Abstract
Aim: This study investigated the effect of reactive oxygen species (ROS) on transforming growth factor (TGF)-β-mediated epithelial-to-mesenchymal transition (EMT) in order to clarify the influence of ROS and TGFβ on the induction of dysplasia and ultimately, tumorigenesis. Materials and Methods: Confluent MCF-10A human mammary epithelial cells were treated with H2O2 for 1 h, then reseeded at low density in the presence of TGFβ and cultured until confluence. Results: Hydrogen peroxide (H2O2, 250 μM) enhanced TGFβ-mediated EMT, as evidenced by increased expression of EMT-associated transcription factors, which was accompanied by increased nuclear localization of phosphorylated SMAD family member 2 (SMAD2) and up-regulation of the TGFβ signaling pathway components mitogen-activated protein kinase kinase (MEK)/extracellular signal-regulated kinase (ERK). Pharmacological inhibition of MEK/ERK signaling partly reversed the effects of H2O2. Conclusion: H2O2 enhances TGFβ-mediated EMT via SMAD and MEK/ERK signaling.
Epithelial-to-mesenchymal transition (EMT) is involved in many biological processes, including embryogenesis, tissue fibrosis and cancer metastasis. EMT is characterized by loss of cell polarity and adhesion, resulting in changes in cell morphology and increased cell migration and size (1, 2). Activation of EMT-inducing transcription factors including SNAIL family transcriptional repressor 1 (SNAIL1), SNAIL2, and TWIST family bHLH transcription factor 1 (TWIST), among others, is thought to promote tumor initiation via negative regulation of the tumor suppressor p53 (3). Transgenic mice overexpressing resistance to audiogenic seizures (Ras) oncogene alone in mammary glands do not develop mammary carcinoma, while mice overexpressing Twist1 in addition to Ras are susceptible to cancer (4). In addition, some EMT features, including loss of E-cadherin expression, are observed in some types of epithelial dysplasias (5). Expression of EMT-inducing transcription factors may also confer tumor cells with an advantage under conditions of stress by enabling oncogene-induced cell senescence and escape from apoptosis (6, 7).
Reactive oxygen species (ROS) are generated by normal cellular metabolism (mitochondrial electron transport chain) and in response to cytokines and environmental factors (ultraviolet or ionizing radiation); an increase in intracellular ROS production can lead to oxidative stress (8), which underlies degenerative diseases such as obesity, arteriosclerosis, rheumatoid arthritis, and aging. Increased ROS levels have been observed in many types of tumor cells, suggesting that they contribute to tumorigenesis and tumor metastasis (9, 10). Mesothelioma is caused by inhalation of asbestos fibers, which results in higher ROS levels and EMT induction (11, 12). These observations suggest that tumorigenesis is associated with the interaction of ROS and EMT.
Transforming growth factor (TGF)-β is a cytokine that plays an important role in cell growth inhibition, cell migration, tumor metastasis, extracellular matrix remodeling, and immunosuppression. The biological function of TGFβ is context-dependent; for instance, it inhibits and promotes tumor cell growth at early and late stages of tumorigenesis, respectively (13). TGFβ is activated by oxidative stress in various tissues, including mammary gland (14).
Primer pairs for quantitative reverse transcription-polymerase chain reaction.
We speculated that increased ROS and TGFβ cooperatively induce EMT to cause dysplasia and, ultimately, tumorigenesis. In order to test this hypothesis, the present study investigated the effect of hydrogen peroxide (H2O2), an ROS, on TGFβ-mediated EMT and molecular mechanisms underlying the event, as ionizing radiation has been shown to enhance TGFβ-mediated EMT in human mammary epithelial cells (15, 16).
Materials and Methods
Cell culture. Immortalized MCF-10A human mammary epithelial cells were purchased from the American Type Culture Collection (Manassas, VA, USA) and maintained as previously described (15). Confluent cells were treated with 50 or 250 μM H2O2 in culture medium for 1 h or with γ radiation (1 Gy; Gammacell, Best Theratronics, Ottawa, Canada) at a dose-rate of 0.9 Gy/min. The cells were then detached with a mixture of 0.05% trypsin and cell dissociation solution (Sigma-Aldrich, St. Louis, MO, USA) and counted, then sparsely reseeded with 400 pg/ml human recombinant TGFβ1 (R&D Systems, Minneapolis, MN, USA), except for immunofluorescence analysis (100 pg/ml). Cells were cultured until confluence (about 6 days); the medium containing TGFβ1 was changed every 48 h. To evaluate the effect of inhibitor on expression of E-cadherin, cells were treated with 10 μM of the MEK inhibitor U0126 (Cell Signaling Technology, Danvers, MA, USA).
Immunofluorescence analysis. Immunocytochemistry was carried out as previously described (15). Briefly, cells were grown on a 96-well plate (ibidi GmbH, Martinsried, Germany), and then fixed with ice-cold methanol. The cells were then labeled with antibodies against E-cadherin (610181; BD Biosciences, San Jose, CA, USA), N-cadherin (610921; BD Biosciences), phospho-(p-) SMAD family member 2 (SMAD2) (#3108; Cell Signaling Technology), or p-p44/42 MAPK (extracellular signal-regulated kinase, ERK1/2) (#4370; Cell signaling Technology) followed by Alexa 488-conjugated anti-mouse or -rabbit secondary antibody (Thermo Fisher Scientific, San Jose, CA, USA). Nuclei were counterstained with Hoechst33258 (Thermo Fisher Scientific). Images were acquired using the IN Cell Analyzer 2000 automated fluorescence microscope (GE Healthcare, Indianapolis, IN, USA) or Opera Phoenix (Perkin Elmer, Waltham, MA, USA).
Western blotting. Western blotting was performed as previously described (17). Cells were lysed using Complete Lysis-M, EDTA-free (Roche diagnostics, Mannheim, Germany) according to manufacturer's instructions. Primary antibodies were listed as follows: mitogen-activated protein kinase kinase (MEK1/2) (#4694), p-MEK1/2 (#9154), p38 MAPK (#8690), p-p38 MAPK (#4511), ERK1/2 (#4695), p-ERK1/2, stress-activated protein kinase (SAPK)/JNK (#9252), and p-SAPK/JNK (#4668) (all from Cell Signaling Technology). β-Actin (A5316; Sigma-Aldrich) was used as a loading control.
RNA purification and quantitative reverse transcription (qRT)-polymerase chain reaction (PCR). RNA extraction and qRT-PCR were carried out as previously described (18) using the SYBR Green system (Thunderbird SYBR qPCR Mix; TOYOBO, Osaka, Japan) and specific primers (Table I) (4). Glyceraldehyde 3-phosphate dehydrogenase expression was used as an internal standard. Relative gene expression levels were calculated with the 2−ΔΔCT method (19).
Statistical analysis. Results are presented as mean±SD of at least triplicate measurements of one representative from several independent experiments. Differences between groups were evaluated with Welch's t-test, and a value of p<0.05 was considered statistically significant.
Results
H2O2 enhances TGFβ-mediated EMT. We investigated the effect of H2O2 on TGFβ-mediated EMT. Confluent MCF-10A cells were treated with H2O2 for 1 h, and then trypsinized and replated at a low density in the presence or absence of TGFβ. The highest concentration of TGFβ was 400 pg/ml, which is not sufficient to induce EMT (15). Immunofluorescence analysis revealed that E- and N-cadherin, markers for epithelial and mesenchymal phenotypes, respectively, were down- and up-regulated, respectively, after H2O2 treatment in the presence of TGFβ (Figure 1A and B), although TGFβ alone down-regulated E-cadherin expression to some extent. These changes were induced by H2O2 at concentrations greater than 50 μM, as determined using a high-throughput image analyzer (Figure 1C and D). E-Cadherin transcript levels were similar to the protein levels in each treatment group (Figure 1E).
Immunofluorescence analysis of E-cadherin and N-cadherin expression at 144 h after H2O2 treatment. Representative images of E-cadherin (A) and N-cadherin (B) (green) are shown. Nuclei (blue) are stained with Hoechst33258. Bar, 50 μm. Quantification of E-cadherin (C) and N-cadherin (D) intensity. For each sample, 98 fields from the same area of each well were analyzed and average intensity of E- or N-cadherin per cell was determined using IN Cell Developer Toolbox software (GE Healthcare BioScience). *p<0.05, **p<0.01 vs. corresponding negative control (−) of in the presence or absence of transforming growth factor β (TGFβ). E: Quantitative reverse transcription polymerase chain reaction analysis of E-cadherin gene expression. Expression level was calculated relative to control cells at 144 h of triplicate measurements from one representative of three independent experiments. *p<0.05 vs. corresponding negative control (−) of in the presence or absence of TGFβ.
Up-regulation of EMT-inducing transcription factors contributes to the enhancement of TGFβ-mediated EMT by H2O2. We investigated the molecular basis for enhancement of TGFβ-mediated EMT by H2O2. Several transcription factors are known to inhibit E-cadherin expression and subsequently stimulate that of N-cadherin (20). Here we observed up-regulation of SNAIL1, SNAIL2, zinc finger E-box-binding homeobox 1 (ZEB1) and ZEB2 transcripts 48 h after application of TGFβ irrespective of H2O2 treatment or irradiation (data not shown). SNAIL1, SNAIL2 and ZEB1 levels were significantly increased in TGFβ-treated cells 96 h after application of 250 μM H2O2 (Figure 2), while there was little change in the expression levels in cells that were not exposed to TGFβ. Similar up-regulation was observed upon γ-radiation treatment, consistent with previous reports of increased ZEB1 levels following irradiation (15).
Quantitative reverse transcription polymerase chain reaction analysis of SNAIL family transcriptional repressor 1 (SNAIL1), SNAIL2, zinc finger E-box-binding homeobox 1 (ZEB1) and ZEB2 expression at 96 h after H2O2 or radiation treatment (IR). Expression levels were calculated relative to control cells of triplicate measurements from one representative of three independent experiments. *p<0.05, **p<0.01 vs. corresponding negative control (−) of in the presence or absence of transforming growth factor β (TGFβ).
Nuclear localization of p-SMAD2 and MEK/ERK phosphorylation are involved in the enhancement of TGFβ-mediated EMT by H2O2. Finally, we investigated upstream signaling molecules of these transcription factors (SNAIL1, SNAIL2 and ZEB1). Activation of SMADs and subsequent up-regulation of SNAIL1, SNAIL2 and ZEB1 expression is important for TGFβ-induced EMT signaling (13, 20). In this study, immunofluorescence analysis revealed an increase in nuclear localization of p-SMAD2 associated with the enhancement of TGFβ-mediated EMT at 72 h after H2O2 or radiation treatment (Figure 3).
MEK/ERK, JNK, and p38 are critical components of non-canonical, SMAD-independent TGFβ signaling (13). Of these, ERK1/2 is also known to regulate SNAIL1, SNAIL2 and ZEB1 expression (21, 22). MEK/ERK phosphorylation was enhanced by both H2O2 and irradiation in TGFβ-treated cells (Figure 4), as previously reported (15). On the other hand, both JNK and p38 phosphorylation were unaltered (Figure 4). Moreover, immunofluorescence analysis revealed an increase in p-ERK1/2 expression associated with the enhancement of TGFβ-mediated EMT at 72 h after H2O2 or radiation treatment (Figure 5). The down-regulation of E-cadherin upon H2O2 -induced enhancement of TGFβ-mediated EMT was partly but significantly abrogated by treatment with the MEK inhibitor U0126 (Figure 6A and B), which also blocked ERK1/2 phosphorylation (Figure 6C). These results indicate that both SMAD and MEK/ERK pathways are critical for the enhancement of TGFβ-mediated EMT by H2O2.
Immunofluorescence analysis of phospho-SMAD family member 2 (p-SMAD2) expression at 72 h after H2O2 treatment or irradiation (IR). Representative images of p-SMAD2 expression (green) are shown. Nuclei (blue) are stained with Hoechst33258. Bar, 50 μm. Quantification of nuclear p-SMAD2 intensity. A total of 81 fields from the same area of each well was analyzed for each sample and average intensity of nuclear p-SMAD2 per cell were determined using the Harmony image-analysis software (Perkin Elmer). **p<0.01 vs. corresponding negative control (−) of in the presence or absence of transforming growth factor β (TGFβ).
Discussion
The results presented here demonstrate that TGFβ-mediated EMT in MCF-10A cells was enhanced not only by radiation exposure, which has been previously established, but also by H2O2 treatment at concentrations greater than 50 μM. This was likely associated with the observed up-regulation of SNAIL1, SNAIL2 and ZEB1, which encode EMT-inducing transcription factors. The induction of EMT was accompanied by increased nuclear localization of phosphorylated SMAD2, as well as MEK/ERK activation, which is important for non-canonical, SMAD-independent TGFβ signaling. These results indicate that the enhancement of TGFβ-mediated EMT by H2O2 involves SMAD2 and MEK/ERK activation followed by upregulation of SNAIL1, SNAIL2, and ZEB1.
The binding of TGFβ to its receptor and consequent activation of SMAD2/3, which is phosphorylated and accumulated in the nucleus, initiates EMT (13, 23, 24). It was reported that SMAD2/3 phosphorylation increased in cells exposed to radiation in the presence of TGFβ (15). In the present study, we observed a significant increase in the nuclear localization of phosphorylated SMAD2 (Figure 3). In addition, the TGFβ–SMAD pathway is correlated with other signaling molecules including MEK/ERK, JNK or p38 (25). In fact, MEK/ERK inhibition partly abrogated the enhancement of TGFβ-mediated EMT by H2O2, suggesting that both SMAD2 and MEK/ERK are important.
A: Western blot analysis of mitogen-activated protein kinase kinase (MEK), phospho-(p-) MEK, extracellular signal-regulated kinase (ERK1/2), p-ERK1/2, stress-activated protein kinase (SAPK)/JUN N-terminal kinase (JNK), p-JNK, p38 and p-p38 expression in MCF-10A cells 144 h after H2O2 treatment or irradiation (IR). The level of actin in each sample was used as an internal control. B: Densitometric analysis of the ratio of p-MEK to MEK and p-ERK1/2 to ERK1/2 of triplicate measurements from one representative of three independent experiments. *p<0.05 vs. negative control (−).
It is unclear how H2O2 activates SMAD and MEK/ERK signaling to enhance TGFβ-mediated EMT. However, SMAD activation might be caused via TGFβ, as TGFβ is induced by oxidative stress. Epidermal and platelet-derived growth factor receptors, which act upstream of MEK/ERK, are activated by H2O2 treatment in a ligand-independent manner (26, 27), while SRC and RAS are also required for ERK1/2 activation induced by H2O2 (28). These are possible mechanisms of action that require further investigation.
Our results showed that MEK/ERK signaling is important for the enhancement of TGFβ-mediated EMT by H2O2, in agreement with previous observations that the MAPK pathway is activated by TGFβ via ROS in quiescent renal tubular epithelial cells (29). Most studies have evaluated EMT at relatively early time points and thus report the direct response of cells to oxidative stress. In contrast, our study evaluated the effects of oxidative stress in the daughter cells of those exposed to H2O2; the results suggest that the changes induced by H2O2 are transmitted to subsequent generations, which may ultimately contribute to tumorigenesis.
Immunofluorescence analysis of phospho-extracellular signal-regulated kinase (p-ERK1/2) expression 72 h after H2O2 treatment or irradiation (IR). Representative images of p-ERK1/2 expression (green) are shown. Nuclei (blue) are stained with Hoechst33258. Bar, 50 μm. B: Quantification of p-ERK1/2 intensity. A total 81 fields from the same area of each well were analyzed for each sample and average intensity of p-ERK1/2 per cell were determined using Harmony image-analysis software. **p<0.01 vs. corresponding negative control (−) of in the presence or absence of transforming growth factor β (TGFβ).
In summary, H2O2 was found to enhance TGFβ-mediated EMT in MCF-10A cells, and this involved both SMAD and MEK/ERK signaling. Our finding that oxidative stress can enhance EMT in cooperation with TGFβ suggests that these events (i.e. activation of EMT/EMT-inducing transcription factors) are involved in the induction of many types of cancer, and provides a basis for the development of therapeutic strategies for cancer that target ROS as well as TGFβ.
A: Representative images of E-cadherin expression (green) in cells treated with 10 μM U0126 for 24 h after 144 h of incubation with transforming growth factor β (TGFβ). Nuclei (blue) are stained with Hoechst33258. Bar, 50 μm. B: Quantification of E-cadherin intensity. Thirty-three sections from the same area of each well was analyzed for each sample and average E-cadherin intensity per cell was determined using Harmony image-analysis software. *p<0.05, **p<0.01. C: Western blot analysis of extracellular signal-regulated kinase (ERK1/2) and phospho-(p-)ERK1/2 expression in MCF-10A cells treated with 10 μM U0126 for 8 h after 120 h of incubation with TGFβ. IR: Radiation treatment.
Acknowledgements
The Authors thank Dr. T. Imaoka (National Institutes for Quantum and Radiological Science and Technology) for useful comments; Ms. J. Takemoto for technical assistance; and the staff at the Radiation Research Center for Frontier Science Facilities, Hiroshima University, for managing the radiation sources. This work was supported in part by Grants-in-Aid for both Challenging Exploratory Research and Scientific Research (C) from the Ministry of Education, Culture, Sports, Science, and Technology of Japan (no. 15K12207 to K.K. and no. 26340024 to D.I.).
Footnotes
This article is freely accessible online.
- Received January 26, 2017.
- Revision received February 24, 2017.
- Accepted February 28, 2017.
- Copyright© 2017, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved
