Transforming Growth Factor- 1 Stimulates Vascular Endothelial Growth Factor 164 via Mitogen-activated Protein Kinase Kinase 3-p38 and p38 Mitogen-activated Protein Kinase-dependent Pathway in Murine Mesangial Cells*

Transforming growth factor1 (TGF1) is a potent inducer of extracellular matrix synthesis leading to progressive glomerular fibrosis. The intracellular signaling mechanisms involved in this process remain incompletely understood. The p38 mitogen-activated protein kinase (MAPK) is a major stress signal transducing pathway that is rapidly activated by TGF1 in mesangial cells. We have previously demonstrated MKK3 as the immediate upstream MAPK kinase required for selective activation of p38 MAPK isoforms, p38 and p38 , and stimulation of pro1(I) collagen by TGF1 in murine mesangial cells. In this study, we further sought to determine MAPK kinase 3 (MKK3)-dependent TGF1 responses by gene expression profiling analysis utilizing mesangial cells isolated from Mkk3( / ) mice compared with Mkk3( / ) controls. Interestingly, vascular endothelial growth factor (VEGF) was identified as a TGF1-induced gene affected by deletion of Mkk3. VEGF is a well known endothelial mitogen, whose actions in nonendothelial cell types are still not well understood. We confirmed that TGF1 increased VEGF mRNA and protein synthesis of VEGF164 and VEGF188 isoforms in wild-type mesangial cells. However, in the Mkk3( / ) mesangial cells, both TGF1-induced VEGF mRNA and VEGF164 protein expression were inhibited, whereas TGF1-induced VEGF188 protein expression was unaffected. Furthermore, transfection of dominant negative mutants of p38 and p38 resulted in marked inhibition of TGF1-induced VEGF164 expression but not VEGF188, and treatment with recombinant mouse VEGF164 increased collagen and fibronectin mRNA expression in mesangial cells. Taken together, our findings suggest a critical role for the MKK3-p38 and p38 MAPK pathway in mediating VEGF164 isoform-specific stimulation by TGF1 in mesangial cells. Further, VEGF164 stimulates collagen and fibronectin expression in mesangial cells and thus in turn enhances TGF1induced extracellular matrix and may play an important role in progressive glomerular fibrosis. Mitogen-activated protein kinases (MAPKs) constitute a family of serine/threonine kinases that are central in the signaling cascades regulating a wide array of intracellular processes such as cell growth, differentiation, apoptosis, and cellular responses to external stress signals (1). Three major subfamilies of MAPKs have been identified in mammalian cells and include the extracellular signal-regulated kinase (ERK), the c-Jun N-terminal kinase (JNK), and the p38 MAPK (2). These protein kinases share 60–70% identity, but differ in the sequence and size of their activation loop, and are activated by different stimuli, and each MAPK subfamily consists of several isoforms and members, which often have distinct functions. Activation of the MAPK signaling cascade involves a series of three protein kinases consisting of a MAPK that is activated by the dual phosphorylation of Thr and Tyr residues in a TXY motif by specific upstream MAPK kinases (MKKs), which in turn are phosphorylated by MAPK kinase kinases. The TXY motif is unique to each of the MAPK subfamilies, where X is Glu, Pro, and Gly in ERK, JNK, and p38 MAPK, respectively (2). In general, ERK1 and ERK2, also known as p44 and p42 MAPKs, are prototypically activated by mitogenic stimuli and growth factors, whereas JNK and p38 MAPK are activated predominantly by environmental stresses such as osmotic changes, ultraviolet light, heat shock, and inflammation (2–5). More recent investigations including ours have revealed that p38 MAPK is also activated by a variety of cytokines including transforming growth factor1 (TGF1) (6–9). TGF1 is a pleiotropic cytokine that regulates multiple cellular functions including cell proliferation, differentiation, and apoptosis (10, 11). TGF1 also plays a key role in progressive diseases as a potent inducer of extracellular matrix (ECM) protein synthesis and progressive tissue fibrosis, including the kidney (12–14). There is now increasing body of evidence to provide support for the involvement of p38 MAPK pathway in pathological processes and implicate an important role in mediating TGF1 signals. For instance, activation of p38 MAPK has been demonstrated in various disease models such as inflammation, septic shock, ischemia, ischemia-reperfusion, vascular injury, and pulmonary fibrosis (15–17). Moreover, specific inhibitors of p38 MAPK were shown to attenuate disease severity in these studies. In the kidney, increased p38 MAPK activation has been demonstrated in ischemic and ischemic* This work was supported in part by NIDDK, National Institutes of Health Grant R01 DK57661 and Grant-in-aid 0355746U from the American Heart Association (to M. E. C.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ‡ To whom correspondence should be addressed: Dept. of Medicine, Renal-Electrolyte Division, University of Pittsburgh School of Medicine, E1158 Biomedical Science Tower, 200 Lothrop St., Pittsburgh, PA 15213. Tel.: 412-648-8617; Fax: 412-648-7010; E-mail: choim@pitt.edu. 1 The abbreviations used are: MAPK, mitogen-activated protein kinase; ERK, extracellular signal-regulated kinase; JNK, c-Jun N-terminal kinase; TGF, transforming growth factor; MKK, MAPK kinase; ECM, extracellular matrix; VEGF, vascular endothelial growth factor; FBS, fetal bovine serum; MOPS, 4-morpholinepropanesulfonic acid; PAI, plasminogen activator inhibitor; dnm, dominant negative mutant. THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 279, No. 32, Issue of August 6, pp. 33213–33219, 2004 © 2004 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A.

Transforming growth factor-␤1 (TGF-␤1) is a potent inducer of extracellular matrix synthesis leading to progressive glomerular fibrosis. The intracellular signaling mechanisms involved in this process remain incompletely understood. The p38 mitogen-activated protein kinase (MAPK) is a major stress signal transducing pathway that is rapidly activated by TGF-␤1 in mesangial cells. We have previously demonstrated MKK3 as the immediate upstream MAPK kinase required for selective activation of p38 MAPK isoforms, p38␣ and p38␦, and stimulation of pro-␣1(I) collagen by TGF-␤1 in murine mesangial cells. In this study, we further sought to determine MAPK kinase 3

(MKK3)-dependent TGF-␤1 responses by gene expression profiling analysis utilizing mesangial cells isolated from Mkk3(؊/؊) mice compared with Mkk3(؉/؉) controls. Interestingly, vascular endothelial growth factor (VEGF) was identified as a TGF-␤1-induced gene affected by deletion of Mkk3.
VEGF is a well known endothelial mitogen, whose actions in nonendothelial cell types are still not well understood. We confirmed that TGF-␤1 increased VEGF mRNA and protein synthesis of VEGF 164 and VEGF 188 isoforms in wild-type mesangial cells. However, in the Mkk3(؊/؊) mesangial cells, both TGF-␤1-induced VEGF mRNA and VEGF 164 protein expression were inhibited, whereas TGF-␤1-induced VEGF 188 protein expression was unaffected. Furthermore, transfection of dominant negative mutants of p38␣ and p38␦ resulted in marked inhibition of TGF-␤1-induced VEGF 164 expression but not VEGF 188 , and treatment with recombinant mouse VEGF 164 increased collagen and fibronectin mRNA expression in mesangial cells. Taken together, our findings suggest a critical role for the MKK3-p38␣ and p38␦ MAPK pathway in mediating VEGF 164

isoform-specific stimulation by TGF-␤1 in mesangial cells. Further, VEGF 164 stimulates collagen and fibronectin expression in mesangial cells and thus in turn enhances TGF-␤1induced extracellular matrix and may play an important role in progressive glomerular fibrosis.
Mitogen-activated protein kinases (MAPKs) 1 constitute a family of serine/threonine kinases that are central in the signaling cascades regulating a wide array of intracellular processes such as cell growth, differentiation, apoptosis, and cellular responses to external stress signals (1). Three major subfamilies of MAPKs have been identified in mammalian cells and include the extracellular signal-regulated kinase (ERK), the c-Jun N-terminal kinase (JNK), and the p38 MAPK (2). These protein kinases share ϳ60 -70% identity, but differ in the sequence and size of their activation loop, and are activated by different stimuli, and each MAPK subfamily consists of several isoforms and members, which often have distinct functions. Activation of the MAPK signaling cascade involves a series of three protein kinases consisting of a MAPK that is activated by the dual phosphorylation of Thr and Tyr residues in a TXY motif by specific upstream MAPK kinases (MKKs), which in turn are phosphorylated by MAPK kinase kinases. The TXY motif is unique to each of the MAPK subfamilies, where X is Glu, Pro, and Gly in ERK, JNK, and p38 MAPK, respectively (2). In general, ERK1 and ERK2, also known as p44 and p42 MAPKs, are prototypically activated by mitogenic stimuli and growth factors, whereas JNK and p38 MAPK are activated predominantly by environmental stresses such as osmotic changes, ultraviolet light, heat shock, and inflammation (2)(3)(4)(5). More recent investigations including ours have revealed that p38 MAPK is also activated by a variety of cytokines including transforming growth factor-␤1 (TGF-␤1) (6 -9).
TGF-␤1 is a pleiotropic cytokine that regulates multiple cellular functions including cell proliferation, differentiation, and apoptosis (10,11). TGF-␤1 also plays a key role in progressive diseases as a potent inducer of extracellular matrix (ECM) protein synthesis and progressive tissue fibrosis, including the kidney (12)(13)(14). There is now increasing body of evidence to provide support for the involvement of p38 MAPK pathway in pathological processes and implicate an important role in mediating TGF-␤1 signals. For instance, activation of p38 MAPK has been demonstrated in various disease models such as inflammation, septic shock, ischemia, ischemia-reperfusion, vascular injury, and pulmonary fibrosis (15)(16)(17). Moreover, specific inhibitors of p38 MAPK were shown to attenuate disease severity in these studies. In the kidney, increased p38 MAPK activation has been demonstrated in ischemic and ischemic-reperfused rat kidneys and in glomeruli isolated from rats with experimental proliferative glomerulonephritis and diabetic nephropathy, and administration of a specific p38 MAPK inhibitor ameliorated acute glomerulonephritis (18 -21).
We have previously demonstrated that TGF-␤1 rapidly and strongly activated p38 MAPK in cultured glomerular mesangial cells from the rat and mouse (8,9). Activation of the p38 MAPK involves phosphorylation by upstream MKKs in the protein kinase cascade and includes MKK3, MKK6, and possibly MKK4 (22). Our previous studies have also demonstrated MKK3 as the immediate upstream MAPK kinase required for selective activation of p38 MAPK isoforms, p38␣ and p38␦, and stimulation of pro-␣1(I) collagen by TGF-␤1 in mouse mesangial cells (9). In the present study, we further sought to determine MKK3-dependent TGF-␤1 responses by gene expression profiling analysis utilizing mesangial cells isolated from MKK3-null (Mkk3Ϫ/Ϫ) mice compared with wild-type (Mkk3ϩ/ϩ) control mice. Interestingly, vascular endothelial growth factor (VEGF) was identified as a TGF-␤1-induced gene affected by the deletion of Mkk3 in mesangial cells. VEGF is a well known endothelial mitogen that plays an important role in angiogenesis, wound repair, tumorigenesis, and vascular diseases (23,24). However, its actions in nonendothelial cell types are still not well understood. Moreover, the molecular mechanism responsible for isoform-specific VEGF gene regulation are not known. Our data provide novel evidence that this occurs through activation of distinct intracellular signaling molecules. Here we report that the MKK3-p38␣ and p38␦ MAPK pathway is required in mediating VEGF 164 isoform-specific stimulation by TGF-␤1 in mouse mesangial cells. Further, VEGF 164 induces collagen and fibronectin expression in mouse mesangial cells and thus in turn enhances TGF-␤1-induced ECM and may play an important role in progressive glomerular fibrosis.

EXPERIMENTAL PROCEDURES
Reagents-Recombinant human TGF-␤1 was purchased from R & D Systems (Minneapolis, MN). VEGF (147) rabbit polyclonal antibodies and ␣-tubulin antibodies were obtained from Santa Cruz Biotechnology. Mouse recombinant VEGF 164 was purchased from Sigma. The specific inhibitor of p38␣ MAPK, SB203580, was purchased from Calbiochem (San Diego, CA). Geneticin® (G418 sulfate) and LipofectAMINE Plus TM reagent were obtained from Invitrogen.
Murine Mesangial Cell Culture-Glomerular mesangial cells were isolated and characterized as previously described, from glomeruli of MKK3-null (Mkk3Ϫ/Ϫ) mice and wild-type (Mkk3ϩ/ϩ) control mice and from C57BL/6 mice, using differential sieving technique with the following modifications (9). Following collagenase digestion, the cells were plated in RPMI 1640 medium (Mediatech) supplemented with 20% FBS (Bio-Whittaker), insulin (10 g/ml), 5 units/ml penicillin, and 5 g/ml streptomycin and incubated in a humidified atmosphere of 5% CO 2 and 95% air at 37°C. Using this technique, we have been successful in establishing homogeneous cultures of glomerular mesangial cells that immunostain for anti-vimentin (Dako) and anti-myosin antibodies (Zymed Laboratories Inc. and negative staining for cytokeratin (Roche Applied Science) and von Willebrand's factor (Dianova) as well as negative fluorescent acetylated low density lipoprotein uptake (Biomedical Technologies Inc.). Cells between 7 and 16 passages were used for the experiments. The targeted disruption of the Mkk3 gene by homologous recombination and the generation of Mkk3(Ϫ/Ϫ) mice were as previously described (9).
Once the cells were established in culture, they were maintained in RPMI 1640 medium with 15% FBS, 5 units/ml penicillin, and 5 g/ml streptomycin. In experiments involving treatment with exogenous TGF-␤1 (2 ng/ml) or VEGF (100 ng/ml), the cells grown to subconfluence were rendered quiescent in serum-free medium for 24 h, followed by treatment with human TGF-␤1 (2 ng/ml) (R & D Systems) or VEGF (100 ng/ml) (Sigma). In experiments using the p38 MAPK inhibitors, the cells were preincubated for 1 h in the absence or presence of 10 -30 M SB203580, prior to treatment with or without exogenous TGF-␤1.
cDNA Array Analysis-Quiescent MKK3-deficient (Mkk3Ϫ/Ϫ) and wild-type (Mkk3ϩ/ϩ) mesangial cells were incubated in the absence or presence of exogenous TGF-␤1 (2 ng/ml) for 6 h, and total RNA was isolated by cell lysis with TRIzol (Invitrogen) according to the manufacturer's instructions. DNase-treated RNA (4 g) was reverse-transcribed to cDNA and labeled with [␣-32 P]dATP and then hybridized to the Atlas TM mouse 1.2 arrays containing 1176 mouse genes (Clontech). After exposure to a PhosphorImager screen (Molecular Dynamics), the signals were analyzed using the ImageQuant software (Molecular Dynamics). The relative levels of gene expression were quantified after subtraction of the background, and signal values were normalized by dividing by the mean filter hybridization signal. Three independent experiments were performed. Genes were considered differentially expressed when they exhibited a 2-fold or greater increase in the TGF-␤1-treated cells compared with the untreated cells. The relative expression of housekeeping genes (glyceraldehyde-3-phosphate dehydrogenase, ␤-actin, and 40 S ribosomal protein S29) served to normalize gene expression levels and did not differ by more than 10% between the Mkk3Ϫ/Ϫ and Mkk3ϩ/ϩ mesangial cells.
Stable Transfection of Murine Mesangial Cells-To generate clones that stably expressed p38␣ and p38␦ dominant negative mutant (dnm), cells were transfected using LipofectAMINE Plus TM reagent (Invitrogen) as follows. Cells grown to ϳ60% confluence on 100-mm plates were incubated with 5 g of DNA, p38␣, and p38␦ dnm constructs ligated in pcDNA3.1, as previously described (25), in RPMI and LipofectAMINE Plus reagent complex at 37°C in 5% CO 2 atmosphere. Control cells were incubated with empty vector pcDNA3.1 (not containing p38␣ and p38␦ dnm) in RPMI and Plus-LipofectAMINE TM reagent complex. After 3-4 h of incubation, the transfection media were replaced with RPMI containing 20% FBS (no antibiotics), and cells were incubated for another 24 h. To select for stable transfectants, the cells were treated with 400 g/ml Geneticin® (G418 Sulfate) in RPMI medium containing 20% FBS, and the medium was changed every 3-4 days. G418-resistant clones emerging ϳ14 days after transfection were subcloned using ring cylinders, expanded, and maintained in RPMI medium containing 15% FBS, 200 g/ml G418, 5 units/ml penicillin, and 5 g/ml streptomycin.
Western Blot Analysis-The cells were washed with ice-cold phosphate-buffered saline, followed by lysis in radioimmune precipitation assay buffer (1ϫ phosphate-buffered saline, 1% Igepal CA-630, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM phenylmethylsulfonyl fluoride, 1 g/ml aprotinin, 1 g/ml leupeptin, 1 g/ml pepstatin, 1 mM Na 3 VO 4 , 1 mM NaF). The cell lysates were passed through 21-gauge needles several times and then centrifuged for 15 min at 14,000 ϫ g at 4°C. The protein concentration was determined by BCA protein assay reagent kit (Pierce). 80-g protein samples for each group were loaded onto 12% SDS-PAGE gels and transferred to polyvinylidene difluoride membranes. The membranes were blocked with 5% nonfat milk for 1 h and incubated with anti-VEGF antibodies (1:1000 dilution) overnight on a rocker at 4°C followed by four or five washes (30 min/each) with buffer containing 10 mM Tris, pH 7.5, 50 mM NaCl, and 0.1% Tween 20 and then incubated incubation with horseradish peroxidase-conjugated anti-rabbit antibodies for 1 h at room temperature. Signal development was carried out using LumiGLO (New England Biolabs) and exposure x-ray film. Three independent experiments were performed with essentially the same results, and representative blots are shown.
Northern Blot Analysis-Total RNA in each group was isolated by cell lysis with TRIzol (Invitrogen) according to the manufacturer's instructions and was size-fractionated (15 g/lane) on a 1% agarose, 2% formaldehyde gel in 20 mM MOPS, 5 mM sodium acetate, and 1 mM EDTA (pH 7.2). mRNA was transferred and UV cross-linked to nylon membranes (Gene Screen Plus; Dupont). The blots were prehybridized for 2 h in Church Gilbert's hybridization buffer (Quality Biological, Inc.) and hybridized overnight in same solution containing the appropriate 32 P-labeled probe at 65°C. The VEGF cDNA probe used was a 0.65-kb EcoRI insert of the human full-length VEGF cDNA cloned in pGEMT (Promega) and verified by sequencing. The human pro-␣1(I)collagen, fibronectin 1, and plasminogen activator inhibitor (PAI)-1 cDNA probes used were previously described (8,9). The blots were then washed two times in solution containing 0.5% bovine serum albumin, 5% SDS, 40 mM phosphate buffer (pH 7.0), and 1 mM EDTA (pH 8.0) for 30 min each at 65°C, followed by 15-min washes with solution containing 1% SDS, 40 mM phosphate buffer (pH 7.0), and 1 mM EDTA (pH 8.0) at 65°C. The blots were exposed to Kodak X-AR film. Similar results were obtained from three independent experiments, and representative blots are shown. To control for the relative equivalence of RNA loading, the same blots were hybridized with 32 P-labeled oligonucleotide probe corresponding to the 18 S rRNA as previously described (9).
TGF-␤1 Induces VEGF Gene Expression in Murine Mesangial Cells-We first sought to confirm whether TGF-␤1 stimulated VEGF mRNA expression in cultured wild-type murine mesangial cells by Northern blot analyses. As shown in Fig. 2, VEGF mRNA was detected in wild-type mesangial cells by bands of ϳ3.7 kb in size. Exposure to exogenous TGF-␤1 (2 ng/ml) resulted in increased VEGF mRNA levels as early as after 4 h of treatment in mesangial cells and up to 24 h after TGF-␤1 treatment.
TGF-␤1 Induces VEGF 164 and VEGF 188 Isoforms in Murine Mesangial Cells-We next determined whether TGF-␤1 stimulated the synthesis of VEGF protein by Western blot analyses using polyclonal VEGF. As shown in Fig. 3, two bands, ϳ19 -21 and 28 -30 kDa in size, were observed in cultured wild-type murine mesangial cells corresponding to VEGF 164 and VEGF 188 isoforms, respectively. Exposure to exogenous TGF-␤1 (2 ng/ml) resulted in time-dependent increases in the levels of VEGF 164 and VEGF 188 isoforms up to 24 h after TGF-␤1 treatment.
MKK3 Is Required for Induction of VEGF 164 Isoform by TGF-␤1 in Murine Mesangial Cells-We also examined the effects of MKK3 deficiency on TGF-␤1-stimulated protein synthesis of VEGF isoforms. As shown in Fig. 5, treatment with exogenous TGF-␤1 (2 ng/ml) resulted in increased protein expression of VEGF isoforms VEGF 164 and VEGF 188 as expected in wild-type (Mkk3ϩ/ϩ) mesangial cells. In contrast, in MKK3deficient (Mkk3Ϫ/Ϫ) mesangial cells, TGF-␤1 treatment resulted in increased expression of the VEGF 188 protein but not the VEGF 164 isoform. The previously observed TGF-␤1 stimulated VEGF 164 isoform expression in the wild-type mesangial cells was completely abrogated in MKK3-deficient (Mkk3Ϫ/Ϫ) mesangial cells. A higher concentration of exogenous TGF-␤1 (10 ng/ml) also failed to induce VEGF 164 isoform expression in MKK3-deficient (Mkk3Ϫ/Ϫ) mesangial cells (data not shown).

FIG. 3. Time course of TGF-␤1-stimulated VEGF protein synthesis in murine mesangial cells.
Wild-type mouse mesangial cells were incubated in the absence (Ctl) or in the presence of exogenous TGF-␤1 (2 ng/ml) for the indicated time periods. Total cell lysates were isolated and subjected to Western blot analyses using polyclonal anti-VEGF antibodies, as described under "Experimental Procedures." As loading controls, the same cell lysates were subjected to immunoblotting with mouse monoclonal anti-␣-tubulin antibodies.

FIG. 4. Inhibition of TGF-␤1-stimulated VEGF gene expression in MKK3-deficient murine mesangial cells.
Total RNA isolated from wild-type (Mkk3ϩ/ϩ) and MKK3-deficient (Mkk3Ϫ/Ϫ) mouse mesangial cells incubated in the absence (Ctl) or in the presence of exogenous TGF-␤1 (2 ng/ml) for the indicated time periods were subjected to Northern blot hybridization with 32 P-labeled cDNA probe corresponding to VEGF. 18 S rRNA hybridization signals served as normalization for RNA loading.

TGF-␤1 Stimulates VEGF 164 via MKK3-p38-dependent Pathway Dominant Negative Mutants of p38␣ and p38␦ Inhibits Induction of VEGF 164 Isoform by TGF-␤1 in Murine Mesangial
Cells-Our previous study had demonstrated that MKK3 is the upstream MAPK kinase for the selective activation of p38␣ and p38␦ MAPK isoforms by TGF-␤1 in murine mesangial cells (9). Here we further investigated whether p38␣ and/or p38␦ were involved in mediating the TGF-␤1-stimulated VEGF expression in murine mesangial cells. We employed an approach of stable transfection of dominant negative mutants of p38␣ or p38␦ (p38␣ dnm and p38␦ dnm) to block the p38 MAPK signaling pathway, in an isoform-specific fashion, in wild-type murine mesangial cells. As shown in Fig. 6, following treatment with exogenous TGF-␤1 (2 ng/ml) increased protein levels of VEGF 188 were observed in mesangial cells stably transfected with p38␣ dnm or p38␦ dnm, as well as in empty vector (pcDNA3.1) transfected control cells. In contrast, TGF-␤1stimulated VEGF 164 isoform observed in the empty vector (pcDNA3.1) control cells was markedly inhibited in mesangial cells stably transfected with p38␣ dnm or p38␦ dnm.
p38 MAPK Inhibitor SB 203580 Inhibits Induction of VEGF 164 isoform by TGF-␤1 in Murine Mesangial Cells-We next utilized an alternative approach to block the p38 MAPK signaling pathway through chemical inhibition using SB203580, an inhibitor of p38␣ and p38␤ activation. As seen in Fig. 7, in wild-type murine mesangial cells, dose-dependent inhibition of the TGF-␤1-stimulated VEGF 164 isoform was observed after pretreatment with p38 MAPK inhibitor SB203580. Given that SB203580 specifically inhibits p38␣ and p38␤, but the other p38 isoforms including p38␦ are insensitive to inhibition by SB203580, we examined the combined effects of SB203580 and overexpression of dominant negative mutant p38␦ in cultured wild-type murine mesangial cells. Fig. 8 demonstrates that pretreatment of SB203580 in mesangial cells stably transfected with p38␦ dnm resulted in a more complete inhibition of VEGF 164 isoform induction by TGF-␤1, compared with similar pretreatment of SB203580 in empty vector (pcDNA3.1)-transfected control cells.
Recombinant Mouse VEGF 164 Induces ECM Genes in Murine Mesangial Cells-It is well known that VEGF is an endothelial mitogen, but little is known about its actions in nonendothelial cell types, including glomerular mesangial cells. Here we used recombinant mouse VEGF 164 to investigate its effects in murine mesangial cells. The effects of VEGF 164 on the expression of ECM genes, namely collagen ␣1(I) and fibronectin and PAI-1, were examined by Northern analyses. As shown in Fig.  9, treatment with recombinant mouse VEGF 164 (100 ng/ml) resulted in increased expression of collagen ␣1(I) and fibronectin mRNA in cultured wild-type murine mesangial cells. Both collagen and fibronectin are known to be potently induced by TGF-␤1 in mesangial cells. DISCUSSION Although it is known that TGF-␤1 is a potent inducer of ECM synthesis and is largely regarded as a key mediator in the pathogenesis of renal fibrosis, the intracellular signaling mechanisms involved in this process remain incompletely understood. The MKK3-p38 MAPK is emerging as a major stress signal transducing pathway activated by TGF-␤1 and implicating an important role in mediating TGF-␤1 signals. Our previous studies have demonstrated that TGF-␤1 rapidly and strongly activated p38 MAPK and its immediate upstream MAPK kinase, MKK3, in cultured glomerular mesangial cells, a major target cell type is a variety of renal glomerular injury (8,9). In the present study, gene expression profiling analysis, utilizing mesangial cells isolated from Mkk3(Ϫ/Ϫ) mice compared with Mkk3(ϩ/ϩ) controls, have identified VEGF as a MKK3-dependent TGF-␤1 target gene in mesangial cells (Fig.  1).
VEGF was initially described as a vascular permeability factor and subsequently as an endothelial mitogen (26 -29). VEGF has been shown to act a key regulator of physiological angiogenesis, such as in embryogenesis, as well as pathological FIG. 5. Effects of MKK3 deficiency on TGF-␤1-stimulated protein synthesis of VEGF isoforms, VEGF 164 and VEGF 188 , in murine mesangial cells. Total cell lysates from wild-type (Mkk3ϩ/ϩ) and MKK3-deficient (Mkk3Ϫ/Ϫ) mouse mesangial cells incubated in the absence (Ϫ) or in the presence (ϩ) of exogenous TGF-␤1 (2 ng/ml) for 24 h were subjected to Western blot analyses using polyclonal anti-VEGF antibodies, as described under "Experimental Procedures." As loading controls, the same cell lysates were subjected to immunoblotting with mouse monoclonal anti-␣-tubulin antibodies.
FIG. 6. Inhibition of TGF-␤1-stimulated the VEGF 164 isoform by dominant negative mutants of p38␣ and p38␦ in murine mesangial cells. Wild-type mouse mesangial cells transfected with empty vector (pcDNA3.1), dominant negative mutant of p38␣ (p38␣ dnm), or dominant negative mutant of p38␦ (p38␦ dnm), were incubated in the absence (Ϫ) or in the presence (ϩ) of exogenous TGF-␤1 (2 ng/ml) for 24 h. Total cell lysates were isolated and subjected to Western blot analyses using polyclonal anti-VEGF antibodies, as described under "Experimental Procedures." As loading controls, the same cell lysates were subjected to immunoblotting with mouse monoclonal anti-␣-tubulin antibodies.
FIG. 7. Effects of SB203580 on TGF-␤1-stimulated protein synthesis of VEGF isoforms, VEGF 164 and VEGF 188 , in murine mesangial cells. Wild-type mouse mesangial cells, pretreated without (Ϫ) or with 10 -30 M SB203580, were incubated in the absence (Ϫ) or in the presence (ϩ) of exogenous TGF-␤1 (2 ng/ml) for 24 h. Total cell lysates were isolated and subjected to Western blot analyses using polyclonal anti-VEGF antibodies, as described under "Experimental Procedures." As loading controls, the same cell lysates were subjected to immunoblotting with mouse monoclonal anti-␣-tubulin antibodies.
angiogenesis, most notably neovascularization in tumorigenesis (24,30). However, besides these well documented activities, VEGF can also exert other certain effects such as promoting atherosclerotic plaque development and inflammatory response through monocyte activation and migration (31,32). Indeed, although endothelial cells are generally thought to be the primary target of VEGF, nonendothelial cell types express VEGF and its receptors and can be targets of VEGF actions. Aortic smooth muscle cells, which play a pivotal role in the pathogenesis of atherosclerosis, express and secrete VEGF, and hypoxia and hypoglycemia have been demonstrated to stimulate VEGF expression in these cells (33,34). Enhanced VEGF expression was also observed in retinal pigment epithelial cells stimulated by hypoxia and by advanced glycation products and in osteoblasts that are thought to promote bone formation (35)(36)(37). Renal glomerular expression of VEGF has also been demonstrated, for instance, in glomerular visceral epithelial cells where VEGF may have a role in the induction of proteinuria in renal diseases (38). Both VEGF and VEGF receptors were up-regulated in kidneys of diabetic rats, and administration of monoclonal anti-VEGF antibodies improved early renal dysfunction, as assessed by decreased hyperfiltration, albuminuria, and glomerular hypertrophy, in experimental diabetes, suggesting a role in the pathogenesis of diabetic nephropathy (39,40). Up-regulation of VEGF expression in the glomerular mesangium has also been reported in human and experimental mesangioproliferative glomerulonephritis (41,42). It is noteworthy that in both of these pathological processes, TGF-␤1 is thought to be a key mediator. Our present findings confirm that glomerular mesangial cells express VEGF mRNA and, further, that stimulation with exogenous TGF-␤1 strongly induces the expression of VEGF in a time-dependent fashion (Fig. 2).
VEGF, also known as VEGF-A, in its native form is a disulfide-linked homodimeric glycoprotein of 34 -46 kDa in molecular size (30). At least six human VEGF isoforms (VEGF 121 , VEGF 145 , VEGF 165 , VEGF 183 , VEGF 189 , and VEGF 206 ) generated from alternative splicing of the VEGF mRNA have been identified (30,43). VEGF 121 is a freely soluble protein; VEGF 165 , the major isoform, is partially secreted, whereas VEGF 189 and VEGF 206 are almost completely sequestered in the extracellular matrix. VEGF 145 and VEGF 183 are less frequent splice variants (44). Murine VEGF is shorter than human VEGF by one amino acid. Analysis of exons suggests the generation of three isoforms, VEGF 120 , VEGF 164 , and VEGF 188 , in mice (26,43,44). In our studies, we clearly demonstrate increases in protein expression of VEGF isoforms, VEGF 164 and VEGF 188 , in glomerular mesangial cells (Fig. 3). In agreement with our results of gene expression profiling analysis, Northern analyses confirmed that in MKK3-deficient (Mkk3Ϫ/Ϫ) mesangial cells the TGF-␤1-induced VEGF mRNA was notably inhibited (Fig. 4), and moreover, TGF-␤1 treat- TGF-␤1 Stimulates VEGF 164 via MKK3-p38-dependent Pathway ment failed to induce the VEGF 164 isoform (Fig. 5), suggesting a critical role of the MKK3 pathway. We have previously established the specific absence of activated MKK3 protein in the MKK3Ϫ/Ϫ mesangial cells and that in the absence of MKK3 activation, TGF-␤1 was unable to induce downstream p38 MAPK phosphorylation, specifically the isoforms p38␣ and p38␦ MAPK (9). Accordingly, we examined whether either or both of the downstream p38 MAPK isoforms were involved in inducing VEGF expression by TGF-␤1 in mesangial cells. By isoform-specific dominant negative inhibition of p38␣ or p38␦, and additionally by chemical inhibition of p38␣ MAPK pathway using SB203580, we show the involvement of both isoforms in mediating TGF-␤1-stimulated VEGF 164 expression in mesangial cells (Figs. 6 and 7). However, dominant negative mutant of p38␣ or p38␦, or SB203580, individually did not completely inhibit TGF-␤1-stimulated VEGF 164 expression but required simultaneous blockade of both p38␣ and p38␦ (Fig. 8), indicating that both isoforms contributed to TGF-␤1-stimulated VEGF 164 expression in mouse mesangial cells.
Although the molecular mechanism of regulation of VEGF gene expression remains incompletely understood, an emerging body of evidence suggests that, indeed, the gene expression of VEGF isoforms derived from specific splice variants is cell type-and stimulus-specific. For example, VEGF isoforms are differentially transcribed by specific cell types within the mammary gland during mouse mammary gland development (45). Analysis of VEGF transcription by reverse transcriptase-PCR revealed mRNA for all three isoforms VEGF 120 , VEGF 164 , and VEGF 188 within the mammary gland of nulliparous females, but during pregnancy VEGF 188 levels declined and became undetectable during lactation in association with the increased abundance of VEGF 120 and VEGF 164 mRNAs. VEGF 188 mRNA transcription occurs as a late event during lipogenesis distinct from earlier induction of VEGF 120 and VEGF 164 mRNAs during differentiation (45). In the human uterus, 17␤-estradiol increases endometrial expression of all VEGF isoforms, whereas progesterone selectively increases the expression of the VEGF 189 isoform (46). Studies in hyperoxic acute lung injury by Corne et al. (47) show that interleukin-13 selectively stimulates VEGF 164 , whereas interleukin-13 plus hyperoxia stimulate VEGF 120 and VEGF 188 . Although the precise molecular mechanism responsible for such isoform-specific differential VEGF gene regulation is not known, our data provide novel evidence that this occurs through activation of distinct intracellular signaling molecules. In glomerular mesangial cells, TGF-␤1 selectively stimulates VEGF 164 (and not VEGF 188 ) via the MKK3-p38␣ and p38␦ MAPK-dependent pathway.
The expression of multiple VEGF isoforms in mammalian cells, and the different profiles of VEGF isoforms expressed in different cell types suggest that these VEGF isoforms may differ in their physiological function. Compared with VEGF 164 , VEGF 120 binds with lower affinity to VEGFR-1 and is less potent in stimulating endothelial growth (44). VEGF 164 is at least twice as potent as VEGF 120 in inducing intercellular adhesion molecule 1 (ICAM-1)-mediated leukocyte stasis within the retinal vasculature and blood-retinal barrier breakdown in diabetic retina (48). To date, little is known regarding the role of the distinct VEGF isoforms in glomerular mesangial cells. Demonstration of all three VEGF receptor types, VEGFR-1 (Flt-1), VEGFR-2 (KDR), and neurophilin-1, expressed in cultured glomerular mesangial cells indicate that the mesangial cells are potential targets of VEGF actions (49). In the present study, we have demonstrated that MKK3 mediates isoform-specific stimulation of VEGF 164 by TGF-␤1 in mesangial cells, and given our previous studies demonstrating role of MKK3 in TGF-␤1-stimulated collagen expression, we explored the effects of VEGF 164 , on ECM gene expression. Interestingly, treatment with recombinant mouse VEGF 164 increased the expression of two major ECM genes, collagen ␣1(I) and fibronectin, known to be potently induced by TGF-␤1 in mesangial cells (Fig. 9). Other known inducers of ECM such as angiotensin II, high glucose, and mechanical stretch also have been reported to stimulate VEGF in mesangial cells (50 -52). Thus, taken together, our findings suggest a critical role of MKK3-p38␣ and p38␦ MAPK pathway in mediating VEGF 164 isoform-specific stimulation by TGF-␤1 in mesangial cells. Further, VEGF 164 stimulates collagen and fibronectin expression in mesangial cells. Both collagen and fibronectin are known to be potently induced by TGF-␤1 in mesangial cells. Thus, TGF-␤1 stimulation of VEGF 164 can in turn further enhance the effects of TGF-␤1-induced ECM and may play an important role in progressive glomerular fibrosis.