmid pTFWT is the wild-type minimal TF promoter from2111 to114; pTF (Sp1m), pTF

(Egr-1m), and pTF (Egr-1m/Sp1m) have been described.26The

hatched boxes represent Egr-1 binding sites and the filled boxes Sp1 binding sites. B, Effect of shear stress on Sp1, Egr-1, and Sp1/Egr-1 mutations in the minimal TF promoter. HUVECs were transfected and subjected to shear stress as described in Figure 4. Results show the average of at least 3 experiments.

Figure 6.A, Superactivation of the minimal TF promoter (pTFWT) by Egr-1 transfection and shear stress. After promoter reporter gene andtrans-activator plasmid titration (data not shown), 5mg of minimal TF promoter was transfected into HUVECs in the absence or presence of an increasing amount of p930trans-activator plasmid. Cells were exposed to shear stress of 1.5 N/m2_{for 3 hours. Results show the average of 3}

experiments. B, Shear stress–activated minimal synthetic pro-moter harboring an Egr-1 binding site and a TATA box. Bg indi-catesBglII and H,HindIII restriction enzyme sites. The arrow denotes the transcription start site in pGL3 basic luciferase reporter vector. The Egr-1 binding site is shown in a filled box and the TATA in a hatched box. HUVECs were transfected with 5mg of plasmid pSP13 and the cells exposed to shear stress as described above. Results show the average of 3 experiments compared with cells transfected with pSP13 but not exposed to shear stress.

(8)

First, our in vitro EMSA data support previous findings16

that prebound Sp1 is competed off an oligonucleotide spanning the overlapping Sp1/Egr-1 binding site by Egr-1 protein at a far lower excess of competitor protein than the prebound Egr-1 in competition with Sp1. Second, we show that a mutation in the Egr-1 binding site, but not in the Sp1 binding site, reduces the response of the TF promoter to shear stress. Third, we show that under conditions of titrated reporter and Egr-1 trans-activator plasmids, it is possible to superactivate the wild-type minimal TF pro-moter by shear stress. Our data showing reduced

superac-tivation of the minimal TF promoter with an increasing quantity of p930 DNA are intriguing. One possible expla-nation for this is that overproduction of Egr-1 may act in an inhibitory fashion, suppressing egr-1 promoter activity and ultimately its own production.18_{Finally, data obtained}

from a rat model of arterial stenosis support the observa-tion that arterial occlusion, and hence an increase in laminar shear stress, induces the synthesis of both Egr-1 and TF proteins in the vascular endothelium.

Our data would appear to be at odds with a recently published study suggesting that Sp1 and not Egr-1 mediates Figure 7.Egr-1 (A), TF (B), and vWF (C) immunological analysis in arterial sections after occlusion. Positive staining is indicated by brown coloration. The instrumented but unoccluded right carotid artery acted as the control. Histological analysis showed that no phys-ical damage to the arterial wall was detectable. Scale bars510mm, magnification3400. D, Immunoblot analysis of TF from peripheral blood mononuclear cells stimulated (S) or unstimulated (U) with LPS (100 ng/mL) for 6 hours. TF, migrating at 42 kD is arrowed. Protein molecular-weight markers are shown in kilodaltons.

(9)

shear response of the TF promoter in bovine aortic ECs.14

Although our data do not rule out a minor involvement of Sp1 in the shear response of the TF promoter (Figure 5B), we conclude that Egr-1 is the most significant transcription factor in shear stress trans-activation in human ECs. It is possible that the requirement for Sp1 is to set the basal level of the promoter and that Egr-1 allows maximal activation of the TF promoter under shear stress stimulation. There are 3 Egr-1 binding sites within the TF promoter26_{; 2 constitute}

low-affinity binding sites and 1 a high-low-affinity binding site. These sites overlap with binding sites for Sp1. Under shear stress stimulation, Sp1 may occupy all the Sp1/Egr-1 sites, whereas Egr-1 may displace Sp1 at the high-affinity site studied in our report. In addition, there may be subtle differences between the signaling cascades that confer shear stress activation in both bovine and human ECs that dictate the use of disparate sets of transcription factors. Interestingly, the SV40 promoter is a classical Sp1-dependent promoter,39_{which was not shear}

responsive but whose response to shear was augmented by the addition of a functional GAGACC (NF-kB binding) sequence upstream of the promoter.10_{In addition to the data}

showing clear shear stress response of a minimal promoter containing an Egr-1 binding site, recent data from our group suggest that Egr-1 constitutes a powerful SSRE when located downstream of a number of core transcription factor binding sites (P.H. et al, unpublished data, 1999).

The findings from this study demonstrate that an acute change in shear stress may confer a procoagulant phenotype on the vascular endothelium. However, it has been suggested that chronic shear stress may confer an antiatherogenic phenotype on the endothelium,40 _{and indeed it has been}

demonstrated that chronic shear stress may protect the endo-thelium from apoptosis.41_{We propose that an acute change in}

shear stress rather than exposure to a constant chronic shear stress may be an initiating factor for the stimulation of, for example, immediate early gene expression via ERK1/2 signaling.

This early growth response gene product has been shown to be a key player in regulating target genes of the immune system.42_{In addition, its role in regulating both PDGF A and}

PDGF B,43_{and now TF, suggests that Egr-1 is a significant}

factor in modulating vascular EC biology. An investigation into the molecular events surrounding Egr-1 upregulation by shear stress and the signaling cascades surrounding Egr-1 production will enhance our understanding of the role of this pluripotent transcription factor in shear stress–mediated events in vascular biology.

Acknowledgments

We are grateful to our colleagues in vascular diseases for their constructive criticisms and to Dr Philip Green for establishing the laminar flow system.

References

1. Davies PF, Tripathi SC. Mechanical stress mechanisms and the cell: an endothelial paradigm. Circ Res. 1993;72:239 –245.

2. Nerem RM, Harrison DG, Taylor WR, Alexander RW. Hemodynamics and vascular endothelial biology. J Cardiovasc Pharmacol. 1993;21: S6 –S10.

3. Gimbrone MA Jr, Cybulsky MI, Kume N, Collins T, Resnick N. Vascular endothelium: an integrator of pathophysiologic stimuli in atherosclerosis.

Ann N Y Acad Sci. 1995;748:122–131.

4. Patrick CW Jr, McIntire LV. Shear stress and cyclic strain modulation of gene expression in vascular endothelial cells. Blood Purif. 1995;13: 112–124.

5. Shyy JY-J, Li Y-S, Lin M-C, Chen W, Yuan S, Usami S, Chien S. Multiple cis-elements mediate shear-stress induced gene expression.

J Biomechanics. 1995;28:1451–1457.

6. Ando J, Kamiya A. Flow-dependent regulation of gene expression in vascular endothelial cells. Jpn Heart J. 1996;37:19 –32.

7. Malek AM, Izumo S. Control of endothelial gene expression by flow. J

Biomech. 1995;28:1515–1528.

8. Braddock M, Schwachtgen J-L, Houston P, Dickson MC, Lee MJ, Campbell CJ. Fluid shear stress modulation of gene expression in endo-thelial cells. News Physiol Sci. 1998;13:241–246.

9. Resnick N, Collins T, Atkinson W, Bonthron DT, Dewey CF Jr, Gimbrone MA Jr. Platelet-derived growth factor B chain promoter contains a cis-acting shear-stress-responsive element. Proc Natl Acad Sci

U S A. 1993;90:4591– 4595.

10. Khachigian LM, Resnick N, Gimbrone MA Jr, Collins T. Nuclear factor-kB interacts functionally with the platelet-derived growth factor B-chain shear-stress response element in vascular endothelial cells exposed to fluid shear stress. J Clin Invest. 1995;96:1169 –1175. 11. Ohno M, Cooke JP, Dzau VJ, Gibbons GH. Fluid shear stress induces

endothelial transforming growth factor beta-1 transcription and produc-tion: modulation by potassium channel blockade. J Clin Invest. 1995;95: 1363–1369.

12. Lan Q, Mercurius KO, Davies PF. Stimulation of transcription factors NF-kB and AP1 in endothelial cells subjected to shear stress. Biochem

Biophys Res Commun. 1994;201:950 –956.

13. Shyy JY-J, Lin M-C, Han JH, Lu YJ, 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.

14. Lin M-C, Almus-Jacobs F, Chen H-H, Parry GCN, Mackman N, Shyy JY-J, Chien S. Shear stress induction of the tissue factor gene. J Clin

Invest. 1997;99:737–744.

15. Khachigian LM, Anderson KR, Halnon NJ, Gimbrone MA Jr, Resnick N, Collins T. Egr-1 is activated in endothelial cells exposed to fluid shear stress and interacts with a novel shear stress response element in the PDGF A-chain promoter. Arterioscler Thromb Vasc Biol. 1997;17: 2280 –2286.

16. Khachigian LM, Lindner V, Williams AJ, Collins T. Egr-1-induced endothelial gene expression: a common theme in vascular injury. Science. 1996;271:1427–1431.

17. Khachigian LM, Collins T. Inducible expression of Egr-1 dependent genes: a paradigm of transcriptional activation in vascular endothelium.

Circ Res. 1997;81:457– 461.

18. Schwachtgen J-L, Houston P, Campbell CJ, Sukhatme VP, Braddock M. Fluid shear stress activation of egr-1 transcription in cultured human endothelial and epithelial cells is mediated via the ERK1/2 MAP kinase pathway. J Clin Invest. 1998;101:2540 –2549.

19. Grabowski EF, Zuckerman DB, Nemerson Y. The functional expression of tissue factor by fibroblasts and endothelial cells under flow conditions.

Blood. 1993;12:3265–3270.

20. Van’t Veer C, Hackeng TM, Delahaye C, Sixma JJ, Bouma BN. Acti-vated factor X and thrombin formation triggered by tissue factor on endothelial cell matrix in a flow model: effect of the tissue factor pathway inhibitor. Blood. 1994;84:1132–1142.

21. Grabowski EF, Lam FP. Endothelial cell function, including tissue factor expression under flow conditions. Thromb Haemost. 1995;74:123–128. 22. Schwachtgen JL, Ferreira V, Meyer D, Kebiriou-Nabias D. Optimization

of the transfection of human endothelial cells by electroporation.

Bio-techniques. 1994;17:882– 885.

23. Lawrence MB, Smith CW, Eskin SG, McIntire LV. Effect of venous shear stress on CD-18 mediated neutrophil adhesion to cultured endothe-lium. Blood. 1990;75:227–237.

24. Nightingale J, Brophy G, Braddock M. Quan-T-CAT: A New and Rapid

Method for the Enhanced Detection of CAT Reporter Gene Expression.

Arlington Heights, Ill: Amersham Life Sciences; 1994.

25. Su S, Vivier RG, Dickson MC, Thomas N, Kendrick MK, Williamson NM, Anson JG, Houston JG, Craig FF. High through-put RT-PCR anal-ysis of multiple transcripts using a microplate RNA isolation procedure.

Biotechniques. 1997;22:1107–1113.

26. Cui M-Z, Parry GCN, Oeth P, Larson H, Smith M, Huang R-P, Adamson ED, Mackman NJ. Transcriptional regulation of the tissue factor gene in human endothelial cells is mediated by Sp1 and Egr-1. J Biol Chem. 1996;271:2731–2739.

(10)

27. Nathwani AC, Gale KM, Pemberton KD, Crossman DC, Tuddenham EGD, McVey J. Efficient gene transfer into human umbilical vein endo-thelial cells allows functional analysis of the human tissue factor gene promoter. Br J Haematol. 1994;88:122–128.

28. Igawa T, Nagamura Y, Ozeki Y, Itoh H, Unemi F. Stenosis enhances role of platelets in growth of regional thrombus and intimal wall thickening in rat carotid arteries. Circ Res. 1995;77:310 –316.

29. Ward-Smith AJ. Internal Fluid Flow: The Fluid Dynamics of Flow in

Pipes and Ducts. Oxford, UK: Clavendon Press; 1980.

30. Matsuki Y, Yamamoto T, Hara K. Detection of inflammatory cytokine mes-senger RNA (mRNA)-expressing cells in human inflamed gingiva by combined in situ hybridization and immunohistochemistry. Immunology. 1992;76:42–47. 31. Mackman N. Regulation of the tissue factor gene. FASEB J. 1995;9:

883– 889.

32. Oeth PA, Parry GCN, Mackman N. Role of AP1, NF-kB/Rel, and Sp1 proteins in uninduced and lipopolysaccharide-induced expression.

Arte-rioscler Thromb Vasc Biol. 1997;17:365–374.

33. Oeth PA, Parry GCN, Kunsch C, Nantermet P, Rosen CA, Mackman N. Lipopolysaccharide induction of tissue factor gene expression in monocytic cells is mediated by binding of c-Rel/p65 heterodimers to a kappa B-like site. Mol Cell Biol. 1994;14:3772–3781.

34. Parry GCN, Mackman N. A set of inducible genes expressed by activated human monocytic and endothelial cells contain kappa B-like sites that specifically bind c-Rel-65 heterodimers. J Biol Chem. 1994;269: 20823–20825.

35. Mackman N. Regulation of the tissue factor gene. Thromb Haemost. 1997;78:747–754.

36. Mackman N, Fowler BJ, Edgington TS, Morrissey JH. Functional anal-ysis of the human tissue factor promoter and induction by serum. Proc

Natl Acad Sci U S A. 1990;87:2254 –2258.

37. Parry GCN, Mackman N. Transcriptional regulation of tissue factor expression in human endothelial cells. Arterioscler Thromb Vasc Biol. 1995;15:612– 621.

38. Bierhaus A, Zhang Y, Quehenberger P, Luther T, Haase M, Muller M, Mackman N, Ziegler R, Nawroth PP. The dietary pigment curcumin reduces endothelial tissue factor gene expression by inhibiting binding of AP-1 to the DNA and activation of NF-kB. Thromb Haemost. 1997;77: 772–782.

39. Takahashi K, Vigneron M, Mathes H, Wildeman A, Zenke M, Chambon P. Requirements of stereospecific alignments for initiation from the simian virus 40 early promoter. Nature. 1986;319:121–126.

40. Gimbrone MA Jr, Nagel T, Topper JN. Biomechanical activation: an emerging paradigm in endothelial adhesion biology. J Clin Invest. 1997; 99:1809 –1813.

41. Kaiser D, Freyberg M-A, Friedl P. Lack of haemodynamic forces triggers apoptosis in vascular endothelial cells. Biochem Biophys Res Commun. 1997;231:586 –590.

42. McMahon SB, Monroe JG. The role of early growth response gene (egr-1) in regulation of the immune response. J Leukoc Biol. 1996;60: 159 –166.

43. Khachigian LM, Collins T. Inducible expression of Egr-1– dependent genes: a paradigm of transcriptional activation in vascular endothelium.

Circ Res. 1997;81:457– 461.