Am J Physiol Lung Cell Mol Physiol 290: L661-L673, 2006.
First published October 14, 2005; doi:10.1152/ajplung.00269.2005
1040-0605/06 $8.00
Transforming growth factor-
1 induces Nox4 NAD(P)H oxidase and reactive oxygen species-dependent proliferation in human pulmonary artery smooth muscle cells
Anne Sturrock,1
Barbara Cahill,1
Kimberly Norman,1
Thomas P. Huecksteadt,1
Kenneth Hill,1
Karl Sanders,1
S. V. Karwande,2
James C. Stringham,2
David A. Bull,2
Martin Gleich,1,
Thomas P. Kennedy,1 and
John R. Hoidal1
Divisions of 1Respiratory, Critical Care and Occupational Pulmonary Medicine and 2Cardiovascular Surgery, University of Utah Health Sciences Center and Veterans Affairs Medical Center, Salt Lake City, Utah
Submitted 22 June 2005
; accepted in final form 13 October 2005
 |
ABSTRACT
|
|---|
Transforming growth factor-
1 (TGF-1) is abundantly expressed in pulmonary hypertension, but its effect on the pulmonary circulation remains unsettled. We studied the consequences of TGF-1 stimulation on freshly isolated human pulmonary artery smooth muscle cells (HPASMC). TGF-1 initially promoted differentiation, with upregulated expression of smooth muscle contractile proteins. TGF-1 also induced expression of Nox4, the only NAD(P)H oxidase membrane homolog found in HPASMC, through a signaling pathway involving Smad 2/3 but not mitogen-activated protein (MAP) kinases. TGF-1 likewise increased production of reactive oxygen species (ROS), an effect significantly reduced by the NAD(P)H oxidase flavoprotein inhibitor diphenylene iodonium (DPI) and by Nox4 siRNAs. In the absence of TGF-1, Nox4 was present in freshly cultured cells but progressively lost with each passage in culture, paralleling a decrease in ROS production by HPASMC over time. At a later time point (72 h), TGF-1 promoted HPASMC proliferation in a manner partially inhibited by Nox4 small interfering RNA and dominant negative Smad 2/3, indicating that TGF-1 stimulates HPASMC growth in part by a redox-dependent mechanism mediated through induction of Nox4. HPASMC activation of the MAP kinases ERK1/2 was reduced by the NAD(P)H oxidase inhibitors DPI and 4-(2-aminoethyl)benzenesulfonyl fluoride, suggesting that TGF-1 may facilitate proliferation by upregulating Nox4 and ROS production, with transient oxidative inactivation of phosphatases and augmentation of growth signaling cascades. These findings suggest that Nox4 is the relevant Nox homolog in HPASMC. This is the first observation that TGF-1 regulates Nox4, with important implications for mechanisms of pulmonary vascular remodeling.
Smad 2/3; pulmonary arterial hypertension; hypertrophy; superoxide anion; rac 1; ERK 1/2; phosphatases
DESPITE SIGNIFICANT ADVANCES in diseases of the systemic circulation, pulmonary arterial hypertension (PAH) remains an important cause of cardiopulmonary morbidity. PAH is characterized by prominent proliferation and hypertrophy of medial pulmonary artery smooth muscle and development of prominent medial smooth muscle in distal parts of the pulmonary arterial circulation that are normally nonmuscular. Transforming growth factor (TGF)- plays an important role in development of this process by signaling through a transmembrane TGF- superfamily of serine/threonine receptor kinases that include five type II (TRII) and seven type I (TRI) receptors (10, 26). TGF-1 binds primarily to TRII, which in turn recruits TRI, also called activin receptor-like kinase (ALK). The type I receptor acts downstream to phosphorylate a family of Smad transcription factors, stimulating their formation of heteromeric complexes with Smad4. These complexes then translocate to the nucleus where they mediate TGF- effects determined by which families of ALKs and Smads are transactivated. Recently, familial primary pulmonary hypertension has been linked to genetic mutations encoding the type II bone morphogenetic protein receptor (BMPR-II), a member of the TGF- superfamily receptors (19).
TGF-1 has been previously thought to inhibit the growth of SMC, but under conditions producing PAH, TGF-1 may be proliferative (6) and have a direct pathogenic role in pulmonary vascular disease. TGF-1 is released by human pulmonary artery smooth muscle cells (HPASMC) exposed to hypoxia and mediates hypoxia-induced HPASMC cyclooxygenase-2 expression (35). TGF-1 also stimulates proliferation in HPASMC from patients with idiopathic primary pulmonary hypertension (30). Activity of TGF- is elevated in the lung lymph of sheep with chronic PAH induced by continuous air embolization (32), and TGF-1 expression is increased in pulmonary arteries of lambs with PAH from surgically increased pulmonary blood flow (27). In addition, pulmonary artery expression of the TRI ALK5, which is antiangiogenic (14), is decreased in hypertensive lambs (27), but expression of the proangiogenic type I receptor, ALK1 (14), is significantly increased (27), suggesting a change in the balance of ALK5 to ALK1 in the face of PAH.
Reactive oxygen species (ROS) also play a prominent role in pulmonary vascular remodeling. Superoxide anion (O2–)-scavenging antioxidants protect against hypoxic pulmonary vascular remodeling (28). Also, pulmonary arteries from fetal lambs with flow-induced PAH exhibit increased O2– production likely pathogenic in the PAH disease process (9). On the basis of work in porcine (31) and bovine (15) vessels, the ROS source mediating hypoxic vasoconstriction and HPASMC remodeling is postulated to be NAD(P)H oxidase with Nox2 (gp91phox) or Nox4 as its major membrane subunit, but the homolog in humans remains unsettled.
Evidence favors roles for both TGF-1 and NAD(P)H oxidase in PAH, but their relative importance and potential relationship have been incompletely explored. We have therefore established early passages of normal HPASMC to study their response to TGF-1 and identify their major Nox components. We report that Nox4 is the major homolog of the HPASMC NAD(P)H oxidase. TGF-1 initially stimulates HPASMC differentiation through a Smad 2/3 signaling pathway and increases ROS production from enhanced expression of Nox4. As a delayed secondary response, TGF-1 then stimulates HPASMC proliferation dependent on Nox4 expression. This is mediated in two ways: by TGF-1-stimulated autocrine production of platelet-derived growth factor BB (PDGF-BB) and by enhancement of Nox4-derived ROS signaling cascades.
|
EXPERIMENTAL PROCEDURES
|
|---|
Procurement of pulmonary artery tissue.
Tissue collection was approved by the University of Utah Institutional Review Board. Pulmonary arteries from donors without cardiopulmonary disease were collected at the time of thoracic organ procurement usually within 8 h of the declaration of clinical brain death. At the time of organ procurement, hearts of donors were still beating and their major organs were adequately perfused to maintain viability required for subsequent organ transplantation. Upon acquisition, pulmonary artery segments were placed in ice-cold normal saline and transported to the laboratory where they were either processed for immunohistochemistry or dissected and digested for cell culture.
Cell culture.
HPASMC were obtained by collagenase-elastase (Roche Biochemicals, Indianapolis, IN) digestion of donor pulmonary arteries. In brief, the adventitia was physically removed, and the cleaned artery was incubated in collagenase (1 mg/ml) for 20 min at 37°C, after which we removed the endothelial layer by scraping with the blunt edge of a scalpel. The remaining tissue consisting of the smooth muscle layer was minced in a solution containing collagenase A (2 mg/ml) and elastase grade II (0.5 mg/ml), incubated at 37°C in a shaking water bath for 1 h, pipetted several times to disrupt the cells, and incubated for an additional hour. Undigested tissue was removed by straining through sterile gauze. The cell pellet was washed with sterile phosphate-buffered saline (PBS), and the cells were plated on gelatin-coated dishes in SMC growth medium (Cascade Biologicals, Portland, OR). The initial culture is referred to as the "primary culture." Human pulmonary artery endothelial cells (HPAEC), human microvascular lung endothelial cells (HMVLEC), and human aorta smooth muscle cells (HAoSMC; Cambrex Bioproducts, East Rutherford, NJ) were grown in commercial media containing growth supplements specific for each cell type. Human airway smooth muscle cells (HAwSMC, passage 3–6) established by collagenase digestion and grown under conditions identical to HPASMC were harvested and grown as previously detailed (7). Lung fibroblasts (IMR-90) were purchased from ATCC (Manassas, VA) and grown in DMEM plus 10% fetal calf serum (FCS). Polymorphonuclear leukocytes were isolated from freshly drawn blood of normal volunteers by standard procedures. Human TGF-1, human activin A (Act A), human bone morphogenic protein-7 (BMP-7; R&D Systems, Minneapolis, MN), and mitogen-activated protein (MAP) kinase inhibitors (Calbiochem, San Diego, CA) were added to HPASMC as described in the figure legends.
Immunohistochemistry.
Donor pulmonary arteries were frozen in Tissue-Tek optimum cutting temperature embedding medium (Canemco Marivac, Quebec, Canada) immediately upon receipt, and 7-µm sections were cut and labeled with specific antibodies. Immunohistochemistry of HPASMC cultures was performed on cells fixed with 4% paraformaldehyde for 10 min. A monoclonal antibody to gp91phox (clone 54.1) was a generous gift from Dr. M. Quinn (Montana State University, Bozeman, MT). Human Nox4 and Nox1 antibodies against the peptides SKPAEFTQHKFVKICMEE (256–273) and EMWDDRDSHCRRPKFEGH (248–265), respectively, were produced in rabbits and affinity purified before use. A mouse monoclonal 
-smooth muscle actin (-sma, clone 1A4) antibody was obtained from Zymed-Invitrogen (Carlsbad, CA). Antibody dilutions were empirically derived for each antibody. The dilution used is reported in the results.
Immunoblot for proteins.
Whole cell lysates were prepared for Western blotting by disrupting cells in radioimmunoprecipitation assay buffer containing a broad-spectrum metallo-, serine-, and cysteine-proteinase inhibitor cocktail (Roche Biochemicals). The protein content of the lysate was determined by Bio-Rad Protein Assay. Equal amounts of protein were loaded onto preformed SDS-PAGE gels (Gradipore; Life Therapeutics, Frenchs Forest, Australia). After transfer to a nitrocellulose membrane, the protein of interest was detected using a specific primary antibody and a secondary horseradish peroxidase-conjugated antibody raised against the species in which the primary antibody was developed. The signal was developed using the West Pico chemiluminescent system (Pierce Biotechnology, Rockford, IL).
Reverse transcriptase-polymerase chain reaction.
Total cellular RNA was isolated from cultured cells using Tri-Reagent (MRC, Cincinnati, OH) with the addition of a second ethanol precipitation step. First-strand cDNA was reverse transcribed from 2 µg of total RNA using mouse Maloney reverse transcriptase (Promega, Madison, WI). cDNA (1 µg) was amplified using gene-specific primers selected on the basis of published sequences or selected using Oligo version 6. Primers used to identify the Nox homologs were those reported by Cheng et al. (11). Polymerase chain reaction (PCR) products were electrophoresed on a 1.5% agarose gel for 30 min at 100 volts, and bands were visualized by UV light after staining with ethidium bromide.
Real-time PCR.
Nox1, Nox2, Nox4, Nox5, Duox1, and Duox2 transcripts were quantitated by RT-PCR using an ABI PRISM 7900HT Sequence Detection System (Applied Biosystems, Foster City, CA). cDNA (50 ng) was mixed with ABI TaqMan Universal PCR Master Mix and ABI TaqMan Gene Expression Assay for human Nox4 and human GAPDH. Results are expressed as fold-induction over control values after correcting for GAPDH.
Intracellular superoxide production.
Intracellular O2– production was measured by nitroblue tetrazolium (NBT) reduction as described by Rook et al. (33). Immediately before starting the assay, we washed the cells in a 24-well plate once with sterile PBS and incubated them in 1 ml of serum-free DMEM without phenol red. Thirty minutes before the addition of NBT, cells were treated with the NAD(P)H oxidase inhibitor diphenylene iodonium (DPI) from a 1,000x solution in dimethyl sulfoxide (DMSO) made immediately before use. Equivalent amounts of DMSO were added to control wells. Stock NBT (25 mg/ml) was prepared in methanol and stored at 4°C. NBT (10 µl) was added to each well, and cells were incubated for 1 h at 37°C. The medium was removed, and cells were fixed with ice-cold methanol for 5 min. The supernatant NBT was removed by several washes with methanol at room temperature. After the wells were air-dried, the insoluble blue formazan was dissolved by addition of 440 µl of 1 M KOH to each well, incubation for 5 min, and then addition of 560 µl of spectrophotometric-grade DMSO. The contents of each well were mixed with a pipette to solubilize the purple formazan, and absorbance was determined at 630 nm.
Extracellular superoxide production.
Extracellular superoxide production was measured by ferricytochrome c reduction. Cells in a 24-well plate at 80% confluence were washed once with sterile PBS and incubated in DMEM 0.2% FCS for 48 h with and without TGF-1 (1 ng/ml). The assay was performed in serum- and phenol red-free DMEM containing 80 µM ferricytochrome c. After 2-h incubation at 37°C, ferricytochrome c reduction was measured spectrophotometrically at 550 nm.
Extracellular hydrogen peroxide generation.
Extracellular hydrogen peroxide (H2O2) was measured by the homovanillic acid (HVA) assay (25). Cells in a six-well plate at 80% confluence were washed once with sterile PBS and in DMEM 0.2% FCS for 48 h with and without TGF-1 (1 ng/ml). The assay was performed in 2 ml of serum- and phenol red-free DMEM containing 100 µM HVA and 4 U/ml horseradish peroxidase. After 1- and 2-h incubation at 37°C, 1-ml aliquots were removed, and 75 µl of HVA stop buffer (0.1 M glycine/0.1 M NaOH, pH 12, and 25 mM EDTA in PBS) were added. Fluorescence was read at 312-nm excitation and 420-nm emission and normalized to cell protein.
Proliferation assay.
Proliferation of HPASMC was determined by a previously reported colorimetric method based upon metabolic reduction of the soluble yellow tetrazolium dye 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) to its insoluble purple formazan (8). In brief, cells were seeded in a 24-well plate at 30,000 cells per well and allowed to attach overnight in growth medium. The wells were washed once with PBS and incubated in DMEM containing 0.2% FCS, which maintains viability without stimulating growth. Agonists were added to this medium, and cultures were incubated for 72 h. Medium and agonists were removed, cells were washed twice with PBS, and 1 ml of serum-free DMEM containing 100 µg/ml MTT was added to each well. After 1-h incubation at 37°C, MTT-containing medium was removed, plates were washed with PBS, DMSO (0.5 ml) was added to each well, and the solubilized purple formazan dye was measured at 540 nm. Preliminary studies were performed with 50–200 µg/ml MTT incubated for 15 min–3 h to determine the optimum concentration and incubation time at which the rate of conversion was linear and proportional to the number of cells present. The absorbance of the MTT formazan reduction product (A540) correlated with cell numbers counted by hemocytometer with an R2 = 0.99.
Transfection of HPASMC.
Primary cells are difficult to transfect. We exhaustively tested most commercial lipid and nonlipid transfection reagents and found all to be inefficient, giving a maximum transfection efficiency of 
20%. High transfection efficiency (>90%) of primary and low-passage HPASMC was obtained with the Amaxa Nucleofector system (Gaithersburg, MD), which allows nonviral gene transfer directly into the nucleus using proprietary Nucleofector solutions. We employed Nucleofector solution MC 59 optimized for primary mammalian SMC from various organs and the electrical parameters supplied by program A-33. These parameters were used to transfect HPASMC with dominant negative (dn) Smad2/3 plasmids (a generous gift from Dr. J. Wrana, Mount Sinai Hospital, Toronto, Canada). Adenoviral gene transfer was used to express the dn Rac1 protein (Ad-N17Rac1; a generous gift from Dr. Beata Wojciak-Strothard, University College School of Medicine, London, UK) in HPASMC to determine the role of the small GTPase Rac1 in Nox4-dependent ROS production. Ad-N17Rac1 at multiplicity of infections (1:1,000 and 1:2,000) was added to HPASMC. Experiments were performed 16–18 h after infection.
RNA interference.
Nox4 was inhibited by RNA interference technology. A SMARTpool consisting of four short or small interfering RNAs (siRNA) for Nox4 was obtained from Dharmacon (Lafayette, CO). Experiments were carried out according to the Dharmacon protocol using 50 and 100 nM of Nox4 SMARTpool siRNAs. The Amaxa system was used to transfect the siRNA into HPASMC. Transcript inhibition was determined by RT-PCR performed 24 h after siRNA transfection and measurement of ROS production 72 h after siRNA transfection.
Laser scanning confocal microscopy.
HPASMC were grown on Thermanox coverslips (Nunc, Rochester, NY), fixed with 4% freshly prepared paraformaldehyde, and permeabilized with 0.1% Triton X-100. Cells were incubated overnight with anti-Nox4 and anti--sma antibodies (Chemicon, Temecula, CA) in a humidified chamber at 4°C and then incubated with FITC- and Cy5-conjugated secondary antibodies (Jackson ImmunoResearch, West Grove, PA) for 1 h at room temperature. Cells on coverslips were mounted on a glass slide using ProLong gold antifade reagent (Molecular Probes, Eugene, OR) and visualized through the glass slide. Confocal microscopy that utilized argon and red HeNe lasers to acquire images from two fluorochromes was performed using a Nikon Personal Microscopy PCM-2000. The Simple Personal Confocal Image program (PCI, Compix; Hamamatsu Photonics, Hamamatsu City, Japan) was used to acquire digital images. The FITC label was detected with the argon laser at 488 nm and Cy5 with the red argon laser at 633 nm. The cells were individually scanned with each respective laser filter, and the two images were merged together to determine colocalization.
Statistical analysis.
Data are expressed as means ± SE for a minimum of four observations, unless otherwise indicated. Differences between two groups were compared with the unpaired Student’s t-test. Two-tailed tests of significance were used. Differences between multiple groups were compared with one-way analysis of variance. If significant interactions were found, Newman-Keuls post hoc multiple comparisons test was applied to locate the sources of differences. Levels of significance are stated.
|
RESULTS
|
|---|
Nox4 is the primary NADPH oxidase homolog in human proximal pulmonary artery.
To determine the major Nox homologs used as catalytic subunits of NAD(P)H in HPASMC, we examined normal, mature human proximal pulmonary arteries from six donors free of cardiopulmonary disease ex vivo using immunohistochemistry and antibodies specific to Nox1, Nox2, and Nox4. The same profile was observed in each artery. A representative pulmonary artery is shown in Fig. 1. The medial contractile smooth muscle layer showed intense staining throughout for Nox4 (Fig. 1A), with no staining for Nox2 (Fig. 1B) or Nox1 (Fig. 1C). The small amount of staining for Nox2 in the medial layer was associated with neutrophils and macrophages. In contrast, both the endothelial and the adventitial layers showed staining for all three Nox homologs (Fig. 1, A–C). This observation implies that nondividing contractile smooth muscle cells in the medial layer of normal human proximal pulmonary artery utilize Nox4, but not Nox1 nor Nox2, as the catalytic NAD(P)H oxidase subunit to produce ROS.
View larger version (71K):
[in this window]
[in a new window]
|
Fig. 1. Localization of Nox homologs in human donor pulmonary artery. Arteries were frozen in OCT freezing medium immediately upon receipt. Frozen sections (10 µm) were stained using primary antibodies at a 1:200 dilution, and the appropriate horseradish peroxidase-conjugated secondary antibody at 1:500. Magnification is x40. The endothelium, smooth muscle, and adventitia are labeled. A: intense staining for Nox4 is shown throughout the vessel. B: the endothelial and adventitial layers show intense staining for Nox2, but sparse staining in the medial layer appears to be associated with neutrophils and macrophages. C: Nox1 shows a similar pattern of staining to Nox2.
|
|
Nox4 is the primary NAD(P)H oxidase homolog in cultured HPASMC but is progressively lost in culture.
We established cultures of HPASMC to determine whether culture conditions, known to result in a loss of contractile and a gain in synthetic phenotype, affect Nox4 NAD(P)H oxidase expression. Proteolytic digestion of "cleaned" (see EXPERIMENTAL PROCEDURES) surgical fragments of human proximal pulmonary artery was chosen to establish HPASMC culture, rather than an explant method, because explants requires a longer growth period (as long as 21 days) and may select for a subpopulation of SMC. Seven days after digestion, primary cultures of HPASMC showed heterogeneity in staining for -sma and Nox4 (Fig. 2A). This was not a result of contamination with either endothelial cells or fibroblasts, as primary cultures were routinely stained and found to be negative for both von Willebrand factor as a marker for endothelial cells and for fibroblast surface protein marker. The lack of staining for -sma is consistent with a change toward the synthetic phenotype. The staining pattern for Nox4 mirrored that of -sma. Therefore, it is probable that the phenotypic state of HPASMC also affects NAD(P)H oxidase expression. We routinely examined primary HPASMC for Nox1 and Nox2 protein and saw no staining for either despite the emerging synthetic phenotype.
Because -sma and Nox4 were not expressed in a fraction of primary HPASMC, we predicted a further decrease in transcripts for -sma and Nox4 with increasing passage number. This is shown in Fig. 2B. Transcripts for -sma and Nox4 were reduced at the first passage (p1) and either absent (-sma) or very low (Nox4) by the fourth passage (p4). Presuming that increased passage led to an increase in HPASMC with a synthetic phenotype, we determined whether the shift from contractile to synthetic phenotype resulted in a change from expression of Nox4 to either a Nox1 or Nox2 as the NAD(P)H oxidase homolog. No transcripts or staining for either Nox1 or Nox2 occurred in HPASMC up to p8. The lack of Nox1 and Nox2 is further evidence that the HPASMC cultures were not contaminated, since human pulmonary artery endothelial and fibroblast cells were positive for both homologs (Fig. 1). Expression of p22phox, the presumed membrane catalytic subunit partner for Nox4, was less affected by growth in culture (Fig. 2B). Small amounts of transcript and protein were also variably expressed for both p47phox and p67phox, the cytosolic regulatory components of the classical oxidase (data not shown). By p4, HPASMC possessed only low levels of Nox4 and no other homologs with which p22phox might partner to form an active NAD(P)H oxidase. Consistent with their low levels of Nox4 expression, ROS production was greatly reduced in HPASMC from p4 compared with primary cultures (Fig. 2C). Thus loss of Nox4 expression during progressive HPASMC passages may be a manifestation of evolution to the synthetic phenotype.
TGF-1 induces Nox4 expression and ROS production in HPASMC.
In contrast with the synthetic phenotype, differentiation of SMC to the contractile phenotype is associated with increased expression of smooth muscle contractile proteins. TGF-1, which stimulates smooth muscle differentiation and thus contractile protein expression (1, 16, 17, 22), strongly induced Nox4. HPASMC (p4) produced greater than 70-fold more Nox4 transcripts when grown in low-serum (0.2%) medium containing TGF-1 (1 ng/ml) compared with control (Fig. 3A). This effect was not vessel specific, as it also occurred in HAwSMC and HAoSMC (Fig. 3B). TGF-1 induction of Nox4 was also not a smooth muscle specific effect, as TGF-1 induced Nox4 transcripts in HPAEC, in HMVLEC, and in human lung fibroblasts (IMR-90) (Fig. 3C). The increase in Nox4 transcripts in response to TGF-1 was accompanied by an increase in Nox4 protein (Fig. 3D). Among NAD(P)H oxidase homologs, TGF-1 induction was specific to Nox4. HPASMC RNA from TGF-1-treated cultures was analyzed for all other known Nox homologs expressed in nonfetal tissues. TGF-1 failed to induce transcripts for Nox1, Nox2, Nox5, Duox1, or Duox2 at concentrations as high as 10 ng/ml (Fig. 3E). Nox4 induction was also specific for TGF-1. No induction of Nox4 transcripts occurred in HPASMC (p4) stimulated with PDGF-BB or human interferon-
(Fig. 3F), nor was Nox4 induced by treatment of HPASMC with heparin or retinoic acid (data not shown), both known inducers of the SMC contractile phenotype.
View larger version (44K):
[in this window]
[in a new window]
|
Fig. 3. Induction of Nox4 expression in HPASMC by transforming growth factor (TGF)-1. A: real-time PCR of Nox4 transcripts using 50 ng of cDNA transcribed from total RNA. HPASMC (p4) at 80% confluence were grown in low-serum (0.2%) medium with and without TGF-1 (1 ng/ml) for 24 h. Results are expressed as fold-induction over control, are corrected for GAPDH transcript levels, and represent means ± SE of 3 experiments. *P < 0.01. B: human airway (HAwSMC) and aortic smooth muscle cells (HAoSMC) show induction of Nox4 transcripts upon addition of TGF-1 (1 ng/ml) but not when grown in low serum (0.2%) for 24 h. Conversely, no induction of p22phox was observed. C: human pulmonary artery endothelial cells (HPAEC), human microvascular lung endothelial cells (HMVLEC), and human lung fibroblasts (IMR-90) show induction of Nox4 transcripts when grown in the presence of TGF-1 (1 ng/ml) for 24 h. No increase in p22phox was observed. D: HPASMC (p4) show increase in Nox4 protein by Western blot when grown in the presence of TGF-1 (1 ng/ml) for 72 h. Human embryonic kidney cells (HEK-293) were used as a positive control for Nox4 protein. Polymorphonuclear leukocytes (PMN) were used as a negative control. E: TGF-1 (1 ng/ml) stimulation of HPASMC for 24 h failed to induce transcripts for Nox1, Nox2, Nox5, Duox1, or Duox2. Fold-induction for control and TGF-1-stimulated cells is shown in the table at bottom. F: growth factor induction of Nox4 expression is specific for TGF-1. HPASMC (p4) were treated 24 h in low serum (0.2%) with TGF-1, platelet-derived growth factor BB (PDGF-BB), or human interferon- (huIFN-) at the indicated concentrations, and real-time PCR was performed to quantitate fold-induction relative to untreated control. Results are corrected for GAPDH transcript levels and represent means ± SE of 3 experiments. Only TGF-1 showed a significant increase compared with control (*P < 0.01). cont, Control.
|
|
By the p4 HPASMC showed an approximate sevenfold reduction in ROS production (Fig. 2C). This loss was reversed by TGF-1 stimulation. After incubation with TGF-1 (1 ng/ml) for 48 h, there was a sixfold increase in ROS production compared with control HPASMC (p4) in medium alone (Fig. 4A). This increase was maintained for at least 6 days (Fig. 4A). That TGF-1-increased ROS production involved NADPH oxidase was suggested from partial inhibition of ROS generation by DPI (10 µM), a flavoprotein inhibitor of Nox oxidases (23, Fig. 4B). TGF-1 only increased intracellular O2– production. Extracellular H2O2 generation measured by HVA oxidation and extracellular O2– measured by ferricytochrome c reduction were not affected by TGF-1 (data not shown). Further evidence that TGF-1-stimulated ROS are produced by Nox4 NAD(P)H oxidase was obtained using RNA interference to silence Nox4 (siNox4, Fig. 4C). This resulted in a complete inhibition of TGF-1 induction of ROS (Fig. 4D). ROS levels were lower in siNox4-treated than in HPASMC (p4) control cells, which possess low levels of Nox4 (Fig. 2B), suggesting that siNox4 inhibited both basal and induced Nox4-mediated ROS production.
In addition to TGF-1, we studied the responsiveness of HPASMC to other agonists previously demonstrated to stimulate vascular ROS generation (23). Although the protein kinase C activator phorbol myristate acetate (PMA) failed to increase ROS, the amphiphile arachidonic acid (AA) substantially enhanced ROS generation in HPASMC (Fig. 5A). Furthermore, AA-induced ROS generation by HPASMC was not inhibited by infection of cells with Ad-N17Rac1, the dn inhibitor of Rac1 (Fig. 5B), suggesting that small GTPases are not necessary for ROS production by smooth muscle in the pulmonary circulation.
View larger version (24K):
[in this window]
[in a new window]
|
Fig. 5. Evaluation of classical NAD(P)H oxidase agonists in stimulating ROS production in HPASMC. A: the amphibole arachidonic acid (AA) but not the protein kinase C agonist phorbol myristate acetate (PMA) stimulates ROS production as measured by NBT reduction. Agonists were added to cell cultures concomitantly with NBT. *P < 0.01 compared with control. B: control and AA-enhanced ROS production are unaffected by inhibition of the small GTPase Rac1 by infection of HPASMC with Ad-N17Rac1. MOI, multiplicity of infection.
|
|
TGF-1 induces Nox4 expression via the Smad signaling pathway.
Using a CpG island search in UCSC Genome Browser, we identified the human Nox4 promoter. Analysis indicated a cis-element with a 1-bp mismatch to the classic TGF- control element and a cis-element with a 2-bp mismatch to the novel TGF- control element. Identification of these sites suggests the possibility that in addition to regulating SMC contractile genes (1, 16, 17, 22), TGF-1 can also directly regulate Nox4 gene expression. Supporting a direct regulation of Nox4 gene by TGF-1 was the observation that protein synthesis inhibitor cycloheximide at 10 and 30 µM failed to inhibit TGF-1 induction of Nox4 transcripts in HPASMC (Fig. 6), indicating that new protein synthesis was not required for Nox4 transcript induction. Somewhat surprisingly, cycloheximide resulted in an increase in TGF-1 induction of Nox4 transcripts from 50-fold to over a 100-fold increase (Fig. 6), suggesting that cycloheximide inhibits the synthesis of a repressor protein(s) influencing Nox4 expression posttranscriptionally. The classic pathway by which TGF- directly controls gene expression is via phosphorylation of Smad 2/3, which then translocates into the nucleus and assembles into gene regulatory complexes (10, 26). As shown in Fig. 7, when HPASMC are transfected with dn Smad 2/3 plasmids to inhibit the Smad 2/3 signaling pathway, TGF-1 induction of Nox4 is reduced by at least 45% and as much as 90% (Fig. 7A). Inhibition of TGF-1 signaling with dn Smad 2/3 also substantially decreased TGF-1-induced ROS production (Fig. 7B).
Other TGF- superfamily members are less effective in Nox4 induction.
We also explored the response of Nox4 gene expression to the other TGF-1 subfamily members Act A, which signals through Smad 2/3 (10, 26), and BMP-7, which signals through Smad 1/5/8 (10, 26). Addition of Act A to HPASMC increased Nox4 transcript levels, although much less than TGF-1 despite a 100-fold higher concentration (Fig. 8). BMP-7 (100 ng/ml) failed to increase Nox4 transcripts in HPASMC (Fig. 8). Thus prominent induction of Nox4 expression appears specific for TGF-1.
TGF-1 induction of Nox4 does not involve MAP kinase signaling pathways.
Because transfection of HPASMC with dn Smad 2/3 did not result in a total inhibition of TGF-1-induced Nox4 transcripts and TGF-1 can signal through MAP kinases (26) or cross talk between Smad 2/3 and MAP kinases (1, 18), we determined whether one or more of the three MAP kinase pathways were involved in Nox4 regulation. Levels of Nox4 transcripts were measured by RT-PCR in HPASMC stimulated with TGF-1 (1 ng/ml), with or without pretreatment with inhibitors (at 5 x IC50 reported by Calbiochem) specific for the p38 stress-activated kinase (SB-202190), the p42/44 MAP kinase ERK1/2 (U0126), and Jun NH2-terminal kinase (JNK II inhibitor). Inhibitors were added to HPASMC 30 min before the addition of TGF-1, and the cells were incubated for a further 24 h. The inhibitors by themselves had no effect on Nox4 transcript levels, and there was no inhibition of the TGF-1 induction of Nox4 (Fig. 9).
TGF-1 induces expression of contractile proteins in HPASMC.
To confirm that TGF-1 also induces the contractile phenotype in the pulmonary circulation, we studied the effect of TGF-1 on expression of contractile protein transcripts in HPASMC. HPASMC (p5) possessed low or absent transcript levels for contractile proteins including -sma, smooth muscle myosin heavy chain (sm-MHC), myosin-specific protein 2, vinculin, and calponin. All showed strong induction with the addition of TGF-1 (1 ng/ml) for 24 h (Fig. 10A). Myocardin, a recently discovered early-onset gene that affects the expression of SMC contractile proteins (24) was also induced in HPASMC upon treatment with TGF-1. Protein content of -sma was likewise increased in TGF-1-treated HPASMC compared with control cells (data not shown). ROS production has been shown to induce an increase in SMC contractile proteins, but the source of the ROS has not been established (37). We therefore determined whether the TGF-1-induced increase in contractile protein transcripts in HPASMC was mediated by ROS. The antioxidant N-acetyl cysteine (10 mM) and O2– scavenger 4,5-dihydroxy-1,3-benzene-disulfonic acid (Tiron, 10 mM) significantly decreased TGF-1 induction of smooth muscle contractile protein transcripts in HPASMC (Fig. 10B). Antioxidants were added to the cells 30 min before the addition of TGF-1 and did not induce toxicity as shown by a constant level in GAPDH expression (Fig. 10B, top). Despite reduction of the TGF-1-induced increase in Nox4, transfection with siNox4 failed, however, to inhibit TGF-1-induction of either -sma or sm-MHC transcripts (Fig. 10C). These results suggest that ROS also mediate induction of the contractile phenotype by TGF-1 in HPASMC, but the source of ROS for this signaling is not Nox4 NAD(P)H oxidase.
View larger version (33K):
[in this window]
[in a new window]
|
Fig. 10. TGF-1 induces smooth muscle contractile protein transcripts in a redox dependent- but not Nox4-mediated manner. A: TGF-1 induces transcripts for early and late marker for the smooth muscle contractile phenotype. HPASMC (p5) were treated in low serum (0.2%) for 24 h with TGF-1 (1 ng/ml), and transcripts levels were determined by PCR for the contractile proteins -smooth muscle actin (-sma), smooth muscle myosin heavy chain (sm-MHC), myocardin, vinculin, myosin-specific protein 2 (sm-2), and calponin. Transcript levels for all contractile proteins studied were absent from p5 HPASMC and were induced upon addition of TGF-1. B: ROS inhibitors prevent contractile protein induction in HPASMC by TGF-1. HPASMC (p5) were treated for 3 days with TGF-1 (1 ng/ml) in low serum (0.2%) with and without addition of the antioxidant N-acetyl cysteine (NAC, 10 mmol/l) or the O2– scavenger 4,5-dihydroxy-1,3-benzene-disulfonic acid (Tiron, 10 mm), added 30 min before TGF-1. GAPDH levels remained constant with all treatments, implying lack of toxicity. C: RNA silencing of Nox4 does not prevent TGF-1 induction of smooth muscle differentiation. HPASMC (p5) were transfected with 50 or 100 nmol/l siNox4 (Dharmacon SMARTpool). After 6 h, cells were stimulated for an additional 24 h with TGF-1 (1 ng/ml) in low serum (0.2%). Transcript levels were determined for Nox4, -sma, and sm-MHC. siNox4 prevented TGF-1 induction of Nox4 transcripts but did not affect induction of smooth muscle contractile proteins, indicating that redox-dependent induction of the smooth muscle contractile phenotype is dependent on an ROS source other than Nox4.
|
|
Nox4 NAD(P)H oxidase mediates TGF-1-induced HPASMC proliferation.
TGF-1 has been previously shown to promote proliferation of cultured SMC within 48–72 h of initial stimulation by two additive mechanisms: 1) autocrine secretion of PDGF (accounting for 10–40% of TGF-1-induced growth) and 2) a second mechanism that enhances the proliferative SMC responses to other growth factors but has not been defined (36). HPASMC treated with TGF-1 (1 ng/ml) in low serum for 72 h (Fig. 11, A and B) experienced even greater proliferation than that induced by PDGF-BB (10 ng/ml), the major isoform of PDGF produced by these cells. Similar to results in systemic vascular SMC (36), the delayed proliferative effects of TGF-1 in HPASMC may also involve two mechanisms. First, TGF-1 prominently induced expression of transcripts for PDGF-BB (Fig. 11C), indicating promotion of autocrine PDGF production as previously observed (36). Second, TGF-1 also significantly increased ROS production by HPASMC, but PDGF-BB did not (Fig. 11A). Treatment of HPASMC with siNox4 significantly reduced TGF-1-induced cell proliferation (Fig. 11D), suggesting a relationship between Nox4-derived ROS and TGF-1 induction of cellular growth. There was no observable toxicity due to siNox4, and the control cultures were transfected with an equal amount of Nucleofector as treated cells. Reduction of growth in the control cultures not treated with TGF-1 may have occurred from silencing residual low levels of Nox4 NAD(P)H oxidase present in untreated HPASMC (p4) (see Fig. 2B). Consistent with a Nox4-mediated mechanism, TGF-1-induced proliferation was also substantially reduced by transfection of dn Smad 2/3, but the proliferative response to exogenous PDGF-BB was not (Fig. 11E). Finally, immunolocalization by confocal microscopy demonstrated enhanced staining for Nox4 protein in TGF-1-treated HPASMC in a perinuclear and nuclear distribution, with no staining in focal adhesions or at the plasma membrane surface (Fig. 11F). Localization of Nox4 in these regions suggests a role for Nox4 NAD(P)H oxidase in mediating redox regulation of transcriptional events important for HPASMC proliferation.
One mechanism by which ROS from Nox oxidases might positively influence proliferative signaling is through transient oxidative inactivation of phosphatases in kinase-based growth signaling cascades (29), including inactivation of MAP kinase phosphatases (21), with induction of proliferation through positive regulation of cell cycle activity (5, 34). Using antibodies specific for their phosphorylated active forms, we therefore studied the influence of ROS on TGF-1-induced activation of ERK1/2 in HPASMC. TGF-1 stimulated prompt activation of ERK1/2 in HPASMC within 5 min (Fig. 12A). To enhance activity of NAD(P)H oxidase, we costimulated with 30 µM AA, which significantly activates ROS production (Fig. 5A) and ERK1/2 activity (13) by a phospholipase A2- and Nox4-mediated mechanism (13). AA-induced ROS generation was nearly doubled in cells exposed to TGF-1 (Fig. 12B). ERK1/2 activation measured 24 h after TGF-1 and 20 min after AA addition was substantially reduced (Fig. 12C) by the NAD(P)H oxidase inhibitors DPI or 4-(2-aminoethyl)benzenesulfonyl fluoride (12). This suggests that a large part of TGF-1-stimulated SMC proliferation is mediated by TGF-1 induction of Nox4, which enhances ROS generation and overall "feed-forward" activation of growth factor-induced MAP kinase signaling, possibly from oxidative inhibition of counterbalancing phosphatases.
|
DISCUSSION
|
|---|
We hypothesized that TGF-1 plays a major role in PAH. In this study, the major goals were to determine the effects of TGF-1 on cultured HPASMC and to identify the important components of this cell’s NAD(P)H oxidase. We have shown that the major Nox oxidase homolog in HPASMC is Nox4 (Fig. 1), which is present in freshly cultured cells but progressively lost with each passage in culture (Fig. 2B), paralleling a decrease in ROS production by HPASMC over time (Fig. 2C). Expression of Nox4 is induced by TGF-1 (Fig. 3) through a signaling pathway involving Smad 2/3 (Fig. 7A) but not MAP kinases (Fig. 9). TGF-1 also increases ROS production (Fig. 4A), an effect significantly reduced by the NAD(P)H oxidase flavoprotein inhibitor DPI (Fig. 4B) and by Nox4 siRNAs (Fig. 4D). Nox4 transcript induction by TGF-1 is not prevented by the protein synthesis inhibitor cycloheximide (Fig. 6), indicating that TGF-1 directly signals transcriptional events through preformed factors, such as Smad 2/3 (Fig. 7). Moreover, TGF-1 stimulates HPASMC proliferation (Fig. 11A) in a manner partially inhibited by Nox4 siRNA (Fig. 11D) and dn Smad 2/3 (Fig. 11E), indicating that TGF-1 produces HPASMC growth largely by a redox-dependent mechanism mediated by induction of Nox4. Cellular proliferation in response to many growth factors is mediated in part by production of ROS (5, 34) that may act through transient oxidative inactivation of signaling pathway phosphatases (29), resulting in augmentation of tyrosine and serine-threonine kinase signaling cascades. Therefore, TGF-1 induction of Nox4 might serve to prime proliferative responses in HPASMC overall by increasing the source of ROS to mediate phosphatase inactivation and feed-forward growth signaling enhancement. These findings suggest that Nox4 is the relevant Nox homolog in HPASMC, that it is directly regulated by TGF-1, and that this relationship has important implications for mechanisms pulmonary arterial vascular remodeling during disease states in vivo.
TGF-1 has previously been demonstrated to increase extracellular ROS release by human fetal lung fibroblasts in a delayed, transcriptionally mediated fashion 16 h after initial stimulation, but the source of ROS in these experiments is postulated to be an NADH oxidase activity that directly produces H2O2 through two-electron reduction of O2 without interval generation of O2– (38, 39). Furthermore, this oxidase is not believed mechanistically important in mitogenesis, since catalase fails to impair fibroblast proliferation (39). Our findings in HPASMC are similar in that TGF-1-induced O2– production is delayed and is mediated by transcriptional induction of Nox4 expression. However, in contrast to effects reported in lung fibroblasts (38, 39), in HPASMC TGF-1 stimulation induced intracellular generation of O2– derived from Nox4 (Fig. 4), a NAD(P)H oxidase (11, 23), localized to endoplasmic reticulum (Fig. 11F), not the plasma membrane, and did not result in extracellular release of either O2– or H2O2. Furthermore, in our experiments, TGF-1-induced ROS production plays an important role in HPASMC proliferation (Fig. 11). These disparities in TGF-1 effects on ROS production in fibroblasts compared with HPASMC may be explained by differences in ROS production between cell types or between ROS sources important in the fetal vs. adult condition.
Elements of the vascular NAD(P)H oxidase are well characterized in the systemic circulation and have recently been reviewed (23), but the important components of the pulmonary arterial SMC NAD(P)H oxidase are less well explored. Nox2 (gp91phox) has been reported in porcine (31) and bovine (15) pulmonary vascular SMC, and Nox4 (15) has been identified in bovine pulmonary artery SMC, but the elements of the HPASMC NAD(P)H oxidase have not been previously defined. In early-passage (p0) cells, we readily identified p22phox and Nox4 in HPASMC (Fig. 2B) but no other membrane homologs, including Nox2. Nox4 expression was progressively lost as part of the dedifferentiation apparent in later cell passages (Fig. 2B). We have also found small amounts of transcript and protein expressed for p47phox and p67phox, the cytosolic regulatory components of the classical oxidase, consistent with the variable expression of these elements in systemic vascular SMC (23). We found no transcripts for the p47phox and p67phox homologs, Noxo1 and Noxa1 (data not shown). Unlike that of Nox4, expression of p22phox, p47phox, and p67phox was unaffected by TGF-1 treatment of cultures. Our HPASMC were obtained from large lung vessels rather than resistance pulmonary arteries. Therefore, our results must be qualified. Nevertheless, available evidence suggests that Nox4 is the major regulated NAD(P)H homolog in HPASMC.
Our experience with Nox4 regulation in nontransfected HPASMC is similar to that recently reported from transfection of artificial constructs in HEK-293 cells (3, 25). In these studies, Nox4 expression colocalized with grp-78, a marker for endoplasmic reticulum (3), and PTP1B, a tyrosine phosphatase localized to the outside of endoplasmic reticular membranes (25). Also, Nox4 activity was unaffected by stimulation with PMA (25), inhibition of Rac1 by transfection of expression constructs for dn N17Rac1 (25), or transfection of expression constructs for the cytosolic regulatory components p47phox and p67phox (25) or their respective homologs Noxo1 or Noxa1 (3, 25). Only p22phox was required for Nox4-mediated ROS production (3, 25), suggesting that, unlike Nox1 and Nox2 (23), neither small membrane G proteins nor cytosolic NAD(P)H oxidase components are necessary for Nox4 activity. Our finding that TGF-1 prominently induces both Nox4 (Fig. 3) and ROS production (Fig. 4) in HPASMC in a manner unaffected by stimulation with PMA (Fig. 5) or inhibition by dn N17Rac2 (Fig. 5) raises the possibility that as long as the membrane partner p22phox is available (3), Nox4 is sufficient for O2– production without need for membrane or cytosolic regulatory elements and that Nox4 activity is primarily regulated by its transcription. This scenario would make Nox4 similar to another cellular source of reactive oxidant species important in pathological conditions, inducible nitric oxide synthase, which is exclusively regulated by transcription and substrate availability (4).
The NAD(P)H oxidase flavoprotein inhibitor DPI (Fig. 5B) and transfection of Nox4 siRNA (Fig. 5D) both inhibited ROS production by HPASMC, but Nox4 siRNA failed to prevent TGF-1-induced SMC differentiation (Fig. 10C). This is despite the fact antioxidants did prevent induction of contractile markers (Fig. 10B). Redox regulation of vascular smooth muscle differentiation has been previously recognized (37), but the source for ROS mediating this activity was not determined. TGF-1 is known to stimulate ROS production from other intracellular sources, including mitochondria (2, 40) and cytochrome P-450 (2). Our findings (Fig. 10, B and D) are consistent with the observation that sites other than NAD(P)H oxidase exist for TGF-1-responsive ROS production and suggest that while Nox4 plays a role in TGF-1-induced HPASMC proliferation (Fig. 11), other sites for ROS generation, perhaps mitochondrial, are the likely source for ROS mediating TGF-1-induced SMC contractile differentiation.
At present Nox4-derived ROS may have many undefined functions in HPASMC. Because it is the only Nox homolog identified, it is tempting to ascribe a role to Nox4 in producing ROS important for mediating hypoxic pulmonary vasoconstriction (15, 31), but final determination of the source of ROS for this physiological process in humans will require additional studies. Our demonstration of delayed stimulation of HPASMC proliferation by TGF-1 at 72 h is consistent with the previous observation (36) that, while it is not initially mitogenic within the first 24 h, TGF-1 induces a delayed proliferative response at 48–72 h in vascular SMC. This work represents the first demonstration of a Nox4-mediated proliferative function for TGF-1 (Fig. 11) in HPASMC and elucidates the PDGF-independent mechanism of TGF-1-induced proliferation previously observed in vascular SMC by others (64). TGF-1 (27, 30, 32) and ROS (9, 20, 28) production is elevated in a number of pathophysiological circumstances resulting in PAH, including mechanical stress. Therefore, a TGF-1- and Nox4-mediated signaling relationship may play an important pathogenic role in the development of PAH in a number of unresolved diseases of the human pulmonary circulation.
|
GRANTS
|
|---|
This work was supported by National Heart, Lung, and Blood Institute Grant HL-67281 (J. R. Hoidal) and VA salary support for T. P. Huecksteadt, K. Hill, and Drs. K. Sanders, T. P. Kennedy, and J. R. Hoidal.
|
ACKNOWLEDGMENTS
|
|---|
We gratefully acknowledge Dr. Martin Gleich (deceased) who initiated the research.
We thank our colleagues from Intermountain Donor Service for assistance in obtaining tissue samples.
|
FOOTNOTES
|
|---|
Address for reprint requests and other correspondence: John R. Hoidal, Div. of Respiratory, Critical Care and Occupational Pulmonary Medicine, Wintrobe 701, Univ. of Utah Medical Center, 26 North 1900 East, Salt Lake City, UT 84132 (e-mail: John.Hoidal{at}hsc.utah.edu)
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Deceased 
|
REFERENCES
|
|---|
- Abecassis L, Rogier E, Vazquez A, Atfi A, and Bourgeade MF. Evidence for a role of MSK1 in transforming growth factor-beta-mediated responses through p38 and Smad signaling pathways. J Biol Chem 279: 30474–30479,2004.[Abstract/Free Full Text]
- Albright CD, Salganik RI, Craciunescu CN, Mar MH, and Zeisel SH. Mitochrondrial and microsomal derived reactive oxygen species mediate apoptosis induced by transforming growth factor-1 in immortalized rat hepatocytes. J Cell Biochem 89: 254–261,2003.[CrossRef][ISI][Medline]
- Ambasta RK, Kumar P, Griendling KK, Schmidt HHHW, Busse R, and Brandes RP. Direct interaction of the novel Nox proteins with p22phox is required for the formation of a functionally active NADPH oxidase. J Biol Chem 279: 25935–25941,2004.[Free Full Text]
- Beck KF, Eberhardt W, Frank S, Huwiler A, Meßmer UK, Muhl H, and Pfeilschifter J. Inducible NO synthase: role in cellular signaling. J Exp Biol 202: 645–653,1999.[Abstract]
- Berasi SP, Xiu M, Yee AS, and Paulson KE. HBP1 repression of the p47phox gene: cell cycle regulation via the NADPH oxidase. Mol Cell Biol 24: 3011–3024,2004.[Abstract/Free Full Text]
- Berk BC. Vascular smooth muscle growth: autocrine growth mechanisms. Physiol Rev 81: 999–1030,2001.[Abstract/Free Full Text]
- Brar SS, Kennedy TP, Sturrock AB, Huecksteadt TP, Quinn MT, Murphy TM, Chitano P, and Hoidal JR. NADPH oxidase promotes NF-
B activation and proliferation in human airway smooth muscle. Am J Physiol Lung Cell Mol Physiol 282: L782–L795,2002.[Abstract/Free Full Text] - Brar SS, Kennedy TP, Whorton AR, Murphy TM, Chitano P, and Hoidal JR. Requirement for reactive oxygen intermediates in serum-induced and platelet derived growth factor-induced proliferation of airway smooth muscle. J Biol Chem 274: 20017–20026,1999.[Abstract/Free Full Text]
- Brennan LA, Steinhorn RH, Wedgwood S, Mata-Greenwood E, Roark EA, Russell JA, and Black SM. Increased superoxide generation is associated with pulmonary hypertension in fetal lambs. A role for NADPH oxidase. Circ Res 92: 683–691,2003.[Abstract/Free Full Text]
- Byfield SD and Roberts AB. Lateral signaling enhances TGF- response complexity. Trends Cell Biol 14: 107–111,2004.
TOP
ABSTRACT
EXPERIMENTAL PROCEDURES
P < 0.01 TGF-