NADPH Oxidase Is Involved in Prostaglandin F2α-induced Hypertrophy of Vascular Smooth Muscle Cells INDUCTION OF NOX1 BY PGF2α

Prostaglandin (PG) F2α, one of the primary prostanoids generated in vascular tissue, is known to cause hypertrophy in vascular smooth muscle cells. To clarify the molecular mechanisms underlying PGF2α-induced hypertrophy, the involvement of reactive oxygen species was examined in a rat vascular smooth muscle cell line, A7r5. PGF2α and (+)-fluprostenol, a selective agonist of the PGF receptor, significantly increased intracellular O 2 − in A7r5. The PGF2α-induced O 2 − increase was suppressed by diphenyleneiodonium (DPI), an inhibitor of NADPH oxidase that has been reported to be the major source of O 2 − in vascular cells. The augmented synthesis of the protein induced by PGF2α or (+)-fluprostenol was suppressed in the presence of DPI. In PGF2α or (+)-fluprostenol-treated cells, a dose-dependent increase in the expression of NOX1, a homolog of the catalytic subunit of the phagocyte NADPH oxidase gp91 phox , was demonstrated by Northern blot analysis. Finally, depletion of NOX1 mRNA in the cells transfected with ribozymes targeted for three independent cleavage sites on the mRNA sequence significantly reduced the PGF2α-induced increase in protein synthesis. Taken together, these results suggest that hypertrophy of vascular smooth muscle cells caused by PGF2α is mediated by NOX1 induction and the resultant overproduction of O 2 − by NADPH oxidase.

Prostaglandin (PG) F 2␣ , one of the primary prostanoids generated in vascular tissue, is known to cause hypertrophy in vascular smooth muscle cells. To clarify the molecular mechanisms underlying PGF 2␣ -induced hypertrophy, the involvement of reactive oxygen species was examined in a rat vascular smooth muscle cell line, A7r5. PGF 2␣ and (؉)-fluprostenol, a selective agonist of the PGF receptor, significantly increased intracellular O 2 ؊ in A7r5. The PGF 2␣ -induced O 2 ؊ increase was suppressed by diphenyleneiodonium (DPI), an inhibitor of NADPH oxidase that has been reported to be the major source of O 2 ؊ in vascular cells. The augmented synthesis of the protein induced by PGF 2␣ or (؉)-fluprostenol was suppressed in the presence of DPI. In PGF 2␣ or (؉)-fluprostenol-treated cells, a dose-dependent increase in the expression of NOX1, a homolog of the catalytic subunit of the phagocyte NADPH oxidase gp91 phox , was demonstrated by Northern blot analysis. Finally, depletion of NOX1 mRNA in the cells transfected with ribozymes targeted for three independent cleavage sites on the mRNA sequence significantly reduced the PGF 2␣ -induced increase in protein synthesis. Taken together, these results suggest that hypertrophy of vascular smooth muscle cells caused by PGF 2␣ is mediated by NOX1 induction and the resultant overproduction of O 2 ؊ by NADPH oxidase.
Prostanoids are metabolites of arachidonic acid that exert a variety of biological actions in tissue. Prostaglandin (PG) 1 F 2␣ is a vasoactive factor that causes constriction and hypertrophy of vascular smooth muscle cells (VSMC) and cardiac myocytes (1)(2)(3). PGF 2␣ exerts its biological actions through binding to its specific receptor, FP, on plasma membranes (4). FP is coupled to phospholipase C and elicits mobilization of cytosolic Ca 2ϩ . Although several protein kinases are reported to be involved in the PGF 2␣ -induced hypertrophy of VSMC or myocytes, the signaling pathway(s) mediating this effect is still unknown (5)(6)(7).
Reactive oxygen species (ROS) including superoxide (O 2 Ϫ ) and hydrogen peroxide (H 2 O 2 ) are recognized as important signaling molecules in cardiovascular tissues. It has been shown that NADPH oxidases are the major source of O 2 Ϫ in vascular cells and myocytes (8 -10), and their activities are increased by vasoactive factors such as angiotensin II and thrombin (11)(12)(13). The phagocyte NADPH oxidase is composed of two plasma membrane-spanning subunits, gp91 phox (NOX2) and p22 phox ; two cytosolic subunits, p47 phox and p67 phox ; and a small G protein, Rac. The catalytic subunit NOX2 contains binding sites for NADPH, FAD, and hemes. To date, six homologs of the NOX2, NOX1, NOX3, NOX4, NOX5, DUOX1, and DUOX2 have been reported (14 -16). Among them NOX1, NOX4, and NOX5 have been detected in VSMC (16 -18). NOX1 is also highly expressed in a colon carcinoma cell line Caco-2 (17,19), and increased cell growth along with a transformed appearance were reported in NOX1-transfected NIH3T3 cells (17,20). Platelet-derived growth factor (PDGF), angiotensin II, phorbol ester, and fetal bovine serum (FBS) were shown to induce the expression of NOX1 mRNA in VSMC (17,18). Accordingly, NOX1/NADPH oxidase-derived O 2 Ϫ may be implicated in the pathogenesis of vascular lesions mediating the proliferation and hypertrophy of VSMC.
Since PGF 2␣ is known to elicit hypertrophy of VSMC we examined whether this effect of PGF 2␣ is mediated by O 2 Ϫ generated by NOX1/NADPH oxidase.

EXPERIMENTAL PROCEDURES
Materials-Ribozyme expression vector pPUR-KE, which contains the human tRNA Val promoter between the EcoRI and BamHI sites of pPUR (CLONTECH), was a gift from Professor K. Taira   Protein Labeling Mix and 100 nM PGF 2␣ or 100 nM (ϩ)-fluprostenol in the presence or absence of 100 nM DPI, 100 M Mn(III)tetrakis(4benzoic acid)porphyrin chloride (MnTBAP), or 500 units/ml of polyethylene glycol-conjugated-catalase. Labeled cells were washed with phosphate-buffered saline, trypsinized, and precipitated with an equal volume of 10% trichloroacetic acid. After freezing and thawing, the mixture was passed through a glass fiber filter (GF/C, Whatman). The filter was washed three times with 5% trichloroacetic acid, and the radioactivity on the filter was measured with a liquid scintillation counter.
Northern Blot Analysis-Total RNA was isolated from A7r5 cells by the acid guanidinium thiocyanate/phenol/chloroform method (21). Ten micrograms of RNA were separated by electrophoresis on a 1.5% agarose gel and transferred onto a nylon membrane (Hybond-N ϩ , Amersham Biosciences).
Rat cDNA fragments for NOX1, NOX4, and p22 phox were amplified by RT-PCR. The fragments were cloned into a vector and linearized with an appropriate restriction enzyme. Using these DNAs as templates, antisense RNA probes were synthesized with SP6 or T7 RNA polymerase in the presence of [␣-32 P]UTP. Hybridization was carried out at 70°C for 15 h in 5ϫ SSC, 50% formamide, 5ϫ Denhardt's solution, 0.2% SDS, 0.05 M sodium phosphate (pH 6.5), 250 g/ml of heat-denatured salmon sperm DNA, 200 g/ml of yeast tRNA, and the 32 P-labeled probe. Membranes were washed twice at 70°C in 0.1ϫ SSC containing 0.1% SDS for 20 min. Hybrids were detected with a Fujix BAS 2000 Bioimaging Analyzer (Fuji photo film, Tokyo, Japan). Blots were then rehybridized with a 32 P-labeled DNA probe for glyceraldehyde-3-phosphate dehydrogenase. Hybridization was carried out at 68°C for 15 h in 6ϫ SSC containing 5ϫ Denhardt's solution and 0.5% SDS. Membranes were washed twice at 68°C for 30 min in 2ϫ SSC containing 1% SDS. The radioactivity on the membrane was quantitated with the bioimaging analyzer, and the levels of NOX1 mRNA were normalized based on the levels of glyceraldehyde-3-phosphate dehydrogenase mRNA.
Synthesis of Anti-NOX1 Ribozymes-Hammerhead ribozymes against rat NOX1 mRNA were designed using the MFOLD program in the following way. Rzm168 is targeted at the GUU triplet located at nucleotides 166 -168. Rzm243 is targeted at the GUA located at nucleotides 241-243, and Rzm603 is targeted at the GUC spanning nucleotides 601-603 of the rat NOX1 mRNA sequence (17). Schematic diagrams of these ribozymes are shown in Fig. 5. The construction of plasmids for ribozyme expression was performed essentially as described previously (22,23).
Establishment of Clones Stably Expressing Anti-NOX1 Ribozymes-Ribozyme expression plasmids (pPUR-KE containing NOX1 ribozyme sequence) were transfected into A7r5 cells using GenePORTER2 transfection reagent (Gene Therapy Systems). Stable transfectants were selected by single cell cloning in the presence of puromycin (10 g/ml). For mock transfection the pPUR-KE vector was transfected and selected with puromycin. The expression of ribozymes was verified as described previously (24). Briefly, total RNA (10 g) was separated by electrophoresis on a 2.4% agarose gel and transferred onto a nylon membrane. Oligonucleotides complementary to the respective ribozyme sequences (see Fig. 5) were labeled with T4 polynucleotide kinase in the presence of [␥-32 P]ATP. Hybridization was carried out at 60°C for 4 h in 6ϫ SSC, 5ϫ Denhardt's solution, 0.5% SDS, 0.01 M sodium phosphate (pH 6.5), 100 g/ml of heat-denatured salmon sperm DNA, and the 32 P-labeled probe. Membranes were washed twice at 60°C in 2ϫ SSC containing 1% SDS for 15 min.
Statistical Analysis-Values are expressed as the mean Ϯ S.E. Statistical analysis was performed with Student's t test. For multiple treatment groups one-way analysis of variance followed by Bonferroni's t test was applied.

Measurements of O 2
Ϫ in A7r5 Cells-To examine whether PGF 2␣ induces production of O 2 Ϫ in A7r5 cells, oxidation of intracellular hydroethidine by O 2 Ϫ was measured by flow cytometry. As shown in Fig. 1A, significant increases in mean fluorescence on oxidation of hydroethidine to fluorescent ethidium were observed in the cells stimulated with PGF 2␣ or (ϩ)-fluprostenol, a selective PGF receptor (FP) agonist. DPI, an inhibitor of NADPH oxidase, significantly suppressed the PGF 2␣ -induced increase in fluorescence (Fig. 1B).
Protein Synthesis in A7r5 Cells-Next, the effects of FP agonists and DPI on protein synthesis in A7r5 cells were in-vestigated (Fig. 2). Growth-arrested cells were stimulated for 24 h with 100 nM PGF 2␣ or 100 nM (ϩ)-fluprostenol in the presence or absence of 100 nM DPI. Increased synthesis of protein as determined from the incorporation of [ 35 S]methionine was observed in the cells stimulated with PGF 2␣ or (ϩ)-fluprostenol. On the other hand, the incorporation of [ 35 S]methionine induced by these FP agonists was significantly suppressed in the presence of DPI. Mn(III)tetrakis(4-benzoic acid)porphyrin chloride (MnTBAP), a cell-permeable superoxide dismutase mimetic, significantly suppressed PGF 2␣ -induced [ 35 S]methionine incorporation at 100 M, whereas polyethylene glycol-conjugated catalase did not affect the incorporation even at the concentration of 500 units/ml (data not shown). These results suggest that the PGF 2␣ -induced hypertrophy of VSMC is mediated by O 2 Ϫ generated by NADPH oxidase.
Expression of NOX1 mRNA in A7r5 Cells-To further clarify the mechanism underlying the FP agonist-induced increase in O 2 Ϫ generation, effects of FP agonists on the expression of NOX1, a member of the catalytic subunit superfamily of NADPH oxidase, was examined. Growth-arrested A7r5 cells were stimulated with 10% FBS, 20 g/ml PDGF-AB, or 1 M PGF 2␣. The expression of NOX1 mRNA was determined by Northern blot analysis. As shown in Fig. 3, a small NOX1 mRNA signal was detected at ϳ2.6 kb in growth-arrested A7r5 cells. When cells were stimulated with 10% FBS or PDGF, the level of NOX1 mRNA was markedly increased as previously reported in VSMC isolated from rat aorta (17,18). In line with these findings, PGF 2␣ significantly increased the level of NOX1 mRNA.
A dose-dependent induction of NOX1 mRNA was demonstrated in the cells stimulated with either PGF 2␣ or (ϩ)-fluprostenol for 24 h (Fig. 4A). The level of NOX1 mRNA began to increase at 10 Ϫ10 M and reached a plateau at 10 Ϫ9 M of PGF 2␣ . When stimulated with (ϩ)-fluprostenol, the level of NOX1 mRNA first detected at 10 Ϫ12 M increased dose dependently and reached a maximum at 10 Ϫ9 M. By contrast, no noteworthy alteration in the expression levels of NOX4 (another homolog of NOX2) and p22 phox was detected in the cells stimulated with PGF 2␣ . As shown in Fig. 4B, an increase in NOX1 expression was clearly observed 3 h after stimulation with 10 Ϫ7 M PGF 2␣ . The level of NOX1 mRNA reached a maximum at 12 h and remained high until 48 h after the stimulation.
Effects of Other Prostanoids on NOX1 Expression-We also examined the effects of other prostanoids on NOX1 mRNA expression. PGE 2 , U-46619 (a thromboxane A 2 receptor agonist), and carbaprostacyclin (a PGI 2 receptor agonist) induced expression of NOX1 mRNA at concentrations higher than 10 Ϫ7 M. However, thromboxane A 2 receptor (TP) antagonists SQ 29,548 and I-SAP did not affect induction of NOX1 expression by TP agonists U-46619 and I-BOP (data not shown).
Effects of Ribozymes Targeted at NOX1 mRNA-To verify the involvement of NOX1 in PGF 2␣ -induced hypertrophy in VSMC, a ribozyme that cleaves the targeted sequence of mRNA was designed to control the expression of NOX1 in A7r5 cells. Hammerhead ribozymes targeting three independent sites of the NOX1 mRNA sequence are illustrated in Fig. 5. Ribozyme expression plasmids were constructed and transfected into A7r5 cells that gave cell clones stably expressing Rzm168, 243, and 603 (Fig. 6A). Almost complete suppression in the expression of NOX1 mRNA was demonstrated in these clones. When stimulated with PGF 2␣ the increase in NOX1 mRNA level was effectively suppressed in these clones compared with the mocktransfected cells (Fig. 6B). Furthermore, a PGF 2␣ -induced increase in O 2 Ϫ production as well as the basal level of cellular O 2 Ϫ was significantly reduced in these clones compared with the mock-transfected cells (Fig. 6C). As demonstrated in Fig. 7, PGF 2␣ elicited more than a 2-fold increase in [ 35 S]methionine incorporation in mock-transfected cells. On the other hand, the extent of the PGF 2␣ -induced increase in [ 35 S]methionine incorporation was significantly reduced in the ribozyme-expressing cells. These results denoted the involvement of NOX1/NADPH oxidase in PGF 2␣ -induced hypertrophy of VSMC.

The present findings indicate that PGF 2␣ -induced hypertrophy of VSMC is mediated by O 2
Ϫ generated by NADPH oxidase through induction of NOX1 expression. The following four major lines of evidence are provided in this study. 1) PGF 2␣ evoked generation of intracellular O 2 Ϫ in A7r5 vascular smooth muscle cells.
2) The PGF 2␣ -induced increase in protein synthesis was attenuated by DPI, an inhibitor of NADPH oxidase. 3) A dosedependent induction of NOX1, a catalytic subunit of NADPH oxidase, was produced by FP agonists PGF 2␣ and (ϩ)-fluprostenol. 4) The extent of the PGF 2␣ -induced increase in protein synthesis was significantly suppressed when NOX1 mRNA was depleted by transfection of ribozymes specifically targeted at the NOX1 mRNA sequence.
Augmented production of O 2 Ϫ caused by overexpression of NOX1 constituting NADPH oxidase appears to be important to the action of PGF 2␣ leading to hypertrophy of VSMC. PGF 2␣ is synthesized from arachidonic acid, which is converted to PGH 2 through oxygenation and reduction by prostaglandin H syn- thase (cyclooxygenase). Subsequent conversion of PGH 2 (or via PGE 2 ) by PGF synthase generates PGF 2␣ , one of the primary prostanoids produced by vascular cells in response to various stimuli. Increased expression of COX-2, an inducible isoform of cyclooxygenase, is reported in VSMC after balloon angioplasty or pinching injury of the carotid artery (25) as well as in endothelial cells exposed to tumor necrosis factor-␣ (TNF-␣) (26). It has also been demonstrated that PGF 2␣ is released from endothelial cells exposed to hypoxia (27). Thus PGF 2␣ is synthesized and released from vascular lesions under inflammatory conditions following various injuries including ischemiareperfusion. Our findings indicate that the hypertrophic effects of PGF 2␣ on VSMC are mediated by O 2 Ϫ generated by cellular NOX1/NADPH oxidase activity.
Among the NOX superfamily, NOX1, NOX4, and NOX5 have been detected in VSMC (16 -18). In the present study, no alteration in the expression levels of NOX4 and p22 phox was detected in A7r5 cells stimulated with PGF 2␣ . As stimulation of FP leads to a cytosolic mobilization of Ca 2ϩ , NOX5, which produces O 2 Ϫ in a Ca 2ϩ -dependent manner, could be activated indirectly by PGF 2␣ . NOX5 has not been identified in rodents however, and we could not detect the expression of NOX5 homolog in A7r5 cells by RT-PCR using synthetic primers designed from the reported human sequence (16). In addition, stimulation with a calcium ionophore A23187 did not affect the level of O 2 Ϫ generated in the cells (data not shown). Accordingly, the contribution of other members of the NOX family to the Ϫ demonstrated in A7r5 cells seems to be minimal.
The induction of NOX1 expression by PGF 2␣ seems to be mediated specifically by FP, not by other prostanoid receptors. There are several possible explanations for this. First, (ϩ)fluprostenol, which increased NOX1 mRNA in a dose-dependent fashion, is a highly selective FP agonist. It was reported that this compound does not react with other prostanoid receptors (28). Second, expression of FP was demonstrated in A7r5 cells (2), and we also verified the expression by RT-PCR (data not shown). Third, the concentration of PGF 2␣ that elicited NOX1 expression was 10 Ϫ10 M, much lower than for other vasoactive prostanoids investigated such as PGE 2 , U-46619 (a thromboxane A 2 receptor agonist), and carbaprostacyclin (a PGI 2 receptor agonist). They induced NOX1 expression at concentrations greater than 10 Ϫ7 M suggesting that these prostanoids exerted their action by cross-reacting with FP. Thromboxane (TX) A 2 is also known as a prostanoid causing proliferation and hypertrophy of VSMC (1,29). Yet binding or functional responsiveness to TXA 2 receptor (TP) agonist I-BOP was not observed in A7r5 cells (30). When examined by RT-PCR the level of TP mRNA in our A7r5 cells was also found to be very low (data not shown). Furthermore, TP antagonists SQ 29,548 and I-SAP did not affect the induction of NOX1 expression by TP agonists U-46619 and I-BOP. Thus, as far as A7r5 cells are concerned, induction of NOX1 by prostanoids seems to be mediated specifically by FP. We cannot exclude the possibility, however, that the mitogenic effects of TXA 2 on VSMC are mediated by TP expressed in the cells, and its activation leads to NOX1 induction and O 2 Ϫ production. It has been reported that several protein kinases including tyrosine kinase, c-Jun-N-terminal kinase, and extracellular signal-regulated kinase (ERK) are involved in the PGF 2␣ -induced hypertrophy of VSMC or cardiac myocytes (5-7). On the other hand, activation of p38 MAPK and Akt, not of ERK, by PDGF or angiotensin II is reported to be mediated by NOX1 (18). Although the exact signal transduction pathway(s) leading to the induction of NOX1 expression on stimulation of FP remains to be elucidated, the cell clones expressing anti-NOX1 ribozyme established in this study will be useful to clarify the pathophysiological significance of NOX1-derived O 2 Ϫ in vascular cells.
In summary, we demonstrated that PGF 2␣ -induced hypertrophy of VSMC is mediated by NOX1 induction and resultant O 2 Ϫ production. The findings indicate the possibility of using FP antagonists as novel therapeutic agents for preventing vascular hypertrophy following inflammatory responses associated with diverse vascular injury.