Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Free Radic Biol Med. Author manuscript; available in PMC 2013 October 1.
Published in final edited form as:
PMCID: PMC3448829

Nox4 involvement in TGF-beta and SMAD3-driven induction of the epithelial-to-mesenchymal transition and migration of breast epithelial cells


The epithelial-to-mesenchymal transition (EMT) is the development of increased cell plasticity that occurs normally during wound healing and embryonic development and can be coopted for cancer invasion and metastasis. TGF-beta induces EMT but the mechanism is unclear. Our studies suggest Nox4, a member of the NADPH oxidase (Nox) family, is a source of reactive oxygen species (ROS) affecting cell migration and fibronectin expression, an EMT marker, in normal and metastatic breast epithelial cells. We found TGF-beta induces Nox4 expression (mRNA and protein) and ROS generation in normal (MCF10A) and metastatic (MDA-MB-231) human breast epithelial cells. Conversely, cells expressing a dominant-negative form of Nox4 or Nox4-targeted shRNA showed significantly lower ROS production upon TGF-beta treatment. Expression of a constitutively active TGF-beta receptor type I significantly increased Nox4 promoter activity, mRNA and protein expression, and ROS generation. Nox4 transcriptional regulation by TGF-beta was SMAD3-dependent based on the effect of constitutively active SMAD3 increasing Nox4 promoter activity, whereas dominant-negative SMAD3 or SIS3, a SMAD3-specific inhibitor, had the opposite effect. Furthermore, Nox4 knockdown, dominant-negative Nox4 or SMAD3, or SIS3 blunted TGF-beta induced wound healing and cell migration, whereas cell proliferation was not effected. Our experiments further indicate Nox4 plays a role in TGF-beta regulation of fibronectin mRNA expression, based on the effects of dominant-negative Nox4 in reducing fibronectin mRNA in TGF-beta treated MDA-MB-231and MCF10A cells. Collectively, these data indicate Nox4 contributes to NADPH oxidase-dependent ROS production that may be critical for progression of the EMT in breast epithelial cells, and thereby has therapeutic implications.

Keywords: NADPH oxidase 4 (Nox4), Cell Migration, TGF-beta signaling, Epithelial-to-Mesenchymal Transition (EMT)


Recent observations suggest reactive oxygen species (ROS) play important roles in TGF-beta-induced epithelial-to-mesenchymal transition (EMT) and cell mobility of many cell types; however, little is known about redox-dependent TGF-beta-induced mechanisms in breast epithelial cells [1, 2]. EMT is a common process during embryonic development, wound healing, and organ fibrosis in which epithelial cells undergo morphological changes resulting in increased cell plasticity and mobility as they transition into a mesenchymal-like cell phenotype. Recently, the EMT has been a focus of tumor cell migration and invasion. Hallmarks of cells undergoing EMT include cytoskeletal reorganization, loss of polarization, disruption of cell-cell junctions, and alterations in ECM production. Tumor microenvironments secrete soluble growth factors, inflammatory cytokines (TGF-beta, HGF, VEGF, PDGF), and matrix proteins that activate epithelial cells to induce specific EMT transcriptional programs. TGF-beta is established as a contributor to EMT progression. Smooth muscle actin (SMA) and fibronectin expression are established indicators of EMT progression, both of which are modulated in a redox-dependent manner [2, 3].

Redox mediated signaling has become a focus toward understanding EMT progression. Oxidative stress is primarily associated with cytotoxicity, however, ROS are now recognized as important regulators of signaling pathways and gene transcription [4, 5]. ROS can act through oxidative modification of protein tyrosine phosphatases, redox-dependent transcription factors, and DNA [69]. Major sources of cellular ROS include NADPH oxidase enzymes (Nox), mitochondrial electron transport, xanthine oxidase, and nitric oxide synthase [3]. Evidence suggests the Nox family of ROS generating enzymes (Nox1–5 and Duox1–2) are significant players in redox-mediated signaling [10]. In particular, Nox4 was identified as a TGF-beta-responsive oxidase implicated in cytoskeletal alterations in endothelial cells, osteoblast differentiation, pulmonary fibroblast proliferation, idiopathic pulmonary fibrosis, and enhanced oxidative stress in hepatitis C virus infection [1116].

Nox4 is a 578 amino acid, six transmembrane domain flavocytochrome that functions in generating superoxide by transporting electrons from cytosolic NADPH across biological membranes to molecular oxygen. Nox4 has 39% homology to Nox-2 (gp91phox), the Nox family prototype. Originally described in the kidney, Nox4 mRNA is also detected in other human and murine tissues including bone, vascular tissue, heart, liver, and lung [1722]. Nox4 is primarily localized in perinuclear regions and the ER but is also detected at the plasma membrane, focal adhesions, and associated with mitochondria [9, 2227]. Although Nox4 expression is regulated by TGF-beta in many cell types, little is known about its role in the EMT of breast epithelial cells.

In a previous study Nox family genes were surveyed in several human cancer cell lines and in patient tumors and adjacent normal tissues [28]. Their results showed increased Nox4 in breast cancer cell lines as well as in patient tumor samples. In addition, Graham, et at., reported high expression of Nox4 in invasive and non-invasive cell lines as well as in patient breast tumor samples [29]. Furthermore, their report indicates overexpression of Nox4 in MCF12A or MDA-MB-435 cells resulted in cellular senescence, resistance to apoptosis, and tumorigenic transformation. Interestingly, several studies reported the presence of senescent cells in tumor microenvironments. Although resistance to mitogenic stimuli is characteristic of senescent cells, these cells remain metabolically active and capable of secreting growth factors that activate tumor cells [30].

Here we identify Nox4 as the principle source of TGF-beta-induced superoxide production in MCF10A and MDA-MB-231 epithelial cells, and we show Nox4 functions as an important player in EMT-related events including cell migration and fibronectin gene regulation.

Materials and Methods

Cell culture

Immortalized human MCF10A and human metastatic MDA-MB-231 breast epithelial cell lines were obtained from the ATCC (Rockville, MD). MCF10A cells were maintained in Dulbecco’s minimal essential medium/F-12 (DMEM/F12) (Invitrogen, Carlsbad, CA) supplemented with 5% horse serum (Sigma, St. Louis, MO), 500ng/ml hydrocortisone (Sigma), 20ng/ml EGF (R and D Systems, Minneapolis, MN), 100ng/ml cholera toxin (Sigma), 10µg/ml insulin (Sigma), and 100 mg/ml of penicillin-streptomycin (Invitrogen). MDA-MB-231 cells were maintained in Dulbecco’s minimal essential medium (DMEM) (Invitrogen) supplemented with 10% fetal bovine serum (HyClone/Thermo Scientific, Logan, UT) and 100 mg/ml of penicillin-streptomycin. All cells were grown in a humidified atmosphere of 5% CO2 and 95% air at 37°C.

Plasmids and primers

Vectors encoding truncated, dominant-negative Nox4 (Nox4-DN; residues 1–305), pGL3-Nox4 (−1848), and Nox4 shRNA plasmids were previously described [16]. Human Nox4 was generated using full-length Nox4 (GenBank accession number NM_016931) cDNA subcloned into pcDNA3.3. The HA-tagged constitutively active TGF-beta receptor type-1 (T204D) cDNA was subcloned into pcDNA3.1 TOPO (Invitrogen). The full-length receptor was amplified from MCF10A cDNA using the following primers: forward primer 5’-CACCATGGAGGCGGCGGTCGCTGCTCC-3’, and reverse primer 5’-TTACTAGGCGTAATCGGGGACATCATAGGGGTACATTTTGATGCCTTCCTGTTGACT-3’ (HA epitope tag underlined in reverse primer). Site directed mutagenesis was performed for mutation of TGF-beta receptor-1 residue 204 from threonine to aspartate with the following primers: forward primer 5’-AGAACAATTGCGAGAGATATTGTGTTACAAGAA -3’, and reverse primer 5’-TTCTTGTAACACAATATCTCTCGCAATTGTTCT – 3’ (mutated sequence underlined). All constructs were verified by automated DNA sequencing (Macrogen, Rockville, MD).

Transient transfections

MCF10A cells were transfected with Fugene 6 (Roche, Indianapolis, IN, USA) using manufacturer’s protocols. Briefly, 2.5 × 105 cells were seeded in 6-well tissue culture dishes (BD Biosciences) 24 hours before transfection. Transfection mixtures were incubated at RT for 20 minutes in serum-free DMEM and then added drop-wise to cells. MDA-MB-231 cells were transfected with Lipofectamine 2000 (Invitrogen) according to manufacturer’s protocols. Briefly, 2 × 105 cells were seeded on 6-well plates 24 hours before transfection. The transfection mixtures were incubated at RT for 20 minutes in serum-free DMEM and added drop-wise to cells containing serum-free medium. After 4 hours, the medium was changed to DMEM containing serum.

ROS detection

Cells were collected by trypsinzation and washed twice with Hank’s Balanced Salt Solution (HBSS; Invitrogen). Cells were treated 10 min at 37°C with or without 10 µM diphenylene iodonium chloride (DPI) (Sigma). Kinetic ROS measurements were performed by chemiluminescence in 96-well plates at 37°C over a 1 hr time course using a Luminoskan luminometer (Thermo, Waltham, MA, USA). Extracellular superoxide production was measured as superoxide dismutase-inhibitable chemiluminescence detected using Diogenes reagent (National Diagnostics, Atlanta, GA, USA). Extracellular H2O2 was measured by a luminol/HRP-based chemiluminescence assay. Briefly, the cells were collected as above and treated with or without 10µM DPI (37°C for 10 min). An equal volume of HBSS containing 1 mM luminol and 20 U/ml HRP was added following DPI treatment. Luminescence was measured in 96-well plates at 37°C over a 1 hr time course using a Luminoskan luminometer.

RNA isolation and RT-PCR

Total RNA from MCF10A or MDA-MB-231 cells was extracted from cells with Trizol (Invitrogen). One microgram of total RNA was used for Thermoscript RT-PCR. Both were conducted according to manufacturer’s protocols (Invitrogen).

Quantitative real-time RT-PCR

Gene expression was quantified by real-time PCR using an ABI Prism 7500 RT-PCR System (Applied Biosystems). One microgram of total cellular RNA was reverse transcribed with ThermoScript RT-PCR kit (Invitrogen). SYBR Green PCR mix (Invitrogen) was used to detect mRNA expression with the following human specific primers: (Fibronectin) forward 5’- CGAGCTTCCCCAACTGGTAACCC-3’, reverse 5’-GGTGGCACCTCTGGTGAGGC-3’; (GAPDH) forward 5’-AGCCACATCGCTCAGACAC-3’, reverse 5’- GCCCAATACGACCAAATCC-3’. Primers designed to detect human Nox1–5, Duox1 or 2, and p22phox were used as previously described [16].

Antibodies and immunoblotting analysis

MCF10A cell lysates were processed for Western blotting as previously described [16] and probed with the following antibodies: anti-rabbit polyclonal Nox4 (Novus Biologicals); mouse monoclonal anti-V5 (Invitrogen); mouse monoclonal anti-HA (Sigma); rabbit monoclonal anti-phospho-SMAD3, rabbit polyclonal anti-phospho-SMAD2, rabbit monoclonal anti-SMAD3, and mouse monoclonal anti-SMAD2 (Cell Signaling); rabbit polyclonal anti-beta Tubulin H-235 (Santa Cruz).

Luciferase assay

The human Nox4 promoter (−1848) was subcloned into pGL3 reporter plasmid as previously described [16]. Luciferase assays were performed according to manufacturer’s protocols (Promega). Briefly, 1 × 105 MCF10A cells were seeded per well in six-well dishes. 24 hours later, each well was transfected with 2µg total DNA using 4µL FuGENE. Luciferase expression was assayed 48 hours later by lysing cells in 200uL cell lysis buffer, aliquoting 20µL lysate in triplicate in 96-well plates, and injecting 100µL luciferase assay buffer per well; luciferase activities from multiple transfections were normalized to cells transfected with pGL-basic vector, which lacked the Nox4 reporter sequence. Assays were performed using a Luminoskan luminometer.

Wound Closure Assay

2 × 105 MDA-MB-231 cells were seeded on 6-well tissue culture plates 24 hours before transfection. The cells were transfected with 2ug of either control vector (empty or GFP-expressing) or vectors expressing Nox4-DN, Nox4-targeted or control shRNA, TGFBR1(T204D), SMAD3-CA, or SMAD3-DN proteins plasmids. Wounding was performed by scratching cell monolayers with a 200 µl plastic pipette tip 48 hours post-transfection. For treated cells, either 5 ng/mL TGFbeta or 10 uM SIS3 (SMAD3 inhibitor) were added immediately following the scratch. Images of scratched monolayers were monitored under phase contrast microscopy just before addition of reagents (0 hr) and 6 hours (MDA-MB-231) or 24 hours (MCF10A) later. Experiments were repeated three to five times.

Matrigel Cell Migration Assay

MDA-MB-231 cells were seeded and transfected as described above, then trypsinized and washed 24 hours later. 1.5 × 105 transfected cells were seeded in upper chambers of 6-well BD Biocoat matrigel transwell culture plates (BD Biosciences) and incubated with lower chambers containing 1.5ml of DMEM medium with TGF-beta (5ng/ml). After 24 hours, non-migrating cells were scraped away and migrating cells were stained with Diff Stain (IMEB, San Marcos, CA). Invading cells were counted from 10 random fields. Matrigel experiments were repeated three times.

Cell proliferation assay

Cell proliferation was assessed by the Click-iT EdU Imaging Kit according to manufacturer’s instructions (Invitrogen). Briefly, cells after EdU-labeling (1hr exposure), fixation and permeabilization were stained with anti-EdU Alexa Fluor 555. Cell densities were determined by counting DAPI-stained nuclei in fields, using a Leica fluorescence microscope (average n=6–10 fields/sample), after 24-hour treatments with TGF-beta and normalized relative to densities of samples before TGF-beta treatments. The proportion of proliferating cells was calculated based on the number of EdU-positive cells divided by the number of DAPI-stained nuclei (20× magnification; n=6–10).

Statistical Analysis

Data are represented as the means ± standard deviations of the results of at least three independent experiments. Student's t test was used to calculate significance values. Significant values are indicated as (*) P value of <0.05 or (**) P value of <0.01.


TGF-beta induces NADPH oxidase-dependent superoxide production in normal and in metastatic human breast epithelial cells

To investigate the effect of TGF-beta on NADPH oxidase-dependent ROS production in human breast epithelial cells, we treated immortalized (MCF10A) or metastatic (MDA-MB-231) cell lines with TGF-beta for 24 hours. We found MCF10A cells produced a significant amount of extracellular NADPH oxidase-dependent superoxide when treated with increasing concentrations of TGF-beta. Superoxide production was inhibited with DPI, a selective inhibitor of Noxes and other flavoenzymes (Fig. 1A, left). Interestingly, detection of extracellular H2O2 by luminol/HRP was not inhibited by DPI in cells treated with TGF-beta (Fig. 1A, right), suggesting the prevalent H2O2 source is not a Nox enzyme. Similarly, the effect of TGF-beta on MDA-MB-231 cells resulted in increased DPI-sensitive superoxide production, whereas, H2O2 generation was unaffected by TGF-beta or DPI treatment (Fig. 2B). These results indicate TGF-beta treatment of both normal and metastatic breast epithelial cells results in NADPH-oxidase dependent superoxide production at the plasma membrane.

Figure 1
TGF-beta induces Nox4-dependent superoxide generation in normal and metastatic human breast epithelial cells
Figure 2
Dominant negative Nox4 and shRNA-mediated Nox4 gene knockdown suppress TGF-beta-induced superoxide production in normal and metastatic breast epithelial cells

To determine if induction of a member of Nox family enzymes is involved in TGF-beta-mediated ROS production, we measured Nox gene mRNA levels by quantitative PCR. Nox4 mRNA was increased upon TGF-beta treatment of MCF10A and MDA-MB-231 cells (Fig. 1C), while expression of other Nox family members showed no significant change. Increased Nox4 mRNA was further confirmed by semi-quantitative PCR in both cell lines (Fig. 1D). Nox4 protein expression was also upregulated in TGF-beta treated cells compared to untreated control cells (Fig. 1E). Activation of TGF-beta signaling was confirmed by immuno-detection of phosphorylated SMAD2 and SMAD3. Basal levels of SMAD proteins were not affected by TGF-beta. Our results indicate TGF-beta increases NADPH oxidase-dependent superoxide release, which correlates with a significant up-regulation of Nox4 mRNA and protein expression in both normal and metastatic breast epithelial cells.

Dominant-negative Nox4 and Nox4-specific gene knockdown reduce TGF-beta-induced superoxide generation

To determine the role of Nox4 as a TGF-beta responsive superoxide generator, we constructed a Nox4 dominant-negative cDNA with deleted C-terminal FAD and NADPH domains. We observed that overexpression of Nox4-DN diminished superoxide production compared to cells overexpressing empty vector control in MCF10A or MDA-MB-231 cells (Fig 2A left and right panels). These results where further confirmed by knockdown of endogenous Nox4 with Nox4-specific shRNA. Cells transfected with Nox4-shRNA plasmids showed a significant reduction of Nox4 expression at both the mRNA and protein level (Fig. 2B). Furthermore, decreased expression of endogenous Nox4 resulted in reduced TGF-beta-mediated superoxide generation in MCF10A and MDA-MB-231 cells (Fig. 2C, left and right panels). Collectively, these data indicate Nox4 is the primary source of TGF-beta-induced ROS in both normal and metatstatic breast epithelial cells.

Constitutive activation of TGF-beta receptor-1 modulates Nox4 expression and superoxide generation

To further validate the role of TGF-beta mediated Nox4 induction, MCF10A cells were transfected with an HA epitope-tagged constitutively active TGF-beta type 1 receptor (T204D) cDNA construct. Interestingly, MCF10A cells expressing T204D produced a significant level of DPI-sensitive superoxide in the absence of TGF-beta treatment compared to control (Fig. 3A). Superoxide generation was further increased upon TGF-beta stimulation in both T204D and vector control cells. Constitutive activation of TGF-beta receptor-1 also affected Nox4 protein levels. TGF-beta stimulated MCF10A cells increased Nox4 protein in control and T204D transfected cells (Fig. 3B). Untreated T204D cells also resulted in an up-regulation of Nox4 expression. These results correlate with the T204D-dependent superoxide production in panel A. Subsequent immunoblotting with phospho-SMAD3, total SMAD3, and HA antibodies confirmed TGF-beta signaling activation and T204D expression. Real-time quantitative PCR of MDA-MB-231 cells expressing T204D also revealed Nox4 mRNA was substantially upregulated in the presence or absence of TGF-beta in this metastatic cell line (Fig. 3C).

Figure 3
Constitutively active type 1 TGF-beta receptor enhances superoxide production and Nox4 expression in MCF10A and MDA-MB-231 cells

TGF-beta/SMAD3 signaling modulates the human Nox4 promoter in breast epithelial cells

We previously reported TGF-beta treated HepG2 hepatocytes significantly increased Nox4 promoter activity indicating Nox4 expression is regulated by TGF-beta at the transcriptional level [16]. Here, we further confirmed our results in MCF10A cells expressing the human Nox4 promoter (−1848 bp) linked to a luciferase reporter plasmid co-transfected with empty vector or T204D. Nox4 promoter-driven luciferase activity was significantly increased when stimulated by TGF-beta in both control and T204D cells (Fig. 4A). To demonstrate the activating contribution of SMAD3 on the Nox4 promoter, we selectively inhibited SMAD3 by pretreating MCF10A cultures transfected with T204D or vector control with SIS3, a SMAD3 specific inhibitor, 24 hours before assaying luciferase activity. As expected, SIS3 treatment significantly reduced luciferase activity compared to untreated cells (Fig. 4B). Similar inhibition was observed by transfecting MCF10A cultures with a SMAD3 dominant-negative (SM3-DN) construct compared to SMAD3 wild-type (SM3-WT), whereas transfection of a constitutively active variant (SM3-CA) induced significantly higher expression compared to both wild-type and vector control (Fig. 4C). Collectively, these data indicate the TGF-beta/SMAD3 signaling pathway positively regulates the Nox4 promoter. However, we cannot rule out involvement of other transcription factors involved in TGF-beta regulation of the Nox4 promoter.

Figure 4
TGF-beta/SMAD3 signaling modulates the human Nox4 promoter in breast epithelial cells

Nox4 contributes to TGF-beta-mediated cell migration

Excess ROS production is known to play a role in EMT gene regulation and cell migration. In order to determine if Nox4 is involved in TGF-beta mediated migration of breast epithelial cells, we performed a wound-healing assay on confluent MDA-MB-231 and MCF10A cells and examined the effects of Nox4-DN or Nox4-shRNA. TGF-beta treated vector control cells displayed an advanced stage of wound closure at 6 hrs (MDA-MB-231 cells; upper panel) or 24 hrs (MCF10 cells; lower panel), whereas cells expressing Nox4-DN (MDA-MB231) or Nox4-shRNA (MCF-10A) displayed minimal wound closure in the presence of TGF-beta (Fig. 5A).

Figure 5
Nox4 is involved in TGF-beta/SMAD3-induced cell migration

To further confirm involvement of Nox4 in cell migration, we utilized the Matrigel migration assay on MDA-MB-231 cells. Vector control or Nox4-DN-transfected cells were seeded in the upper chambers of transwells atop a porous membrane coated with a thin layer of Matrigel. The chamber was then placed in serum-free medium with or without TGF-beta. Vector-transfected cells showed enhanced migration through the Matrigel and porous membrane after 24 hours when incubated with TGF-beta containing medium in comparison to untreated, whereas cells over-expressing Nox4-DN did not show a significant increase in migration in response to TGF-beta (Fig. 5B).

TGF-beta has also been shown to enhance cell proliferation [31]; therefore, we examined whether Nox4 is involved in TGF-beta-mediated cell proliferation of MDA-MB-231 cells. Cell proliferation after 24 hours of TGF-beta treatment was determined either by measuring relative changes in cell density (DAPI-stained nuclei/ surface area) or by monitoring EdU incorporation for 1 hour after 24 hour TGF-beta treatment. Expression of dominant-negative Nox4 did not affect the increases in cell density or the number of EdU-positive cells (Fig. 5C).

Lastly, we examined whether SMAD3 inhibition would abrogate cell migration in MDA-MB-231 cells in a manner similar to transfection of the Nox4-DN construct. Cultures transfected with the T204D and empty vector or with T204D and constitutively active Smad3 (SM3-CA) showed enhanced wound closure after 6 hours compared to cultures transfected with T204D and dominant negative Smad3 (SM3-DN) (Fig. 5D). Treatment with SIS3 similarly inhibited wound closure in cells transfected with T204D, whereas no treatment resulted in wound closure after 6 hours, and cells transfected with vector alone showed minimal cell migration (Fig. 5E). Vector control cells treated without vs. with SIS3 provide an indication of the basal SMAD3 activity that contributes to cell migration in MDA-MB-231 cells. Together, these results indicate Nox4 via Smad3 is involved in TGF-beta-mediated cell migration but not cell proliferation.

Nox4 modulates TGF-beta-mediated fibronectin expression

It has been well established that TGF-beta induces fibronectin in migrating cells and in cells undergoing EMT [32]. Nox4 has been implicated in the modulation of matrix protein gene regulation including fibronectin and smooth muscle actin in cardiac fibroblasts, mesangial cells, and pulmonary fibroblasts [12, 3335]. Recently, Nox4 was shown to be involved in TGF-beta/SMAD3-dependent gene regulation of fibronectin and smooth muscle actin in human lung mesenchymal cells [35]. We investigated the involvement of Nox4 in breast epithelial cell migration by analyzing fibronectin expression in TGF-beta treated cells transfected with Nox4-DN or Nox4-shRNA. Compared to the vector control, Nox4-DN expression decreased TGF-beta-induced fibronectin mRNA in both MDA-MB-231 cells (left panel) and MCF10A cells (right panel) as indicated by quantitative real-time PCR (Fig. 6A). Similarly, shRNA-mediated knockdown of endogenous Nox4 significantly reduced TGF-beta-mediated fibronectin expression compared to control shRNA transfected cells (Fig. 6B) in both cell models. Finally, we examined whether Nox4 cDNA overexpression can induce fibronectin expression in the absence of TGF-beta treatments (Fig. 6C). The overexpressed Nox4 mRNA levels exceeded those of endogenous Nox4 induced by TGF-beta in both cell lines, however fibronectin mRNA levels were only marginally affected in the MCF10A cells. Collectively, our data indicate an important role for Nox4 in TGF-beta-induced breast epithelial cell migration and fibronectin expression, but that Nox4 overexpression itself does not appear to be sufficient to mediate these effects.

Figure 6
Nox4 modulates TGF-beta-induced fibronectin expression in breast epithelial cells


TGF-beta-induced oxidative stress has been implicated in regulating EMT progression, including gene expression, cytoskeletal rearrangement, and cell mobility. Here we show Nox4 contributes to TGF-beta-mediated cell migration and fibronectin expression of normal (MCF10A) and metastatic (MDA-MB-231) breast epithelial cells. We found TGF-beta induced NADPH oxidase-dependent superoxide release in both MCF10A and MDA-MB-231 breast epithelial cells (Fig. 1A and B). We show Nox4 mRNA and protein expression was substantially increased upon TGF-beta stimulation (Fig. 1C–E). Furthermore, expression of constitutively active TGF-beta receptor type-1 increased NADPH oxidase-dependent superoxide release, and Nox4 mRNA, protein, and promoter activity (Figs. 3 and and4A).4A). We further show that wild type and constitutively active SMAD3 significantly increase Nox4 promoter activity, whereas dominant-negative SMAD3 and SIS3, a SMAD3-specific inhibitor, decreased promoter activity (Fig. 4B and C). In contrast, TGF-beta-mediated superoxide was significantly reduced in MCF10A and MDA-MB-231 cells expressing a dominant-negative form of Nox4 or Nox4-specific shRNA (Fig. 2), indicating Nox4 is the primary source of superoxide generation. Moreover, wound closure and cell migration assays of epithelial cells expressing dominant-negative Nox4 or Nox4-shRNA were significantly reduced in response to TGF-beta treatment (Fig. 5A and B), indicating Nox4 is involved in TGF-beta regulation of breast epithelial cell migration. Inhibition of endogenous SMAD3 by transfection of a dominant-negative construct (Fig. 5D) or SIS3 treatment (Fig. 5E) also substantially reduced cell migration further suggesting SMAD3 is a key regulator of Nox4-mediated cell migration. Interestingly, dominant-negative Nox4 had no effect on TGF-beta-mediated cell proliferation (Fig. 5D), indicating TGF-beta modulation of Nox4 plays important roles in cell migration rather than cell proliferation. We also found TGF-beta up-regulation of fibronectin, an indicator of EMT progression, was reduced in dominant-negative Nox4 or Nox4 shRNA transfected breast epithelial cells (Fig. 6), suggesting Nox4 is involved in redox-dependent fibronectin gene regulation during EMT. Taken together, our results indicate Nox4 plays an important role in breast epithelial cell migration and fibronectin expression driven by a TGF-beta/SMAD3 signaling mechanism, as summarized in Figure 7.

Figure 7
Nox4 modulates TGF-beta/SMAD3-mediated breast epithelial cell migration

Previous reports indicate TGF-beta modulates Nox4TGF-beta in many cell types including vascular endothelial cells, kidney fibroblasts, hepatocytes, and lung epithelial cells [3638], however, little is know about TGF-beta mediated ROS generation in breast epithelial cells. Breast cancer epithelial cells are known to secrete high levels of soluble TGF-beta into the tumor microenvironment allowing for paracrine and autocrine TGF-beta signaling [39, 40]. While TGF-beta can have pro-apoptotic effects on specific cell types, MDA-MB-231 and MCF10A breast epithelial cells are resistant to TGF-beta-mediated growth inhibition [31, 39, 41]. Rather, breast epithelial cells respond to TGF-beta by increasing EMT gene expression and enhancing their migratory and metastatic potential. Tobar et al. recently reported Nox4-dependent ROS in migration of MCF-7 cells, a mildly invasive breast epithelial line [36]. They demonstrated that TGF-beta is a soluble factor secreted by RMF-EG mammary fibroblasts that increased Nox4 mRNA, ROS production and cell migration of co-cultured MCF-7 cells. Here we show TGF-beta-mediated cell migration is activated through a TGFBR1→ SMAD3→Nox4 signaling pathway in both normal and highly metastatic breast epithelial cells.

Nox4 has been implicated in EMT gene regulation and migration of other cell types. Bondi et al. reported TGF-beta induced Nox4-dependent modulation of EMT markers alpha-smooth muscle actin (SMA) and fibronectin during conversion of rat kidney fibroblast to a myofibroblast phenotype [42]. Nox4 was also reported to have a promigratory effect on angiotensin-II-treated vascular adventitial myofibroblasts [43]. Recently, Nam et al. described ROS generated from Nox1 and Nox4 involved in migration of human keratinocytes co-treated with HGF and TGF-beta [37], and Pendyala et al. indicated Nox4 is involved in hyperoxia-induced pulmonary artery endothelial cell migration [44]. Also, Meng et al. reported Nox4 and Rac1 mediate IGF-1-induced ROS production and cell migration in vascular smooth muscle cells [45]. They also demonstrated IGF-1-induced matrix metalloproteinase (MMP)-2 and 9 activities, indicative of EMT progression, were inhibited by knockdown of Nox4 or Rac1 inactivation through independent mechanisms.

Several ECM proteins are modulated under the control of TGF-beta including increased fibronectin expression. Fibronectin is important for directing cell attachment and migration during the cell reorganization and migratory process. Previous reports implicated Nox4 in cytoskeletal alterations and regulation of fibronectin expression. Under high glucose conditions, the diabetic rat kidney showed Nox4-derived ROS can mediate renal hypertrophy and fibronectin expression [33, 34]. Also, Hu et al. reported expression of dominant-negative Nox4 in HUVEC cells blocked the effect of TGF-beta on filipodia formation and F-actin assembly [13]. In vascular smooth muscle cells, p22phox association with poldip2 appears to increase Nox4 activity, which was correlated with cytoskeletal remodeling and focal adhesion turnover [46]. Our study not only demonstrates Nox4 is involved in TGF-beta-induced breast epithelial cell migration by wound healing and matrigel assays, but also provides evidence that Nox4 plays a role in EMT and cell mobility by modulating fibronectin expression.

The role of SMAD proteins, particularly SMAD3, during TGF-beta-induced EMT has been well established. Previous observations showed in SMAD3 knockout mice that EMT of lens epithelium after injury is blocked in the absence of SMAD3, along with decreased expression of EMT markers including, snail, SMA, and collagen [47]. Another study using selective knockdown of SMAD2 and SMAD3 demonstrated that TGF-β–induced EMT is SMAD3 dependent in lung cancer cells [48]. Moreover, Sturrock et al. showed TGF-beta-induced Nox4 mRNA and ROS were decreased in pulmonary smooth muscle cells overexpressing either dominant-negative SMAD2 or SMAD3 [49]. By comparing constitutively active SMAD3 (C-terminal pseudo-phosphorylated) to dominant-negative SMAD3 (deletion of the MH2 domain), we are the first to demonstrate that phosphorylation of the C-terminus of SMAD3 positively regulates the human Nox4 promoter in breast epithelial cells. While we cannot rule out involvement of other transcriptions factors in response to TGF-beta, our findings provide an important foundation for identifying SMAD binding elements (SBE) and other key components involved in TGF-beta regulation of the Nox4 promoter.

The precise mechanisms of Nox4-derived ROS in TGF-beta-mediated cell migration and EMT progression of breast epithelial cells warrants further investigation. It is possible that Nox4-dependent ROS alter the activation states of signaling components or affect ECM proteins involved in EMT processes by direct protein oxidation. ROS have been shown to modulate cell signaling by altering oxidation states of proteins such as tyrosine phosphatases (PTP1B, SHP2, PTEN, and LMW-PTP), redox sensitive transcription factors (NFκ-B, p53, and AP-1), receptor and non-receptor tyrosine kinases (EGFR, PDGFR, and Src), and cytoskeletal proteins (beta-actin or associated proteins) [5053]. Moreover, a previous study reporting a proteomic screen for cysteine glutathionylated proteins revealed several cytoskeletal proteins including actin, cofilin, vimentin, and profilin are redox sensitive[54]. Nox4 subcellular localization has been described in many places including the plasma membrane, ER, perinuclear, and nuclear compartments. Nox4 was recently reported to co-localization with F-actin at sites of invadopodia; and Nox4 knockdown resulted in reduced invadopodia formation in Src transformed 3T3 cells [55]. Determination of TGF-beta-induced Nox4 localization will provide further insight into Nox4-dependent redox mechanisms involved in EMT progression and cell motility.


  • -
    We show that Nox4 in involved in the epithelial-to-mesenchymal transition.
  • -
    TGF-beta induces Nox4-dependent ROS generation and migration of epithelial cells.
  • -
    Inhibition of Nox4 or SMAD3 attenuates TGF-beta-mediated cell migration.
  • -
    Inhibition of SMAD3 significantly reduces TGF-beta-induced Nox4 promoter activity.
  • -
    Nox4 is involved in TGF-beta-induced fibronectin expression.


This work was supported by funds from the Intramural Research Program of the NIH, National Institute of Allergy and Infectious Diseases (ZO1-AI-000614).


Reactive oxygen species
Transforming growth factor-beta
Epithelial-to-mesenchymal transition
Extracellular matrix


Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.


The authors declared no conflict of interest.


1. Xu J, Lamouille S, Derynck R. TGF-beta-induced epithelial to mesenchymal transition. Cell Res. 2009;19:156–172. [PubMed]
2. Cannito S, Novo E, di Bonzo LV, Busletta C, Colombatto S, Parola M. Epithelial-mesenchymal transition: from molecular mechanisms, redox regulation to implications in human health and disease. Antioxidants & redox signaling. 12:1383–1430. [PubMed]
3. Mori K, Shibanuma M, Nose K. Invasive potential induced under long-term oxidative stress in mammary epithelial cells. Cancer Res. 2004;64:7464–7472. [PubMed]
4. Jiang F, Zhang Y, Dusting GJ. NADPH Oxidase-Mediated Redox Signaling: Roles in Cellular Stress Response, Stress Tolerance, and Tissue Repair. Pharmacol Rev [PubMed]
5. Brown DI, Griendling KK. Nox proteins in signal transduction. Free Radic Biol Med. 2009;47:1239–1253. [PMC free article] [PubMed]
6. Kwon J, Lee SR, Yang KS, Ahn Y, Kim YJ, Stadtman ER, Rhee SG. Reversible oxidation and inactivation of the tumor suppressor PTEN in cells stimulated with peptide growth factors. Proc Natl Acad Sci U S A. 2004;101:16419–16424. [PubMed]
7. Liu H, Colavitti R, Rovira II, Finkel T. Redox-dependent transcriptional regulation. Circ Res. 2005;97:967–974. [PubMed]
8. Gordillo G, Fang H, Park H, Roy S. Nox-4-dependent nuclear H2O2 drives DNA oxidation resulting in 8-OHdG as urinary biomarker and hemangioendothelioma formation. Antioxidants & redox signaling. 12:933–943. [PMC free article] [PubMed]
9. Chen K, Kirber MT, Xiao H, Yang Y, Keaney JF., Jr Regulation of ROS signal transduction by NADPH oxidase 4 localization. J Cell Biol. 2008;181:1129–1139. [PMC free article] [PubMed]
10. Bedard K, Krause KH. The NOX family of ROS-generating NADPH oxidases: physiology and pathophysiology. Physiol Rev. 2007;87:245–313. [PubMed]
11. Carnesecchi S, Deffert C, Donati Y, Basset O, Hinz B, Preynat-Seauve O, Guichard C, Arbiser JL, Banfi B, Pache JC, Barazzone-Argiroffo C, Krause KH. A key role for NOX4 in epithelial cell death during development of lung fibrosis. Antioxidants & redox signaling. 2011;15:607–619. [PMC free article] [PubMed]
12. Amara N, Goven D, Prost F, Muloway R, Crestani B, Boczkowski J. NOX4/NADPH oxidase expression is increased in pulmonary fibroblasts from patients with idiopathic pulmonary fibrosis and mediates TGFbeta1-induced fibroblast differentiation into myofibroblasts. Thorax. 2010;65:733–738. [PMC free article] [PubMed]
13. Hu T, Ramachandrarao SP, Siva S, Valancius C, Zhu Y, Mahadev K, Toh I, Goldstein BJ, Woolkalis M, Sharma K. Reactive oxygen species production via NADPH oxidase mediates TGF-beta-induced cytoskeletal alterations in endothelial cells. Am J Physiol Renal Physiol. 2005;289:F816–F825. [PMC free article] [PubMed]
14. Mandal CC, Ganapathy S, Gorin Y, Mahadev K, Block K, Abboud HE, Harris SE, Ghosh-Choudhury G, Ghosh-Choudhury N. Reactive oxygen species derived from Nox4 mediate BMP2 gene transcription and osteoblast differentiation. Biochem J. 2011;433:393–402. [PubMed]
15. Li S, Tabar SS, Malec V, Eul BG, Klepetko W, Weissmann N, Grimminger F, Seeger W, Rose F, Hanze J. NOX4 regulates ROS levels under normoxic and hypoxic conditions, triggers proliferation, and inhibits apoptosis in pulmonary artery adventitial fibroblasts. Antioxidants & redox signaling. 2008;10:1687–1698. [PubMed]
16. Boudreau HE, Emerson SU, Korzeniowska A, Jendrysik MA, Leto TL. Hepatitis C virus (HCV) proteins induce NADPH oxidase 4 expression in a transforming growth factor beta-dependent manner: a new contributor to HCV-induced oxidative stress. Journal of virology. 2009;83:12934–12946. [PMC free article] [PubMed]
17. Geiszt M, Kopp JB, Varnai P, Leto TL. Identification of renox, an NAD(P)H oxidase in kidney. Proc Natl Acad Sci U S A. 2000;97:8010–8014. [PubMed]
18. Geiszt M, Leto TL. The Nox family of NAD(P)H oxidases: host defense and beyond. J Biol Chem. 2004;279:51715–51718. [PubMed]
19. Sancho P, Bertran E, Caja L, Carmona-Cuenca I, Murillo MM, Fabregat I. The inhibition of the epidermal growth factor (EGF) pathway enhances TGF-beta-induced apoptosis in rat hepatoma cells through inducing oxidative stress coincident with a change in the expression pattern of the NADPH oxidases (NOX) isoforms. Biochim Biophys Acta. 2009;1793:253–263. [PubMed]
20. Cheng G, Cao Z, Xu X, van Meir EG, Lambeth JD. Homologs of gp91phox: cloning and tissue expression of Nox3, Nox4, and Nox5. Gene. 2001;269:131–140. [PubMed]
21. Yang S, Zhang Y, Ries W, Key L. Expression of Nox4 in osteoclasts. Journal of cellular biochemistry. 2004;92:238–248. [PubMed]
22. Kuroda J, Ago T, Matsushima S, Zhai P, Schneider MD, Sadoshima J. NADPH oxidase 4 (Nox4) is a major source of oxidative stress in the failing heart. Proc Natl Acad Sci U S A. 2010;107:15565–15570. [PubMed]
23. Hilenski LL, Clempus RE, Quinn MT, Lambeth JD, Griendling KK. Distinct subcellular localizations of Nox1 and Nox4 in vascular smooth muscle cells. Arteriosclerosis, thrombosis, and vascular biology. 2004;24:677–683. [PubMed]
24. Takac I, Schroder K, Zhang L, Lardy B, Anilkumar N, Lambeth JD, Shah AM, Morel F, Brandes RP. The E-loop is involved in hydrogen peroxide formation by the NADPH oxidase Nox4. J Biol Chem. 2011;286:13304–13313. [PubMed]
25. Block K, Gorin Y, Abboud HE. Subcellular localization of Nox4 and regulation in diabetes. Proc Natl Acad Sci U S A. 2009;106:14385–14390. [PubMed]
26. Kuroda J, Nakagawa K, Yamasaki T, Nakamura K, Takeya R, Kuribayashi F, Imajoh-Ohmi S, Igarashi K, Shibata Y, Sueishi K, Sumimoto H. The superoxide-producing NAD(P)H oxidase Nox4 in the nucleus of human vascular endothelial cells. Genes to cells : devoted to molecular & cellular mechanisms. 2005;10:1139–1151. [PubMed]
27. Weyemi U, Lagente-Chevallier O, Boufraqech M, Prenois F, Courtin F, Caillou B, Talbot M, Dardalhon M, Al Ghuzlan A, Bidart JM, Schlumberger M, Dupuy C. ROS-generating NADPH oxidase NOX4 is a critical mediator in oncogenic H-Ras-induced DNA damage and subsequent senescence. Oncogene. 2011 [PMC free article] [PubMed]
28. Juhasz A, Ge Y, Markel S, Chiu A, Matsumoto L, van Balgooy J, Roy K, Doroshow JH. Expression of NADPH oxidase homologues and accessory genes in human cancer cell lines, tumours and adjacent normal tissues. Free radical research. 2009;43:523–532. [PMC free article] [PubMed]
29. Graham KA, Kulawiec M, Owens KM, Li X, Desouki MM, Chandra D, Singh KK. NADPH oxidase 4 is an oncoprotein localized to mitochondria. Cancer biology & therapy. 2010;10:223–231. [PMC free article] [PubMed]
30. Chuaire-Noack L, Sanchez-Corredor MC, Ramirez-Clavijo SR. The Dual Role of Senescence in Tumorigenesis. Int J Morphol. 2010;28:37–50.
31. Muraoka-Cook RS, Shin I, Yi JY, Easterly E, Barcellos-Hoff MH, Yingling JM, Zent R, Arteaga CL. Activated type I TGFbeta receptor kinase enhances the survival of mammary epithelial cells and accelerates tumor progression. Oncogene. 2006;25:3408–3423. [PubMed]
32. Thiery JP. Epithelial-mesenchymal transitions in tumour progression. Nat Rev Cancer. 2002;2:442–454. [PubMed]
33. Gorin Y, Block K, Hernandez J, Bhandari B, Wagner B, Barnes JL, Abboud HE. Nox4 NAD(P)H oxidase mediates hypertrophy and fibronectin expression in the diabetic kidney. J Biol Chem. 2005;280:39616–39626. [PubMed]
34. Block K, Eid A, Griendling KK, Lee DY, Wittrant Y, Gorin Y. Nox4 NAD(P)H oxidase mediates Src-dependent tyrosine phosphorylation of PDK-1 in response to angiotensin II: role in mesangial cell hypertrophy and fibronectin expression. J Biol Chem. 2008;283:24061–24076. [PubMed]
35. Hecker L, Vittal R, Jones T, Jagirdar R, Luckhardt TR, Horowitz JC, Pennathur S, Martinez FJ, Thannickal VJ. NADPH oxidase-4 mediates myofibroblast activation and fibrogenic responses to lung injury. Nat Med. 2009;15:1077–1081. [PMC free article] [PubMed]
36. Tobar N, Guerrero J, Smith PC, Martinez J. NOX4-dependent ROS production by stromal mammary cells modulates epithelial MCF-7 cell migration. British journal of cancer. 2010;103:1040–1047. [PMC free article] [PubMed]
37. Nam HJ, Park YY, Yoon G, Cho H, Lee JH. Co-treatment with hepatocyte growth factor and TGF-beta1 enhances migration of HaCaT cells through NADPH oxidase-dependent ROS generation. Exp Mol Med. 2010;42:270–279. [PMC free article] [PubMed]
38. Caja L, Sancho P, Bertran E, Iglesias-Serret D, Gil J, Fabregat I. Overactivation of the MEK/ERK pathway in liver tumor cells confers resistance to TGF-{beta}-induced cell death through impairing up-regulation of the NADPH oxidase NOX4. Cancer Res. 2009;69:7595–7602. [PubMed]
39. Lei X, Bandyopadhyay A, Le T, Sun L. Autocrine TGFbeta supports growth and survival of human breast cancer MDA-MB-231 cells. Oncogene. 2002;21:7514–7523. [PubMed]
40. Tobin SW, Douville K, Benbow U, Brinckerhoff CE, Memoli VA, Arrick BA. Consequences of altered TGF-beta expression and responsiveness in breast cancer: evidence for autocrine and paracrine effects. Oncogene. 2002;21:108–118. [PubMed]
41. Tang B, Vu M, Booker T, Santner SJ, Miller FR, Anver MR, Wakefield LM. TGF-beta switches from tumor suppressor to prometastatic factor in a model of breast cancer progression. The Journal of clinical investigation. 2003;112:1116–1124. [PMC free article] [PubMed]
42. Bondi CD, Manickam N, Lee DY, Block K, Gorin Y, Abboud HE, Barnes JL. NAD(P)H oxidase mediates TGF-beta1-induced activation of kidney myofibroblasts. Journal of the American Society of Nephrology : JASN. 2010;21:93–102. [PubMed]
43. Haurani MJ, Cifuentes ME, Shepard AD, Pagano PJ. Nox4 oxidase overexpression specifically decreases endogenous Nox4 mRNA and inhibits angiotensin II-induced adventitial myofibroblast migration. Hypertension. 2008;52:143–149. [PubMed]
44. Pendyala S, Gorshkova IA, Usatyuk PV, He D, Pennathur A, Lambeth JD, Thannickal VJ, Natarajan V. Role of Nox4 and Nox2 in hyperoxia-induced reactive oxygen species generation and migration of human lung endothelial cells. Antioxidants & redox signaling. 2009;11:747–764. [PMC free article] [PubMed]
45. Meng D, Lv DD, Fang J. Insulin-like growth factor-I induces reactive oxygen species production and cell migration through Nox4 and Rac1 in vascular smooth muscle cells. Cardiovascular research. 2008;80:299–308. [PubMed]
46. Lyle AN, Deshpande NN, Taniyama Y, Seidel-Rogol B, Pounkova L, Du P, Papaharalambus C, Lassegue B, Griendling KK. Poldip2, a novel regulator of Nox4 and cytoskeletal integrity in vascular smooth muscle cells. Circ Res. 2009;105:249–259. [PMC free article] [PubMed]
47. Saika S, Kono-Saika S, Ohnishi Y, Sato M, Muragaki Y, Ooshima A, Flanders KC, Yoo J, Anzano M, Liu CY, Kao WW, Roberts AB. Smad3 signaling is required for epithelial-mesenchymal transition of lens epithelium after injury. The American journal of pathology. 2004;164:651–663. [PubMed]
48. Reka AK, Kurapati H, Narala VR, Bommer G, Chen J, Standiford TJ, Keshamouni VG. Peroxisome proliferator-activated receptor-gamma activation inhibits tumor metastasis by antagonizing Smad3-mediated epithelial-mesenchymal transition. Molecular cancer therapeutics. 2010;9:3221–3232. [PMC free article] [PubMed]
49. Sturrock A, Cahill B, Norman K, Huecksteadt TP, Hill K, Sanders K, Karwande SV, Stringham JC, Bull DA, Gleich M, Kennedy TP, Hoidal JR. Transforming growth factor-beta1 induces Nox4 NAD(P)H oxidase and reactive oxygen species-dependent proliferation in human pulmonary artery smooth muscle cells. American journal of physiology. Lung cellular and molecular physiology. 2006;290:L661–L673. [PubMed]
50. Chiarugi P, Cirri P. Redox regulation of protein tyrosine phosphatases during receptor tyrosine kinase signal transduction. Trends in biochemical sciences. 2003;28:509–514. [PubMed]
51. Fiaschi T, Cozzi G, Raugei G, Formigli L, Ramponi G, Chiarugi P. Redox regulation of beta-actin during integrin-mediated cell adhesion. J Biol Chem. 2006;281:22983–22991. [PubMed]
52. Hsu TC, Young MR, Cmarik J, Colburn NH. Activator protein 1 (AP-1)-and nuclear factor kappaB (NF-kappaB)-dependent transcriptional events in carcinogenesis. Free Radic Biol Med. 2000;28:1338–1348. [PubMed]
53. Giannoni E, Buricchi F, Raugei G, Ramponi G, Chiarugi P. Intracellular reactive oxygen species activate Src tyrosine kinase during cell adhesion and anchorage-dependent cell growth. Mol Cell Biol. 2005;25:6391–6403. [PMC free article] [PubMed]
54. Fratelli M, Demol H, Puype M, Casagrande S, Eberini I, Salmona M, Bonetto V, Mengozzi M, Duffieux F, Miclet E, Bachi A, Vandekerckhove J, Gianazza E, Ghezzi P. Identification by redox proteomics of glutathionylated proteins in oxidatively stressed human T lymphocytes. Proc Natl Acad Sci U S A. 2002;99:3505–3510. [PubMed]
55. Diaz B, Shani G, Pass I, Anderson D, Quintavalle M, Courtneidge SA. Tks5-dependent, nox-mediated generation of reactive oxygen species is necessary for invadopodia formation. Science signaling. 2009;2:ra53. [PMC free article] [PubMed]