Characterization of a Rac1 Signaling Pathway to Cyclin D1 Expression in Airway Smooth Muscle Cells*

We examined the importance of the Rho family GTPase Rac1 for cyclin D1 promoter transcriptional activation in bovine tracheal myocytes. Overexpression of active Rac1 induced transcription from the cyclin D1 promoter, whereas platelet-derived growth factor (PDGF)-induced transcription was inhibited by a dominant-negative allele of Rac1, suggesting that Rac1 functions as an upstream activator of cyclin D1 in this system. Rac1 forms part of the NADPH oxidase complex that generates reactive oxygen species such as H2O2. PDGF stimulated a substantial increase in intracellular reactive oxygen species, as measured by the fluorescence of dichlorofluorescein-loaded cells, and this was blocked by the glutathione peroxidase mimetic ebselen. Pretreatment with ebselen, catalase, and the flavoprotein inhibitor diphenylene iodonium each attenuated PDGF- and Rac1-mediated cyclin D1 promoter activation, while having no effect on the induction of cyclin D1 by mitogen-activated protein kinase/extracellular signal-regulated kinase (ERK) kinase-1 (MEK1), the upstream activator of ERKs. Antioxidant treatment also inhibited PDGF-induced cyclin D1 protein expression and DNA synthesis. Overexpression of an N-terminal fragment of p67 phox , a component of NADPH oxidase which interacts with Rac1, attenuated PDGF-induced cyclin D1 promoter activity, whereas overexpression of the wild-type p67 did not. Finally, Rac1 was neither required nor sufficient for ERK activation. Taken together, these data suggest a model by which two distinct signaling pathways, the ERK and Rac1 pathways, positively regulate cyclin D1 and smooth muscle growth.

Excess airway smooth muscle cell proliferation is thought to contribute to airflow obstruction in patients with asthma (1). The signaling mechanisms underlying airway smooth muscle proliferation are not completely understood. We have investigated the role of extracellular signal regulated kinases (ERKs), 1 cytosolic serine/threonine kinases of the mitogen-activated protein kinase superfamily, in bovine tracheal myocyte DNA synthesis. ERK activation is required for platelet-derived growth factor (PDGF)-induced DNA synthesis (2), and also regulates the transcriptional activation of cyclin D 1 (3), a critical regulator of G 1 progression in these cells (4).
Studies in immortalized cell lines (5,6) and primary hepatocytes (7) have demonstrated a requirement for Rho family GTPases in G 1 progression. The mechanisms underlying this requirement are not precisely known. Rac constitutes part of the phagocyte NADPH oxidase complex that generates reactive oxygen species such as H 2 O 2 (8,9). This enzyme, by donating an electron, catalyzes the reaction: 2 O 2 ϩ NADPH 3 2 O 2 . ϩ NADP ϩ H ϩ . The superoxide produced is subsequently converted to H 2 O 2 . The human NADPH oxidase consists of at least seven components: two membrane spanning polypeptides, p22 phox and gp91 phox (which comprise cytochrome b 558 ), three cytoplasmic polypeptides, p47 phox , p67 phox , and p40 phox , Rap1A and Rac, the last of which is required for oxidase activation. Increasing evidence suggests that reactive oxygen species may play a role in mitogen-activated cell signaling. Recent data suggest that the generation of H 2 O 2 by NADPH oxidase occurs in tissues other than phagocytes, including aortic adventitia, kidney, liver, vascular smooth muscle cells, and fibroblasts (10 -13). Production of reactive oxygen species has been noted upon growth factor stimulation in arterial smooth muscle cells (14) and chondrocytes (15). Antioxidants have been shown to inhibit growth factor-induced ERK activation and DNA synthesis in arterial smooth muscle cells (14), growth factor-induced c-fos expression in chondrocytes (15), and phorbol ester-induced cyclin D 1 expression and DNA synthesis in a subclone of NIH3T3 cells (16). Activation of Rac1 has been noted to increase intracellular reactive oxygen species in HeLa cells (17), NIH3T3 cells (18), and rabbit synovial fibroblasts (19). Finally, it has recently been demonstrated that a Rac1 effector site critical for the activation of NADPH oxidase is also required for the mitogenic effect of Rac1 in rat embryonic fibroblasts (20).
In NIH3T3 cells, Rac1 activates transcription from the cyclin D 1 promoter (21,22), suggesting a mechanism by which Rac1 signaling may regulate G 1 progression. However, the requirement of reactive oxygen species for transcriptional activation of cyclin D 1 has not been tested.
In the present study, we examined the importance of the Rho family GTPase Rac1 for airway smooth muscle cyclin D 1 expression. Overexpression of active Rac1 induced transcription from the cyclin D 1 promoter. PDGF-induced transcription was also inhibited by a dominant-negative allele of Rac1, suggesting that Rac1 is a critical upstream activator of the cyclin D 1 promoter in these cells. PDGF stimulated a substantial increase in the fluorescence of 2Ј,7Ј-dichlorofluorescein diacetate (DCFH-DA)-loaded cells which was blocked by the glutathione peroxidase mimetic ebselen, implying that growth factor treatment increases the concentration of intracellular reactive oxygen species. Furthermore, antioxidants (ebselen, catalase) and inhibitors of NADPH oxidase (the flavoprotein inhibitor diphenylene iodonium and the N-terminal fragment of p67 phox ) attenuated PDGF-and Rac1-mediated cyclin D 1 promoter activation, but had no effect on the response induced by MEK1, the upstream activator of ERKs. Finally, Rac1 was neither required nor sufficient for ERK activation. Taken together, these data suggest that Rac1-induced cyclin D 1 expression is mediated by the generation of intracellular reactive oxygen species. Furthermore, these data suggest a model by which two distinct signaling pathways, the ERK and Rac1 pathways, positively regulate smooth muscle growth.
Cell Culture-Bovine tracheal smooth muscle cells were isolated as described previously (30). Myocytes of passage number 5 or less were studied. Confluent cultures exhibited the typical "hill and valley" appearance and showed specific immunostaining for ␣-smooth muscle actin. Cells were cultured in Dulbecco's minimum essential medium (DMEM) containing 10% fetal bovine serum (FBS), 1% non-essential amino acids, penicillin (100 units/ml), and streptomycin (100 g/ml).
Determination of Cyclin D 1 Promoter Transcriptional Activity-Cells were seeded into 60-mm dishes at 50 -80% confluence and incubated in 10% FBS/DMEM overnight. After rinsing, cells were incubated with a liposome solution consisting of serum-and antibiotic-free medium, plasmid DNA (total of 1.8 g/plate) and LipofectAMINE (Life Technologies, Gaithersburg, MD; 12 l/plate). Cells were transiently co-transfected with plasmids encoding the human cyclin D 1 promoter subcloned into a luciferase reporter and either the relevant expression vector or control vector. After 5 h, the liposome solution was replaced with 10% FBS/ DMEM. The next day, cells were serum-starved in DMEM. After 24 h of serum starvation, selected cultures were treated with PDGF (30 ng/ml).
Finally, 8 h after PDGF treatment, cells were harvested for analysis of luciferase activity using lysis buffer provided with the Promega Luciferase Assay system (Madison, WI). Luciferase activity was measured at room temperature using a luminometer (Turner Designs, Sunnyvale, CA). Luciferase content was assessed by measuring the light emitted during the initial 30 s of the reaction and the values expressed in arbitrary light units. The background activity from cell extracts was typically less than 0.02 units, compared with signals on the order of 10 2 -10 3 units.
The transfection of primary cells holds certain limitations that should be noted here. First, transfection efficiency, as assessed by ␤-galactosidase staining, was less than 50%. Second, co-transfection with viral promoter-driven expression vectors tended to suppress cyclin D 1 promoter activity. Concentration-response curves were therefore generated for each expression vector to determine optimal concentration. In all cases, concentrations of 30 -100 ng/plate were used. Also, due to unacceptably low cyclin D 1 promoter activity levels, we were unable to co-transfect cells with more than two expression vectors (␤-galactosidase plus one signaling intermediate of interest).
Preparation of Cell Extracts for Immunoblotting-Cells were cultured in 6-well plates and serum starved for 24 h prior to PDGF treatment (30 ng/ml for 16 h). Cells were washed in phosphate-buffered saline (150 mM NaCl, 0.1 M phosphate, pH 7.5) and extracted in a lysis buffer containing 50 mM Tris, pH 7.5, 40 mM ␤-glycerophosphate, 100 mM NaCl, 2 mM EDTA, 50 mM sodium fluoride, 200 M Na 3 VO 4 , 200 M phenylmethylsulfonyl fluoride, and 1% Triton X-100. Lysates were centrifuged (13,000 rpm for 10 min at 4°C) and the supernatant transferred to fresh microcentrifuge tubes.
Western Analysis of Cyclin D 1 Protein Levels-Extracts (10 g) were resolved on a 10% SDS-polyacrylamide gel and transferred to nitrocellulose by semidry transfer (Hoefer, San Francisco, CA). After incubation with a polyclonal antibody against cyclin D 1 (Santa Cruz Biotechnology, Santa Cruz, CA), signals were amplified and visualized using anti-rabbit IgG and enhanced chemiluminescence.
Measurement of ERK Activation-Cells were transiently co-transfected with cDNAs encoding HA-tagged ERK2 and the expression vector of interest. Cells were seeded into 100-mm plates at a density of 5 ϫ 10 5 cells/plate and incubated in 10% FBS/DMEM overnight. After rinsing, cells were incubated in a solution consisting of serum-and antibiotic-free medium, plasmid DNA (10 g/plate) and LipofectAMINE (40 l/plate). After 5 h, the solution was replaced with 10% FBS/DMEM. Forty-eight h after transfection, cells were serum-starved in DMEM. The next day, selected cultures were treated with PDGF (30 ng/ml for 10 min). Activation of ERK was then assessed by immunoprecipitation of the epitope tag, followed by an in vitro phosphorylation assay using MBP as a substrate, as described (27). Treated cells were washed twice with phosphate-buffered saline and incubated in a lysis buffer consisting of 50 mM Tris-HCl, pH 7.5, 1% Triton X-100, 40 mM ␤-glycerophosphate, 100 mM NaCl, 50 mM sodium fluoride, 2 mM EDTA, 200 M Na 3 VO 4 , and 0.2 mM phenylmethylsulfonyl fluoride (30 min at 4°C). Insoluble materials were removed by centrifugation (13,000 rpm for 10 min at 4°C). Cell lysates were then incubated for 3 h with 30 l of protein A-Sepharose beads precoupled with the 12CA5 anti-hemagglutinin antibody. Immunoprecipitates were washed three times with lysis buffer and twice with kinase buffer containing 20 mM Hepes, pH 7.4, 10 mM MgCl 2 , 1 mM dithiothreitol, 200 M Na 3 VO 4 , and 10 mM p-nitrophenyl phosphate. Immune complexes were resuspended in a final volume of 30 l of kinase buffer and incubated (20 min at 30°C) with 5 Ci of [␣-32 P]ATP and 0.25 mg/ml MBP. Reactions were terminated by adding Laemmli buffer and boiling. Samples were resolved on a 10% sodium dodecyl sulfate gel and the proteins transferred to a nitrocellulose membrane by semi-dry transfer. After Ponceau staining, the mem-FIG. 2. Rac1 and ERK function on distinct signaling pathways leading to cyclin D 1 transcriptional activation. A, Rac1 is not required for PDGFinduced ERK activation. Left, cells were co-transfected with hemagglutinin-tagged ERK2 (pCDNA3-HA-ERK2) and either a dominant-negative Rac1 (pEXV-N17Rac1) or empty vector. Selected cultures were treated with PDGF (30 ng/ml). ERK activation was assessed by immunoprecipitating cell extracts with an anti-hemagglutinin antibody (12CA5) and measuring the in vitro phosphorylation of MBP (upper panel). To verify that changes in MBP phosphorylation were related to differences in HA-ERK activity, not expression, immunoprecipitated proteins were transferred to nitrocellulose and probed for HA-ERK2 using the 12CA5 antibody (lower panel). Right, group mean data demonstrating the absence of an inhibitory effect of N17Rac1 on ERK activation. Data shown represent mean Ϯ S.E. for four independent experiments. PDGF treatment significantly increased ERK activation compared with the relevant vector controls (p Ͻ 0.05, one way analysis of variance). B, activation of Rac1 is insufficient for ERK activation. Cells were co-transfected with a hemagglutinin-tagged ERK2 and either empty vector or active Rac1 (pEXV-Myc-V12Rac1). Upper panel, activation of transfected ERK was assessed by in vitro phosphorylation assay using MBP as a substrate. Ϫ1745CD1LUC and Ϫ163CD1LUC, PDGF, active Rac1 and active MEK1 each significantly increased promoter activity relative to untreated empty vector (p Ͻ 0.05, one way analysis of variance). For Ϫ66CD1LUC, only PDGF and active Rac1 significantly increased promoter activity. brane was exposed to film and substrate phosphorylation assessed by optical scanning (Jandel Scientific, San Rafael, CA).
To confirm that apparent differences in ERK activity were not related to alterations in expression of the epitope-tagged ERK, nitrocellulose membranes were probed with the anti-hemagglutinin antibody 12CA5. Signals were amplified and visualized using peroxidase-linked rat anti-mouse light chain IgG and enhanced chemiluminescence.
Anti-phospho-ERK Immunoblots-In some experiments, the ERK activity of whole cell lysates was estimated by determining the level of phosphorylated ERKs. These experiments utilized a phospho-specific antibody (32) which recognizes ERKs only when phosphorylated at Thr 183 and Tyr 185 , which are required for full enzymatic activity (33). Cell extracts were resolved on a 10% SDS-polyacrylamide gel and transferred to nitrocellulose, as described above. After incubation with antibody, signals were amplified and visualized using anti-rabbit IgG and enhanced chemiluminescence.
Measurement on Intracellular Hydrogen Peroxide Levels-Cells were grown on coverslips in 35-mm 2 tissue culture-treated dishes. Coverslips were placed inside a perfusion chamber and imaged with an inverted phase/epifluorescent microscope. Fluorescence was measured with a cooled slow-scanning PC-controlled camera coupled to imaging software for quantification of changes in emission fluorescence. Cells were loaded with DCFH-DA (10 M) and dichlorofluorescein (DCF) fluorescence was determined by imaging with an inverted phase/epifluorescence microscope, as described (34). After a 1-h stabilization period, selected cultures were treated with PDGF (30 ng/ml). In some instances, cells were pretreated with ebselen (30 M).
Fractional Labeling with Bromodeoxyuridine-Subconfluent bovine tracheal myocytes were serum starved in DMEM for 24 h. Eight h following PDGF treatment, cells were incubated with bromodeoxyuridine (10 M) and fluorodeoxyuridine (1 M). In some wells, catalase (100 -1000 units/ml) was added 45 min prior to growth factor treatment. Sixteen h later, myocytes were fixed in periodate lysine paraformaldehyde buffer and the DNA precipitated with 2 M HCl. After acid neutralization with 0.1 M borate buffer, pH 9.0, cells were permeabilized with 0.2% Triton X-100. Cells were then immunostained with fluorescein isothiocyanate-labeled anti-BrdUrd and counterstained with propidium iodide. cyclin D 1 promoter activity. To determine the requirement of Rac1 for PDGF-induced cyclin D 1 promoter activity, cells were transiently co-transfected with the dominant-negative Rac1 and the luciferase-tagged cyclin D 1 promoter. N17Rac1 attenuated the PDGF-induced transcription from the cyclin D 1 promoter (Fig. 1). Taken together, these experiments suggest that Rac1 is an important upstream activator of the cyclin D 1 promoter in airway smooth muscle cells.

The Rac1 Signaling Pathway to Cyclin D 1 Transcriptional Activation Is Distinct from the ERK Pathway in Cultured Airway Smooth
Muscle Cells-In HEK293 human kidney fibroblasts, co-expression of Raf-1 and Rac1 caused a synergistic increase in both MEK1 and ERK activation which was greater than the sum of their effects alone (35,36). Since both active Rac1 and MEK1 induce cyclin D 1 promoter activity (above), it is conceivable that activation of Rho family kinases enhances cyclin D 1 transcriptional activation via activation of ERK. We therefore assessed the requirement and sufficiency of Rac1 for ERK activation. Cells were transiently co-transfected with hemagglutinin-tagged ERK2 and either V12Rac1 or N17Rac1. ERK activity was assessed by immunoprecipitation with an anti-hemagglutinin antibody followed by in vitro phosphorylation using MBP as the substrate. Rac1 was neither required nor sufficient for activation of ERK2 (Fig. 2, A and B). We also tested the effects of a synthetic MEK inhibitor, PD98059, on Rac1-induced responses. Pretreatment with PD98059 attenuated PDGF-induced cyclin D 1 promoter activity, but did not decrease Rac1-activated transcription (Fig. 2C). These data strongly suggest that activation of ERK is not required for Rac1-induced transcription from the cyclin D 1 promoter.
Since inhibition of either Rac1 or MEK1 attenuated PDGFinduced cyclin D 1 promoter activity, we also sought to determine whether Rac1 might function downstream of MEK1 in this pathway. We reasoned that if this model were correct, then the downstream transcription factor targets of each intermediate would activate the same site on the cyclin D 1 promoter. To test this, we examined the response of two 5Ј cyclin D 1 promoter fragments, Ϫ163CD1LUC and Ϫ66CD1LUC, to PDGF, Rac1, and MEK1 stimulation. Although cyclin D 1 promoter basal activity was somewhat decreased in these 5Ј deletion mutants (data not shown), PDGF and Rac1 responsiveness was maintained, whereas MEK1 responsiveness was lost upon deletion of base pairs Ϫ163 to Ϫ66 (Fig. 2D). These data imply that the downstream transcription factor targets of the MEK1 and Rac1 pathways act at different sites in the cyclin D 1 promoter, and provide evidence that Rac1 and ERK function on distinct signaling pathways.
Growth Factor Treatment Increases the Generation of Reac-tive Oxygen Species-Production of reactive oxygen species has been noted upon growth factor stimulation of arterial smooth muscle cells (14) and chondrocytes (15). To test whether PDGF treatment induces the intracellular generation of reactive oxygen species in bovine tracheal myocytes, cells were loaded with DCFH-DA (10 M). Upon loading, DCFH-DA enters the  (1:1000 dilution). C, DPI (10 M) does not decrease ERK phosphorylation. ERK phosphorylation was assessed by immunoblotting with an ERK phospho-specific antibody (1:10,000 dilution). For both B and C, immunoblots shown are representative of three independent experiments. D, p67 phox is required for PDGF-induced activation of the cyclin D 1 promoter. Cells were transiently co-transfected with cDNAs encoding Ϫ1745CD1LUC and either p67 phox (full-length) or p67 phox   (N-terminal deletion). Data are expressed as luciferase/␤-galactosidase/h normalized to the control vector, and represent the mean Ϯ S.E. of three independent experiments. cell and the acetate group is cleaved by cellular esterases, trapping the non-fluorescent DCFH inside. Subsequent oxidation by reactive oxygen species, particularly H 2 O 2 and hydroxyl radical, yields the fluorescent product DCF. PDGF (30 ng/ml) increased DCF fluorescence almost 2-fold over baseline (Fig. 3). This increase was similar in magnitude to that observed after hypoxic exposure of cardiac myocytes (34), and was blocked by ebselen (30 M), a glutathione peroxidase mimetic (37).
Cyclin D 1 Expression Is Sensitive to Antioxidants-To examine the potential role of reactive oxygen species in growth factor and Rac1-induced cyclin D 1 promoter activity, we tested the effect of ebselen on PDGF-, Rac1-, and MEK1-mediated responses. PDGF-and V12Rac1-induced cyclin D 1 promoter activities were attenuated by ebselen pretreatment, whereas MEK-2E-induced responses were not (Fig. 4A). Pretreatment with catalase also attenuated PDGF-induced cyclin D 1 promoter activity (Fig. 4B). Finally, ebselen and catalase each decreased cyclin D 1 protein abundance (Fig. 4C). Antioxidant pretreatment had no effect on PDGF-induced ERK phosphorylation, however (Fig. 4D). Taken together, these data suggest that there are antioxidant-sensitive (Rac1-mediated) and antioxidant-insensitive (mediated by MEK1/ERK) pathways to cyclin D 1 expression in bovine tracheal myocytes, and provide further evidence that Rac1 and ERK function on distinct signaling pathways.
We have previously demonstrated in bovine tracheal myocytes that cyclin D 1 is required for DNA synthesis (4). We therefore reasoned that antioxidant pretreatment would decrease DNA synthesis in these cells. Pretreatment of cells with catalase abolished PDGF-induced fractional BrdUrd labeling (Fig. 5).
Role of NADPH Oxidase Complex in Rac1-mediated Cyclin D 1 Promoter Activation-Recent data suggest that the generation of H 2 O 2 by NADPH oxidase occurs in tissues other than phagocytes, including vascular smooth muscle cells (10 -13). The human NADPH oxidase consists of at least seven components including two membrane spanning polypeptides, p22 phox and gp91 phox (which comprise the flavoprotein cytochrome b 558 ), p67 phox , and Rac, the last of which is required for oxidase activation. We tested the effects of a flavoprotein inhibitor, diphenylene iodonium (DPI), on transcription from the cyclin D 1 promoter. PDGF and V12Rac1-stimulated cyclin D 1 promoter activity were attenuated by DPI, whereas MEK2E-induced responses were unaffected (Fig. 6A), consistent with the notion that Rac1 and ERK function on distinct pathways to cyclin D 1 promoter activation. DPI also decreased PDGF-induced cyclin D 1 protein abundance (Fig. 6B). Pretreatment of cells with DPI did not alter ERK phosphorylation (Fig. 6C).
The further examine the role of the NADPH oxidase complex in growth factor-induced cyclin D 1 promoter activity, we transiently co-transfected cells with cDNAs encoding the fulllength cyclin D 1 promoter and either wild-type p67 phox or an N-terminal fragment of this protein, p67 phox  . Since Rac1 interacts directly with p67 phox , previous studies have shown the N-terminal fragment to function as a specific inhibitor of the Rac1 signaling pathway (26). PDGF stimulation increased cyclin D 1 promoter activity in control and wild-type p67 phoxtransfected cells, whereas those transfected with the N-terminal fragment no longer responded to PDGF (Fig. 6D). Taken together, these data suggest that NADPH oxidase function is required for PDGF-and Rac1-induced transcription from the cyclin D 1 promoter in bovine tracheal myocytes. DISCUSSION Studies in NIH 3T3 cells (5, 6) and rat hepatocytes (7) have demonstrated a requirement for Rac1 in G 1 progression. The mechanisms underlying this requirement are not precisely known. Activation of Rac1, a constituent of the NADPH oxidase complex, has been noted to increase intracellular reactive oxygen species (17)(18)(19). It has also been demonstrated that a Rac1 effector site critical for the activation of NADPH oxidase is required for the mitogenic effect of Rac1 in rat embryonic fibroblasts (20). In NIH3T3 cells, Rac1 activates transcription from the cyclin D 1 promoter (21,22), suggesting a mechanism by which Rac1 signaling may regulate G 1 progression. However, the requirement of reactive oxygen species for transcriptional activation of cyclin D 1 has not been tested.
In this study, we found in primary bovine tracheal myocytes that (i) overexpression of dominant-negative Rac1 inhibits PDGF-induced promoter activity, whereas active Rac1 induces transcription from the cyclin D 1 promoter; (ii) growth factor treatment increases the intracellular generation of reactive oxygen species; (iii) antioxidant pretreatment attenuates PDGF-and Rac1-induced cyclin D 1 promoter activity, as well as PDGF-induced cyclin D 1 protein abundance and DNA synthesis; and (iv) inhibitors of NADPH oxidase decrease PDGFand Rac1-induced responses. Taken together, these data strongly suggest that Rac1 is an important upstream activator of the cyclin D 1 promoter in airway smooth muscle cells. Furthermore, these findings suggest that Rac1-induced transcription from the cyclin D 1 promoter is dependent on the generation of reactive oxygen species by NADPH oxidase.
Our findings that dominant-negative and constitutively active forms of Rac1 regulate cyclin D 1 promoter activity in cultured bovine tracheal myocytes imply that Rac1 is required and sufficient for transcription from the cyclin D 1 promoter in airway smooth muscle. However, because these experiments depend on the transient overexpression of mutant alleles of Rac1 in the cell, their results should be interpreted with caution. First, dominant-negative proteins may bind upstream signaling intermediates, thereby blocking the activity of other proteins that bind to the same site. An alternative approach for confirming the requirement of Rac1 for growth factor-induced cyclin D 1 promoter activity would be to use other inhibitors. However, reagents such as Clostridium botulinum exoenzyme C3 transferase, Clostridium sordellii lethal toxin, Clostridium difficile toxin B, and lovastatin are each poorly specific for Rac1 (38 -41). Second, it is possible that overexpression of a constitutively active protein could induce supraphysiologic outcomes. Thus, activation of Rac1 may be insufficient for transcription from the cyclin D 1 promoter under normal physiologic conditions. Nevertheless, when evaluated in the context of additional experiments employing inhibitors of NADPH oxidase, the function of which depends on Rac1, we believe that our results strongly suggest that Rac1 is an upstream activator of cyclin D 1 promoter activity in airway smooth muscle cells. Whether it is sufficient to do so under physiologic conditions, or requires the activation of other pathways (such as the ERK pathway, see below) will require further investigation.
Transcription from the cyclin D 1 promoter has been demonstrated to be a consequence of ERK activation in several cell types, including bovine tracheal myocytes (3,28,29,42). In HEK293 human kidney fibroblasts, co-expression of the serine threonine kinase Raf-1 and Rac1, Rac2, Cdc42, RhoA, or RhoB caused a synergistic increase in both MEK1 and ERK activation which was greater than the sum of their effects alone (35,36), suggesting that Rac1 activation of cyclin D 1 promoter activity might be mediated by the activation of the ERK pathway. The enhancement of ERK activity by Rac1 was mediated by p21 (Cdc42/Rac)-activated kinase-1, or PAK1, which phosphorylates MEK1 on a site important for Raf-1-MEK1 interaction. Using a panel of Rac1 mutants, Westwick and colleagues (21) demonstrated that PAK1 binding was required for Rac1-induced transcription from the cyclin D 1 promoter, consistent with this model. We therefore examined potential interactions between the Rac1 and ERK pathways in our system. Rac1 was neither required nor sufficient for ERK activation. Furthermore, inhibition of ERK activation by pretreatment with the chemical inhibitor PD98059 did not attenuate Rac1-induced responses. These data suggest that, in cultured airway smooth muscle cells, ERK activation is not the major mechanism of Rac1-induced cyclin D 1 transcription. We also found that Rac1but not MEK1-induced transcription was sensitive to pretreatment with antioxidants or inhibitors of NADPH oxidase, and that Rac1 and MEK1 target different response elements in the cyclin D 1 promoter. Taken together, these data strongly suggest that Rac1 and MEK1 stimulate transcription from the cyclin D 1 promoter through distinct signaling pathways.
Once generated, reactive oxygen species may influence a number of signaling processes, including activation of NF-B (43)(44)(45) and p38 mitogen-activated protein kinase (46). Treatment with antioxidants has been demonstrated to attenuate Rac1-induced activation of NF-B in HeLa cells (17), and PDGF treatment induces NF-B activity in mouse fibroblasts (47), consistent with the notion that NF-B may mediate the Rac1/reactive oxygen species signal. The smallest 5Ј promoter fragment we studied (Ϫ66CD1LUC) holds consensus sequences for both NF-B and cAMP response element/activation transcription factor family proteins, the downstream phosphorylation targets of p38 (48,49). Since this deletion fragment remained both PDGF and Rac1 responsive, it is possible that one or both of these transcription factors is involved in growth factor-induced cyclin D 1 promoter activation.
We and others have previously shown in cultured airway smooth muscle cells that activation of the ERK pathway is required for cyclin D 1 expression (3) and DNA synthesis (2,50). In the present study, we demonstrate the importance of the Rho family GTPase Rac1 for airway smooth muscle cyclin D 1 expression. Taken together, these data suggest a model whereby two distinct signaling pathways, the ERK pathway and Rac1 pathway, regulate airway smooth muscle G 1 progression.