Journal Logo


EGF-Genistein Inhibits Neointimal Hyperplasia After Vascular Injury in an Experimental Restenosis Model

Trieu, Vuong N.*†; Narla, Rama Krishna; Myers, Dorothea E.†‡; Uckun, Fatih M.†‡

Author Information
Journal of Cardiovascular Pharmacology: April 2000 - Volume 35 - Issue 4 - p 595-605
  • Free


Human epidermal growth factor (EGF) is a 53-aminoacid, single-chain, 6,216-daltons polypeptide, which exerts its biologic effects by binding to a specific 170-kDa cell membrane epidermal growth factor receptor (EGF-R/ErbB-1) (1-4). The human EGF-R consists of an extracellular domain with a high cysteine content and N-linked glycosylation, a single transmembrane domain, and a cytoplasmic domain with protein tyrosine kinase (PTK) activity. Binding of EGF to the EGF-R/ErbB-1 results in receptor dimerization with itself or other members of the Erb-B (subtype I) transmembrane PTK family (e.g., Erb-B2, Erb-B3), resulting in activation with autophosphorylation of the PTK domain (5,6). EGF-R is physically and functionally associated with Src protooncogene family PTK, including p60src(5-7). This association is believed to be an integral part of the signaling events mediated by the EGF-R (7-9). We recently reported that targeting genistein (Gen) (4′,5,7-trihydroxyisoflavone), a naturally occurring tyrosine kinase inhibitor present in soybeans, to the EGF-R-PTK complexes in breast cancer cells by using the EGF-genistein bioconjugate inhibited in vitro proliferation of breast cancer cells and significantly improved tumor-free survival in a severe combined immune deficiency (SCID) mouse xenograft model of human breast cancer (10).

Given the effectiveness of EGF-Gen against breast cancer (10), we explored its use against other proliferative diseases, such as restenosis. Restenosis is a proliferative response of smooth muscle cells to angioplasty-associated vascular injuries (11,12). Angioplasty opens up a stenosed blood vessel by compressing the atherosclerotic lesion with an expanding balloon, and the vessels are kept open mechanically by stents. However, damage to the vascular endothelial lining and fissuring of the underlying atherosclerotic lesion trigger thrombotic events and wound-healing response culminating in narrowing of the vessel in >30% of the patients (11,12). Proliferating smooth muscle cells at the site of injury contribute heavily to the cellular build up in the restenotic lesion (12). Like breast cancer cells, these proliferating VSMCs express high levels of EGF-R in vitro (13). We therefore postulated that the EGF-R on smooth muscle cells may be a suitable target for restenosis prophylaxis by using EGF-Gen. Due to the limited quantity of EGF-Gen, we chose to develop a murine model for restenosis. At the time, there were three other murine models of angioplastic injury: the perivascular electric injury model, the wire-injured aortic model, and the carotid artery ligation model (14-16). These models rely heavily on physical trauma to the blood vessel. We chose to use a gentler method in which photoactivation of rose bengal causes only limited oxygen singlet-induced damage to the vascular endothelial lining. In this model system, we found EGF-R overexpression among the elements of neointima-a condition similar to that of human breast cancer (17,18). We therefore evaluated the effect of EGF-Gen against neointima formation in this model system. Here we first report the characterization of our murine model for restenosis, the overexpression of EGF on murine neointimal cells, and the successful inhibition of neointima formation by the EGF-R-directed tyrosine kinase inhibitor, EGF-Gen.


Murine restenosis model

Three- to four-week-old C57B1/6 male mice (Taconic, Germantown, NY, U.S.A.) were kept in microisolator cages on a 12-h day/night cycle and fed the Paigen's cocoa butter diet (15.8% fat, 1.25% cholesterol, and 0.5% sodium cholate; Harlan Teklad, Madison, WI, U.S.A.) for 2 weeks before induction of restenosis to induce high plasma cholesterol levels similar to those of patients with lipid disorders. Plasma cholesterol increased from 125 ± 21 mg/dl (n = 12) to 346 ± 62 mg/dl (n = 5; p < 0.0001) after 1 week on the cocoa butter diet and remained constant throughout the experiment (370 ± 60 mg/dl, n = 30, at 2 weeks; 371 ± 64 mg/dl, n = 29, at 3 weeks; 267 ± 91 mg/dl, n = 29, at 4 weeks; 337 ± 101 mg/dl, n = 27, at 5 weeks; and 353 ± 71 mg/dl, n = 26, at 6 weeks. Mice were injected with 300 μl of a 3 mg/ml rose bengal solution in sterile PBS via the tail vein and anesthetized with a ketamine/xylazine solution (100 mg ketamine/kg and 5 mg xylazine/kg). Photoactivation of rose bengal by using percutaneous irradiation with a green light source (300 Watt xenon are lamp equipped with a 550-nm broadband interference filter; Oriel Scientific, Stratford, CT, U.S.A.) was performed on the shaved left leg by placing the 3.2-mm glass fiber optic light guide directly onto the left femoral vein/artery for 10 min. We did not expose the femoral artery by a cut-down to avoid desiccation injury. After irradiation, the mice were treated for 28 days with intraperitoneal injections of (a) EGF-Gen at a dose level of 0.1 mg/kg/dose once in the morning and once in the evening, (b) 2 mg/kg/dose of unconjugated genistein dissolved in 10% DMSO once per day, or (c) with PBS. All mice tolerated the 10% DMSO intraperitoneal injection with no signs of toxicity associated with higher doses of DMSO.

At the times indicated, the animals were killed with ketamine/xylazine and perfused with PBS followed by 4% phosphated buffered formalin. PBS and formalin were pumped through the left ventricle and allowed to exit through a 3-mm incision through the anterior wall of the right ventricle. The brain, thymus, heart, lung, liver, pancreas, kidney, spleen, intestine, stomach, muscle, skin, adrenal, and testes were collected for pathological evaluation. Tissue blocks containing the femoral artery/vein were excised and postfixed in 4% phosphate-buffered formalin overnight and processed for hematoxylin and eosin (H&E) or Masson's trichrome staining. Six-micron-thick serial sections spanning the entire injured area (∼600 sections per mouse, 40 sections per slide) were prepared for immunostaining and histochemical analysis. Serial sections were examined for areas of maximal neointimal hyperplasia, for which the ratios of neointima/media were determined. This was necessary to compensate for angular differences among specimens. A Pixera camera (Pixera Corp., CA, U.S.A.) was used for image capture, and the NIH image 1.61 program was used for histomorphometric analysis. Serial sections also were scored for presence/nonpresence of clot, dead media, or neointima at a clearly obvious observational level (i.e., >25% of the lumen occluded by blood clots, >25% of the media dead, and >25% of the arterial surface covered with neointima, respectively).


Immunohistochemistry was performed as described (19). Paraffin-embedded sections were washed 3 times at 10 min each in Hemo De (Fisher Scientific, Pittsburgh, PA, U.S.A.), rinsed in absolute ethanol, and then treated with 0.5% hydrogen peroxide in methanol for 30 min. The slide was then washed in water, incubated in PBS-Tween (PBS with 0.1% Tween-20) for 10 min, followed by blocking in 10% FBS in DMEM for 1 h. For visualization of the EGF receptor, the section was incubated with a 10-μg/ml solution of rabbit polyclonal antibody against mouse EGF receptor (epitope mapping at residues 1005-1016; Santa Cruz Biotech, Santa Cruz, CA, U.S.A.). For visualization of α-actin, the section was incubated with a 1/100 dilution of clone 1A4 mouse ascites fluid, monoclonal anti-α-actin raised against the aminoterminal decapeptide of human α-actin. 1A4 crossreacted with murine α-actin and does not reacted with actin from fibroblast (β- and τ-cytoplasmic), striated muscle (α-sarcomeric), and myocardium (α-myocardial). After 1 h, the sections were washed twice in PBS-Tween for 5 min each, and then incubated an additional hour with a 1:100 dilution of anti-rabbit or anti-mouse horseradish peroxidase conjugated antibody (Pierce, Rockford, IL, U.S.A.). Sections were washed in PBS-Tween and incubated for 10 min in the substrate solution (100 μl of a stock solution of 3-amino-9-ethylcarbazole in N′N′-dimethyl formamide at 2.4 mg/ml, 1 ml of acetate buffer, pH 5.2, and 5 μl of 30% wt/wt hydrogen peroxide). Sections were counterstained with Mayer's hematoxylin and mounted by using Crystal Mount (BioMeda Corp.). The EGF-R and α-actin immunostained slides were quantitated by using a scoring system of negative (0), weak positive (1), positive (2), and strong positive (3).

Preparation of the EGF-Gen

EGF-Gen was produced by conjugating recombinant human EGF (rhEGF) to genistein (Gen) as previously described (7). rhEGF was produced in Escherichia coli harboring a genetically engineered plasmid that contains a synthetic gene for human EGF fused at the NH2 terminus to a hexapeptide leader sequence for optimal protein expression and folding. The rhEGF fusion protein precipitated in the form of inclusion bodies, and the mature protein was recovered by trypsin cleavage, followed by purification with ion exchange chromatography and HPLC. rhEGF was 99% pure by reverse-phase HPLC and sodium dodecylsulfate-polyacrylamide gel electrophoresis (SDS-PAGE) with an isoelectric point of 4.6 ± 0.2. The endotoxin level was 0.172 EU/mg. The recently published photochemical conjugation method (20) using the heterobifunctional photoreactive cross-linking agent, sulfosuccinimidyl 6-(4′azido-2′-nitrophenylamino)hexanoate (Sulfo-SANPAH; Pierce Chemical Co., Rockford, IL, U.S.A.) has been used in the synthesis of the EGF-Gen conjugate. Sulfo-SANPAH-modified rhEGF was mixed with a 10:1 molar ratio of Gen (LC Laboratories, Woburn, MA, U.S.A.; 50 mM solution in DMSO) and then irradiated with gentle mixing for 10 min with UV light at wavelengths of 254-366 nm with a multiband UV light emitter (model UVGL-15 Mineralight; UVP, San Gabriel, CA, U.S.A.). Photolytic generation of a reactive singlet nitrene on the other terminus of EGF-SANPAH in the presence of a 10-fold molar excess of Gen resulted in the attachment of Gen via its available C7-hydroxyl group to lysine 28 or lysine 48 residues of EGF. Excess Gen in the reaction mixture was removed by passage through a PD-10 column, and MI 12,000 EGF-EGF homoconjugates, with or without conjugated Gen, as well as higher molecular weight reaction products were removed by size-exclusion HPLC. Reverse-phase HPLC with a Hewlett-Packard 1100 series HPLC instrument was used for separation of EGF-Gen from EGF-SANPAH. After the final purification, analytic HPLC was performed by using a Spherisorb ODS-2 reverse-phase column (250 × 4-mm; Hewlett-Packard). Before the HPLC runs, a Beckman DU 7400 spectrophotometer was used to generate a UV spectrum for each of the samples to ascertain the λ max for EGF-Gen, EGF-SANPAH, and unmodified EGF. Each HPLC chromatogram was subsequently run at wavelengths of 214,265, and 480 nm by using the multiple wavelength detector option supplied with the instrument to ensure optimal detection of the individual peaks in the chromatogram. Analysis was achieved by using a gradient flow consisting of 0-100% eluent in a time interval of 0-30 min. Five-microliter samples applied to this column were run by using the following gradient program: 0-5 min, 0-20% eluent; 5-20 min, 20-100% eluent; 25-30 min, 100% eluent; and 30-35 min, 100%-0 eluent. The eluent was a mixture of 80% acetonitrile (CH3CN), 20% H2O, and 0.1% TFA.

Electrospray ionization mass spectrometry (21,22) was performed by using a PE SCIEX API triple quadruple mass spectrometer (Norwalk, CT, U.S.A.) to determine the stoichiometry of Gen and EGF in EGF-Gen. 125I-Gen also was used to confirm the stoichiometry of Gen and EGF in EGF-Gen and to verify the removal of free genistein and genistein-labeled EGF-EGF homoconjugates by the described purification procedure. Gen (in 65% ethanol, 35% PBS, pH 7.5; LC Laboratories, Woburn, MA, U.S.A.) was radioiodinated at room temperature in Reacti-Vials containing Iodo-beads (Pierce Chemical Co., Rockford, IL, U.S.A.) and 125I (Na, carrier-free, 17.4 Ci/mg; NEN, Boston, MA, U.S.A.) as per manufacturer's instructions (20,23). The purity of EGF-125I-Gen was assessed by SDS-PAGE (10% separating gels, nonreducing conditions) and autoradiography using intensifying screens and Kodak XAR-5 film. EGF-125I-Gen was also used for in vitro ligand-binding assays (20,23) and EGF-Gen internalization studies to confirm the functionality of EGF-Gen.

Confocal laser scanning microscopy

Differentiated vascular smooth muscle cells (VSMCs; Cascade Biologics, Inc, Portland, OR, U.S.A.) were maintained in Medium 231 supplemented with SMDS, a differentiation supplement containing fetal bovine serum and heparin. Immunofluorescence was used to study the expression of EGF receptors on human VSMCs by using a monoclonal antibody to EGF receptor, previously described in detail (24). α-Actin was visualized by using a monoclonal antibody against α-actin, clone asm-1 from Boehringer Mannheim GmbH, Philadelphia, PA, U.S.A. Cells were counterstained with propidium iodide (PI) and the coverslips were mounted with Vectashield (Vector Labs, Burlingame, CA, U.S.A.). Cells were viewed with a confocal laser scanning microscope (Bio-Rad MRC 1024) mounted in a Nikon Labhophot upright microscope. Digital images were saved on a Jaz disk and processed with Adobe Photoshop software (Adobe Systems, Mountain View, CA, U.S.A.).

Cytotoxicity/cell proliferation assays

MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide) assays (Boehringer Mannheim Corp., Indianapolis, IN, U.S.A.) (25) were used to determine the effect of EGF-Gen on the viability of VMSCs. In brief, VMSCs were seeded into a 96-well plate at a density of 2.5 × 104 cells/well and incubated for 36 h at 37°C before drug exposure. On the day of treatment, culture medium was carefully aspirated from the wells and replaced with fresh medium containing unconjugated EGF, or EGF-Gen at concentrations ranging from 0.1 to 100 μg/ml. Triplicate wells were used for each treatment. The cells were incubated with the drugs for 24 h at 37°C in a humidified 5% CO2 atmosphere. To each well, 10 μl of MTT (0.5 mg/ml final concentration) was added, and the plates were incubated at 37°C for 4 h to allow MTT to form formazan crystals by reacting with metabolically active cells. The formazan crystals were solubilized overnight at 37°C in a solution containing 10% SDS in 0.01 M HCl. The absorbance of each well was measured in a microplate reader (Labsystems) at 540 nm and at a reference wavelength of 690 nm. To translate the OD540 values into the number of live cells in each well, the OD540 values were compared with those on a standard curve of OD540 versus VSMC cell number. The percentage survival was calculated by using the formula: % survival = Live cell number[test]/Live cell number [control] × 100.

In vitro migration assays

The in vitro migration of human VSMCs was assayed by using a previously published method, which uses Matrigel-coated Costar 24-well transwell cell-culture chambers ("Boyden chambers") with a 8.0-μm pore polycarbonate filter insert (24). The chamber filters were coated with 50 μg/ml of Matrigel matrix, incubated overnight at room temperature under a laminar flow hood and stored at 4°C. On the day of the experiment, the coated inserts were rehydrated with 0.5 ml serum-free DMEM containing 0.1% bovine serum albumin for 1-2 h. To study the effects of EGF-Gen on migration of VSMCs, VSMCs were incubated in triplicate with EGF-Gen at various concentrations ranging from 0.25 to 1 μg/ml or unconjugated EGF at 1 μg/ml overnight. The cells were trypsinized, washed twice with serum-free DMEM containing BSA, counted, and resuspended at 1 × 105 cells/ml; 5 × 104 cells in a serum-free DMEM containing EGF-Gen or PBS were added to the Matrigel-coated and rehydrated filter inserts. Next, 750 μl of the NIH fibroblast-conditioned medium was placed as a chemoattractant in 24-well plates, and the inserts were placed in wells and incubated at 37°C for 48 h. After the incubation period, the filter inserts were removed, the medium was decanted off, and the cells on the top side of the filter that did not migrate were scraped off with a cotton-tipped applicator. The cells that migrated to the lower side of the filter were fixed, stained with Hema-3 solutions, and counted under the microscope. Five to 10 random fields per filter were counted by a blinded individual to determine the mean (±SEM) values for the migrating fraction. The migrating fractions of cells invading Matrigel treated with EGF-Gen were compared with those with PBS-treated or EGF-treated control cells.

Statistical analysis

All data are shown as mean ± SEM. Statistical analysis between groups was performed by Mann-Whitney t test or Fisher's exact test by using Instat (GraphPad Software, San Diego, CA, U.S.A.). Cumulative percentage plots were analyzed by using the Wilcoxon statistic by Statview (Abacus Concepts, Inc., Berkeley, CA, U.S.A.).


Establishment of a murine model of vascular injury-induced neointimal hyperplasia

Our experimental model of vascular injury-induced neointimal hyperplasia is schematically described in Fig. 1A. Within 10 min after irradiation, the femoral artery was occluded by a visible 3-mm-long thrombus in eight of eight mice (100%), and occlusive thrombus was histologically confirmed in two of two mice (100%) examined at 1 h after irradiation (Fig. 1B and C). Even 16 h after irradiation, the femoral artery was clot-occluded in seven (86%) of eight mice examined (Fig. 1B and C). This occlusive thrombus in the femoral artery underwent fibrinolysis with complete restoration of arterial blood flow in the majority of mice. Occlusive thrombi were found in none of the three mice examined at 1 week, only one (7%) of 15 mice examined at 2 weeks, and none of the 45 mice examined at 4 weeks (Fig. 1B and C). At 16 h, concomitant with the fibrinolysis of the occluding clot, the media of the injured femoral artery became necrotic, with disappearance of the smooth muscle cells in six (75%) of eight mice examined (Fig. 1). The media was eventually repopulated by an influx of highly proliferative myofibroblasts and the frequency of microscopically detectable media necrosis progressively decreased from 100% (three of three mice) at 1 week to 80% (12 of 15 mice) at 2 weeks, and 33% (15 of 45 mice) at 4 weeks (Fig. 1; p = 0.0024, 2 vs. 4 weeks). Neointima formation was excessive, and a pericentric neointimal hyperplasia was observed in 13 (87%) of 15 mice at 2 weeks and 45 of 45 mice at 4 weeks after vascular injury (Fig. 1).

FIG. 1
FIG. 1:
An experimental mouse model of vascular injury-induced neointimal hyperplasia. A: Schematic illustration of neointima formation after vascular injury. Photoactivation of rose bengal, a xanthene-type dye structurally related to fluorescein, results in generation of reactive oxygen intermediates. Percutaneous irradiation of the left femoral artery with a green light source after the intravenous administration of rose bengal results in blood vessel injury, evidenced by destruction of the vascular endothelium. The denuded vessel wall triggers platelet aggregation and thrombus formation at the site of vascular injury. B: The injured femoral arteries were scored for the presence of neointima, dead media, or clot at 1 h (n = 2), 16 h (n = 8), 1 week (n = 4), 2 weeks (n = 15), and 4 weeks (n = 45), and the percentages of animals scored positive are shown. C: Microscopic analysis of murine restenosis. Top: Views through a dissecting microscope showing normal unclotted femoral artery at 0 h, clotted artery at 1 h, and at 16 h after irradiation. The arrows indicate the 3-mm thrombus induced by irradiation. Middle and bottom: H&E-stained slides at ×400 magnification showing the artery at 0 h, 1 h, 16 h, 1 week, 2 weeks, and 4 weeks. Arrows indicate the internal elastic lamina. Neointima (n) and thrombus (t) also are indicated.

Upregulated EGF-receptor expression on myofibroblasts of the hyperplastic neointima after vascular injury

The cellular elements of the hyperplastic neointima generated after vascular injury were positive for α-actin, a marker for VSMCs (i.e., myofibroblasts) in eight of eight mice analyzed, with a mean score of 2.5 ± 0.2, similar to that of uninjured right femoral arteries (mean score of 2.6 ± 0.2, n = 8; Fig. 2). Among these immunocytochemically identified myofibroblasts of the neointima, those closest to the lumen were spindle-shaped, longitudinally aligned with the lumen, and showed more intense staining with the anti-α-actin antibody. Myofibroblasts closest to the media were stellate, aligned randomly, and stained weakly for α-actin. The myofibroblasts of the neointima were EGF-receptor positive in eight of eight mice (100%) analyzed, with a mean score (2.6 ± 0.2, n = 8) significantly higher than that of the underlying media (0.9 ± 0.2, n = 8, p < 0.0001) and that of the media from uninjured femoral arteries (0.8 ± 0.1, n = 9, p < 0.0001; Fig. 2).

FIG. 2
FIG. 2:
Amplified epidermal growth factor (EGF)-receptor expression in neointimal hyperplasia. The neointima mass stained positively with a rabbit polyclonal anti-EGF receptor antibody (D). The cellular elements of the hyperplastic neointima generated after vascular injury were positive for α-actin, a marker for vascular smooth muscle cells (i.e., myofibroblasts) in seven of seven mice (100%) analyzed (C). The neointima stained negatively with control rabbit antibody (B) raised against human EGF peptide or control mouse monoclonal antibody raised against human collagen IV (A). The uninjured right femoral artery strongly stained with anti α-actin (E) and only weakly with anti-EGF-R antibody (F). The arrows indicate the internal elastic lamina.

Effects of EGF-Gen on neointimal hyperplasia after vascular injury

We next examined the effects of EGF-Gen on obstructive neointimal hyperplasia after vascular injury. As reflected by the distribution patterns of the cumulative percentage versus stenosis index plots, EGF-Gen significantly inhibited the development of neointimal hyperplasia after vascular injury (p = 0.0005, Wilcoxon statistics; Fig. 3, Table 1). In 75% of the EGF-Gen-treated mice, the maximal stenosis index was 0.44 ± 0.13, whereas in 75% of PBS-treated mice, the maximal stenosis index was 1.20 ± 0.25. The mean neointima/media ratios for areas of maximal neointimal hyperplasia were 0.59 ± 0.16 (n = 24) for the EGF-Gen-treated group, 0.99 ± 16 (n = 45) for the PBS group (EGF-Gen vs. PBS; p = 0.0017, two-tailed Mann-Whitney t test) and 1.03 ± 18 (n = 8) for group treated with unconjugated genistein (EGF-Gen vs. genistein, p = 0.0088, two-tailed Mann-Whitney t test). EGF-Gen treatment of mice with vascular injury to the left femoral artery also was not associated with any clinical signs of toxicity, and no histopathologic lesions were found in any of the organs of the EGF-Gen-treated mice. Notably, the uninjured right femoral artery of EGF-Gen-treated mice (n = 20) did not exhibit any necrosis of the media or other histopathologic changes (data not shown).

FIG. 3
FIG. 3:
Inhibition of neointima formation by epidermal growth factor-genistein (EGF-Gen) at 4 weeks. Cumulative percentage plot of the neointima/media ratio of the untreated group (open circles), the EGF-Gen-treated group (solid circles), and genistein-treated group (solid triangles) are shown. The insets are representative H&E-stained arteries from animals treated with PBS, unconjugated genistein, or EGF-Gen. Arrows indicate the internal elastic lamina.
Effect of EGF-Gen on vascular injury-injuced neointimal hyperplasia

Effects of EGF-Gen on vascular smooth muscle cell migration in vitro

Analysis of VSMCs stained with a monoclonal antibody to EGF-receptor by confocal laser scanning microscopy showed membrane-associated bright green fluorescence consistent with surface expression of EGF receptors (Fig. 4). The incubation of VSMCs with EGF-Gen at a concentration range of 0-100 μg/ml did not have any significant effect on VSMC viability or proliferation in vitro; and stimulation of VSMC proliferation by unconjugated EGF was only marginal (Fig. 5). However, incubation of VSMCs with EGF-Gen at a concentration of 2.4 μg/ml disrupted the α-actin stress fibers, changing them from organized linear actin cables into a random actin net (Fig. 6). As the α-actin stress fibers are necessary for VSMC migration (26,27), we evaluated the effect of EGF-Gen on VSMC migration in vitro. As shown in Fig. 7, whereas unconjugated EGF significantly promoted the migration of VSMCs, EGF-Gen inhibited the migration of VSMCs in a concentration-dependent fashion with complete cessation of in vitro migration at 1 μg/ml. These findings provide direct evidence that the EGF-R plays a pivotal role in the migration capacity of VSMCs.

FIG. 4
FIG. 4:
Cell-surface expression of epidermal growth factor (EGF) receptor (EGF-R) on vascular smooth muscle cells. Cells were fixed in 2% paraformaldehyde and immunostained with a monoclonal antibody to EGF-R (A; green fluorescence). The immunostained cells were analyzed with a laser scanning confocal microscope. Red fluorescence represents nuclei stained with PI. Controls (CON) in which primary antibody was omitted showed no green fluorescence (B).
FIG. 5
FIG. 5:
Effect of epidermal growth factor (EGF) or EGF-Gen on vascular smooth muscle cells (VSMCs). VSMCs were incubated with either unconjugated EGF (open circles) or EGF-Gen (solid circles) at a concentration range of 0-100 μg/ml for 24 h and cell survival determined by MTT assay. No significant inhibition or stimulation of VSMC proliferation was observed after EGF-Gen treatment.
FIG. 6
FIG. 6:
Disruption of α-actin linear bundles by epidermal growth factor-genistein (EGF-Gen). Cells treated with EGF-Gen at a concentration of 2.4 μg/ml for 48 h exhibited a random network of α-actin (B). In contrast, PBS-treated cells exhibited orderly linear bundles of α-actin (A).
FIG. 7
FIG. 7:
Effects of epidermal growth factor-genistein (EGF-Gen) on vascular smooth muscle cell migration through Matrigel matrix. Vascular smooth muscle cells were incubated with 0.25 μg/ml, 0.5 μg/ml, or 1 μg/ml EGF-Gen or 1 μg/ml human recombinant EGF overnight, trypsinized, and then processed for the migration assay by using Matrigel matrix-coated Boyden chambers. A-E: Morphologic features of migrated vascular smooth muscle cells through Matrigel matrix-coated Boyden chamber filters. F: The cells that migrated through the Matrigel were counted and expressed as percentage of controls (CON) treated with PBS. EGF-Gen inhibited the migration of vascular smooth muscle cells.


Because of its remarkably low morbidity and mortality, revascularization of obstructed coronary arteries by percutaneous transluminal coronary angioplasty (PTCA) has become an integral component of front-line treatment programs for patients with ischemic heart disease (28). Although acute complications of PTCA have markedly declined with optimized use of anticoagulants, antispasmodic agents, and intravascular stents, the incidence of coronary artery restenosis has remained at 30-50% and represents the major obstacle to a more successful outcome of PTCA (29). Therefore, the development of effective strategies for restenosis prophylaxis has become a focal point for translational cardiovascular research (29). The pathogenesis of restenosis has been compared with an exaggerated wound-healing response with migration of smooth muscle cells from the media to the intima of the revascularized coronary artery, where they proliferate and cause an obstructive neointimal hyperplasia (30). Inhibition of VSMC proliferation by a platelet-derived growth factor (PDGF)-antagonist and PDGF-receptor neutralizing antibody have generated promising results in rodent and monkey, thereby confirming the biologic importance of VSMCs in the pathophysiology of restenosis (31,32). EGF-Gen is a targeted biotherapeutic agent directed against the EGF-R, which inhibits both EGF-R tyrosine kinase and the EGF-R-associated Src family PTK (7,10). The specificity of EGF-Gen for EGF-R and its effects on biochemical signal-transduction events in EGF-R-expressing cells have been previously published in detail (7,10). In our model of restenosis, neointimal myofibroblasts were shown to express high levels of EGF-R, and EGF-Gen was effective as an inhibitor of neointimal hyperplasia after vascular injury. The results presented herein indicate that EGF-R is upregulated in murine neointimal hyperplasia and that the EGF-R-directed protein tyrosine kinase (PTK) inhibitor, EGF-Gen, might be useful as a prophylactic agent for prevention of restenosis in clinical settings. Our findings also provide unprecedented evidence that the function of the EGF-R tyrosine kinase affects the ability of the VSMCs to migrate to intima of the injured blood vessels and cause neointimal hyperplasia. The dose level for EGF-Gen in our study was identical to the effective dose level of EGF-Gen as an antiproliferative agent against xenografted human breast cancer in SCID mice (10). This dose level is 25-fold lower than the highest tested and nontoxic dose level of EGF-Gen in mice (7), and it is not associated with any toxicity in cynomolgus monkeys (10).

Unlike EGF-Gen, unconjugated genistein administered at a dose level of 270-fold molar excess of EGF-Gen dose level did not inhibit neointima formation. We have previously reported that targeting genistein to surface receptors associated with tyrosine kinases results in markedly improved biologic activity (7,10,20,23). EGF-Gen is >1,000-fold more potent than unconjugated genistein and inhibits the in vitro proliferation of EGF-R-positive MDA-MB-231 cells at nanomolar concentration (IC50 values for inhibition of proliferation: EGF-Gen vs. Gen, 30 ± 3 nM, vs. 120 ± 18 μM; p < 0.001) (10). This may in part due to the delivery of more Gen molecules to target cells, thereby increasing the intracellular Gen concentration, by this targeted biotherapy approach. We further postulated that the binding of EGF-Gen to the EGF-R brings Gen in direct contact with EGF-R tyrosine kinase as well as Src family PTK associated with the EGF-R. The inhibitor is held in close proximity to the EGF-R and associated PTK because of its covalent attachment to EGF. Localization of the Gen molecule in close proximity to the ATP-binding domains of the EGF-R-associated PTK may increase the effective binding constant by both reducing entropy and providing additional linker binding contacts and lead to sustained inhibition of the PTK. Decreasing the effective off-rate of Gen by conjugating it to EGF also may promote covalent modification of the EGF-R-associated PTK, reminiscent of the oxidative inactivation of CD19-associated Src family PTK by B43-Gen, an anti-CD19 antibody-Gen immunoconjugate (20). Therefore, it is not surprising that only EGF-Gen, but not unconjugated genistein, was able to inhibit neointima formation.

Our mouse model for restenosis has many features amenable to biologic screening of drugs and drug combinations for the prevention of restenosis. In this model system, irradiation of the right femoral artery with cold green light after the intravenous administration of rose bengal results in generation of reactive oxygen singlets, which destroy the vascular endothelium, triggering platelet aggregation and thrombus formation at the site of vascular injury followed by neointimal formation: a condition akin to PTCA-induced vascular injury (33). This is the first description of this model in a hypercholesterolemic mouse. The application of this model to transgenic and knockout mice should allow the genetic dissection of factor(s) involved in restenosis. One common thread among restenosis literature is the failure of agents shown to be effective against restenosis in animal models to exhibit comparable activity in humans (34,35). This may be due to the choice of the animal model used for screening new agents (36). The commonly used animal model, the rat, is primarily a proliferative model whereby the response to vascular injury is a rapid proliferation of medial smooth muscle cells (37). In contrast, neointima formation in clinical settings is a slow process, and mitotic figures are rare in human neointima, suggesting that cellular recruitment to the site of injury is more important than cellular proliferation. In patients, angioplastic injury is frequently followed by thrombus formation. The thrombus serves as a scaffold for the accumulation of fibroblasts, leukocytes, and smooth muscle cells that form the neointima (38). The most striking finding of our study is the observed inhibition of neointima formation by an antimigratory agent (EGF-Gen), which has no antiproliferative or cytotoxic effects on VSMCs. These data indicate that migration of VSMCs plays a prominent role in our mouse model of restenosis, making this model comparable to human restenosis.


1. Flx SB. The epidermal growth factor receptor as a prognostic marker: results of 370 patients and review of 3009 patients. Breast Cancer Res Treat 1994;29:41-9.
2. Earp HS, Dawson TL, Li X, Xu H. Heterodimerization and functional interaction between EGF receptor family members: a new signaling paradigm with implications for breast cancer research. Breast Cancer Res Treat 1995;35:115-32.
3. Wright JD, Reuter CW, Weber MJ. An incomplete program of cellular tyrosine kinase phosphorylation by kinase-defective epidermal growth factor receptors. J Biol Chem 1995;270:12085-93.
4. Broome MA, Hunter T. Requirement for c-Src catalytic activity and the SH3 domain in platelet-derived growth factor BB and epidermal growth factor mitogenic signaling. J Biol Chem 1996;271:16798-806.
5. Muthuswamy SK. Mammary tumors expressing the neu protooncogene possess elevated cSrc tyrosine kinase activity. Mol Cell Biol 1994;14:735-43.
6. Ottenhoff-Kalff AK, Rijksen G, van Beurden EA, Hennipman A, Michels AA, Staal GE. Characterization of protein tyrosine kinases from human breast cancer: involvement of the cSrc oncogene product. Cancer Res 1992;52:4773-8.
7. Uckun FM, Narla RK, Jun X, et al. Cytotoxic activity of EGF-genistein against human breast cancer cells. Clin Cancer Res 1998;4:901-12.
8. Uckun FM, Evans WE, Forsyth CJ, et al. Biotherapy of B-cell precursor leukemia by targeting genistein to CD19-associated tyrosine kinases. Science 1995;267:886-91.
9. Aikyama T, Ishida J, Nakagawa S, et al. Genistein: a specific inhibitor of tyrosine-specific protein kinases. J Biol Chem 1987;262:5592-5.
10. Uckun FM, Narla R, Zeren T, et al. In vivo toxicity, pharmacokinetics, and anti-cancer activity of genistein linked to recombinant human epidermal growth factor. Clin Cancer Res 1998;4:1125-34.
11. Schwartz RS. Pathophysiology of restenosis: interaction of thrombosis, hyperplasia, and/or remodeling. Am J Cardiol 1998;81:14E-7E.
12. Schwartz RS. The vessel wall reaction in restenosis. Semin Interv Cardiol 1997;2:83-8.
13. Saltis J, Thomas AC, Agrotis A, Campbell JH, Campbell GR, Bobik A. Expression of growth factor receptors in arterial smooth muscle cells: dependency on cell phenotype and serum factors. Atherosclerosis 1995;118:77-87.
14. Carmeliet P, Moons L, Stassen JM, et al. Vascular wound healing and neointima formation induced by perivascular electric injury in mice. Am J Pathology 1997;150:761-76.
15. Kumar A, Lindner V. Remodeling with neointima formation in the mouse carotid artery after cessation of blood flow. Arterioscler Thromb Vasc Biol 1997;17:2238-44.
16. Lindner V, Fingerle J, Reidy MA. Mouse model of arterial injury. Circ Res 1993;73:792-6.
17. Okamura K, Kobayashi I, Matsuo K, et al. Immunohistochemical localization of cathepsin D, proliferating cell nuclear antigen and epidermal growth factor receptor in human breast carcinoma analysed by computer image analyser: correlation with histological grade and metastatic behaviour. Histopathology 1997;31:540-8.
18. Schroeder W, Biesterfeld S, Zillessen S, Rath W. Epidermal growth factor receptor: immunohistochemical detection and clinical significance for treatment of primary breast cancer. Anticancer Res 1997;17:2799-802.
19. Trieu VN, Uckun FM. Apolipoprotein(a): a link between atherosclerosis and tumor angiogenesis. Biochem Biophys Res Commun 1999;257:714-8.
20. Uckun FM, Evans WE, Forsyth CJ, et al. Biotherapy of B-cell precursor leukemia by targeting genistein to CD19-associated tyrosine kinase. Science 1995;267:886-91.
21. Feng R, Konishi Y, Bell AW. High accuracy molecular weight determination and variation, characterization of proteins up to 80 KU by ion spray mass spectrometry. J Am Soc Mass Spectrom 1991;2:387-401.
22. Covay TR, Bonner RS, Shusan BI, Henion JD. Determination of protein oligonucleotides and peptide molecular weights by ion spray mass spectrometry. Rapid Commun Mass Spectrom 1988;2:249-56.
23. Myers DE, Jun X, Waddick KG, et al. Membrane-associated CD19-LYN complex is an endogenous p53-independent and bcl-2 independent regulator of apoptosis in human B-lineage lymphoma cells. Proc Natl Acad Sci U S A 1995;92:9575-9.
24. Narla RK, Liu X-P, Klis D, Uckun FM. Inhibition of human glioblastoma cell adhesion and invasion by 4-(4′-hydroxylphenyl)-amino-6,7-dimethoxyquinazoline (WHI-P131) and 4-(3′-bromo-4′-hydroxylphenyl)-amino-6,7-dimethoxyquinazoline (WHI-P154). Clin Cancer Res 1998;4:2463-71.
25. Narla RK, Liu X, Myers D, Uckun FM. 4-(3′-Bromo-4′hydroxylphenyl)-amino-6,7-dimethoxyquinazoline: a novel quinazoline derivative with potent cytotoxic activity against human glioblastoma cells. Clin Cancer Res 1998;4:1405-14.
26. Rocnik EF, Chan BMC, Pickering JG. Evidence for a role of collagen synthesis in arterial smooth muscle cell migration. J Clin Invest 1998;101:1889-98.
27. Engel L, Ryan U. TGF-beta 1 reverses PDGF-stimulated migration of human aortic smooth muscle cells in vitro. In Vitro Cell Dev Biol Anim 1997;33:443-51.
28. Vaitkus PT. The continuing evolution of percutaneous transluminal coronary angioplasty in the treatment of coronary artery disease. Coron Artery Dis 1995;6:429-39.
29. Landzberg BR, Frishman WH, Lerrick K. Pathophysiology and pharmacological approaches for prevention of coronary artery restenosis following coronary artery balloon angioplasty and related procedures. Progr Cardiovasc Dis 1997;39:361-298.
30. Ueda M, Becker AE, Naruko T, Kojima A. Smooth muscle cell de-differentiation is a fundamental change preceding wound healing after percutaneous transluminal coronary angioplasty in humans. Coron Artery Dis 1995;6:71-81.
31. Mullins DE, Hamud F, Reim R, Davis HR. Inhibition of PDGF receptor binding and PDGF-stimulated biological activity in vitro and of intimal lesion formation in vivo by 2-bromomethyl-5-chlorobenzene sulfonylphthalimide. Arterioscler Thromb Vasc Biol 1994;14:1047-55.
32. Giese NA, Marjianowski MM, McCook O, et al. The role of alpha and beta platelet-derived growth factor receptor in the vascular response to injury in nonhuman primates. Arterioscler Thromb Vasc Biol 1999;19:900-9.
33. Trieu VN, Uckun FM. Genistein is neuroprotective in murine models of familial amyotrophic lateral sclerosis and stroke. Biochem Biophys Res Commun 1999;258:685-8.
34. Vaitkus PT. Assessment of factors important in atherosclerotic occlusion and restenosis. Coron Artery Dis 1995;6:429-39.
35. Landzberg BR, Frishman WH, Lerrick K. Pathophysiology and pharmacological approaches for prevention of coronary artery balloon angioplasty and related procedures. Prog Cardiovasc Dis 1997;39:361-298.
36. Schwartz SM, Reidy MA, O'Brien ERM. Assessment of factors important in atherosclerotic occlusion and restenosis. Thromb Haemost 1995;74:541-51.
37. Schwartz RS, Chu A, Edwards WD, et al. A proliferation analysis of arterial neointimal hyperplasia: lessons for antiproliferative restenosis therapies. Int J Cardiol 1996;53:71-80.
38. Miller DD, Karim MA, Edwards WD, Schwartz RS. Relationship of vascular thrombosis and inflammatory leukocyte infiltration to neointimal growth following porcine coronary artery stent placement. Atherosclerosis 1996;124:145-55.

EGF-Gen; Neointimal hyperplasia; Restenosis

© 2000 Lippincott Williams & Wilkins, Inc.