Platelet-derived Growth Factor-induced H2O2 Production Requires the Activation of Phosphatidylinositol 3-Kinase*

From the ‡Center for Cell Signaling Research, Division of Molecular Life Sciences, and Department of Biological Sciences, Ewha Womans University, Seoul 120-750, Korea, the ¶Schepens Eye Research Institute, Harvard Medical School, Boston, Massachusetts 02114, the iInternational Joint Research Laboratory of Center for Cell Signaling Research and Laboratory of Cell Signaling, NHLBI, National Institutes of Health, Bethesda, Maryland 20892

The mechanism of receptor-mediated generation of H 2 O 2 has been studied extensively in phagocytic cells, in which O 2 . is produced via the one-electron reduction of O 2 by a multicomponent NADPH oxidase system (28). The O 2 . is then spontaneously or enzymatically dismutated to H 2 O 2 . The NADPH oxidase complex includes two cytosolic components (p47 phox and p67 phox ) and two transmembrane flavocytochrome b components (gp91 phox and p22 phox ). In addition, the small GTP-binding protein Rac (either Rac1 or Rac2) is required for activation of NADPH oxidase. In contrast, the mechanism of H 2 O 2 generation in nonphagocytic cells remains unclear. Evidence suggests that the system responsible for H 2 O 2 (O 2 . ) production in nonphagocytic cells is structurally and genetically distinct from, but functionally similar to, the NADPH oxidase system of phagocytes (29,30). Consistent with this notion, overproduction of Rac1 in fibroblasts was associated with increased production of H 2 O 2 (31, 32) and a homolog of gp91-phox was found in several nonphagocytic cells (33). The binding of growth factors to their receptors results in receptor autophosphorylation on specific tyrosine residues. These phosphotyrosine residues initiate cellular signaling by acting as high affinity binding sites for the Src homology 2 domains of various effector proteins. In the PDGF ␤ receptor * This work was supported in part by Center of Excellence Grant 1998G0202 and International Joint Research Laboratory Grant 1999L0001 from the Korean Science and Engineering Foundation. 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  (PDGF␤R), seven autophosphorylation sites have been identified as specific binding sites for Src family tyrosine kinases (Tyr 579 and Tyr 581 ), phosphatidylinositol 3-kinase (PI3K) (Tyr 740 and Tyr 751 ), the GTPase-activating protein of Ras (GAP) (Tyr 771 ), Src homology 2 domain-containing protein-tyrosine phosphatase-2 (SHP-2) (Tyr 1009 ), and phospholipase C-␥1 (PLC-␥1) (Tyr 1021 ). A series of PDGF␤R mutants was previously constructed that includes a kinase-deficient [kinase(Ϫ)] mutant and receptors in which the binding sites for PI3K, GAP, SHP-2, and PLC-␥1 were changed individually or in various combinations to phenylalanine (34).
To characterize the mechanism of H 2 O 2 production in nonphagocytic cells, we have now measured PDGF-dependent H 2 O 2 generation in human hepatoma HepG2 cells expressing thesevariousPDGF␤Rmutants.OurdataindicatethatPDGF␤Rdependent H 2 O 2 production requires both the intrinsic kinase activity of the receptor as well as the activation of PI3K. Experiments with a dominant negative mutant of Rac1 (N17Rac1) also suggest that Rac1 participates in the PDGF␤R-induced production of H 2 O 2 in HepG2 cells.
HepG2 Cell Lines Expressing Wild-type or Mutant PDGF␤Rs-Construction of the various PDGF␤R mutants shown in Fig. 1 and of the HepG2 cells expressing these mutants has been described previously (34). The cells were maintained at 37°C under an atmosphere of 5% CO 2 in culture dishes containing Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum.
Assay of Intracellular H 2 O 2 -Intracellular production of H 2 O 2 was assayed as described previously (8). In brief, at various times after stimulation with PDGF-BB (100 ng/ml), dishes of confluent cells were washed with MEM lacking phenol red and then incubated in the dark for 5 min in Krebs-Ringer solution containing 5 M 2Ј,7Ј-dichlorofluorescin diacetate. This compound is converted by intracellular esterases to 2Ј,7Ј-dichlorofluorescin, which is then oxidized by H 2 O 2 to the highly fluorescent 2Ј,7Ј-dichlorofluorescein (DCF). Culture dishes were transferred to a Zeiss Axiovert 135 inverted microscope that was equipped with a x20 Neofluor objective and Zeiss LSM 410 confocal attachment, and DCF fluorescence was measured with an excitation wavelength of 488 nm and emission at 515-540 nm. To avoid photooxidation of 2Ј,7Јdichlorofluorescin, the fluorescence image was collected by a single rapid scan (4-line average, total scan time of 4.33 s), with identical parameters (such as contrast and brightness) for all samples. After collection of the fluorescence image, the cells were imaged by digital interference contrast. Five groups of 10 -20 cells each were randomly selected from each sample, and the mean relative fluorescence intensity for each group of cells was then measured by Karl Zeiss vision system (KS400, version 3.0) and averaged. All experiments were repeated at least five times.
Expression of N17Rac1 in HepG2 Cells-HepG2 cells were plated at a density of 2 ϫ 10 5 /well in culture dishes and allowed to recover for 24 h. Adenovrus infection was performed at an multiplicity of infection of 200. After adenovirus infection for 72 h, the cells were then washed with MEM without phenol red, exposed for 30 min to PDGF-BB (100 ng/ml), and subjected to assay of H 2 O 2 production. Western blot analysis of Myc-tagged form of N17Rac was performed using an antibody to rac1 (35). (34) previously constructed the series of PDGF␤R mutants shown in Fig. 1 (Fig. 1). In addition, the ability of F5 to signal via PI3K was selectively restored by changing Phe 740 and Phe 751 in this mutant back to Tyr, thereby generating Y740/751 (Fig. 1). Similarly, the ability of F5 to signal via PLC-␥1 was restored by changing Phe 1021 in this mutant back to Tyr, generating Y1021 (Fig. 1). The wild-type and mutant PDGF␤R proteins were expressed, with the use of a retroviral expression vector, in HepG2 cells, which do not express detectable amounts of endogenous PDGF␤ receptors. The level of PDGF␤R expression was approximately 5 ϫ 10 5 receptors/cell for both wild-type and mutant receptors (34).

Vallius and Kazlauskas
The production of H 2 O 2 by HepG2 cells was measured with a fluorescence-based assay with 2Ј,7Ј-dichlorofluorescin diacetate and laser-scanning confocal microscopy. Stimulation of HepG2 cells expressing wild-type PDGF␤R with PDGF resulted in a time-dependent increase in the intensity of DCF fluorescence, with the maximal, 3.5-fold increase apparent 10 -30 min after stimulation; fluorescence had returned to the baseline value after 60 min ( Fig. 2A). In contrast, cells expressing the kinase(Ϫ) or F5 mutant receptors failed to produce H 2 O 2 in response to PDGF (Fig. 2, B and C). These results suggest that PDGF-induced H 2 O 2 production requires the kinase activity of PDGF␤R as well as at least one of the five tyrosine phosphorylation sites mutated in F5.
We next measured PDGF-induced H 2 O 2 production in cells expressing other PDGF␤R mutants (Fig. 3). Whereas PDGF had no effect on H 2 O 2 production in cells expressing F740/751 or F740/751/771 mutants, it increased H 2 O 2 production in cells expressing the F771 or F1021 mutants to an extent similar to that observed in cells expressing the wild-type receptor. Cells expressing Y740/751 (equivalent to F5 in which the PI3K binding site has been restored) showed a response to PDGF that was slightly greater than that of cells expressing the wild-type receptor. However, cells expressing Y1021 (equivalent to F5 in which the PLC-␥1 binding site has been restored) did not generate H 2 O 2 in response to PDGF. These results thus indicate that activation of PI3K, but not of GAP, SHP-2, or PLC-␥1, is necessary for H 2 O 2 production in response to PDGF.
The essential role of PI3K was further supported by the observation that prior incubation with 10 M LY294002, a specific inhibitor of PI3K, completely blocked the PDGF-induced increase in H 2 O 2 production in HepG2 cells expressing wild-type or Y740/751 PDGF receptors (Fig. 4). Similar inhibition was also observed with wortmannin, another inhibitor of PI3K (data not shown). Prior incubation with the inhibitors had no effect on the basal level of H 2 O 2 production. PDGF also induced a transient increase in H 2 O 2 production, with the maximal, 6-fold increase apparent 5-10 min after stimulation, in NIH 3T3 cells, and this effect was also abolished by 10 M LY294002 (data not shown).
We further examined whether activation of PI3K is sufficient to produce H 2 O 2 by transiently expressing in HepG2 cells the catalytic p110 subunit of PI3K with c-Myc epitope and farnesylation signal (CAAX) sequences at the NH 2 and COOH termini, respectively. It was previously shown that even modest expression of the membrane-targeted p110 (p110-CAAX) is sufficient to trigger activation of downstream events (36,37). Expression of p110-CAAX in HepG2 cells expressing F5 mutant receptor resulted in a 3-fold increase in H 2 O 2 production, and the effect was blocked by prior incubation of cells with LY294002 or wortmannin (Fig. 5).
To investigate the role of Rac1 in PI3K-dependent H 2 O 2 production, we infected HepG2 cells expressing the Y740/751 receptor either with recombinant adenovirus encoding N17Rac1, a dominant negative mutant of Rac1 (Ad.N17rac), or control adenovirus (35). Expression of N17Rac1 in the infected HepG2 cells was confirmed by immunoblot analysis (Fig. 6A). Whereas HepG2 cells expressing the Y740/751 receptor infected control adenovirus showed an increase in H 2 O 2 production in response to PDGF, those infected with recombinant adenovirus encoding N17Rac1 (Ad.N17rac) did not (Fig. 6B). Similar results were observed after infection of cells expressing wild-type PDGF␤R with the recombinant adenovirus encoding N17Rac1 (data not shown). In contrast, expression of dominant negative mutants of other Rho-family members (N19RhoA or N17Cdc42) did not affect the PDGF-induced increase in H 2 O 2 production in HepG2 cells expressing wild-type PDGF␤R (data not shown).

DISCUSSION
The binding of peptide growth factors to their specific receptors results in receptor autophosphorylation at several tyrosine residues, thereby triggering the recruitment of Src homology 2 domain-containing effector enzymes that include Src, PI3K, GAP, SHP-2, and PLC-␥1 (34,38). The interaction of growth factors with their receptors also induces a transient increase in the intracellular concentration of H 2 O 2 (5)(6)(7)(8)(9). In the present study, the role of the receptor-associated effectors in H 2 O 2 production was investigated with HepG2 cells expressing various PDGF␤R mutants. The PDGF␤R was well suited for this study because the effector enzymes bind only when the receptor is phosphorylated at specific tyrosine residues (34). Mutation of these tyrosine residues individually prevents association of a specific effector without affecting the binding of other enzymes to the receptor mutant. In the case of other growth factor receptors, either the effector enzymes bind less specifically at multiple tyrosine residues or their binding sites are not well defined.
Our results suggest that, among the effector enzymes PI3K, GAP, SHP-2, and PLC-␥1, only the binding of PI3K to PDGF␤R is necessary for PDGF-induced H 2 O 2 production. The absence of the binding sites for GAP, SHP-2, and PLC-␥1 in the Y740/ 751 receptor mutant was actually associated with a slight, but reproducible, increase in the extent of PDGF-induced H 2 O 2 production relative to that apparent with the wild-type receptor. This observation is consistent with the results of a previous study showing that PLC-␥1 and GAP negatively regulate the PDGF-induced activation of PI3K (38). Given that phosphorylation of Tyr 740 and Tyr 751 is required to generate the binding site for PI3K, it follows that the intrinsic kinase activity of PDGF␤R is also required for PDGF-induced generation of H 2 O 2 . We previously also showed that an epidermal growth factor receptor mutant lacking kinase activity did not induce an increase in H 2 O 2 production in response to epidermal growth factor (8). In contrast, the observation that a membrane-bound NADPH oxidase prepared from human fat cells produced H 2 O 2 in the absence of ATP in response to insulin suggested that this growth factor is able to induce H 2 O 2 production independently of the kinase activity of its receptor (39).
Activation of PI3K appears to be sufficient to produce H 2 O 2 as indicated by the observation that expression of p110-CAAX induced H 2 O 2 production. One of the major events following PI3K activation is the phosphorylation and activation of Akt, a serine/threonine kinase (40). Therefore, we determined whether Akt is involved in PDGF-induced H 2 O 2 production. Transient expression of either a dominant negative Akt mutant in HepG2 cells expressing Y740/751 receptor or a constitutively active Akt in HepG2 cells expressing F5 receptor had no effect on H 2 O 2 production (data not shown). These results suggest that Akt is not involved in the PDGF-induced H 2 O 2 production.
It was previously shown that the blockage of the growth factor-induced H 2 O 2 increase by catalase in vascular smooth muscle and A431 cells resulted in a marked decrease in tyrosine phosphorylation of various proteins including the growth factor receptors (7,8). At the present time, we do not know why PDGFRs (F740/751, F740/751/771, and F5) that fail to induce H 2 O 2 production were fully autophosphorylated in HepG2 cells (34,38). It is possible that in HepG2 cells, H 2 O 2 production makes a smaller contribution to receptor tyrosine phosphorylation as compared with vascular smooth muscle and A431 cells.
An essential role for Rac in activation of the NADPH oxidase complex has been well established in phagocytes (28). Rac also appears important in this regard in nonphagocytic cells, as indicated by the observations that transient expression of an active form of Rac1 (V12Rac1) in NIH 3T3 cells itself resulted in an increase in H 2 O 2 production and that expression of N17Rac1 inhibited the increase in H 2 O 2 concentration induced by PDGF, epidermal growth factor, tumor necrosis factor-␣, or interleukin-1 (31). Our observation that N17Rac1 blocked the Y740/751 receptor-induced generation of H 2 O 2 indicates that Rac1 acts downstream of PI3K in the signaling pathway that leads to activation of NADPH oxidase. Moreover, signaling by this pathway appears independent of activation of GAP, SHP-2, and PLC-␥1.
Additional evidence suggests that Rac functions downstream of PI3K (41)(42)(43)(44). Thus, the exchange of Rac-bound GDP for GTP catalyzed by guanine nucleotide exchange factors (GEFs) is stimulated by phosphatidylinositol 3,4,5-trisphosphate (PI(3,4,5)P 3 ), a product of the action of PI3K. A family of GEF proteins that mediate the activation of Rac-related proteins has been identified. All members of this family, including Vav, Sos, and Pix, contain a pleckstrin homology domain that binds inositol-containing phospholipids such as PI(4,5)P 2 and PI(3,4,5)P 3 (44 -47). PI(4,5)P 2 , when bound to the pleckstrin homology domain of Vav, inhibited activation of Vav GEF activity by the protein-tyrosine kinase Lck, whereas PI(3,4,5)P 3 enhanced phosphorylation and activation of Vav by Lck (42). Thus, the activation of PI3K might serve to convert a Vav inhibitor to an activator, resulting in a rapid transformation of inactive, GDP-bound Rac to its active, GTP-bound form. Sos possesses two distinct domains that allow it to function as a GEF for both Ras and Rac; the Rac GEF activity requires the binding of activated Ras to the Ras GEF domain of Sos as well as the binding of PI(3,4,5)P 3 to the pleckstrin homology domain (43).
Although not yet demonstrated, the GEF activity of Pix is also likely dependent on a product of PI3K.
The mechanism by which GTP-bound Rac increases the production of superoxide (and thus H 2 O 2 ) in nonphagocytic cells is not clear. In phagocytes, activated Rac1 associates directly with p67-phox to stimulate NADPH oxidase (48,49). Although the NADPH oxidase system in nonphagocytic cells is not well characterized, a similar interaction might occur.
In summary, we have shown that the activation of PI3K is required for PDGF-induced H 2 O 2 production. The essential role of PI3K is likely to provide PI(3,4,5)P 3 for the activation of Rac, a component of the H 2 O 2 -generating system in both phagocytes and nonphagocytic cells.