Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Abnormal vascular SMC proliferation is a pathological feature of restenosis that limits the efficacy of PTCA and of transplantation (Shirotani et al., 1993). This increased SMC growth contributes to the formation of a neointimal layer found in restenotic vessels following balloon or immune injury (Schwartz and Laiw, 1993). Growth factors such as PDGF and bFGF have been shown to control specific events in the neointima development. Jackson and Schwartz (1992) have proposed that bFGF regulates the initial mitogenic response of medial SMCs after endothelial denudation although PDGF controls the migratory response of medial SMCs into the intima. Lindner and Reidy (1991) showed that antibodies to bFGF decreased the proliferative response seen after balloon injury by ~80%. Antibodies to PDGF also inhibited intimal formation after injury; suggesting that PDGF has a role in the VSMC proliferation as well as migration (Ferns et al., 1991). With PDGF and bFGF controlling a significant part of the vascular neointima development, EGF, which is released from platelets, may also have a mitogenic effect on VSMC in the lesion (Klagsbrun and Edelman, 1989).Understanding the mechanistic roles of PDGF, bFGF and EGF in the formation of the neointima are important steps in the development of therapeutic agents that will abrogate abnormal proliferation.

We and others have discovered specific inhibitors of the PDGF and FGF receptor PTK (Connolly et al., 1996; Schroeder et al., 1996; Klutchko et al., 1996; Levitzki and Gazit, 1995). Activity of the FGF receptor tyrosine kinase has been difficult to demonstrate due to the lack of specific FGF receptor antibodies and also the low copy number of FGF receptors in VSMC (Zhan et al., 1993). Tyrosine kinase inhibitors have been shown to impair bFGF signaling in cultured coronary endothelial cells by blocking DNA synthesis (Hawker and Granger, 1994). Inhibition of PDGF receptor tyrosine kinase activity has been reported for several classes of protein tyrosine kinase inhibitors including quinoxalines (Kovalenko et al., 1994), tyrphostins (Bilder et al., 1993) and 2-phenylaminopyrimidines (Buchdunger et al., 1995). The specificity or potency of these PTK inhibitors, except for the 2-phenylaminopyrimidine class, has been limited. We describe the effects of one of the 2-phenylaminopyrimidines, CGP 53716 (fig. 1), on growth responses of cultured RASMC and 3T3 fibroblasts and its relative growth factor selectivity. CGP 53716 is shown to nonselectively inhibit DNA synthesis by PDGF-BB, bFGF or EGF in RASMC with nearly similar potency for each growth factor. A nonselective tyrosine kinase inhibitor may have potential clinical application for the multi- growth factor-mediated vascular proliferation following vascular injury or atherosclerosis.

View larger version (6K): [in this window] [in a new window]   Fig. 1.   Structure of CGP 53716; 2-phenylaminopyridine class of protein tyrosine kinase inhibitors.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Reagents. CGP 53716, N-[4-methyl-3-(4-pyridin-3-yl-pyrimidin-2-ylamino)-phenyl]- benzamide), was prepared by the Parke-Davis Pharmaceutical Research Division, Chemistry Department according to previously described methods (Zimmermann et al., 1996). CGP 53716 was freshly prepared each experimental day by dissolving in dimethyl sulfoxide. The final concentration of dimethyl sulfoxide in the experimental cell cultures was 1% for DNA synthesis and 0.5% for cell proliferation studies. Controlled process serum replacement-2 was obtained from Sigma Chemical Company (St. Louis, MO). [³H]-Thymidine (specific activity 6.7 Ci/mM), was obtained from New England Nuclear (Boston, MA). The rabbit polyclonal antibodies to phospho-MAPK and p42/p44 proteins were obtained from New England Biolabs (Beverly, MA). The monoclonal antibody to c-Fos protein was from Oncogene Science, Inc (Cambridge, MA). PDGF-BB, bFGF, EGF, the sheep polyclonal antibody to the EGF receptor, rabbit polyclonal antibody to the PDGF-AB receptor, the neutralizing goat polyclonal antibody to the PDGF-AB ligand and the monoclonal antibody for phosphotyrosine were obtained from Upstate Biotechnology, Inc (Lake Placid, NY). All other reagents used for these experiments were of the highest commercial purity available.

Cells and culture conditions. Smooth muscle cells were isolated from the thoracic aorta of rats (rat aortic smooth muscle cell) and explanted according to previous methods (Ross, 1971). Balb/3T3 mouse fibroblast cells (3T3; unknown passage) were originally obtained from ATCC cell lines (ATCC, Rockville, MD). Both RASMC and 3T3 cells were grown in DMEM (Gibco BRL, Gaithersburg, MD) containing 10% FBS (HyClone, Logan, UT), 1% glutamine (Gibco, Gaithersburg, MD) and 1% penicillin/streptomycin (Gibco). Cells were identified as smooth muscle cells by their "hill and valley" growth pattern and by immunocytochemical staining with a monoclonal antibody specific for SMC -actin (Clone HHF35; Dako Co., Carpinteria, CA). RASMC used in this study were from two isolates and between passages 5 and 16.

Cell growth assay. RASMC and 3T3 cells were plated into a 24-well plate (10,000 cells/well) in DMEM with 10% FBS. Cells were maintained in DMEM/10% FBS throughout the experiment. After a 24-hr attachment period, the following additions of vehicle or CGP 53716 were made daily and cells counted using an automated cell counter (Coulter Electronics, Miami Lakes, FL) on 1, 3, 6 and 8 days. Control cells (10% FBS) were exposed to vehicle (0.5% DMSO) through day 8. A second group (10% FBS + CGP 53716) was exposed to 0.1 to 10 µM CGP 53716 throughout the study. The third group was exposed to 3 µM CGP 53716 and 10% FBS for days 1 to 3, then exposed to vehicle and 10% FBS for days 4 to 8. Triplicate determinations were made for each data point.

DNA synthesis assay. [Methyl-³-H]-thymidine incorporation assays were performed by a modified method previously described (Sachinidis et al., 1990). RASMC and 3T3 cells were plated into a 24-well plate (30,000 cells/well) in DMEM with 10% FBS. After the cells reached confluence, they were made quiescent by incubation in DMEM containing 0.2% FBS for another 24 hr to synchronize cells in G_o/G₁ phase of the cell cycle. The quiescent status of RASMC and 3T3 cells after this 24-hr period was previously verified using flow cytometry; no further reduction in the number of cells entering S-phase of the cell cycle was observed between 24 and 48 hr in quiescent medium (T.C. Major, unpublished results). Incubation with growth factors (i.e., PDGF-BB, EGF and bFGF) in the absence and presence of varying concentrations of CGP 53716 was carried out in 0.5 ml/well serum-substituted medium (DMEM containing 1% CPSR-2; Sigma Chemical Co., St. Louis, MO). CGP 53716 was added at the same time as the growth factor additions. After 18 hr, 0.25 µCi/well [³H]-thymidine was added. Four hours later (22 hr total) the incubation was stopped by removing the radioactive media, washing the cells twice with 1 ml cold phosphate-buffered saline and then washing twice with cold 5% trichloroacetic acid. The acid-insoluble fraction was dissolved in 0.75 ml 0.25 N NaOH and the radioactivity determined by liquid scintillation counting (Tri-Carb, Packard Instrument Co., Downers Grove, IL). Within each assay, triplicate determinations were made. The dpm for CGP 53716-treated cultures were expressed as a percentage of dpm in control wells (growth factor alone).

Preparation of cellular lysates. RASMC and 3T3 cells were seeded in 10% FBS/DMEM at a density of ~14,000 cells/cm² and grown until near confluence. Cultures were then switched to DMEM + 0.2% FBS for 24 hr to achieve quiescence. The medium was removed and replaced with 10 ng/ml PDGF or 20 ng/ml EGF in DMEM + 1% CPSR-2 and incubated for periods ranging from 10 min for PDGFR autophosphorylation and MAPK phosphorylation to 90 min for c-Fos protein levels in the absence and presence of CGP 53716 (added 2 hr earlier). After the indicated time period, flasks were washed with ice-cold PBS and cells were lysed by addition of 0.4 ml lysis buffer/100-mm petri dish. The lysis buffer contained 50 mM Tris-HCl, pH 7.5, 0.5% Nonidet-40, 1 mM EDTA, 1 mM EGTA, 150 mM NaCl, 10% glycerol, 50 mM sodium fluoride, 10 mM sodium pyrophosphate, 1 mM sodium orthovanadate, 80 µM §-glycerophosphate, 1 mM PMSF, 10 µg/ml aprotinin, 100 µg/ml soybean trypsin inhibitor and 10 µg/ml leupeptin. Cells were scraped off bottom of the dishes with a rubber policeman and lysates were incubated on rocker for 1 hr at 4°C. Protein concentrations were measured by using the Bradford reagent (Bio-Rad, Hercules, CA) for lysates containing detergents with bovine serum globulin as a standard.

Measurement of PDGF receptor autophosphorylation and protein levels. Protein extracts (30 µg) were electrophoresed on separate 10% sodium dodecyl sulfate-polyacrylamide gels and transferred onto nitrocellulose for 2 hr at 1 amp. Filters were stained with ponceau red to show equal protein loading and transfer completeness. Filters were blocked with TBS-T and 3% nonfat dry milk for 1 hr at room temperature followed by incubation overnight at 4°C with anti-pTyr antibody (1 µg/ml) or anti-PDGF type AB receptor antibody (1/1000 dilution) with gentle rocking. Filters were then washed three times with TBS-T for 5 min each. The secondary antibody (goat anti-mouse IgG conjugated to HRP for anti-pTyr or goat anti-rabbit IgG-HRP conjugate for anti-PDGF type AB receptor) was incubated with filters on a rocker for 1 hour at room temperature. Filters were then washed 2 times for 30 minutes each and signals detected by the chemiluminescence detection system (Pierce, Rockford, IL) and exposed to Kodak Biomax MS film. Densitometric signals were imaged by scanning the film using a Macintosh computer with Adobe Photoshop (v. 3.0.4).

Determination of EGF receptor phosphorylation and protein level. Protein extracts (100 µg) were diluted to 500 µl with lysis buffer. The EGF receptor was immunoprecipitated by incubating with anti-EGF receptor antibody for 2 hours at 4°C. Protein A- Sepharose beads (Sigma Chemical Co., St. Louis, MO) were added and gently rocked overnight at 4°C. Immunoprecipitates were washed five times with 1 ml of 50 mM tris-HCl (pH 7.5), 10% glycerol, 0.5% Nonidet P-40 and 150 mM NaCl at 4°C. Sepharose complexes were boiled with 30 µl of Laemmli sample buffer for 5 min. Samples (30 µl) were electrophoresed and transferred as mentioned above. Blots were blocked with TBS-T plus 3% nonfat dry milk and incubated with anti-pTyr antibody (1 µg/ml) or anti-EGF receptor antibody (1/2000 dilution) overnight at 4°C. The blots were washed again and then a secondary antibody (goat anti-mouse IgG conjugated to HRP for anti-pTyr or goat anti-rabbit IgG-HRP conjugate for anti-EGF receptor) was incubated with filters on a rocker for 1 hr at room temperature. The filters were then washed again and signals detected as described above.

Measurement of MAPK phosphorylation and c-Fos protein levels. Protein extracts (30 µg) were size-fractionated on separate 10% sodium dodecyl sulfate-polyacrylamide gels and transferred onto nitrocellulose for 2 hr at 1 amp. Filters were stained with ponceau red to show equal protein loading and transfer completeness. Filters were blocked with TBS-T and 3% nonfat dry milk for 1 hr at room temperature followed by incubation overnight at 4°C with anti-MAPK antibody (1/1000 dilution) or anti-c-Fos protein antibody (5 µg/ml) with gentle rocking. Filters were then washed three times with TBS-T for 5 minutes each. The secondary antibody (goat anti-mouse IgG conjugated to HRP for anti-c-Fos or goat anti-rabbit IgG-HRP conjugate for anti-MAPK) was incubated with filters on a rocker for 1 hr at room temperature. The filters were then washed twice for 30 min each and signals detected by the chemiluminescence detection system (Pierce, Rockford, IL) and exposed to Kodak Biomax MS film. Densitometric signals were imaged as described earlier.

Cellular cytotoxicity. To determine the cytotoxic effects of CGP 53716 on RASMC and 3T3 cell lines, the following assay was performed. Cytotoxicity was evaluated after 22 hr of CGP 53716 exposure by determining the trypan blue exclusion from cells; dead cells were stained blue although live cells were not. The percent viable cells were determined by dividing the number of live cells by the total number of cells in a high powered field. Five high powered fields were counted for each treatment.

Data analysis. Data were expressed as mean ± S.E.M. Statistical analysis utilized the computer-assisted software JMP (SAS Institute, Cary, NC). An analysis of variance was used to test for interaction of the CGP 53716 treatments on the growth factor-induced DNA synthesis curves with Tukey's t test for comparison of treatment means. Values were considered significant at the P < .05 level.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Effects of CGP 53716 on cellular proliferation. To determine the inhibitory effects of CGP 53716 on RASMC and 3T3 cell proliferation, cell counts were determined in the presence of 10% FBS after 1, 3, 6 and 8 days in culture. CGP 53716 inhibited RASMC proliferation (fig. 2A) in a concentration-dependent manner; however, there was no effect on 3T3 cell proliferation (fig. 2B). The IC₅₀ value for CGP 53716 against serum-stimulated RASMC growth was 0.29 ± 0.01 µM.

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Time course for CGP 53716 inhibition of RASMC (top panel) and Balb/3T3 fibroblast cells (bottom panel) proliferation with 10% serum. Cells in all the treatment groups contained 0.5% DMSO. The results are expressed in mean ± S.E.M. of triplicate wells; an IC50 value was determined on day 8. CGP 53716 inhibited serum-induced RASMC growth but did not inhibit growth in the Balb/3T3 fibroblast cells.

Effect of CGP 53716 on growth factor-induced DNA synthesis. Concentration response curves for growth factor-induced DNA synthesis in cultured RASMC and 3T3 fibroblast cells are shown in figure 3. Near maximal stimulation for PDGF-BB (fig. 3A), bFGF (fig. 3B) and EGF (fig. 3C) were achieved at concentrations of 10, 5 and 10 ng/ml, respectively, in both RASMC and 3T3 cells. All subsequent DNA synthesis studies with the TK inhibitor, CGP 53716, used these growth factor concentrations. To characterize the growth inhibitory effects of CGP 53716, concentration-response curves for CGP 53716 were obtained against PDGF, bFGF and EGF-induced DNA synthesis in RASMC and 3T3 cells (fig. 4). CGP 53716 elicited a concentration-dependent inhibition of PDGF-, bFGF- and EGF-induced DNA synthesis in both cell types. CGP 53716 potently inhibited PDGF-induced DNA synthesis in RASMC (table 1). The rank order of potency for CGP 53716 at inhibiting growth factor-induced DNA synthesis in these cells was PDGF-BB > bFGF > EGF; CGP 53716 was significantly more potent vs. PDGF-BB than EGF in RASMC. In 3T3 cells the IC₅₀ value for CGP 53716 against PDGF-BB-induced DNA synthesis was similar to that observed for PDGF-BB stimulation in the RASMC (table 1). The IC₅₀ values for CGP 53716 inhibition of bFGF- and EGF-induced DNA synthesis were 6.5- and 4-fold higher, respectively. CGP 53716 was significantly more potent vs. PDGF-BB than bFGF in 3T3 cells.

View larger version (25K): [in this window] [in a new window]   Fig. 3.   Concentration-response curves for 22 hr PDGF-BB, bFGF and EGF stimulation of DNA synthesis as measured by 3H-thymidine incorporation in RASMC (top panel) and Balb/3T3 fibroblast cells (bottom panel). The results are mean ± S.E.M. of three to five experimental runs done in triplicate. Based on these results, concentrations of PDGF-BB, bFGF and EGF to be used in the CGP 53716 inhibitory studies are 10, 5 and 10 ng/ml, respectively.

View larger version (29K): [in this window] [in a new window]   Fig. 4.   Effect of tyrosine kinase inhibitor, CGP 53716, on the incorporation of 3H-thymidine in cultured RASMC (top panel) and Balb/3T3 fibroblast cells (bottom panel) after 22 hr stimulation with 10 ng/ml PDGF-BB, 5 ng/ml bFGF or 10 ng/ml EGF. Responses are mean ± S.E.M. for three to eight experimental runs in triplicate. VH bars are 1% DMSO vehicle in the absence of growth factors. The control dpm ± S.E.M. for PDGF-BB-, bFGF- and EGF-stimulated RASMC (top panel) were 2382 ± 507, 2920 ± 533 and 2644 ± 1220, respectively. The control dpm ± S.E.M. for PDGF-BB-, bFGF- and EGF-stimulated Balb/3T3 cells (bottom panel) were 389 ± 80, 918 ± 183 and 564 ± 161, respectively. As shown, CGP 53716 inhibited DNA synthesis stimulated by PDGF-BB, bFGF or EGF in both RASMC and 3T3 fibroblast cells.

Effects of CGP 53716 on PDGF and EGF signaling pathways. To determine where in the PDGF or EGF transduction pathways CGP 53716 inhibition of DNA synthesis occurs, phosphorylation levels of PDGF and EGF receptor proteins, MAPK phosphorylation and c-Fos protein levels were measured. Figure 5 depicts the PDGF-§ receptor autophosphorylation (fig. 5A) and phosphorylation of the 44 kDa/42 kDa MAPK proteins (fig. 5B), which were determined 10 min after growth factor stimulation, in both RASMC and 3T3 cells. After PDGF-BB stimulation PDGF receptor autophosphorylation in both cell types was markedly inhibited by CGP 53716 (IC₅₀ < 1 µM); consistent with inhibition of PDGF-BB-induced DNA synthesis. Because CGP 53716 had an inhibitory effect on PDGF receptor phosphorylation, all subsequent downstream protein phosphorylation events would be inhibited. In PDGF-BB-stimulated cells, MAPK phosphorylation was also inhibited although CGP 53716 appeared less potent (IC₅₀ value  3 µM) (fig. 5B). In RASMC, c-Fos protein levels, which were measured 90 min after growth factor addition, were reduced toward baseline by 1 and 10 µM CGP 53716 (fig. 5C). In contrast, CGP 53716 (1, 10 µM) had no measurable effect on EGF receptor autophosphorylation, MAPK phosphorylation and c-Fos protein levels in EGF-stimulated RASMC (fig. 6A, B and C, respectively).

View larger version (31K): [in this window] [in a new window]   Fig. 5.   CGP 53716 effects on 10 ng/ml PDGF-BB-stimulated PDGF receptor autophosphorylation (A, after 10 min), MAPK phosphorylation (B, after 10 min) and c-Fos protein level (C, after 90 min) in RASMC and Balb/3T3 fibroblast cells. CGP 53716 inhibited PDGF receptor autophosphorylation, MAPK phosphorylation and c-Fos protein expression in both RASMC and 3T3 fibroblast cells. Similar results were obtained with another experimental run.

View larger version (25K): [in this window] [in a new window]   Fig. 6.   CGP 53716 effects on 20 ng/ml EGF-stimulated EGF receptor autophosphorylation (A, after 10 min), MAPK phosphorylation (B, after 10 min) and c-Fos protein level (C, after 90 min) in RASMC. CGP 53716 did not inhibit EGF receptor autophosphorylation, MAPK phosphorylation and c-Fos protein expression in both RASMC and 3T3 fibroblast cells. Similar results were obtained with another experimental run.

Role of endogenous PDGF in bFGF-induced DNA synthesis. A polyclonal antibody to PDGF-AB ligand, which has been shown to neutralize all three isoforms of PDGF (Raines et al., 1989), was coadministered with bFGF and DNA synthesis determined. Over a concentration range of 0 to 50 µg/ml, the antibody had no effect on bFGF-induced DNA synthesis (fig. 7). In contrast, this antibody inhibited PDGF-BB-induced DNA synthesis in a concentration-dependent fashion.

View larger version (18K): [in this window] [in a new window]   Fig. 7.   Effects of a neutralizing antibody to PDGF-AB on PDGF-BB and bFGF-induced 3H-thymidine incorporation in RASMC. Data are mean ± S.E.M. from triplicate determinations. PDGF-AB neutralizing antibody blocked the PDGF-BB-induced DNA synthesis but did not inhibit the 4-fold increase in DNA synthesis induced by bFGF.

Cellular morphological and cytotoxic effects of CGP 53716. To determine if CGP 53716 effects on cell growth were due to cell killing, we performed trypan blue exclusion experiments after exposing cells to compound for 22 hr in culture. As shown in table 2, at concentrations up to 50 µM CGP 53716 had no significant effect on percent viable RASMC compared to untreated controls. Similarly, percent viable 3T3 cells in the presence of 10 µM CGP 53716 was not significantly different from control. As the concentration of CGP 53716 increased from 0.1 to 10 µM, the RASMC (fig. 8) and 3T3 cells (fig. 9) morphology changed from a bipolar appearance to a more cubiodal shape. In addition, cytoplasmic vacuoles were observed in both the RASMC and 3T3 cells in the presence of 10 µM CGP 53716 (fig. 8E and 9E, respectively). Although these morphological changes were noted, their significance is unclear because no deleterious effects on cellular viability were observed.

View larger version (83K): [in this window] [in a new window]   Fig. 8.   CGP 53716 effects on RASMC morphology after 22 hr using light microscopy. A-D, magnification = × 200. A, 10 ng/ml PDGF-BB + 1% DMSO; B, 10 ng/ml PDGF-BB + 0.1 µM CGP53716; C, 10 ng/ml PDGF-BB + 1 µM CGP53716; D, 10 ng/ml PDGF-BB + 10 µM CGP53716; E, magnification = × 400. Note the change in shape of RASMC from bipolar to cubical at 1 and 10 µM CGP 53716 as well as the vacuoles in cytoplasm at 10 µM.

View larger version (108K): [in this window] [in a new window]   Fig. 9.   Effects of CGP 53716 on Balb/3T3 fibroblast cell morphology after 22 hr using light microscopy. A-D, magnification = × 200. A, 10 ng/ml PDGF-BB + 1% DMSO; B, 10 ng/ml PDGF-BB + 0.1 µM CGP53716; C, 10 ng/ml PDGF-BB + 1 µM CGP53716; D, 10 ng/ml PDGF-BB + 10 µM CGP53716; E, magnification = ×400. Note the change in shape of RASMC from bipolar to cubical at 1 and 10 µM CGP 53716 as well as the vacuoles in cytoplasm at 10 µM.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

Our study describes the effects of the TK inhibitor, CGP 53716, on PDGF-BB-, bFGF- and EGF-mediated proliferation in cultured RASMC and Balb/3T3 fibroblast cells. CGP 53716 potently inhibited RASMC growth and nonselectively blocked DNA synthesis. Downstream signaling events such as receptor phosphorylation were selectively blocked following PDGF-BB stimulation. The inhibitory effects of CGP 53716 in both RASMC and Balb/3T3 fibroblast cells were not due to general cytotoxicity. The cytoplasmic vacuoles that were observed in the CGP 53716-treated RASMC and 3T3 cells are possibly the accumulation of compound in these cells. These results demonstrate for the first time that CGP 53716 has multi-growth factor inhibitory effects.

Investigators have laboriously catalogued the growth regulating factors involved in the cellular migratory and proliferative events after vascular injury (Clowes et al., 1983; Ross, 1991; Shirotani et al., 1993; Gospodarawicz et al., 1981; Lindner and Reidy, 1993; Jawien et al., 1992). The pharmacology of smooth muscle cell proliferation after angioplasty in models such as the rat balloon-injured carotid artery has been extensively characterized as recently reviewed by Jackson and Schwartz (1992). The complexity of the restenotic process that appears to involve several growth factors offers a difficult challenge in limiting SMC proliferation through specific growth factor antagonists. Because PDGF-BB has been found to bind to the two PDGF receptors ( and §) we used PDGF-BB for this study to ascertain total PDGF-mediated events (Schollmann et al., 1992). The focus of this work was to determine if CGP 53716, a novel TK inhibitor, could effectively block growth factor signaling in vitro. Previous reports indicate that CGP 53716 selectively blocked PDGF receptor activation using human v-sis-transformed Balb/c 3T3 cells in vitro and in vivo (Buchdunger et al., 1995). However, our data demonstrate that CGP 53716 not only inhibited the increase in PDGF-BB-induced DNA synthesis but also abrogated the increases in DNA synthesis observed for bFGF and EGF stimulation in RASMC and wild type Balb/3T3 cells. The ability of CGP 53716 to inhibit multi-growth factor pathways that lead to DNA synthesis would be of great benefit given the complexity of growth factor regulation in vascular injury.

Although CGP 53716 was significantly less potent as an inhibitor of EGF-induced DNA synthesis in RASMC compared with inhibition of PDGF-BB-stimulated DNA synthesis (0.79 vs. 0.23 µM IC₅₀ values, respectively; P < .05), the compound still reduced EGF-mediated DNA synthesis. In comparison, Buchdunger et al. (1995) reported that the IC₅₀ value for CGP 53716 inhibition of EGF receptor TK was > 100 µM; concluding that CGP 53716 was specific for PDGF receptor TK. Our data also showed that CGP 53716 was significantly less potent as an inhibitor of bFGF-induced DNA synthesis in Balb/3T3 cells compared to PDGF-stimulated DNA synthesis (1.10 vs. 0.17 µM, respectively). This 6-fold difference in potency is modest and somewhat surprising in light of the report that CGP 53716 did not inhibit bFGF-stimulated c-fos mRNA expression in Balb/3T3 cells at concentrations of 100 µM (i.e., > 100 fold) (Buchdunger et al., 1995). A potential explanation for this disparity could be bFGF-induced release of PDGF or activation of PDGF pathways in our studies. It has been reported that bFGF increases the PDGF receptor  subtype (Schollman et al.., 1992) and up-regulation of PDGF-A chain mRNA (Winkles and Gay, 1991). In addition, Calara et al. (1996) demonstrate that bFGF stimulates the production of PDGF-AA in RASMC with an increase in DNA synthesis and how an antisense oligonucleotide to PDGF-AA, but not an antibody to PDGF-AA, can inhibit this increase in DNA synthesis. Therefore, precedent has been set for bFGF to modulate an intracrine mediated PDGF-AA DNA synthetic pathway. However, when we used a neutralizing antibody to PDGF-AB, which neutralizes all three isoforms of PDGF including PDGF-AA (Raines et al., 1989), we were unable to modulate bFGF-induced thymidine incorporation. In contrast, this same antibody produced concentration-related decreases in PDGF-BB-stimulated thymidine incorporation. Thus our results indicate the effects of CGP 53716 on bFGF-induced DNA synthesis do not involve a PDGF-BB mediated pathway leading to DNA synthesis. However, as pointed out by Calara et al. (1996), antibodies to PDGF-AA which is produced inside the RASMC by bFGF stimulation, may not inhibit PDGF-AA-induced DNA synthesis and awaits further investigation.

The common mechanism for increased DNA synthesis in the RASMC and Balb/3T3 fibroblast cells involves activation of PTKs which are part of the PDGF (Kovalenko et al., 1994; Claesson-Welsh, 1994; Bilder et al., 1991), bFGF (Ross, 1991; Zhan et al., 1993; Schollmann et al., 1992) and EGF (Posner et al., 1993; Montgomery et al., 1995) signaling pathways. Growth factor receptor signal transduction is initiated by ligand binding, receptor dimerization and inter- and intrareceptor activation of kinase activity, with coincident phosphorylation of tyrosine kinase residues (Ullrich and Schlessinger, 1990). Downstream of the receptor intracellular tyrosine kinases such as Src homology 2 (SH2) and recently SH3 (Erpel et al., 1996) domains on these intracellular kinases take part in transducing these signals within the cell (Levitzki and Gazit, 1995). Several of these intracellular pathways appear to converge at the MAPK with activation of the p42/p44 MAPKs (Pelech and Sanghera, 1992). Activated MAPKs further transduce the signal to the nucleus by increasing mRNA levels of immediate-early genes such as c-fos and c-jun that transform the intracellular signal into cellular growth and/or differentiation (Rothman et al., 1994).

TK inhibitors have been shown to block cellular growth in several cultured cell lines via tyrosine kinase signaling paths activated by PDGF (Kovalenko et al., 1994; Bilder et al., 1993; Buchdunger et al., 1995), bFGF (Hawker and Granger, 1994) and EGF (Posner et al., 1993). CGP 53716, a TKI of the 2-phenylaminopyrimidine class, has been recently shown to selectively inhibit PDGF-induced receptor autophosphorylation, inhibit c-fos mRNA expression and block cellular growth in Balb/c 3T3 cells (Buchdunger et al., 1995). Bilder et al. (1990) reported tyrphostins, another class of TK inhibitors, block PDGF receptor autophosphorylation and c-fos mRNA expression with a potency ratio similar to their anti-mitogenic activity. Our data demonstrated that CGP 53716 inhibited PDGF receptor autophosphorylation and MAPK phosphorylation with a subsequent decrease in c-Fos protein expression in the 3T3 cells and also in RASMC. Interestingly, CGP 53716 did not inhibit the signaling path activated by EGF in RASMC consistent with Buchdunger et al. (1995). EGF receptor autophosphorylation, MAPK phosphorylation and subsequent c-Fos protein levels were not blocked by CGP 53716. These signal transduction results appear to be in contrast to our DNA synthesis data in which CGP 53716 was able to inhibit, albeit with different degrees of potency, the responses to PDGF, bFGF and EGF.

Based on our signal transduction studies, CGP 53716 inhibits the PDGF receptor signaling pathway selectively over EGF. The downstream signaling events beyond the EGF receptor autophosphorylation (i.e., MAPK, c-Fos) were also not affected by CGP 53716. Therefore, CGP 53716 is having its inhibitory effect on EGF-induced DNA synthesis at some point beyond the nuclear activation of c-Fos protein expression or possibly inhibiting an alternate pathway that is not dependent on MAPK activation and c-Fos expression. Marx et al. (1995) showed that the antiproliferative effects of rapamycin, a macrolide antibiotic, in vascular SMC are associated with an inhibition of cell-cycle kinases, cyclins and retinoblastoma protein phosphorylation. These findings demostrate that these are multiple kinase-regulated steps in cell growth. Although we have not specifically investigated the effects of CGP 53716 on cell-cycle kinases, this would be an interesting experimental progression.

The effects of CGP 53716 on serum-induced RASMC and Balb/3T3 cell growth show a discrepancy with the specific PDGF-BB- and bFGF-induced DNA synthesis. We had performed the serum-induced cell growth initially and followed this with studies to try to explain the specific growth factor that drove DNA synthesis and ultimately cell growth. Previous results have shown that PDGF is a major growth factor component in serum (Bernstein et al., 1982); thus it is not surprising that CGP 53716 inhibited serum-induced growth in RASMC with roughly the same potency as inhibition of PDGF-induced DNA synthesis (0.23 vs. 0.29 µM, respectively). However, we were unable to inhibit serum-induced growth in Balb/3T3 cells with CGP 53716 despite inhibition of PDGF-induced DNA synthesis. One possible explanation is that the concentration of CGP 53716 utilized in these assays (1 µM) was insufficient to block the growth factor concentrations in serum. Another possible explanation is that Balb/3T3 cells may respond to a unique growth factor found in serum which is not important for RASMC growth and not responsive to CGP 53716.

In summary, our study provides evidence that CGP 53716 blocks DNA synthesis not only to PDGF but also to bFGF and EGF in RASMC and Balb/3T3 fibroblast cells. The potential therapeutic role for a nonselective tyrosine kinase inhibitor in vascular proliferative disorders such as restenosis and atherosclerosis could prove to be beneficial due to the multiple growth factors involved in this pathology. CGP 53716, with its inhibitory effect on multiple growth factor signaling pathways, may have benefit as an antiproliferative not only on proliferating SMC but also on other cells such as fibroblasts in injured vessels.