Vascular Endothelial Growth Factor Induces Expression of Connective Tissue Growth Factor via KDR, Flt1, and Phosphatidylinositol 3-Kinase-Akt-dependent Pathways in Retinal Vascular Cells

, Fibroblastic proliferation accompanies many angio-genesis-related retinal and systemic diseases. Since connective tissue growth factor (CTGF) is a potent mitogen for fibrosis, extracellular matrix production, and angiogenesis, we have studied the effects and mechanism by which vascular endothelial growth factor (VEGF) regulates CTGF gene expression in retinal capillary cells. In our study, VEGF increased CTGF mRNA levels in a time-and concentration-dependent manner in bovine retinal endothelial cells and pericytes, without the need of new protein synthesis and without altering mRNA stability. VEGF activated the tyrosine receptor phosphorylation of KDR and Flt1 and increased the binding of phosphatidylinositol 3-kinase (PI3-kinase) p85 subunit to KDR and Flt1, both of which could mediate CTGF gene induction. VEGF-induced CTGF expression was mediated primarily by PI3-kinase activation, whereas PKC and ERK pathways made only minimal contributions. Further-more, overexpression of constitutive active Akt was suf-ficient to induce CTGF gene expression, and inhibition of Akt activation by overexpressing dominant negative mutant of Akt abolished the VEGF-induced CTGF expression. These data suggest that VEGF can increase CTGF gene expression in bovine retinal capillary cells via KDR or Flt receptors and the activation of PI3-ki-nase-Akt pathway independently of PKC or Ras-ERK pathway, possibly inducing the fibrosis observed in retinal neovascular diseases.

Angiogenesis and fibrosis are key components in development, growth, wound healing, and regeneration (1). In addition, these processes commonly occur together in many disease states where neovascularization is believed to initiate the pathological cascade. Some of these diseases are proliferative diabetic retinopathy (2), rheumatoid arthritis (3), and age-related macular degeneration (4). Thus, it is possible that the factors that regulate angiogenesis may also induce factors that stimulate extracellular matrix production and fibrosis. Accordingly, we have studied the ability of vascular endothelial growth factor (VEGF), 1 an established angiogenic factor, to regulate the expression of connective tissue growth factor (CTGF), a growth factor with known actions on fibroblast proliferation, matrix production, and associated with fibrotic disorders.
VEGF is expressed as a family of peptides of 121, 145, 165, 189, and 206 amino acid residues (5). Its expression is induced by hypoxia (6) and is essential in the vasculogenesis process during development (7). Several receptors have been shown to mediate the action of VEGF, and most of them belong to the tyrosine kinase receptor family (8). Upon the binding of VEGF to its receptors, multiple signaling cascades are activated, including the tyrosine phosphorylation of phospholipase C␥, elevation of intracellular calcium and diacylglycerol, activation of protein kinase C (PKC), and extracellular signal-regulated kinase (MAPK/ERK) for endothelial cell proliferation (9 -12). In addition, VEGF also stimulates activation of phosphatidylinositol (PI) 3-kinase leading to Akt/PKB activation and possibly enhancing endothelial cell survival (13)(14)(15). However, in non-endothelial cells such as capillary pericytes that predominantly express Flt1 receptor, the action of VEGF is poorly understood.
Connective tissue growth factor (CTGF), a member of CCN family (CYR61, CTGF, and NOV) (16,17), is a potent and ubiquitously expressed growth factor that has been shown to play a unique role in fibroblast proliferation, cell adhesion, and the stimulation of extracellular matrix production (18,19). The 38-kDa protein was originally identified in conditioned medium from human umbilical vein endothelial cells (20), and the expression was shown to be selectively stimulated by transforming growth factor-␤ (TGF-␤) in cultured fibroblasts (21). Due to its mitogenic action on fibroblasts and its ability to induce the expression of the extracellular matrix molecules, collagen type I, fibronectin, and integrin ␣ 5 (18), CTGF is supposed to play an important role in connective tissue cell proliferation and extracellular matrix deposition as one of the mediators of TGF-␤ (22). CTGF also seems to be an important player in the pathogenesis of various fibrotic disorders, since it was shown to be overexpressed in scleroderma, keloids, and other fibrotic skin disorders (23), as well as in stromal rich mammary tumors (24), and in advanced atherosclerotic lesions (25). Recently, the integrin ␣ v ␤ 3 has been reported to serve as a receptor on endothelial cells for CTGF-mediated endothelial cell adhesion, migration, and angiogenesis (26,27).
Besides TGF-␤, the expression of CTGF is reported to be regulated by dexamethasone in BALB/c 3T3 cells (28), high glucose in human mesangial cells (29), kinin in human embryonic fibroblasts (30), factor VIIa, and thrombin in WI-38 fibroblasts (31), tumor necrosis factor ␣ in human skin fibroblast (32), and cAMP in bovine endothelial cells (33). Since many of these cytokines are known to induce VEGF, it is possible that increased VEGF expression can regulate the expression of CTGF. In the present study, we have investigated the regulation of CTGF by VEGF in retinal endothelial cells and pericytes via the PI3-kinase and several other signaling pathways.

EXPERIMENTAL PROCEDURES
Materials-Endothelial cell basal medium was purchased from Clonetics (San Diego, CA). Endothelial cell growth factor was purchased from Roche Molecular Biochemicals. Dulbecco's modified Eagle's medium and fetal bovine serum were obtained from Life Technologies, Inc. VEGF, placenta growth factor (PlGF), TGF-␤1, and anti-CTGF antibody were ordered from R & D Systems (Minneapolis, MN). Anti-KDR (Flk1) and anti-Flt1 antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Protein A-Sepharose was purchased from Amersham Pharmacia Biotech. Anti-phospho-ERK, anti-ERK, antiphospho-Akt, and anti-Akt were purchased from New England Biolabs (Beverly, MA). Anti-p85 and anti-phosphotyrosine were purchased from Upstate Biotechnology, Inc. (Lake Placid, NY). Phosphatidylinositol (PI) was purchased from Avanti (Alabaster, AL), and PD98059, wortmannin, and GF 109203X were obtained from Calbiochem. All other materials were ordered from Fisher and Sigma.
Cell Culture-Primary cultures of bovine retinal endothelial cells (BREC) and pericytes (BRPC) were isolated by homogenization and a series of filtration steps as described previously (34). BREC were subsequently cultured with endothelial cell basal medium supplemented with 10% plasma-derived horse serum, 50 mg/liter heparin, and 50 g/ml endothelial cell growth factor. BRPC were cultured in Dulbecco's modified Eagle's medium with 5.5 mM glucose and 20% fetal bovine serum. Cells were cultured in 5% CO 2 at 37°C, and media were changed every 3 days. Cells were characterized for their homogeneity by immunoreactivity with anti-factor VIII antibody for BREC and with monoclonal antibody 3G5 for BRPC (35). Cells remained morphologically unchanged under these conditions, as confirmed by light microscopy. Only cells from passages 2 through 7 were used for the experiments.
Recombinant Adenoviruses-cDNA of constitutive active Akt (CAAkt, Gag protein fused to N-terminal of wild type Akt) was constructed as described (36). cDNA of dominant negative Akt (DNAkt, substituted Thr-308 to Ala and Ser-473 to Ala) was constructed as described (37). cDNA of dominant negative K-Ras (DNRas, substituted Ser-17 to Asn) was kindly provided by Dr. Takai (Osaka University) (38). cDNA of dominant negative extracellular signal-regulated kinase (DNERK, substituted Lys-52 to Arg in ATP-binding site) was constructed as described (39). cDNA of ⌬p85 was kindly provided by Dr. Kasuga (Kobe University) (40). cDNA of PKC was kindly provided by Dr. Douglas Ways (Lilly). cDNA of dominant negative PKC (DNPKC, substituted Lys-273 to Trp in ATP-binding site) was constructed as described (41). The recombinant adenoviruses were constructed by homologous recombination between the parental virus genome and the expression cosmid cassette or shuttle vector as described (42,43). The adenoviruses were applied at a concentration of 1 ϫ 10 8 plaque-forming units/ml, and adenoviruses with the same parental genome carrying the lacZ gene or enhanced green fluorescein protein gene (CLONTECH, Palo Alto, CA) were used as controls. Expression of each recombinant protein was confirmed by Western blot analysis and increased about 10-fold compared with cells infected with the control adenovirus.
Immunoprecipitation-Cells were washed three times with cold phosphate-buffered saline and solubilized in 200 l of lysis buffer (1% Triton X-100, 50 mmol/liter HEPES, 10 mmol/liter EDTA, 10 mmol/liter sodium pyrophosphate, 100 mmol/liter sodium fluoride, 1 mmol/liter sodium orthovanadate, 1 g/ml aprotinin, 1 g/ml leupeptin, and 2 mmol/liter phenylmethylsulfonyl fluoride). After centrifugation at 12,000 rpm for 10 min, 1.0 mg of protein was subjected to immunoprecipitation. To clear the protein extract, protein A-Sepharose (20 l of a 50% suspension) was added to the cell lysates, after which they were incubated for 1 h, followed by centrifugation and collection of the supernatant. A specific rabbit anti-KDR or Flt1 antibody was added and rocked at 4°C for 2 h; 20 l of protein A-Sepharose was then added, and the sample was rocked for another 2 h at 4°C. For denaturation, protein A-Sepharose antigen-antibody conjugates were separated by centrifugation, washed five times, and boiled for 3 min in Laemmli sample buffer.
Western Blot Analysis-Immunoprecipitated proteins or 30 g of total cell lysates were subjected to SDS-gel electrophoresis and electrotransferred to nitrocellulose membrane (Bio-Rad). The membrane was soaked in blocking buffer (phosphate-buffered saline containing 0.1% Tween 20 and 5% bovine serum albumin) for 1 h at room temperature and incubated with primary antibody overnight at 4°C followed by incubation with horseradish peroxidase-conjugated secondary antibody (Amersham Pharmacia Biotech). Visualization was performed using the enhanced chemiluminescence detection system (ECL, Amersham Pharmacia Biotech) per the manufacturer's instructions.
PI3-Kinase Assay-PI3-kinase activities were measured by the in vitro phosphorylation of PI (10). Cells were lysed in ice-cold lysis buffer containing 50 mM HEPES, pH 7.5, 137 mM NaCl, 1 mM MgCl 2 , 1 mM CaCl 2 , 2 mM Na 3 VO 4 , 10 mM NaF, 2 mM EDTA, 1% Nonidet P-40, 10% glycerol, 1 mM phenylmethylsulfonyl fluoride, 2 g/ml aprotinin, 5 g/ml leupeptin, and 1 g/ml pepstatin. Insoluble material was removed by centrifugation at 15,000 ϫ g for 10 min at 4°C. PI3-kinase was immunoprecipitated from aliquots of the supernatant with antiphosphotyrosine antibodies. After successive washings, the pellets were resuspended in 50 l of 10 mM Tris, pH 7.5, 100 mM NaCl, and 1 mM EDTA. 10 l of 100 mM MgCl 2 and 10 l of PI (2 g/l) sonicated in 10 mM Tris, pH 7.5, with 1 mM EGTA was added to each pellet. The PI3-kinase reaction was initiated by the addition of 5 l of 0.5 mM ATP containing 30 Ci of [␥-32 P]ATP. After 10 min at room temperature with constant shaking, the reaction was stopped by the addition of 20 l of 8 N HCl and 160 l of chloroform/methanol (1:1). The samples were centrifuged, and the organic phase was removed and applied to silica gel TLC plates developing in CHCl 3 /CH 3 OH/H 2 O/NH 4 OH (60:47:11:2). The radioactivity in spots was quantified by PhosphorImager (Molecular Dynamics, Sunnyvale, CA).
Amplification of Human CTGF cDNA Using Reverse Transcriptase-Polymerase Chain Reaction (PCR)-cDNA templates for PCR were synthesized by reverse transcriptase (First Strand cDNA Synthesis Kit, Amersham Pharmacia Biotech) from human fibroblast according to the method recommended by the manufacturer. A standard PCR was performed (PCR optimizer kit, Invitrogen, Carlsbad, CA) using 5Ј-AGGG-CCTCTTCTGTGACTTCG-3Ј (sense primer) and 5Ј-TCATGCCATGTC-TCCGTACATC-3Ј (antisense primer) (20). The PCR products were then subcloned into a vector (pCRII, Invitrogen) and sequenced in their entirety, and comparison with the published human sequences revealed complete sequence identity. This cDNA probe was used for hybridization.
Northern Blot Analysis-Total RNA was isolated using acid-guanidinium thiocyanate, and Northern blot analysis was performed as described previously (44). Total RNA (20 g) was electrophoresed through 1% formaldehyde-agarose gels and then transferred to a nylon membrane. 32 P-Labeled cDNA probes were generated by use of labeling kits (Megaprime DNA labeling systems, Amersham Pharmacia Biotech). After ultraviolet cross-linking using a UV cross-linker (Stratagene, La Jolla, CA), blots were pre-hybridized, hybridized, and washed in 0.5ϫ SSC, 5% SDS at 65°C with 4 changes over 1 h. All signals were analyzed using a PhosphorImager, and lane loading differences were normalized using the 36B4 cDNA probe (45).
Analysis of CTGF mRNA Half-life-CTGF mRNA half-life experiments were carried out using BREC and BRPC. The cells were exposed to vehicle or VEGF (25 ng/ml) for the indicated periods prior to mRNA stability measurements. Transcription was inhibited by the addition of actinomycin D (5 g/ml). For inhibition of protein synthesis, cells were treated with cycloheximide (10 g/ml) for the times indicated.
Statistical Analysis-Determinations were performed in triplicate, and all experiments were repeated at least three times. Results are expressed as the mean Ϯ S.D., unless otherwise indicated. Statistical analysis employed Student's t test or analysis of variance to compare quantitative data populations with normal distributions and equal variance. Data were analyzed using the Mann-Whitney rank sum test or the Kruskal-Wallis test for populations with non-normal distributions or unequal variance. A p value of Ͻ0.05 was considered statistically significant. The dose response to VEGF-induced CTGF mRNA expression was studied after 6 h of VEGF stimulation. As shown in Fig. 1, C and D, the expression of CTGF mRNA was up-regulated in a dose-dependent manner, with significant increases observed at concentrations as low as 0.25 ng/ml in both BREC (Fig. 1C) and BRPC (Fig. 1D). Maximal increases were observed at VEGF concentrations of 25 ng/ml in both BREC and BRPC.

CTGF mRNA Expression by VEGF and
Since BREC and BRPC may express both KDR and Flt1, we examined the effects of PlGF, a Flt1-specific ligand (46 -48), on the induction of CTGF gene expression in vascular cells. As shown in Fig. 1E, CTGF mRNA levels were not affected after stimulation of 25 ng/ml of PlGF in BREC. In contrast, PlGF increased CTGF mRNA after 3 h of stimulation, which peaked after 9 h in BRPC (1.9 Ϯ 0.30-fold, p Ͻ 0.01, Fig. 1F), suggest-ing that VEGF-induced CTGF gene expression was mediated primarily by KDR in BREC and Flt1 in BRPC.
VEGF Induction of CTGF Protein Production-To determine if the effects of VEGF on CTGF mRNA were correlated with its protein level, CTGF protein expression was assessed by Western blot analysis using anti-human CTGF antibody. As shown in Fig. 2A, the detected size of CTGF protein was ϳ38 kDa in both BREC and BRPC. VEGF (25 ng/ml) increased the level of CTGF protein after 10 h in both BREC and BRPC. Comparative studies were performed on the effects of VEGF (25 ng/ml) and TGF-␤1 (10 ng/ml) on the expression of CTGF mRNA and protein. As shown Fig. 2B, VEGF and TGF-␤1 increased CTGF protein expression by a similar amount (2.5 Ϯ 0.4-and 2.8 Ϯ 0.8-fold, respectively, in BREC). CTGF mRNA levels were also increased a similar extent (3.0 Ϯ 0.3-and 3.3 Ϯ 0.5-fold, respectively).
Effects of VEGF on the Half-life of CTGF mRNA-The effects of VEGF on the stability of CTGF mRNA were examined. Northern blot analyses were performed with addition of actinomycin D (5 g/ml) after 6 h of VEGF (25 ng/ml) stimulation. In BREC (Fig. 3A) and BRPC (Fig. 3B), the half-life of CTGF mRNA was 1.7 and 3.6 h, respectively. There was no significant difference between VEGF-treated and -untreated cells.
Effects of Cycloheximide on CTGF mRNA Regulation-In order to examine the possibility that VEGF regulates CTGF mRNA expression through new protein synthesis of cytokines or transcription factors, cells were treated for 6 h with VEGF (25 ng/ml) and a protein synthesis inhibitor, cycloheximide (10 g/ml). Fig. 3, C and D, shows that cycloheximide did not prevent the increase of CTGF mRNA. Addition of both VEGF and cycloheximide increased CTGF mRNA 2.4 Ϯ 0.41-fold in BREC (Fig. 3C) and 2.5 Ϯ 0.40-fold in BRPC (Fig. 3D) after 6 h as compared with cycloheximide alone (p Ͻ 0.01). These data suggest that the stimulation of CTGF mRNA expression by VEGF was not induced by increased synthesis of a regulatory protein.
Involvement of ERK and PI3-Kinase-Akt in VEGF Signal-

FIG. 2. Effect of VEGF on CTGF protein expression.
A, confluent monolayers of BREC (left) or BRPC (right) in serum-free media were studied after 10 h, with and without 25 ng/ml of VEGF. An equal amount of protein for each sample was subjected to 10% SDS-gel electrophoresis and transferred to a nitrocellulose membrane. Signals were detected by Western blot analysis with anti-CTGF antibody. The size of the CTGF protein corresponded to ϳ38 kDa. B, VEGF (25 ng/ml) and TGF-␤1 (10 ng/ml) were studied as described in Figs. 1 and 2A. VEGF and TGF-␤1 increased CTGF protein and mRNA expression by a similar amount.
ing-Since ERK and PI3-kinase-Akt pathways have been reported to play central roles in VEGF signaling and biological actions (9 -15), we investigated whether or not VEGF can activate ERK and PI3-kinase-Akt pathways equally in BREC and BRPC. As shown in Fig. 4A, immunoblot analysis of immunoprecipitates of KDR from BREC stimulated with VEGF or PlGF using an antibody to phosphotyrosine and PI3-kinase p85 subunit demonstrated that VEGF, but not PlGF, promoted the tyrosine phosphorylation of KDR and interactions of KDR and p85 subunit of PI3-kinase. In contrast, as shown in Fig. 4B, immunoblot analysis of immunoprecipitates of Flt1 from BRPC stimulated with VEGF or PlGF demonstrated that both VEGF and PlGF increased the tyrosine phosphorylation of Flt1 and interactions of Flt1 and p85 subunit of PI3-kinase. These data suggest that VEGF can activate the receptor tyrosine phosphorylation and interaction with PI3-kinase p85 subunit in both KDR and Flt1.
To investigate the activation of Akt and ERK, we next performed immunoblot analysis with anti-phosphorylated Akt or anti-phosphorylated ERK antibodies using total cell lysates from BREC or BRPC stimulated with VEGF. As shown in Fig.  4, C and D, VEGF induced phosphorylation of both Akt and ERK in BREC by 3.1-and 5.8-fold (Fig. 4C), but only induced phosphorylation of Akt in BRPC by 2.6-fold (Fig. 4D). No effect on ERK phosphorylation was observed in BRPC. These data suggest that VEGF activated both ERK and PI3-kinase-Akt pathways in BREC, but stimulated only PI3-kinase-Akt pathway in BRPC.
Since the activation of PI3-kinase by VEGF has not been reported in BRPC, we studied the effects of VEGF on PI3kinase activity in BRPC. As shown in Fig. 5, the addition of VEGF (25 ng/ml) increased PI3-kinase activity in a time-dependent manner by 2.1 Ϯ 0.27-fold (p Ͻ 0.01) after 5 min and by 1.6 Ϯ 0.17-fold (p Ͻ 0.05) after 10 min in BRPC.
Role of PKC and Akt/PKB in VEGF-induced CTGF Expression-Since it has been reported that atypical PKC (51-55) and BRPC were incubated with VEGF (25 ng/ml) for the indicated times and harvested. Equal amounts of lysates were immunoprecipitated with anti-phosphotyrosine antibody, and immunocomplexes were assayed for their ability to phosphorylate PI to phosphatidylinositol phosphate (PIP). Representative autoradiogram (right) and quantitation of three experiments in the percentage of intensity of control (left) are shown. Asterisks indicate significant differences at p Ͻ 0.05 (*) and p Ͻ 0.01 (**).
Akt/PKB (13,14,56) have significant roles as signaling molecules downstream of PI3-kinase, we examined the involvement of PKC and Akt in this process. BREC were infected with each adenoviral vector, followed by stimulation with 25 ng/ml VEGF for 6 h. Neither wild type PKC nor DNPKC had significant effects on VEGF-induced increase in CTGF mRNA (Fig. 7A). In contrast, as shown in Fig. 7B, infection with CAAkt increased CTGF mRNA expression 2.1 Ϯ 0.21-fold (p Ͻ 0.01) without VEGF and 2.5 Ϯ 0.40-fold with VEGF. Overexpression with adenoviral vector containing DNAkt inhibited VEGF-induced CTGF expression by 85 Ϯ 13% (p Ͻ 0.01).

DISCUSSION
In this study, we have shown that VEGF can increase the mRNA expression of CTGF in a time-and concentration-dependent manner in both microvascular endothelial cells and contractile cells (capillary pericytes) possibly indicating that the effects of VEGF on CTGF expression may occur in all cells with VEGF receptors. This possibility is supported by the re-sults showing that both Flt1 (VEGFR1) and KDR/Flk1 (VEGFR2) can mediate the increases in CTGF mRNA expression. The ability of Flt1 to induce increases in CTGF mRNA levels is demonstrated in the pericytes that have predominantly Flt1 receptors and where the expression of KDR/Flk1 receptors were not significantly high enough to be determined by Northern blot analysis as reported in a previous publication (57). In addition, PlGF, a Flt1 receptor-specific ligand (48), was able to induce CTGF mRNA levels in BRPC but not in BREC, again supporting the postulate that VEGF can induce CTGF mRNA by activating through Flt1 in pericytes. The KDR/Flk1 receptors in the endothelial cells can also induce CTGF gene expression since KDR/Flk1 receptors are the predominant VEGF receptors in endothelial cells (58), and PlGF was not effective in inducing CTGF mRNA expression in endothelial cells. Further studies will be necessary to determine whether other types of VEGF receptors, such as Flt4 (59) and neuropilin-1 (60), which are present in endothelial cells, can also induce CTGF expression.
The VEGF dose-response curves for CTGF in both BRPC and BREC are similar and suggest that VEGF binds to high affinity receptors, consistent with the known K d values of Flt1 and KDR/Flk1 at 10 -100 pM (58,61). VEGF-induced CTGF mRNA is most likely due to an induction of transcription rather than altering the half-life of CTGF mRNA since the addition of VEGF failed to change the degradation rates of CTGF mRNA. The time course of the action of VEGF on CTGF (which required 6 -9 h) suggests this is potentially a chronic action of VEGF. In addition, the time needed to achieve maximum effect is also consistent with the calculated mRNA half-life of CTGF mRNA of 2-4 h.
From a biological perspective, the effects of VEGF on CTGF mRNA could potentially have important physiological impact for several reasons. First is that the increase in CTGF mRNA results in increased protein levels. Second, the VEGF concentration that was minimally active (0.25 ng/ml) can easily bind and activate a significant percentage of the VEGFR-1, -2 receptors (58, 61). Third, this low level of VEGF may exist even in non-pathological states, suggesting that low levels of VEGF may have physiological actions on maintaining extracellular matrix production via the induction of CTGF. At 2.5-25 ng/ml VEGF which are encountered in hypoxic and angiogenic states (62), the induction of CTGF expression by VEGF could potentially induce the fibrosis that frequently accompanies neovascularization. This possibility is supported further by the demonstration that the protein levels of CTGF expression were increased 10 h after the addition of VEGF that was consistent with the maximum increase in the mRNA levels at 6 -9 h. In addition, the potency of VEGF on CTGF expression appeared to be similar to TGF-␤1, suggesting that both of them could induce fibrosis associated with neovascularization.
The activation of the endogenous tyrosine kinases of KDR/ Flk1s can stimulate multiple signaling pathways, including Ras-ERK (63), PI3-kinase-Akt (13)(14)(15), and phospholipase C␥-PKC (9 -11) cascades. Much less is known regarding the regulation of Flt1 receptors. The results in BREC confirmed previous publications that VEGF can increase the tyrosine phosphorylation of KDR/Flk1 and its interaction with p85 subunit of PI3-kinase. In addition, VEGF also activates the ERK1/2 pathway confirming earlier reports from many laboratories, including ours. In contrast, VEGF was unable to activate ERK1/2 but stimulated the activation of PI3-kinase and phosphorylation of Akt in BRPC. These results have provided further direct evidence that the signaling pathways for Flt1 in vascular cells are different from those for KDR/Flk1. The lack of effect on ERK1/2 activation also supports the hypothesis that Flt1, unlike KDR/Flk1, is not involved in mitogenic actions (64). Further studies will be needed to determine the structural differences responsible for the inability of VEGFR1 to engage the Ras-ERK pathway.
The present report does provide the strong evidence that VEGF is inducing CTGF gene expression in both endothelial cells and pericytes via VEGFR1 or -R2 by the activation of PI3-kinase and Akt. This evidence includes the ability of wortmannin, a PI3-kinase inhibitor, to inhibit the effects of VEGFs in both cell types, whereas PD98059, a MAPK/ERK kinase inhibitor, and GF109203X, a general classical PKC and novel PKC inhibitor (1 M) (49, 50), did not have significant actions. Adenovirus containing dominant negative mutants of p85 subunit of PI3-kinase or Akt inhibited the action of VEGFs, whereas overexpression of dominant negative mutants of Ras and ERK1 by adenovirus vectors did not inhibit CTGF mRNA expression. Conversely, the overexpression of constitutive active Akt increased CTGF mRNA expression by 2.5-fold. The molecular steps between Akt activation and the enhancement of CTGF gene expression in the nucleus remain unclear, although the PKC isoform is most likely not involved since the overexpression of either the wild type or dominant negative of PKC isoform did not alter the effects of VEGF on CTGF mRNA levels.
The molecular processes between Akt phosphorylation and CTGF gene expression in the nucleus have not been studied. However, Pendurthi et al. (31) reported that factor VII and thrombin induced CTGF gene expression through a PI3-kinasedependent pathway. TGF-␤ has been reported to increase the transcription rates of CTGF. A promoter element of CTGF, which is responsive to TGF-␤ stimulation, has been reported to be present between Ϫ162 and Ϫ128 nucleotides in the 5Ј region (65). However, it is unlikely that the effects of VEGF on CTGF mRNA levels are mediated via the expression of TGF-␤ since the addition of cycloheximide did not change these effects.
In summary, these results have provided the first evidence that VEGF can induce the expression of CTGF via both Flt1 and KDR/Flk1 by the selectively activated PI3-kinase-Akt pathway but independent of the Ras-ERK pathway. In addition, the spectrum of signaling pathways may be different between Flt1 and KDR/Flk1, possibly reflecting their physiological roles. Biologically, these results support the conclusion that VEGF, through its effects on CTGF expression, may have physiological roles such as the maintenance of capillary strength and wound healing via the extracellular matrix production. In disease states, VEGF-induced CTGF may cause the proliferation of fibrocellular components in retinal neovascular diseases such as proliferative diabetic retinopathy and agerelated macular degeneration.