A Novel Phosphoinositide 3-Kinase-dependent Pathway for Angiotensin II/AT-1 Receptor-mediated Induction of Collagen Synthesis in MES-13 Mesangial Cells*

Chronic activation of the angiotensin II (ANG II) type 1 receptor (AT-1R) is critical in the development of chronic kidney disease. ANG II activates mesangial cells (MCs) and stimulates the synthesis of extracellular matrix components. To determine the molecular mechanisms underlying the induction of MC collagen, a mouse mesangial cell line MES-13 was employed. ANG II treatment induced an increase in collagen synthesis, which was abrogated by co-treatment with losartan (an AT-1R antagonist), wortmannin (a phosphoinositide 3-kinase (PI3K) inhibitor), an Akt inhibitor, and stable transfection of dominant negative-Akt1. ANG II induced a significant increase in PI3K activity, which was abolished by co-treatment with losartan or 2′,5′-dideoxyadenosine (2′,5′-DOA, an adenylyl cyclase inhibitor) but not by PD123319 (an AT-2R antagonist) or H89 (a protein kinase A (PKA) inhibitor). The Epac (exchange protein directly activated by cAMP)-specific cAMP analog, 8-pHPT-2′-O-Me-cAMP, significantly increased PI3K activity, whereas a PKA-specific analog, 6-benzoyladenosine-cAMP, showed no effect. The ANG II-induced increase in PI3K activity was also blocked by co-treatment with PP2, an Src inhibitor, or AG1478, an epidermal growth factor receptor (EGFR) antagonist. ANG II induced phosphorylation of Akt and p70S6K and EGFR, which was abrogated by knockdown of c-Src by small interference RNA. Knockdown of Src also effectively abolished ANG II-induced collagen synthesis. Conversely, stable transfection of a constitutively active Src mutant enhanced basal PI3K activity and collagen production, which was abrogated by AG1478 but not by 2′,5′-DOA. Moreover, acute treatment with ANG II significantly increased Src activity, which was abrogated with co-treatment of 2′,5′-DOA. Taken together, these results suggest that ANG II induces collagen synthesis in MCs by activating the ANG II/AT-1R-EGFR-PI3K pathway. This transactivation is dependent on cAMP/Epac but not on PKA. Src kinase plays a pivotal role in this signaling pathway between cAMP and EGFR. This is the first demonstration that an AT1R-PI3K/Akt crosstalk, along with transactivation of EGFR, mediates ANG II-induced collagen synthesis in MCs.

Clinical and experimental studies have identified the reninangiotensin system as a key factor in development and progression of glomerular diseases (1)(2)(3). A chronically enhanced effect of the renin-angiotensin system is a central event in the development of proliferative and sclerosing changes in the glomeruli. The octapeptide hormone angiotensin II (ANG II), 2 the major renin-angiotensin system effector, activates mesangial cells (MCs) and increases the synthesis of extracellular matrix components, including collagens, which is a hallmark of the glomerular diseases (4,5). The physiological functions of ANG II are mediated by at least two structurally and pharmacologically distinct seven-transmembrane helices G protein-coupled receptors (GPCRs), ANG II type 1 and 2 receptors (AT-1R and AT-2R). Glomerular cells, including MCs, express primarily the AT-1R, which mediates most of the known effects of ANG II (6,7). Pharmacological blockade of the renin-angiotensin system by angiotensin-converting enzyme inhibitors or AT-1R antagonists attenuates development and progression of glomerular diseases (8 -10). Although the effects of these agents on progression in the glomerular diseases have been studied extensively, detailed cellular mechanisms of the effects have not been fully revealed.
Phosphoinositide 3-kinases (PI3Ks) phosphorylate inositolcontaining lipids at the D-3 position of the inositol ring (11). The mammalian PI3K can be divided into three major classes (classes I-III) based upon their structure and substrate specificity (12). The mammalian class I PI3K can be further divided into two subclasses: class IA PI3K are heterodimers of a 110-kDa catalytic subunit (p110␣, p110␤, or p110␦) and a regulatory subunit of 85 or 55 kDa (p85/p55), whereas the class IB PI3K (PI3K␥) is composed of a p110␥ catalytic subunit and a p101 regulatory subunit (11). The class I PI3Ks can phosphorylate phosphoinositide (PtdIns), PtdIns 4-phosphate, and PtdIns 4,5-bisphoshate in vitro. In the cells, however, class I PI3Ks preferentially convert PtdIns 4,5-bisphosphate to PtdIns 3,4,5-trisphosphate following stimulation by tyrosine kinase or heteromeric GPCRs (13). In general, class IA PI3Ks are activated by receptor tyrosine kinase pathways, whereas class IB PI3K (PI3K␥) is coupled to GPCRs (11). This is best represented by studies performed in leukocytes (14,15). We have recently demonstrated, however, that cardiac PI3K␣ can be activated by stimulation of ␤-adrenergic receptor in vivo (16). It is unclear which isoform of class I PI3Ks is functionally linked to AT-1R in MCs.
Following PI3K activation, PtdIns 3,4,5-trisphosphate recruits and phosphorylates the phosphoinositide-dependent kinase-1 and Akt/protein kinase B, bringing these proteins into proximity at the plasma membrane where Akt is phosphorylated on threonine 308 by phosphoinositide-dependent kinase-1. This is followed by phosphorylation at serine 473 by a yet-to-be identified mechanism (17). Growing evidence indicates that ANG II/AT-1R signaling induces a crosstalk to the PI3K/Akt pathway (18 -21), and a few studies have shown that these changes in PI3K/Akt activities demonstrated altered phenotypes in collagen synthesis of various types of cells, including MCs (22)(23)(24)(25). These findings suggested that ANG II/AT-1R signaling is implicated in hypersynthesis of collagens in MCs through a crosstalk with the PI3K/Akt signaling pathway and is involved in the development and progression of various kinds of glomerular diseases. However, the detailed mechanisms of ANG II-induced collagen synthesis in MCs and a role of ANG II in the signaling pathway have not been fully elucidated. To explore the ANG II/AT-1R to PI3K/Akt pathway, we employed an in vitro culture system by using the well characterized mouse mesangial cell line, MES-13. With pharmacological and genetic signal modification studies we demonstrated that several factors such as cAMP, Src kinase, and epidermal growth factor receptor (EGFR) play critical roles in the pathway.

EXPERIMENTAL PROCEDURES
Materials-All chemicals and reagents were obtained from Sigma unless stated otherwise.
Cell Culture-MES-13 mouse mesangial cells (ATCC no. CRL-1927) were grown in a 3:1 mixture of Dulbecco's modified Eagle's and Ham's F-12 media (Invitrogen) supplemented with 5% (v/v) fetal bovine serum containing 50 units/ml penicillin G and 50 g/ml streptomycin in a humidified atmosphere containing 5% CO 2 at 37°C. Cells were first grown up to 70% confluency and synchronized overnight in serum-free medium prior to treatment. For immunoprecipitation and PI3K assay, the quiescent cells were treated with vehicle or ANG II (0.1 M) for various durations. In some experiments cells were pretreated with selective inhibitors for indicated durations before ANG II treatment.
Immunoprecipitation and Western Blotting-For immunoprecipitation, MES-13 cells were treated with ANG II (0.1 M) for 1-30 min. Protein lysates (0.5 mg) were incubated with protein G-Sepharose beads (GE Healthcare, Piscataway, NJ) for 2 h at 4°C. After centrifugation, the supernatant was transferred to a fresh tube and incubated with an antibody specific for EGFR (1005, Santa Cruz Biotechnology, Santa Cruz, CA) at 4°C for 4 h with continuous rotation. A 20-l packed volume of protein G-Sepharose was added, and the mixture was incubated overnight at 4°C. After washing, 45 l of 2ϫ Laemmli sample buffer was added. The sample was heated in boiling water for 5 min and quenched on ice for 2 min. After vortex and centrifuge, 20 l of the supernatant was resolved on a 7.5% SDS-PAGE gel and immunoblotted with anti-pY (Santa Cruz Biotechnology). The blot was stripped and re-blotted with anti-EGFR. Collagen expressions were measured in cells treated with ANG II (0.1 M) for 24 h by Western blotting using anti-procollagen type I (Y-18, Santa Cruz Biotechnology) and normalized with ␤-actin levels. Protein phosphorylation of Akt and p70S6K was measured in cells treated with ANG II (0.1 M) for 1-30 min by Western blotting using phospho-specific antibodies against Akt (Thr-308) and p70S6K (Thr-389) and normalized with total Akt and p70S6K (Cell Signaling Technology, Beverly, MA) as previously described (26). The results were visualized by ECL chemiluminescence (Pierce).
Stable Transfection-A vector expressing dominant negative Akt1 was engineered by inserting a K179M mutant dominant negative Akt1 cDNA (27) into a multiple cloning site of an eukaryotic expression vector, pUSEamp(ϩ) (Upstate, Charlottesville, VA). A 529F mutant constitutively active Src cDNA (28) was processed similarly to engineer a vector expressing constitutively active Src. Stable transfection of the constructs in MES-13 cells was performed using Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. Individual single cells were isolated and screened for neomycin resistance. The stably transfected cells were treated similarly as described above.
Src Activity Assay-Protein lysates (1 mg) from MES-13 cells were immunoprecipitated with anti-c-Src antibody (N-16, Santa Cruz Biotechnology). Kinase activity was determined by measuring phosphorylation of a specific Src substrate (KVEKIGEGTYGVVYK) using an Src assay kit (Upstate). Briefly, the c-Src immunoprecipitated beads were incubated with a [␥-32 P]ATP-ATP-Mg 2ϩ mix at 30°C for 10 min. A sample without the substrate peptide was included as a background control. Reactions were terminated by adding 40% trichloroacetic acid. After centrifugation the supernatants that include phosphorylated substrate were transferred onto Whatman P81 ion-exchange cellulose chromatography paper circles, washed five times in 0.75% phosphoric acid, washed once in acetone, and then counted in a scintillation counter.
Transfection of siRNA for c-Src-Two complementary hairpin siRNA template oligonucleotides, containing 21-nucleotide target sequences of the mouse c-Src tyrosine kinase (5Ј-AAG TAC AAC TTC CAT GGC ACT-3Ј, GenBank TM BC052006), were annealed and ligated into pScilencer TM 5.1-H1 Retro vector (Ambion, Austin, TX). The vector was then transfected to MES-13 cell using Lipofectamine 2000 (Invitrogen). After the transfection, individual single cells were isolated and screened for puromycin (1000 ng/ml) resistance. As a control of the experiment, the transfection was repeated using scrambled oligonucleotides of the same length with minimum homology to mammalian genes. The knockdown of c-Src was confirmed by RT-PCR and Western blotting.
Statistical Analysis-Statistical significance of the difference among groups was analyzed by the paired Student's t test or one-way analysis of variance followed by Newman-Keuls test. All the statistical analysis was done using Statistica Version 5.0 (StatSoft, Tulsa, OK). All data were represented as the mean Ϯ S.E. of three different experiments unless otherwise noted in the figure legends. A probability of p Ͻ 0.05 was considered to represent a significant difference.

ANG II Induces Increases in Phosphotyrosine-associated PI3K
Activity and Phosphorylation of p70S6K and Akt through AT-1R-In MES-13 cells, acute treatment with ANG II (0.1 M) induced a significant increase in phosphotyrosine-associated PI3K activity (Fig. 1A). The increases in PI3K activity were seen 1 min after ANG II treatment, peaked at 2 min, and disappeared 5 min afterward. To determine the most effective in vitro dosage of ANG II in the experiment, cells were treated with three different concentrations of ANG II (0.01, 0.1, and 1 M) for 2 min. As shown in Fig. 1B, PI3K activity was most strongly induced with 0.1 M ANG II, and this dose was used in all the following experiments.
The ANG II-induced increase in PI3K activity in MES-13 cells was mediated by the AT-1 subtype of ANG receptor, because co-treatment with losartan (1 M, Merck & Co., Inc, Whitehouse Station, NJ), an AT-1R antagonist, abrogated the effect of ANG II, whereas co-treatment with PD123319 (1 M), an AT-2R antagonist, had no effect (Fig. 2). To assess if the downstream signaling molecules in the PI3K signaling pathway are similarly affected by ANG II, MES-13 cells were treated with ANG II for 1, 2, 5, 15, or 30 min, and phosphorylation of Akt and p70S6K was examined by Western blotting. There was a significant increase in Akt phosphorylation (threonine 308) 1 min after ANG II treatment. The effect was peaked at 2 min then tapered down gradually thereafter (Fig.  3A). The phosphorylation of p70S6K (threonine 389) was also significantly increased with the peak effects seen at 2 min after ANG II treatment (Fig. 3B). These results demonstrate that acute AT-1R stimulation induces activation of PI3K and the downstream signaling pathway in MES-13 cells.
AT-1R Stimulation-induced Type I Collagen Synthesis in MES-13 Cells Is Dependent on PI3K Signaling Pathway-As mentioned above, a hallmark of the glomerular diseases is an increase in the synthesis of extracellular matrix components, including collagens. ANG II has been shown to increase the synthesis of collagen. We next examine what subtypes of collagen are affected by acute ANG II treatment. First, quiescent MES-13 cells were treated with ANG II under serum-free condition for various durations (30, 60, 90, 120, 240, and 360 min), and the gene levels of four subtypes of collagen (Coll-1␣1, Coll-1␣2, Coll-3␣1, and Coll-4␣1) were determined by RT-PCR. The quiescent cells expressed only trace amounts of collagen genes. ANG II induced a significant increase in Coll-1␣1 and Coll-1␣2 gene levels 90 min after initiation of treatment, and the effect was still evident after 6 h (Fig. 4A). In contrast, neither of the Coll-3␣1 and Coll-4␣1 gene levels was affected by ANG II.
To further examine if ANG II treatment can increase collagen protein levels and to determine if this effect is mediated by AT-1R, quiescent MES-13 cells were treated with vehicle, ANG II (0.1 M), or ANG II in combination with an inhibitor for PI3K or AT-1R for 24 h. Western blotting results demonstrated a significant increase in the protein levels of procollagen type I in ANG II-treated cells (Fig. 4B). This effect was completely abrogated by co-treatment with either a PI3K inhibitor (wortmannin, 20 nM) or an AT-1R antagonist (losartan, 1 M) (Fig. 4B). These results demonstrate an AT-1R-mediated, PI3K-dependent increase in collagen synthesis in MES-13 cells.
As shown in Fig. 3B, Akt phosphorylation was significantly increased by ANG II treatment. To further determine the role of Akt in the ANG II-mediated increase in collagen synthesis, MES-13 mesangial cells were stably transfected with an empty vector (pUSEamp(ϩ)) or the vector expressing a dominant negative Akt1. As seen in nontransfected cells (Fig. 4B), ANG II  treatment induced a significant increase in procollagen type I levels in cells stably transfected with pUSEamp(ϩ) (Fig. 4C). This effect of ANG II, however, was abrogated when the same pUSEamp(ϩ)-transfected cells were co-treated with an Akt inhibitor (10 M, Calbiochem). In cells stably transfected with dominant negative Akt1, the effect of ANG II on collagen synthesis was also abrogated (Fig. 4C). Taken together, these data suggest that the PI3K signaling pathway plays an important in the AT-1R-mediated activation of collagen gene expression in MES-13 cells.
ANG II-induced Increase in PI3K Activity Requires the Involvement of Adenylyl Cyclase, Src Family Tyrosine Kinases, and EGFR Transactivation-To investigate the intracellular signaling event responsible for ANG II/AT-1R-mediated activation of PI3K, MES-13 cells were treated with vehicle, ANG II alone, or ANG II in combination with activators or inhibitors of several signaling pathways. The effect of acute ANG II treatment on activation of PI3K activity was abolished by co-treatment of 2Ј,5Ј-dideoxyadenosine (100 M; 2Ј,5Ј-DOA), an adenylyl cyclase inhibitor (Fig. 5A). Co-treatment with H-89 (10 M), a cAMP-dependent kinases (PKA) inhibitor, however, failed to diminish the effect of ANG II on PI3K activity. Treatment with forskolin alone (1 M, 5 min), an adenylyl cyclase activator, also increased PI3K activity in MES-13 cells (Fig. 5A). These results suggest that a PKA-independent, G␣ s -cAMP pathway is involved in the ANG II/AT-1R-mediated increase in PI3K activity in MES-13 cells.
The Ca 2ϩ /calmodulin-dependent protein kinases (CaMKs) play important roles in controlling a variety of cellular functions in response to increases in intracellular Ca 2ϩ and has been shown to be a factor in ANG II-induced signaling cascade (49). To explore the role of CaMK in ANG II-induced activation of PI3K, MES-13 cells were treated with the vehicle, ANG II alone, ANG II in combination with KN93 (5 M, Calbiochem), the specific CaMK inhibitor, or ANG II in combination with KN92 (5 M, Calbiochem), an analog of KN93 that dose not affect CaMK activity. The effects of ANG II/AT-1R stimulation on PI3K activation were not significantly changed by either KN93 or KN92 (Fig. 5B). The result suggests that the cAMP-CaMK pathway is not involved in the ANG II/AT-1R-PI3K crosstalk in the MES-13 cells.
Cells were further treated with vehicle, ANG II, cAMP (1 mM, Upstate  6-benzoyladenosine-cAMP showed no significant effect (Fig.  5C). The result suggests that the cAMP-Epac pathway is involved in the ANG II/AT-1R-mediated increase in PI3K activity in MES-13 cells.
We next examined other signaling pathways that may be involved in ANG II-induced activation of PI3K in MES-13 cells. Co-treatment with GF109203X (10 M), an inhibitor for PKC, or farnesylthiosalicylic acid (20 M, Biomol, Plymouth Meeting, PA), an inhibitor for Ras, did not significantly alter the effect of ANG II on PI3K activity. In contrast, co-treatment with PP2 (10 M, Tocris Bioscience, Ellisville, MO), an inhibitor for the Srcfamily tyrosine kinases, completely abolished the effect of ANG II (Fig. 6A). Moreover, co-treatment with an inhibitor for platelet-derived growth factor receptor (AG1296, 10 M, Calbiochem) or insulin-like growth factor-1 receptor (AG1024, 1 M, Calbiochem) did not alter the effect of ANG II, whereas co-treatment with an inhibitor of EGFR (AG1478, 5 g/ml) fully abrogated the effect of ANG II (Fig. 6B). The crosstalk between AT-1R and EGFR was further established in immunoprecipitation experiments. Quiescent MES-13 cells were treated with ANG II for 1-30 min, and cell lysates were subjected to immunoprecipitation with an antibody against EGFR (Santa Cruz Biotechnology), followed by immunoblotting using an antibody specific for EGFR or phosphotyrosine (Santa Cruz Biotechnology). In control cells, the levels of EGFR phosphorylation were only traceable. Acute ANG II treatment significantly increased tyrosine phosphorylation of EGFR (Fig. 7). This effect peaked at 2 min and disappeared at 15 min after ANG II treatment. These data establish EGFR as an important signaling player in AT-1R-mediated signaling pathway in MES-13 cells.
Metalloproteinases have been shown to play important roles in the conversion of heparin-binding EGF-like growth factor (HB-EGF) to activate EGFR. To determine if HB-EGF is involved in ANG II-induced activation of PI3K, quiescent MES-13 cells were preincubated with a nonspecific matrix metalloproteinase inhibitor, 1,10-phenanthroline (1,10-PTL, 300 M) or heparin (10 or 100 g/ml), which inhibits the HB-EGF activity (29), for 30 min before acute treatment with ANG II. Neither 1,10-PTL nor heparin altered the outcome of ANG II-induced increases in PI3K activity (Fig. 8A) or EGFR phosphorylation (Fig. 8B). The results suggest that HB-EGF is not a factor in the AT-1R/EGFR/PI3K crosstalk in MES-13 cells.

c-Src Is Critical for AT-1R-mediated Induction of Collagen Synthesis and Functions Upstream of EGFR and PI3K-We
have shown that pharmacological inhibition of Src-family tyrosine kinases by PP2 effectively abolished the ANG II-mediated increases in PI3K activity (Fig. 6A) and that EGFR plays an important role in AT-1R-mediated signaling in MES-13 cells (Figs. 6B and 7). To strengthen these data, we next investigated if acute ANG II treatment can alter Src activity. Acute stimulation of ANG II/AT-1 (0.1 M for 2 min) induced a significant increase in Src activity in MES-13 cells, and the effect was totally abolished by pretreatment of the cells with 2Ј,5Ј-DOA (100 M) for 1 h (Fig. 9). These results provide strong evidence for ANG II-induced, cAMP-dependent activation of Src in MES-13 cells.
To further establish the critical role of Src in the AT-1R-PI3K signaling pathway and to determine if EGFR functions downstream of Src in this crosstalk, we applied the following two different strategies. First, the Src gene was knocked down in MES-13 cells by stable transfection of a mouse c-Src siRNA. MES-13 cells transfected with a scrambled oligonucleotide were used as a control. Knockdown of c-Src at the gene and protein levels in the c-Src siRNA-transfected cells was confirmed with RT-PCR and Western blotting, respectively. After 25 PCR reaction cycles, there was visible c-Src transcript of predicted size (522 bp) in the scrambled oligonucleotide-transfected cells but not the c-Src siRNA-transfected cells (Fig. 10A). The gene levels of GAPDH were similar between these two transfected cells. Western blotting using an antibody specific for c-Src (N-16, Santa Cruz Biotechnology) showed a significant down-regulation of c-Src protein level in the c-Src siRNAtransfected cells. The protein levels of ␤-actin were not different between these cells (Fig. 10A). These transfected cells were treated with ANG II for 2-15 min, and the cell lysates were immunoprecipitated with an antibody against EGFR (Santa Cruz Biotechnology), followed by immunoblotting using anti-EGFR or an antibody specific for phosphotyrosine (Santa Cruz Biotechnology). As seen earlier in non-transfected MES-13 cells (Fig. 7), in cells transfected with a scrambled oligonucleotide tyrosine phosphorylation of EGFR was induced 2 min after ANG II treatment (Fig. 10B). In contrast, Src knockdown by siRNA completely abrogated the effect of ANG II. Second, MES-13 cells were stably transfected with an empty vector (pUSEamp(ϩ)) or pUSEamp(ϩ) containing a constitutively  active Src cDNA (Y529F mutant). As expected, cells transfected with the empty vector had a significant increase in PI3K activity after acute treatment with ANG II (Fig. 11). The basal PI3K activity in the constitutively active Src-transfected cells was about 7-and 2-fold that of the non-treated and ANG II-treated empty vector-transfected cells, respectively. This high basal PI3K activity was, however, greatly reduced by treatment with an EGFR inhibitor (AG1478) but not by an adenylyl cyclase inhibitor (2Ј,5Ј-DOA) (Fig. 11). We have shown that Src functions downstream of cAMP (Fig. 9). These results further establish that Src is a critical factor in AT-1R-mediated activation of PI3K by functioning downstream of adenylyl cyclase and upstream of the EGFR.
To examine if Src is also critical in AT-1R-mediated increases in collagen synthesis, the c-Src siRNA or scrambled oligonucleotide-transfected cells were cultured with or without ANG II for 24 h, and type I collagen products were evaluated by Western blotting as described previously. As shown in Fig. 12A, ANG II induced type I collagen production in scrambled oligonucleotide-transfected control cells. In contrast, acute ANG II treatment failed to induce synthesis of collagen in c-Src siRNAtransfected MES-13 cells. Treatment with either ANG II alone or an adenylyl cyclase activator (forskolin) alone significantly increased the protein levels of procollagen type I (Fig. 12B). Co-treatment with PP2 abolished the effect of forskolin. In cells stably transfected with the constitutively active Src mutant, the basal levels of procollagen were 5-fold that of cells transfected FIGURE 8. ANG II-induced increases in PI3K activity and EGFR phosphorylation are HB-EGF-independent. A, MES-13 mesangial cells were treated for 2 min with vehicle, ANG II (0.1 M) alone, or ANG II plus pretreatment for 30 min with 1,10-PTL (300 M) or heparin (10 and 100 g/ml). Cell lysates were subjected to PI3K assay as described. PIP (phosphoinositide 3-phosphate), the phosphorylated end-product. The bar graphs show the densitometric scanning results from three individual experiments. Data represent means Ϯ S.E. of percent change in PI3K activity relative to that of vehicle control. *, p Ͻ 0.05 versus vehicle control. B, MES-13 mesangial cells were treated for 2 min with vehicle, ANG II (0.1 M) alone, or ANG II plus pretreatment for 30 min with 1,10-PTL (300 M) or heparin (100 g/ml). Cells lysates were immunoprecipitated with an antibody specific for EGFR and immunoblotted with an antibody specific for phosphotyrosine. The same membrane was stripped and re-blotted with an anti-EGFR antibody for a loading control. The bar graphs show the densitometric scanning results from three individual experiments. Data are normalized with the levels of EGFR and represent means Ϯ S.E. of percent change in tyrosine phosphorylation relative to that of vehicle control. *, p Ͻ 0.05 versus vehicle control. with the empty vector (Fig. 12B). The enhanced collagen levels disappeared when cells were treated with AG1478 but not with 2Ј,5Ј-DOA. These findings demonstrate Src as a critical signal-ing molecule in AT-1R-PI3K-mediated increase in collagen synthesis in MES-13 cells by acting between cAMP and EGFR.

DISCUSSION
Regardless of etiology, most end-stage glomerular diseases are characterized by accumulation of extracellular matrix proteins, including collagens, in mesangium and other areas in glomeruli. The mechanisms underlying this abnormal accumulation of extracellular matrix have not been fully elucidated. ANG II has been implicated in the development of chronic progressive glomerular diseases, and the precise mechanism of this effect has been a subject to be investigated clinically and experimentally. It has been suggested that ANG II induces glomerular injuries through renal hemodynamic effects and stimulation of growth and extracellular matrix production by glomerular cells (30). Several studies on this subject have been published (4,(31)(32)(33), and several factors such as TGF-␤ and endothelin-1 have been identified as crucial components in the mechanism, but the detailed signaling mechanism responsible for the effects are largely unknown. PI3K, in the mesangial context, has been reported to induce proliferation (34), enhance collagen expression (24), and prevent apoptosis (36,37).
During the last decade, crosstalk between ANGII/AT-1R and PI3K has been demonstrated in vascular smooth muscle cells, cardiomyocytes, brain neurons, and choriocarcinoma cells (38 -42). In the present study, we set out to delineate the sig-  naling events that mediate the effect of ANG II on collagen synthesis in MES-13 cells. The results indicate ANG II, via AT-1R, increases collagen synthesis through transactivation of the PI3K signaling pathway. The transactivation of PI3K from ANG II/AT-1R is G␣ s -cAMP-dependent but PKA-independent and involves Src family tyrosine kinases and EGFR. Experiments using stable transfection and siRNA further establish Src and EGFR as critical factors in this signaling pathway. Based on these data, we propose a novel signaling model for ANG II-induced collagen synthesis in MES-13 cells (Fig. 13). In this model, stimulation of AT-1R triggers sequential activation of adenylyl cyclase, Src, and EGFR and leads to transactivation of the PI3K signaling pathway. Activation of downstream signaling molecules within the PI3K signaling pathway, including Akt and p70S6K, results in an increase in collagen synthesis in MES-13 cells.
In response to binding of GPCR to its ligand, the heterotrimeric G protein complex separates into two subunits, G␣ and G␤␥. G␣ s protein, one of the four subfamilies of G␣ protein, is activated by hormone receptors, odor receptors, and taste receptors. G␣ s -mediated signaling involves adenylyl cyclase in all known cases (43). In the present study, we demonstrated that inhibition of adenylyl cyclase by 2Ј,5Ј-DOA abolished ANG IIinduced increase in PI3K activity, whereas H89, a PKA-specific inhibitor, showed no significant effects. These results suggest that ANG II-induced PI3K activation in MES-13 cell is mediated by cAMP and does not require the involvement of PKA. Starting from the 70s and until a few years ago, it was believed that most cAMP effects were mediated by the activation of PKA. Recently, it has become increasingly apparent that PKA are not the only intracellular receptors involved in cAMP signaling in eukaryote (44 -46). Multiple investigations have identified novel cAMP-dependent but PKA-independent signaling pathways in the last several years (47)(48)(49). In addition, novel cAMP analogs such as 8-pHPT-2Ј-O-Me-cAMP and 8-pCPT-2Ј-O-Me-cAMP, have been experimentally tested and shown to activate downstream effectors such as CaMK and Epac, but not PKA (35,48,50). We used CaMK inhibitor KN93 and Epac- Cell lysates were subjected to Western blotting using an antibody against procollagen type I or ␤-actin. The bar graphs show the densitometric scanning results from three individual experiments. Data are normalized with ␤-actin levels and represent means Ϯ S.E. of percent change in protein levels relative to that of pUSEamp(ϩ) control. *, p Ͻ 0.05 versus pUSEamp(ϩ) control. **, p Ͻ 0.05 versus Y529F control. specific cAMP analog 8-pHPT-2Ј-O-Me-cAMP to evaluate the involvement of these two factors and found that the Epac but not CaMK was playing critical roles in the ANG II/AT-1R transactivation of PI3K. Our study is the first to demonstrate such a PKA-independent effect of cAMP in a mesangial context.
Activation of AT-1R induces tyrosine phosphorylation and stimulates various downstream kinases, leading to cell proliferation (51,52). How the AT-1R, which lacks intrinsic tyrosine kinase activity, induces these events is not fully elucidated. Recent evidence suggests that transactivation of EGFR by GPCRs plays a significant role in this signaling mechanism (53). In accordance with this notion, we demonstrated that AT-1R stimulation induces EGFR phosphorylation and that the activation of EGFR is crucial in PI3K transactivation and collagen synthesis in MES-13 cells. Transactivation of EGFR by GPCRs, including AT-1R has been widely established in a variety of cells (54 -56). The mechanisms by which ANG II/AT-1R transactivate EGFR, however, are not well understood. Recent studies have suggested the important role of metalloproteases in the enzymatic conversion of the HB-EGF precursor to soluble ligand that activates EGFR through autocrine or paracrine mechanisms (57,58). In the present study, we used the nonspecific metalloprotease inhibitor 1,10-phenanthroline and heparin, which inhibits the HG-EGF activity competitively (29), to investigate the involvement of HB-EGF in our model. Neither inhibitor, however, abolished the effect of ANG II on inducing PI3K activity, suggesting HB-EGF is not involved in this signaling pathway in MES-13 cells.
Functional signaling interactions between EGFR and nonreceptor tyrosine kinase Src have been documented (59,60). We extensively studied the role of Src in the AT-1R/EFGR/ PI3K signaling pathway in MES-13 cells with four different approaches. First, pretreatment with an Src inhibitor, PP2, completely abolished the effect of acute ANG II treatment on increasing PI3K activity. Second, we measured Src activity and showed that Src activity was significantly increased by acute ANG II treatment, which was effectively abolished by inhibition of cAMP. Third, knockdown of Src by siRNA completely abolished the effect of ANG II on phosphorylation of EGFR and production of type I collagen. Fourth, cells stably transfected with a constitutively active Src cDNA showed a 7-fold increase in basal PI3K activity, which was effectively abrogated by AG1478, an EGFR antagonist, but not by 2Ј,5Ј-DOA, an adenylyl cyclase inhibitor. In the constitutively active Src-transfected cells, collagen synthesis was over 4-fold higher than that of cells transfected with only the empty vector, which again was abrogated by AG1478 but not by 2Ј,5Ј-DOA. These findings establish Src as a critical factor in the AT-1R/EGFR/PI3K signaling pathway-mediated collagen synthesis in MES-13 cells by functioning downstream of cAMP and upstream of EGFR.
In summary, we have investigated the crosstalk of ANG II/AT-1R and PI3K as a crucial factor in inducing collagen synthesis in mesangial cells. We demonstrated that EGFR transactivation is involved in this novel pathway. We further established that Src played a pivotal role in the signaling pathway by functioning between cAMP and EGFR. Although there are other components needed to be identified in this signaling pathway, our findings nonetheless provide important information for a largely unknown mechanism of ANG II-mediated pathogenesis in glomerular diseases. Elucidating further details in this signaling pathway will benefit future choices of treatment for diseases.