Dihydrotestosterone Promotes Vascular Cell Adhesion Molecule-1 ...

0013-7227/04/$15.00/0 Printed in U.S.A.

Endocrinology 145(4):1889 –1897 Copyright © 2004 by The Endocrine Society doi: 10.1210/en.2003-0789

Dihydrotestosterone Promotes Vascular Cell Adhesion Molecule-1 Expression in Male Human Endothelial Cells via a Nuclear Factor-␬B-Dependent Pathway ALISON K. DEATH, KRISTINE C. Y. MCGRATH, MARK A. SADER, SHIRLEY NAKHLA, WENDY JESSUP, DAVID J. HANDELSMAN, AND DAVID S. CELERMAJER Discipline of Medicine (A.K.D., K.C.Y.M., D.S.C.), University of Sydney, Sydney, 2006 New South Wales, Australia; Heart Research Institute (M.A.S., S.N.), Camperdown, Sydney, 2050 New South Wales, Australia; Centre for Vascular Research (W.J.), Faculty of Medicine, University of New South Wales, Sydney, 2031 New South Wales, Australia; ANZAC Research Institute (D.J.H.), Concord, 2139 New South Wales, Australia; and Department of Cardiology (D.S.C.), Royal Prince Alfred Hospital, Sydney, 2050 New South Wales, Australia There exists a striking gender difference in atherosclerotic vascular disease. For decades, estrogen was considered atheroprotective; however, an alternative is that androgen exposure in early life may predispose men to earlier atherosclerosis. We recently demonstrated that the potent androgen, dihydrotestosterone (DHT), enhanced the binding of monocytes to the endothelium, a key early event in atherosclerosis, via increased expression of vascular cell adhesion molecule-1 (VCAM-1). We now show that DHT mediates its effects on VCAM-1 expression at the promoter level through a novel androgen receptor (AR)/nuclear factor-␬B (NF-␬B) mechanism. Human umbilical vein endothelial cells were exposed to 4 – 400 nM DHT. DHT increased VCAM-1 mRNA in a dose- and time-dependent manner. The DHT effect could be blocked by the AR antagonist, hydroxyflutamide. DHT increased VCAM-1 promoter activity via NF-␬B activation without affecting

C

ARDIOVASCULAR DISEASE REMAINS the major cause of mortality in the Western world. Epidemiologic studies show that men are twice as likely as women to die from coronary artery disease, across a wide variety of populations (1). That male gender is an independent risk factor (2) raises important questions concerning the role of sex steroids in promoting or protecting against atherogenesis. Whereas much recent research has focused on the hypothesis that estrogen is cardioprotective, recent findings (3), including the first prospective randomized clinical trials, do not support this concept (4, 5). Much less research has investigated the possible proatherogenic effects of androgens in males. Given the increasing use of androgens in the community for medical therapy (6) as well as anabolic steroid abuse (7, 8) and as an antiaging tonic, it is becoming more necessary to understand

Abbreviations: AP-1, Activator protein-1; AR, androgen receptor; ARE, androgen response element; BAEC, bovine aortic endothelial cell; DHT, dihydrotestosterone; DTT, dithiothreitol; ER, estrogen receptor; HF, hydroxyflutamide; HRP, horseradish peroxidase; HUVEC, human umbilical vein endothelial cell; I␬B, inhibitory protein of the NF-␬B signaling pathway; NF-␬B, nuclear factor-␬B; PBST, PBS plus 0.05% Tween 20; VCAM-1, vascular cell adhesion molecule-1. Endocrinology is published monthly by The Endocrine Society (http:// www.endo-society.org), the foremost professional society serving the endocrine community.

VCAM-1 mRNA stability. Using 5ⴕ deletion analysis, it was determined that the NF-␬B sites within the VCAM-1 promoter region were responsible for the DHT-mediated increase in VCAM-1 expression; however, coimmunoprecipitation studies suggested there is no direct interaction between AR and NF␬B. Instead, DHT treatment decreased the level of the NF-␬B inhibitory protein. DHT did not affect VCAM-1 protein expression and monocyte adhesion when female endothelial cells were tested. AR expression was higher in male, relative to female, endothelial cells, associated with increased VCAM-1 levels. These findings highlight a novel AR/NF-␬B mediated mechanism for VCAM-1 expression and monocyte adhesion operating in male endothelial cells that may represent an important unrecognized mechanism for the male predisposition to atherosclerosis. (Endocrinology 145: 1889 –1897, 2004)

how androgens influence cardiovascular disease. The mechanisms by which androgens may promote atherogenesis have been little studied, and existing data are inconsistent (9, 10). In animal studies, testosterone leads to increased plaque formation in female monkeys and rabbits, male chicks, and ApoE⫺ mice (11–15), although the opposite or neutral effects have also frequently been demonstrated (16 –19). An important feature of testosterone action is its conversion to the bioactive metabolites, estradiol, and dihydrotestosterone (DHT). The atheroprotection that is effected by testosterone treatment requires conversion of testosterone to estradiol (18, 20). The action of its 5-␣ metabolite, DHT, on atherogenesis has not been reported. In humans, androgen use has been associated with premature coronary disease in athletes (7, 8) and impaired vascular reactivity in female-to-male transsexuals (21). Supraphysiological concentrations of testosterone have a direct vasodilator effect in the coronary circulation, producing a small but consistent improvement in cardiac ischemia (22– 24). Low testosterone levels are also associated with coronary artery disease, although in men hypoandrogenemia is also associated with other confounding cardiovascular risk factors so the precise role of endogenous testosterone in atherosclerosis is unclear (25). The classical pathway of testosterone action involves its

1889

1890

Death et al. • Androgens, NF-␬B, and Endothelial Cell Adhesion

Endocrinology, April 2004, 145(4):1889 –1897

binding to a specific receptor, the androgen receptor (AR). This receptor belongs to a large superfamily of nuclear hormone receptors that share a well-conserved DNA-binding domain, a structurally conserved ligand-binding domain, and an N-terminal domain with no homology between the different receptors (26). After androgen binding, the receptor dimerizes and binds to androgen response elements (AREs) located within the promoters of androgen-responsive genes. Atherosclerosis involves interaction between the cells of the arterial wall (endothelial and smooth muscle cells) with those migrating into it (monocyte-macrophages) (27). One of the earliest steps in atherogenesis is the binding of monocytes to the vascular endothelium. Monocyte binding to the endothelium requires endothelial expression of cell adhesion molecules, including vascular cell adhesion molecule-1 (VCAM-1). VCAM-1 plays a major role in atherogenesis because VCAM-1-deficient mice have markedly reduced atherosclerotic plaque development (28). Previous studies in this laboratory have shown that the nonaromatizable androgen, DHT, increases surface expression of VCAM-1 protein in human endothelial cells from male donors and that this leads to enhanced monocyte adherence to endothelial cells (29). The effect could be blocked by cotreatment with hydroxyflutamide (HF), an AR blocker, indicating that DHT’s effects on VCAM-1 protein levels were mediated by AR. Interestingly, whereas activator protein-1 (AP-1), GATA, and nuclear factor-␬B (NF-␬B) transcription factors bind to the VCAM-1 promoter, there is no ARE in the 5⬘-regulatory region of the VCAM-1 gene (30); thus, the mechanism of DHT induction of VCAM-1 expression remains unknown. Therefore, to understand the molecular mechanism underlying the androgen-mediated induction of endothelial VCAM-1 expression, we studied the effects of DHT on VCAM-1 gene expression. Materials and Methods Cell culture Human umbilical vein endothelial cells (HUVECs) from male and female donors were isolated and cultured to passages 2– 4 in phenol red-free M199 (Invitrogen, Melbourne, Australia) supplemented with 20% postmenopausal women serum as described previously (29). Four to six different donors were used for each experiment. Bovine aortic endothelial cells (BAECs; Cell Applications Inc., San Diego, CA) were cultured in phenol red-free RPMI 1640 (Invitrogen) with 10% charcoalstripped fetal calf serum.

RT-PCR HUVECs were seeded in 12-well plates at 1 ⫻ 105 cells/well and allowed to recover overnight. For the time-course experiment, cells were prestimulated with IL-1␤ (10 U/ml) for 24 h and then exposed to 0.1% ethanol (control) or DHT (400 nm; Steraloids, Inc., Newport, RI) for 0 – 48 h. For the dose-response experiment, cells were exposed to 1, 4, 40, or 400 nm DHT for 48 h, with IL-1␤ (10 U/ml) added for the final 24 h. Total RNA from cells was isolated using TRI reagent (Sigma-Aldrich, Sydney, Australia), and 250 ng total RNA were reverse transcribed in a 20-␮l reaction using 2 U Superscript II RNase H⫺ reverse transcriptase (Invitrogen) along with 50 ng random primers (Invitrogen) and 10 mm of each deoxynucleotide triphosphate (Promega Inc., Sydney, Australia). A 2-␮l aliquot of each sample was amplified in reaction mixtures containing the primers [20 pmol each; VCAM-1 (31); ␤-actin (32)] and 2.5 U of RedTaq DNA polymerase (Sigma-Aldrich). Samples were amplified for 28 cycles for VCAM-1 and 24 cycles for ␤-actin. The PCR conditions were

denaturation at 94 C for 15 sec, annealing at 55 C for 20 sec, and extension at 72 C for 45 sec. The PCR products were analyzed on 2% agarose gels and, after staining with ethidium bromide (0.5 ␮g/ml), were directly digitized. Band densities were measured using Phoretix software (Phoretix International, Newcastle Upon Tyne, UK). VCAM-1 mRNA stability was assayed by stimulating HUVECs with IL-1␤ (10 U/ml) for 24 h before blocking transcription by treatment with actinomycin D (5 ␮g/ml). Cells then received DHT (400 nm) or 0.1% ethanol (control) for the times indicated; mRNA levels were compared with those of 0 h cells.

Transient transfections The full-length and 5⬘ truncations of the human VCAM-1 promoter (30) were subcloned into a luciferase reporter vector (pGL3-Enhancer; Promega). For NF-␬B-site-directed mutagenesis, the QuikChange mutagenesis kit (Stratagene, La Jolla, CA) was used with primers: 1) GTT GCA GAG GCG TGA GGG CCC CTC CTT CAA GGG GAA ACC and 2) GGT TTC CCC TTG AAG GAG GGG CCC TCC GCC TCT GCA AC. One day before transfection, BAECs were seeded (1 ⫻ 105 cells/well) onto a 12-well plate. A mixture containing 0.2 ␮g VCAM-1 promoterluciferase, NF-␬B-luciferase (Promega), pAP1-luciferase (Promega), or pI␬B␣-EGFP (BD Biosciences Clontech, San Jose, CA) plasmid DNA, 0.04 ␮g pTK-renilla plasmid DNA (Promega), and Effectene (Qiagen, Melbourne, Australia) was prepared and transfection was performed, following the manufacturer’s protocol. After 4 – 6 h incubation, cells were washed twice with 1⫻ PBS, and 1 ml fresh medium was added supplemented with DHT (400 nm) or 0.1% ethanol (control). For the VCAM-1 promoter experiments only, IL-1␤ (10 U/ml) was added for the final 24 h. After 48 h treatment, cell lysates were prepared by washing the cells with ice-cold PBS twice, followed by the addition of 100 ␮l 1⫻ passive lysis buffer (Promega). To assay for promoter activity, 50 ␮l luciferase solution (Promega) were automatically injected into 10 ␮l cell lysate, and luciferase activity was measured as light emission using a luminometer. Fifty microliters Stop and Glow reagent (Promega) were then added to determine renilla activity (dual-luciferase assay, Promega). For each transfection study, luciferase activity was determined and normalized on the basis of renilla activity.

EMSAs HUVECs were treated with DHT (400 nm) or 0.1% ethanol (control) for 24 h. 3 ⫻ 106 cells were pelleted (200 ⫻ g, 5 min) and resuspended in 30 ␮l buffer A [HEPES 10 mm (pH 7.9), KCl (10 mm), MgCl2 (1.5 mm), dithiothreitol (DTT) (0.5 mm), phenylmethylsulfonyl fluoride (0.5 mm), Nonidet P-40 (0.67%)]. Cells were incubated on ice for 20 min and then centrifuged (10,000 ⫻ g, 4 C). The supernatant was removed and the cell pellet (containing cell nuclei) was resuspended in 30 ␮l buffer B [HEPES 20 mm (pH 7.9), NaCl (0.4 mm), EDTA (0.2 mm), MgCl2 (1.5 mm), DTT (0.5 mm), phenylmethylsulfonyl fluoride (0.5 mm)] and vortexed for 15 sec before being incubated on ice for 15 min. Nuclear extract was centrifuged (10,000 ⫻ g, 30 min) and stored at ⫺80 C. Double-stranded DNA oligonucleotide (5 pmol) encoding the consensus sequence for NF-␬B (Promega) was 32P-labeled with [32P]-ATP and T4 polynucleotide kinase (Promega) in a 10-␮l reaction. The probe (1 ␮l) was then added to a 20 ␮l EMSA reaction mixture containing 1⫻ binding buffer [HEPES (20 mm), EDTA (0.2 mm), EGTA (0.2 mm), NaCl (100 mm), glycerol (5%), DTT (2 mm) (pH 7.9)], 5 ␮g crude nuclear extract, and 1 ␮g poly(dI-dC). Nuclear extract from HeLa cells was used as the positive control (Promega). Control reaction mixtures were also prepared adding unlabeled NF-␬B oligonucleotide or nonspecific SP-1 oligonucleotide. Reactions were incubated for 15 min at room temperature. Protein-DNA complexes were resolved by 5% PAGE (10 ⫻ 10 cm) in 1⫻ TBE [0.09 m Tris-base, 0.09 m boric acid, 2 mm EDTA (pH 8.3)] at 100 V at room temperature for 45 min. Gels were exposed to phosphor-storage screens for 1 h and then directly digitized. Band densities were analyzed using Phoretix software (Phoretix International). Specific DNA-protein complexes were observed because more slowly migrating complexes in the gel.

Western blotting HUVECs were cultured to confluence in 75-cm2 flasks and then treated with 0.1% ethanol (control) or DHT (400 nm) for 48 h. Cells were


harvested by trypsinization, and washed twice with PBS by centrifugation (3,000 rpm, 5 min). Cells were lysed in cell lysis buffer [1% Nonidet P-40, 0.1% sodium dodecyl sulfate, 0.5% deoxycholate, 150 mm NaCl, 50 mm Tris (pH 8.0), 1 ␮l/ml protease cocktail inhibitor (Sigma)]. Denatured proteins (20 ␮g) were resolved by SDS-PAGE (8% acrylamide) at 100 V for 1 h. Proteins were then transferred for 1 h at 13 V to polyvinyl difluoride membranes (Millipore Corp., North Ryde, New South Wales, Australia), blocked using 5% skim milk in 10 mm Tris-HCl and 150 mm NaCl (pH 7.5), washed three times with PBST (1⫻ PBS plus 0.05% Tween 20), and incubated overnight with AR antibody at 1:500 dilution (AR 441/sc7305, Santa Cruz Biotechnology, Santa Cruz, CA) or with ␤-actin antibody at 1:5,000 dilution (MAB1501R, Chemicon, Temecula, CA). The membrane was then washed three times with PBST and incubated with secondary antibody (1:6000 for AR or 1:10,000 for ␤-actin) conjugated to horseradish peroxidase (HRP) for 2 h at room temperature. The membrane was again washed three times with PBST, and immunodetection was accomplished with enhanced chemiluminescence (Amersham Biosciences, Castle Hill, New South Wales, Australia). Films were directly digitized and band densities measured using Phoretix software (Phoretix International).

Immunoprecipitation HUVECs (1 ⫻ 106 cells) were cultured to confluence in a 75-cm2 flask. Cells were harvested by trypsinization, washed twice with PBS by centrifugation (3000 rpm, 5 min), and resuspended in cell lysis buffer (as for Western). Forty micrograms total protein were incubated for 1 h, rotating at room temperature, with 2.5 ␮l mouse IgG1 (Dako) and 20 ␮l protein G Plus-Agarose beads (Upstate Biochemicals, Inc., Lake Placid, NY) and 200 ␮l lysis buffer. The incubation mix was then centrifuged (14,000 ⫻ g; 2 min) and the supernatant transferred to a fresh Eppendorf tube. Then 1.25 ␮l AR antibody [AR 441/sc7305, Santa Cruz) and 20 ␮l protein G Plus-Agarose beads (Upstate) were added and incubated overnight while rotating at 4 C. The mixture was centrifuged and the supernatant discarded. The pellet was washed three times with lysis buffer and resuspended in 5 ␮l sample buffer [12.5 mm Tris-HCl (pH

FIG. 1. The effect of DHT treatment on VCAM-1 mRNA levels. A, Time course of VCAM-1 mRNA induction in prestimulated male HUVECs exposed to DHT (400 nM). Densitometric analysis of VCAM-1 levels expressed as a ratio of VCAM-1 to ␤-actin (BA) band densities. Data are representative of three independent experiments. Inset, Representative RT-PCR gel showing VCAM-1 (top) and ␤-actin (bottom) expression. Samples 0, 3, 6, 12, 24, and 48 h are in duplicate. Last lane is the 1-kb DNA ladder (Invitrogen) showing VCAM-1 PCR product approximately 303 bp as expected. *, P ⬍ 0.05 vs. control (0 h). B, HF blocks the DHT effect on VCAM-1 mRNA levels. Prestimulated male HUVECs were exposed to DHT (400 nM), DHT (400 nM) and HF (4 ␮M), or 0.1% ethanol (control) for 24 h. Quantification is expressed as a percentage of control-treated cells. Results are a summary of four independent experiments. *, P ⬍ 0.05 control vs. DHT.

Endocrinology, April 2004, 145(4):1889 –1897 1891

6.8), 4% glycerol, 0.4% sodium dodecyl sulfate, 0.01% bromophenol blue, 4% ␤-mercaptoethanol] and samples boiled for 5 min. The sample was resolved by SDS-PAGE (8% acrylamide) at 100 V for 1 h and transferred protein to a polyvinyl difluoride membrane for 1 h at 13 V. The membrane was then blocked with 5% skim milk dissolved in PBST, washed three times with PBST, and the membrane incubated overnight with AR antibody (Santa Cruz, AR 441/sc7305, 1:1000 dilution), NF-␬B antibody (Santa Cruz, p65 (C-20) sc 372, diluted 1:1000), or rabbit IgG diluted to 0.2 ␮g/ml. The membrane was then washed three times with PBST for 5 min before incubation with secondary antibody (AR probed with mouse HRP, 1:10,000; NF-␬B; and IgG probed with rabbit HRP, 1:20,000). Band detection was visualized with enhanced chemiluminescence (Amersham Biosciences).

Immunohistochemistry Human mesenteric artery samples were obtained from five postmenopausal women (aged 57– 83 yr, mean 71 yr), two premenopausal women (35 and 43 yr), and seven men (aged 18 –75 yr, mean 51 yr) with no known cardiovascular disease, who underwent bowel operations at Royal Prince Alfred Hospital, Sydney. Samples were fixed in 4% paraformaldehyde for 4 h and then dehydrated through a series of ethanol washes before being embedded in paraffin wax. Sections were probed with an AR monoclonal type IgG1 antibody (AR441, Dako,1:50). Sections were digitally captured and positive cells counted by two independent investigators.

Monocyte-endothelial cell adhesion assay HUVECs (1 ⫻ 104 cells/96-well) from male and female donors were exposed for 48 h to the treatment groups: 1) 0.1% ethanol (control); 2) DHT 40 nm; 3) DHT 400 nm; and 4) DHT 400 nm and HF 4 ␮m. Each treatment group was stimulated with IL-1␤ (10 U/ml) for the final 24 h. After the treatment period, monocyte-endothelial cell adhesion assays were performed as described previously (29).

1892



VCAM-1 ELISA for protein determination HUVECs from four male and three female donors were grown to confluence in 96-well plates and exposed to 0.1% ethanol (control) or 400 nm DHT for 48 h. Cells were stimulated with IL-1␤ (10 U/ml) for the final 24 h. After treatment, an ELISA was performed as described previously (29).

Data analysis Results of the experimental studies are reported as mean ⫾ se, compared with controls. Unpaired Student’s t tests were used to determine the significance of changes between groups. A value of P ⬍ 0.05 was regarded as significant.

Results DHT increases steady-state levels of VCAM-1 mRNA

The results showed that DHT 400 nm increases VCAM-1 mRNA levels in a time-dependent manner in IL-1␤-stimulated endothelial cells from male donors. VCAM-1 mRNA levels increased by 3– 6 h, reaching significance by 12 h (P ⬍ 0.05), peaking by 24 h (P ⬍ 0.05), and then maintained at the higher level for at least 48 h (P ⬍ 0.05) (Fig. 1A). The maximum increase in VCAM-1 mRNA levels was seen in endothelial cells treated with DHT 400 nm; however, endothelial cells treated with 10- and 100-fold lower DHT concentrations also showed increased VCAM-1 mRNA levels (1 nm; 1.50 ⫾ 0.040-fold, P ⬍ 0.05: 4 nm; 1.83 ⫾ 0.099-fold, P ⬍ 0.05: 40 nm; 2.06 ⫾ 0.15-fold, P ⬍ 0.05: 400 nm; 2.04 ⫾ 0.06-fold, P ⬍ 0.005). The effect of DHT 400 nm on VCAM-1 mRNA levels could be blocked by hydroxyflutamide 0.4 ␮m treatment (Fig. 1B). Effect of DHT on VCAM-1 mRNA stability in endothelial cells

To determine whether increased VCAM-1 mRNA levels in DHT 400 nm-treated endothelial cells is because of an effect of DHT on VCAM-1 mRNA stability, we performed an actinomycin D-based stability assay on DHT-treated HUVECs. Figure 2A shows that the degradation rate of VCAM-1 mRNA in control or DHT-treated cells was similar for the 24 h tested. Effect of DHT on VCAM-1 promoter activity in endothelial cells

We next examined the effect of DHT treatment on VCAM-1 transcription. BAECs were transfected with a VCAM-1 promoter-luciferase reporter vector and then exposed to DHT 400 nm or 0.1% ethanol (control) for 48 h. Figure 2B shows DHT treatment induced luciferase activity by 129 ⫾ 8.5% (P ⬍ 0.05) relative to control-treated cells. Hydroxyflutamide 0.4 ␮m treatment blocked the increase in luciferase activity induced by DHT 400 nm. Activation of the VCAM-1 promoter by DHT involves NF-␬B

We next examined the 5⬘-proximal promoter region (⬃2.2 kb) of VCAM-1 for a potential ARE. We could not identify a consensus or even putative ARE in the VCAM-1 promoter. So we used 5⬘-deletion analysis (30) to map the DHTresponsive region of the VCAM-1 promoter. The deletion constructs (F0, F1, F2, F3, F4; Fig. 3A) were designed such that they sequentially lost known transcription factor binding

FIG. 2. The effect of DHT on VCAM-1 mRNA stability and promoter activity. A, VCAM-1 stability assay. Male HUVECs were preincubated with IL-1␤ (10 U/ml) for 24 h before actinomycin D and DHT (400 nM) or 0.1% ethanol (control) was added at time 0 h. VCAM-1 mRNA levels were measured at indicated time points. Data are representative of three independent experiments. B, VCAM-1 promoter activity. BAECs transfected with a VCAM-1-luciferase reporter vector were exposed to DHT (400 nM), DHT (400 nM) and HF (4 ␮M), or 0.1% ethanol (control). Results are expressed as percent of controltreated cells and represent four independent experiments.*, P ⬍ 0.05 vs. control.

sites including AP-1, GATA, and NF-␬B. BAECs were cotransfected with pFfl, pF0, pF1, pF2, pF3, or pF4 and pTKrenilla (to control for transfection efficiency) and exposed to DHT 400 nm for 48 h, with IL-1␤ (10 U/ml) added for the final 24 h. Deletion from 2209 bp to 1549 bp (construct F0) did not significantly affect the DHT-induced 30% increase in promoter activity measured for the full-length promoter region (Ffl, Fig. 3B). The same was true for construct F1 (1412 bp, respectively). Construct F2, which is 260 bp shorter than F1 and lacks the AP-1 site, was also still fully inducible by DHT. Construct F3, which is 160 bp shorter than F2 and contains only the NF-␬B sites, also showed 30% inducibility. Taken together, these results suggest that DHT effects on VCAM-1 promoter activity required only the NF-␬B sites. Construct F4, which lacks all the transcription factor sites and retains only the TATA box, showed no promoter activity in the presence or absence of DHT. Similarly, site-directed mutagenesis of one or both NF-␬B sites also rendered the promoter inactive (Fig. 3B), in keeping with previous studies that NF-␬B is absolutely required for VCAM-1 gene transcription (30).



FIG. 3. NF-␬B is required for the DHT induction of VCAM-1 promoter activity. A, Schematic of VCAM-1 promoter constructs used in the transfection experiments. Ffl, Full-length promoter, F0/F1 shorter constructs that retain the AP-1, GATA, and NF-␬B sites; F2 construct that retains the GATA and NF-␬B sites; F3 construct that retains only the NF-␬B sites; and F4 that has only the TATA box. NF-␬Bm is a site-directed mutagenesis construct in which the NF-␬B sites in the full-length promoter were specifically disrupted. B, BAECs were transfected with the Ffl, F0, F1, F2, F3, F4, and NF-␬Bm constructs and exposed to DHT (400 nM) or 0.1% ethanol (control) for 48 h with IL-1␤ (10 U/ml) added for the final 24 h. Results are expressed as percent of control-treated cells for each construct. Each experiment was repeated three to six times. *, P ⬍ 0.05 control vs. DHT.

DHT increased NF-␬B binding ability and NF-␬B-mediated transactivation in endothelial cells

We next examined whether DHT increases NF-␬B translocation to the nucleus by assaying nuclear NF-␬B levels by EMSA. The results demonstrate that NF-␬B DNA binding levels increased approximately 2-fold (P ⬍ 0.05) in response to DHT 400 nm treatment (Fig. 4A). To confirm whether increased levels of NF-␬B DNA binding ability led to increased NF-␬B-mediated gene transcription, we transfected BAECs with a NF-␬B-luciferase reporter vector. Using pTK-renilla to control for transfection efficiency, we demonstrated that DHT 400 nm exposure for 48 h increased luciferase activity by 200% (P ⬍ 0.05, Fig. 4B). BAECs transfected with an AP-1-luciferase reporter vector showed no response to DHT 400 nm treatment, consistent with our finding that the AP-1 site was not required for DHT’s effects on VCAM-1 promoter activity. These results indicate that DHT does activate NF-␬B in endothelial cells and subsequently increases NF-␬B-mediated gene transcription. AR and NF-␬B interaction

Together the results implicated an interaction between AR and NF-␬B, induced by DHT, that leads to NF-␬B activation and subsequently VCAM-1 expression. To investigate whether there is a direct interaction between these two transcription factors, we performed coimmunoprecipitation studies. Nuclear extracts were prepared from HUVECs treated with DHT 400 nm and immunoprecipitated with AR antibody before Western analysis with either AR antibody (positive control), NF-␬B antibody (for determination of specific interaction), or rabbit IgG (for nonspecific interactions). Our results demonstrate that the coimmunoprecipitation reactions were successful because AR could be detected (Fig. 5A, top panel); however, bands corresponding to NF-␬B were not detected (Fig. 5A, middle panel); only high background

and hints of nonspecific band formation were seen after extended exposure of films to Western blot membrane. IgG blotting similarly showed the same nonspecific band formation. Because no direct interaction between AR and NF-␬B could be shown that could explain NF-␬B activation by DHT, we next examined whether the inhibitory protein of the NF-␬B signaling pathway (I␬B␣) was affected by DHT treatment. In unstimulated cells, I␬B binds NF-␬B in the cytoplasm, but when signaled, I␬B is phosphorylated and rapidly degraded, releasing NF-␬B, which quickly translocates to the nucleus, promoting gene transcription. Endothelial cells were transfected with an I␬B␣-EGFP expression plasmid and then treated with DHT 400 nm or 0.1% ethanol for 48 h. Our results show that DHT treatment reduced I␬B␣ levels by approximately 20% (P ⬍ 0.05), compared with controltreated cells (Fig. 5B). This finding suggests DHT induces I␬B␣ degradation in endothelial cells, activates NF-␬B, and subsequently induces VCAM-1 expression. AR levels are higher in endothelial cells from male vs. female donors

Previously it was demonstrated that DHT had no effect on VCAM-1 expression in female HUVECs (20). Given our findings that AR regulates VCAM-1 gene expression (via NF-␬B), it is possible that gender-specific differences in AR expression could underlie the gender differences in the DHT responses. To test this, we measured AR levels in human endothelial cells. Endothelial cells from male donors had approximately 2-fold higher AR protein levels relative to endothelial cells from female donors, as determined by Western blotting (Fig. 6, A and B). We also found that human artery samples from male donors showed 5-fold more AR-positive cells than

1894


FIG. 4. The effect of DHT on nuclear NF-␬B levels and NF-␬Bmediated gene transcription. A, Representative EMSA. Male HUVECs were treated with DHT 400 nM (D lane) or 0.1% ethanol (C lane, control) for 24 h. Nuclear extracts were prepared and incubated with 32P-labeled NF-␬B oligonucleotides. NF-␬B form-specific complexes with 32P[NF-␬B] (NF-␬B-DNA). Excess unlabeled NF-␬B oligonucleotide abolishes the NF-␬B band (XS lane); negative control (⫺ lane, no extract added) showed no complex, and positive control (⫹ lane, HeLa cell extract) showed strong shifted band. DNA band at bottom of gel (DNA) represents oligonucleotide with no bound protein, and NS is nonspecific band formation. B, Effects of DHT on NF-␬Band AP-1-mediated gene transcription. BAECs were transfected with either pNF-␬B or pAP-1-luciferase reporter vectors and then exposed to 0.1% ethanol (control; black bar) or DHT (400 nM; white bar) for 48 h. Results are reported as the percent of control-treated cells and represent three to four independent experiments. *, P ⬍ 0.05 control vs. DHT.

arteries obtained from female donors (P ⬍ 0.05, Fig. 6, C and D). Consistent with our finding that AR levels regulate VCAM-1 expression, we found that endothelial cells from male donors had significantly higher VCAM-1 mRNA levels than endothelial cells from female donors (Fig. 7A). The higher VCAM-1 mRNA levels in male HUVECs was confirmed by real-time PCR (Fig. 7B). Enhanced monocyte binding to male but not female endothelial cells induced by DHT

Given the gender-specific expression of AR and VCAM-1, we tested the gender specificity of DHT-induced monocyteendothelial cell adhesion and VCAM-1 protein levels. DHT increased monocyte binding in a concentration-dependent manner to IL-1␤-stimulated male endothelial cells (Fig. 8).


FIG. 5. Effect of DHT on NF-␬B/AR interaction and I␬B activation. A, Coimmunoprecipitation of AR and NF-␬B. Total protein from HUVECs (three different male donors, lanes 3– 8) treated with DHT were coimmunoprecipitated with AR antibody and then Western blotted with either NF-␬B antibody (lanes 3–5) or rabbit IgG1 (lanes 6 – 8). Lanes 1 and 2 represent Western blotting (no coimunoprecipitation) of total protein from male HUVECs with NF-␬B antibody (positive control for correct NF-␬B band size indicated by arrow). Only nonspecific band detection was observed for coimmunoprecipitation experiments. B, Effect of DHT on I␬B levels. BAECs were transiently transfected with an I␬B-GFP reporter vector and then exposed to 0.1% ethanol (control) or DHT (400 nM) for 48 h. Results are reported as the percent of control-treated cells and represent three to four independent experiments. *, P ⬍ 0.05 control vs. DHT.

The increase in adhesion induced by DHT was not observed for IL-1␤-stimulated endothelial cells from female donors (Fig. 8). The DHT-mediated increase in adhesion in malederived endothelial cells was completely blocked by cotreatment with the AR antagonist, HF (4 ␮m, Fig. 8). Treatment of HUVECs with DHT (400 nm) also induced VCAM-1 protein levels in male but not female cells (Fig. 8). Discussion

The primary objective of this study was to elucidate the molecular mechanisms by which DHT, the potent 5␣metabolite of testosterone, enhances monocyte binding to the endothelium, a key early event in atherogenesis (29). We describe a novel regulatory mechanism mediated by the AR and the key inflammatory transcription factor, NF-␬B, that ultimately up-regulates the important cell adhesion molecule, VCAM-1. This pathway is activated by exposure to DHT, particularly in male endothelial cells, which demonstrates higher AR expression, relative to female endothelial cells. The net result of this pathway is a higher degree of



FIG. 6. Gender differences in AR protein expression in human vascular cells. A, Western analysis showing gender difference in AR expression in HUVECs from male vs. female donors. Total protein was isolated from three different male and three different female donors and Western blot performed with AR (AR, top panel) or ␤-actin (BA, bottom panel) antibodies. Prostate cancer cells (LnCap) were used as the AR-positive (⫹ve) control. B, Quantification of AR protein levels in male vs. female HUVECs by densitometric analysis. Quantification is expressed as mean ⫾ SE of arbitrary densitometric units obtained from four different donors for each gender. *, P ⬍ 0.05 difference between males and females. C, Representative immunohistochemistry sections showing AR staining in male but relatively little staining in female mesenteric arteries. D, Quantification of AR-positive cells in immunohistochemistry sections. Males had significantly higher AR-positive cells, *, P ⬍ 0.05 male vs. female. Bars represent mean ⫾ SE of seven different male and female artery samples, counting three to four sections for each artery sample by two independent investigators. FIG. 7. VCAM-1 expression in male vs. female endothelial cells. A, Representative picture of a relative RT-PCR gel showing higher VCAM-1 expression in HUVECs from two males (in duplicate, top gel, left four lanes), compared with two females (in duplicate, top gel, right four lanes). ␤2Microglobulin (B2M, bottom gel) was used as the housekeeping control. B, Quantification of VCAM-1 mRNA levels by realtime PCR. Quantitation is expressed as mean ⫾ SE of relative expression value obtained from four different male and female donors. *, P ⬍ 0.05 male vs. female cells.

endothelial cell activation and VCAM-1 expression in male endothelial cells, in turn leading to increased monocyte adhesion to the endothelium and, possibly, early atherogenesis. Our result that VCAM-1 is regulated in endothelial cells by the androgen, DHT, is consistent with other studies examining sex steroid regulation of cell adhesion molecules in human endothelial cells. Testosterone increased VCAM-1 and E-selectin expression in HUVECs (33). However, testosterone inhibited TNF␣-stimulated VCAM-1 expression in human aortic endothelial cells (34) and HUVECs from a female donor (20). In the latter study, testosterone treatment

with high doses (300 nm, 1 ␮m) was shown to decrease VCAM-1 expression in female HUVECs; however, the testosterone effect was blocked by an aromatase inhibitor or estrogen receptor (ER) antagonist (ICI-182780). Thus, this result reflected an inhibitory estradiol effect, confirming previous studies that estradiol down-regulates VCAM-1 (35). In addition, this same study showed DHT had no effect on VCAM-1 expression. We, too, show DHT has very little effect on VCAM-1 expression in endothelial cells from female donors, in contrast to the significant increase in VCAM-1 expression induced by DHT in male endothelial cells. Our

1896


FIG. 8. DHT induces monocyte adhesion to endothelial cells via upregulation of VCAM-1 in a gender-specific manner. Monocyte adhesion to endothelial cells from male (black bars) or female (white bars) donors exposed to control (0.1% ethanol), DHT 40 nM (DHT 40), DHT 400 nM (DHT 400), and DHT 400 nM⫹HF 4 ␮M (DHT ⫹ HF). VCAM-1 protein levels were measured by ELISA. Endothelial cells from male or female donors were exposed to control (0.1% ethanol) or DHT 400 nM (DHT 400). Results are expressed as percent of control-treated cells and represent four to six independent experiments. *, P ⬍ 0.05 control vs. DHT treatment.

findings suggest that the lower expression of AR in female, relative to male, cells underlies the gender specificity of DHT’s effects. Our observations taken together with those of others may explain the lack of coherent findings with regard to the role of testosterone in animal and cellular studies of atherogenesis. First, there are gender-specific effects, and second, it seems that different levels of conversion of testosterone to its metabolites, estradiol or DHT, could result in different effects at the cellular level. Diversification of testosterone to estradiol by aromatase allows protective effects mediated by the ER. By contrast, conversion of testosterone to DHT by either type 1 or 2 5␣-reductase produces proatherogenic effects mediated by AR. Aromatase and 5␣-reductase enzymes are present in HUVECs (18, 36 –38). The nonaromatizable androgen, DHT, rather than testosterone, might then be a clearer marker of androgen action because this androgen acts solely via AR, and thus, results are not subject to the confounding potential interpretations of differing ER- and ARmediated effects. The molecular mechanism that we describe whereby AR and NF-␬B mediate up-regulation of VCAM-1 expression is novel. Steroid hormones usually mediate inhibition rather than stimulation of NF-␬B-mediated transcriptional activation (39). For example, estradiol and glucocorticoids downregulate VCAM-1 expression in endothelial cells (35). By contrast, our data support androgen up-regulation of VCAM-1 in endothelial cells that is linked to AR and NF-␬B action. AR involvement was evidenced by the blockade experiments with the antagonist, HF. That VCAM-1 expression is AR dependent suggested that AR may interact with NF-␬B. However, immunoprecipitation studies showed no evidence


of a direct interaction in endothelial cells, which is consistent with findings in prostate cancer cells (40). Rather than a direct interaction between AR and NF-␬B, our findings suggest DHT induction of AR leads to degradation of the NF-␬B inhibitory protein, I␬B. The exact mechanism by which AR induces I␬B degradation remains to be elucidated. The early pathogenesis of atherosclerosis involves monocyte binding to the endothelium and experiments with VCAM-1-deficient mice have determined that VCAM-1 plays a critical role in this process (28). Importantly, VCAM-1 is specifically up-regulated in arterial endothelial cells in lesion-prone areas (41). Monocyte binding to aortic endothelial cells was higher in male than female rabbits with diet-induced hypercholesterolemia, and this was dependent on VCAM-1 expression (42). Whether the novel pathway that we now describe represents the early stage head start that predisposes males to cardiovascular mortality in later life remains to be determined. Androgens have been shown to have other proatherogenic effects on vascular cells. Testosterone at physiological levels enhances apoptosis damage in human vascular endothelial cells cultured in serum-deprived conditions, possibly related to decreased Bcl-2 protein expression (43). DHT also increases smooth muscle cell proliferation (44) and platelet aggregation (45, 46). Thus, the findings are consistent with a role for testosterone and its 5␣-metabolite in atherosclerosis. We conclude that androgen signaling in endothelial cells after treatment with the pure androgen, DHT, leads to induction of VCAM-1 expression, resulting in increased monocyte binding to the endothelium. The pathway leading to VCAM-1 expression is AR dependent. AR interacts indirectly with NF-␬B signaling to induce VCAM-1 expression. Because there is higher AR expression in male endothelial cells, the pathway is activated to a greater extent in male, relative to female, endothelial cells. This is the first demonstration that AR directly induces proinflammatory events in endothelial cells (via NF-␬B and VCAM-1) and suggests a contributing mechanism for the male predisposition to atherosclerosis. Acknowledgments Received June 25, 2003. Accepted December 8, 2003. Address all correspondence and requests for reprints to: Dr. Alison Death, Department of Medicine (D06), University of Sydney, Sydney, New South Wales 2006, Australia. E-mail: [email protected]. This work was supported in part by the Merck Medical School Grants Program, the National Health and Medical Research Council of Australia, and the National Heart Foundation (Australia).

References 1. Kalin MF, Zumoff B 1990 Sex hormones and coronary disease: a review of the clinical studies. Steroids 55:330 –352 2. Assmann G, Carmena R, Cullen P, Fruchart JC, Jossa F, Lewis B, Mancini M, Paoletti R 1999 Coronary heart disease: reducing the risk: a worldwide view. International Task Force for the Prevention of Coronary Heart Disease. Circulation 100:1930 –1938 3. Mendelsohn ME, Karas RH 1999 The protective effects of estrogen on the cardiovascular system. N Engl J Med 340:1801–1811 4. Hulley S, Grady D, Bush T, Furberg C, Herrington D, Riggs B, Vittinghoff E 1998 Randomized trial of estrogen plus progestin for secondary prevention of coronary heart disease in postmenopausal women. Heart and Estrogen/ progestin Replacement Study (HERS) Research Group. JAMA 280:605– 613 5. Simon JA, Hsia J, Cauley JA, Richards C, Harris F, Fong J, Barrett-Connor


6. 7. 8. 9. 10. 11.

12. 13.

14. 15. 16. 17. 18. 19.

20.

21. 22. 23.

24.

25. 26.

E, Hulley SB 2001 Postmenopausal hormone therapy and risk of stroke: the Heart and Estrogen-Progestin Replacement Study (HERS). Circulation 103: 638 – 642 Bhasin S, Bremner WJ 1997 Clinical review 85: emerging issues in androgen replacement therapy. J Clin Endocrinol Metab 82:3– 8 Rockhold RW 1993 Cardiovascular toxicity of anabolic steroids. Annu Rev Pharmacol Toxicol 33:497–520 Sullivan ML, Martinez CM, Gennis P, Gallagher EJ 1998 The cardiac toxicity of anabolic steroids. Prog Cardiovasc Dis 41:1–15 Liu PY, Death AK, Handelsman DJ 2003 Androgens and cardiovascular disease. Endocr Rev 24:313–340 Wu FC, Von Eckardstein A 2003 Androgens and coronary artery disease. Endocr Rev 24:183–217 Bruck B, Brehme U, Gugel N, Hanke S, Finking G, Lutz C, Benda N, Schmahl FW, Haasis R, Hanke H 1997 Gender-specific differences in the effects of testosterone and estrogen on the development of atherosclerosis in rabbits. Arterioscler Thromb Vasc Biol 17:2192–2199 Toda T, Toda Y, Cho BH, Kummerow FA 1984 Ultrastructural changes in the comb and aorta of chicks fed excess testosterone. Atherosclerosis 51:47–57 von Dehn G, von Dehn O, Volker W, Langer C, Weinbauer GF, Behre HM, Nieschlag E, Assmann G, von Eckardstein A 2001 Atherosclerosis in apolipoprotein E-deficient mice is decreased by the suppression of endogenous sex hormones. Horm Metab Res 33:110 –114 Adams MR, Williams JK, Kaplan JR 1995 Effects of androgens on coronary artery atherosclerosis and atherosclerosis-related impairment of vascular responsiveness. Arterioscler Thromb Vasc Biol 15:562–570 Obasanjo IO, Clarkson TB, Weaver DS 1996 Effects of the anabolic steroid nandrolone decanoate on plasma lipids and coronary arteries of female cynomolgus macaques. Metabolism 45:463– 468 Elhage R, Arnal JF, Pieraggi MT, Duverger N, Fievet C, Faye JC, Bayard F 1997 17 ␤-estradiol prevents fatty streak formation in apolipoprotein E-deficient mice. Arterioscler Thromb Vasc Biol 17:2679 –2684 Alexandersen P, Haarbo J, Byrjalsen I, Lawaetz H, Christiansen C 1999 Natural androgens inhibit male atherosclerosis: a study in castrated, cholesterol-fed rabbits. Circ Res 84:813– 819 Nathan L, Shi W, Dinh H, Mukherjee TK, Wang X, Lusis AJ, Chaudhuri G 2001 Testosterone inhibits early atherogenesis by conversion to estradiol: critical role of aromatase. Proc Natl Acad Sci USA 98:3589 –3593 Fogelberg M, Bjorkhem I, Diczfalusy U, Henriksson P 1990 Stanozolol and experimental atherosclerosis: atherosclerotic development and blood lipids during anabolic steroid therapy of New Zealand white rabbits. Scand J Clin Lab Invest 50:693– 696 Mukherjee T, Dinh H, Chaudhuri G, Nathan L 2002 Testosterone attenuates expression of vascular cell adhesion molecule-1 by conversion to estradiol by aromatase in endothelial cells: implications in atherosclerosis. Proc Natl Acad Sci USA 99:4055– 4060 McCredie RJ, McCrohon JA, Turner L, Griffiths KA, Handelsman DJ, Celermajer DS 1998 Vascular reactivity is impaired in genetic females taking high-dose androgens. J Am Coll Cardiol 32:1331–1335 Webb CM, Adamson DL, de Zeigler D, Collins P 1999 Effect of acute testosterone on myocardial ischemia in men with coronary artery disease. Am J Cardiol 83:437– 439, A9 Thompson PD, Ahlberg AW, Moyna NM, Duncan B, Ferraro-Borgida M, White CM, McGill CC, Heller GV 2002 Effect of intravenous testosterone on myocardial ischemia in men with coronary artery disease. Am Heart J 143: 249 –256 Rosano GM, Leonardo F, Pagnotta P, Pelliccia F, Panina G, Cerquetani E, della Monica PL, Bonfigli B, Volpe M, Chierchia SL 1999 Acute anti-ischemic effect of testosterone in men with coronary artery disease. Circulation 99: 1666 –1670 Alexandersen P, Haarbo J, Christiansen C 1996 The relationship of natural androgens to coronary heart disease in males: a review. Atherosclerosis 125: 1–13 Beato M, Klug J 2000 Steroid hormone receptors: an update. Hum Reprod Update 6:225–236


27. Libby P 2001 Current concepts of the pathogenesis of the acute coronary syndromes. Circulation 104:365–372 28. Dansky HM, Barlow CB, Lominska C, Sikes JL, Kao C, Weinsaft J, Cybulsky MI, Smith JD 2001 Adhesion of monocytes to arterial endothelium and initiation of atherosclerosis are critically dependent on vascular cell adhesion molecule-1 gene dosage. Arterioscler Thromb Vasc Biol 21:1662–1667 29. McCrohon JA, Jessup W, Handelsman DJ, Celermajer DS 1999 Androgen exposure increases human monocyte adhesion to vascular endothelium and endothelial cell expression of vascular cell adhesion molecule-1. Circulation 99:2317–2322 30. Neish AS, Williams AJ, Palmer HJ, Whitley MZ, Collins T 1992 Functional analysis of the human vascular cell adhesion molecule 1 promoter. J Exp Med 176:1583–1593 31. Meagher L, Mahiouz D, Sugars K, Burrows N, Norris P, Yarwood H, BeckerAndre M, Haskard DO 1994 Measurement of mRNA for E-selectin, VCAM-1 and ICAM-1 by reverse transcription and the polymerase chain reaction. J Immunol Methods 175:237–246 32. Melhus H, Gobl A, Ljunghall S 1994 Competitive PCR demonstrates that 9-cis retinoic acid induces cellular retinoic acid-binding protein-II more efficiently than all-trans retinoic acid in human osteosarcoma cells. Biochem Biophys Res Commun 200:1125–1129 33. Zhang X, Wang LY, Jiang TY, Zhang HP, Dou Y, Zhao JH, Zhao H, Qiao ZD, Qiao JT 2002 Effects of testosterone and 17-␤-estradiol on TNF-␣-induced E-selectin and VCAM-1 expression in endothelial cells. Analysis of the underlying receptor pathways. Life Sci 71:15–29 34. Hatakeyama H, Nishizawa M, Nakagawa A, Nakano S, Kigoshi T, Uchida K 2002 Testosterone inhibits tumor necrosis factor-␣-induced vascular cell adhesion molecule-1 expression in human aortic endothelial cells. FEBS Lett 530:129 –132 35. Simoncini T, Maffei S, Basta G, Barsacchi G, Genazzani AR, Liao JK, De Caterina R 2000 Estrogens and glucocorticoids inhibit endothelial vascular cell adhesion molecule-1 expression by different transcriptional mechanisms. Circ Res 87:19 –25 36. Sasano H, Murakami H, Shizawa S, Satomi S, Nagura H, Harada N 1999 Aromatase and sex steroid receptors in human vena cava. Endocr J 46:233–242 37. Diano S, Horvath TL, Mor G, Register T, Adams M, Harada N, Naftolin F 1999 Aromatase and estrogen receptor immunoreactivity in the coronary arteries of monkeys and human subjects. Menopause 6:21–28 38. Milewich L, Kaimal V, Johnson AR 1987 Steroid 5 ␣-reductase activity in endothelial cells from human umbilical cord vessels. J Steroid Biochem 26: 561–567 39. McKay LI, Cidlowski JA 1998 Crosstalk between nuclear factor-␬B and the steroid hormone receptor mechanisms of mutual antagonism. Mol Endocrinol 12:45–56 40. Palvimo JJ, Reinikainen P, Ikonen T, Kallio PJ, Moilanen A, Janne OA 1996 Mutual transcriptional interference between RelA and androgen receptor. J Biol Chem 271:24151–24156 41. Iiyama K, Hajra L, Iiyama M, DiChiara M, Medoff BD, Cybulsky MI 1999 Patterns of vascular cell adhesion molecule-1 and intercellular adhesion molecule-1 expression in rabbit and mouse atherosclerotic lesions and at sites predisposed to lesion formation. Circ Res 85:199 –207 42. Holm P, Andersen HL, Arroe G, Stender S 1998 Gender gap in aortic cholesterol accumulation in cholesterol-clamped rabbits: role of the endothelium and mononuclear-endothelial cell interaction. Circulation 98:2731–2737 43. Ling S, Dai A, Williams MR, Myles K, Dilley RJ, Komesaroff PA, Sudhir K 2002 Testosterone (T) enhances apoptosis-related damage in human vascular endothelial cells. Endocrinology 143:1119 –1125 44. Fujimoto R, Morimoto I, Morita E, Sugimoto H, Ito Y, Eto S 1994 Androgen receptors, 5 ␣-reductase activity and androgen-dependent proliferation of vascular smooth muscle cells. J Steroid Biochem Mol Biol 50:169 –174 45. Ajayi AA, Mathur R, Halushka PV 1995 Testosterone increases human platelet thromboxane A2 receptor density and aggregation responses. Circulation 91:2742–2747 46. Rosenblum WI, el-Sabban F, Nelson GH, Allison TB 1987 Effects in mice of testosterone and dihydrotestosterone on platelet aggregation in injured arterioles and ex vivo. Thromb Res 45:719 –728

Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the endocrine community.