Hypoxia-induced modulation of PTEN activity and EMT phenotypes in lung cancers
Persistent hypoxia stimulation, one of the most critical microenvironmental factors, accelerates the acquisition of epithelial–mesenchymal transition (EMT) phenotypes in lung cancer cells. Loss of phosphatase and tensin homologue deleted from chromosome 10 (PTEN) expression might accelerate the development of lung cancer in vivo. Recent studies suggest that tumor microenvironmental factors might modulate the PTEN activity though a decrease in total PTEN expression and an increase in phosphorylation of the PTEN C-terminus (p-PTEN), resulting in the acquisition of the EMT phenotypes. Nevertheless, it is not known whether persistent hypoxia can modulate PTEN phosphatase activity or whether hypoxia-induced EMT phenotypes are negatively regulated by the PTEN phosphatase activity. We aimed to investigate hypoxia-induced modulation of PTEN activity and EMT phenotypes in lung cancers.
Western blotting was performed in five lung cancer cell lines to evaluate total PTEN expression levels and the PTEN activation. In a xenograft model of lung cancer cells with endogenous PTEN expression, the PTEN expression was evaluated by immunohistochemistry. To examine the effect of hypoxia on phenotypic alterations in lung cancer cells in vitro, the cells were cultured under hypoxia. The effect of unphosphorylated PTEN (PTEN4A) induction on hypoxia-induced EMT phenotypes was evaluated, by using a Dox-dependent gene expression system.
Lung cancer cells involving the EMT phenotypes showed a decrease in total PTEN expression and an increase in p-PTEN. In a xenograft model, loss of PTEN expression was observed in the tumor lesions showing tissue hypoxia. Persistent hypoxia yielded an approximately eight-fold increase in the p-PTEN/PTEN ratio in vitro. PTEN4A did not affect stabilization of hypoxia-inducible factor 1α. PTEN4A blunted hypoxia-induced EMT via inhibition of β-catenin translocation into the cytoplasm and nucleus.
Our study strengthens the therapeutic possibility that compensatory induction of unphosphorylated PTEN may inhibit the acquisition of EMT phenotypes in lung cancer cells under persistent hypoxia.
KeywordsLung cancer Hypoxia Epithelial-mesenchymal transition β-catenin PTEN
The tumor microenvironment, which involves activation of various signal pathways, accelerates the acquisition of epithelial-mesenchymal transition (EMT) in lung cancers cells [1, 2]. Persistent hypoxia induces acquisition of EMT phenotypes, involving de novo EMT-related gene expression, such as twist, and the stabilization of hypoxia-inducible factor 1α (HIF-1α), a significant transcriptional factors in hypoxia [3, 4, 5, 6, 7]. As a result, persistent hypoxia stimulation has been proposed as one of the most critical microenvironmental factors in the development of cancer and tissue fibrosis in vivo [5, 6, 8, 9]. The tumor suppressor gene, phosphatase and tensin homologue deleted from chromosome 10 (PTEN), negatively regulates the activation of various signaling pathways in the tumor microenvironment , but hyperactivation of these pathways is often observed in lung cancers [11, 12, 13]. Loss of PTEN expression might accelerate the development of lung cancer in vivo . Meanwhile, recent studies suggest that phosphorylation of the PTEN C-terminus (p-PTEN) might directly induce a loss of PTEN activity [15, 16]. We recently showed that transforming growth factor β (TGFβ), one of another inducers of EMT, might lead to loss of PTEN activity through a decrease in total PTEN expression level and TGFβ-induced phosphorylation of its C-terminus in lung cancer cells . Nevertheless, it is not known whether or not persistent hypoxia might modulate the PTEN phosphatase activity in lung cancers. Furthermore, whether hypoxia-induced EMT phenotypes are negatively regulated by unphosphorylated PTEN remains elusive.
Based on the knowledge that persistent hypoxia-induced aberrant signaling pathway should be therapeutic target for lung cancer [18, 19], we aimed to determine that persistent hypoxia can modulate the PTEN activity in lung cancers and that unphosphorylated PTEN can inhibit hypoxia-induced EMT.
Effect of persistent hypoxia on total PTEN expression and phosphorylation of the C-terminal tail in PTEN in lung cancers
Effect of persistent hypoxia on modulation of PTEN expression and EMT phenotypes in H358 cells in vitro
Effect of unphosphorylated PTEN on hypoxia-induced EMT in H358 cells
Effect of unphosphorylated PTEN on hypoxia-induced stabilization of HIF-1α expression, translocation of β-catenin, and the EMT-related gene twist expression in H358 cells
Although PTEN mutations and deficiencies might be prevalent at the advanced tumor stage and in therapeutic resistance in many types of human cancer [12, 28, 29, 30], lung cancers rarely show mutation of the PTEN gene [12, 31]. Furthermore, mounting evidence suggests that loss of the PTEN phosphatase activity via the increasing p-PTEN/PTEN ratio might be also crucial for many types of cancer to acquire malignant phenotypes [17, 27, 32, 33]. Therefore, we first evaluated total PTEN expression level and the phosphorylation level of PTEN in lung cancer cells, by assessing five lung cancer cell lines in which the EMT phenotypes had been previously analyzed . The data suggested that, among the five lung cancer cell lines, those with low PTEN expression level and a high p-PTEN/PTEN ratio showed a tendency to have a high degree of the EMT phenotypes . In addition to the above findings, the importance of the tumor microenvironment was suggested by recent studies [1, 2]. Because H358 cells were suitable for evaluating the underlying mechanisms, by which tumor microenvironment factors might induce EMT [6, 17, 20, 27], we utilized H358 cells for further experiments in the present study. Evidence of persistent hypoxia has been shown in tumor implant models and pulmonary fibrosis by several reliable approaches based on the detection of hypoxia-related molecules, such as pimonidazole [6, 34, 35]. Indeed, our immunohistochemical findings in vivo suggested that tumor lesions with persistent hypoxia might show no or little PTEN expression. Because H358ON cells retain endogenous PTEN expression before inoculation into mice (Fig. 1), the decreased expression of PTEN observed in cancer cells indicates that persistent tissue hypoxia might modulate PTEN expression in vivo. It should be noted that phosphorylated PTEN expression was not detected in tumor specimens, possibly owing to our use of an antibody that is not recommended for immunohistochemistry staining. Although the acquisition of EMT phenotypes is modulated by persistent hypoxia [3, 4, 5, 6, 7], persistent hypoxia induces a decrease in total PTEN expression and increases the p-PTEN/PTEN ratio over time in H358 cells incubated under hypoxia. Our in vitro data suggested that persistent hypoxia markedly repressed the expression of total PTEN expression in a time-dependent manner, whereas phosphorylated PTEN levels remained persistent even after hypoxia stimulation, resulting in an approximately eightfold increase in the p-PTEN/PTEN ratio in vitro [36, 37, 38]. These findings indicate that modulation of total PTEN expression and the increasing p-PTEN/PTEN ratio might be induced not only by TGFβ activation  but also by persistent hypoxia stimulation in lung cancer cells. Mounting evidence show that β-catenin is located on the cell membrane via E-cadherin complexes [21, 25, 26]. We previously demonstrated that although β-catenin and E-cadherin were co-localized on the cell membrane when the cells were not treated with TGFβ, TGFβ might induce translocation of β-catenin from the cell membrane to the cytoplasm and nucleus . Taken together with the previous studies [21, 25], double immunostaining for β-catenin and E-cadherin suggested that persistent hypoxia might also induce translocation of β-catenin from the cell membrane to the cytoplasm and nucleus resulting in induction of mesenchymal gene expression. Therefore, a therapeutic strategy to comprehensively regulate various factors derived from tumor microenvironment might be warranted. Nevertheless, whether or not hypoxia-induced EMT might be directly affected by the increasing phosphorylation levels of the PTEN C-terminus has not been fully evaluated. In the present study, compensatory induction of unphosphorylated PTEN blunted hypoxia-induced EMT in lung cancer cells. Although persistent hypoxia-induced HIF-1α stabilization can cause the acquisition of malignant phenotypes [5, 7], PTEN4A did not appear to modulate hypoxia-induced HIF-1α stabilization, indicating that the compensatory induction of PTEN4A inhibited the acquisition of hypoxia-induced EMT phenotypes by mechanisms other than HIF-1α stabilization . Although hypoxia is known to induce de novo twist expression [5, 6], our data suggested that compensatory induction of either PTEN4A or WildPTEN induced the decreasing levels of hypoxia-induced twist gene. De novo expressed GFP-WildPTEN protein decreased the fibronectin/E-cadherin ratio and expression of twist in the cells under persistent hypoxia. Meanwhile, it did not block hypoxia-induced β-catenin translocation from the cell membrane into the cytoplasm and nucleus. A recent study suggests that fibronectin transcription is induced by β-catenin translocation from E-cadherin complexes at the cell membrane into the cytoplasm and nucleus . Indeed, WildPTEN did not appear to repress hypoxia-induced fibronectin expression, whereas repression of hypoxia-induced E-cadherin was restored by induction of WildPTEN. PTEN binds E-cadherin complex via the PDZ binding domain in the PTEN C-terminus . Previous studies suggested that total PTEN expression levels might affect the response to WildPTEN transduction in glioma cells [28, 29, 30]. In the present study, we showed that PTEN protein was abundantly expressed in H358 cells under normoxia, whereas persistent hypoxia induced the repression of total PTEN expression in H358 cells to the same degree as that in H1299 cells. Taken together, compensatory induction of WildPTEN might restore the repression of E-cadherin expression in hypoxia-induced H358 cells showing little expression of PTEN. Thus, exogenous administration of the PTEN gene might be hopeful for treatment of lung cancer under persistent hypoxia. It should be noted that de novo GFP-PTEN4A protein expression showed an approximately 40 % greater potential to block hypoxia-induced EMT in comparison with de novo GFP-WildPTEN protein expression. We demonstrated that unphosphorylated PTEN, but not WildPTEN, inhibits TGFβ-induced EMT . Research into the role of unphosphorylated PTEN in EMT induced by tumor microenvironmental factors remains important .
In summary, our studies strengthen the therapeutic possibility that compensatory induction of unphosphorylated PTEN may inhibit the acquisition of EMT phenotypes in lung cancers under tissue microenvironments involving persistent hypoxia stimulation.
All antibodies used in our experiments were purchased from the commercial companies as follow: monoclonal mouse anti-hypoxia-inducible factor 1α (HIF-1α) antibody was from Novus biologics (Littleton, CO); monoclonal mouse anti-PTEN antibody (clone 6H2.1) was from Cascade Bioscience (Winchester, UK); anti-β-catenin antibody and purified and fluorescein isothiocyanate (FITC)-conjugated mouse anti-E-cadherin antibody were from BD Biosciences (San Diego, CA); purified anti-fibronectin antibody was from Santa Cruz Biotechnology, Inc (Santa Cruz, CA); streptavidin (SAv)-Alexa 594 (SAv-594)-conjugated and SAv-Alexa 488 (SAv-488)-conjugated anti mouse antibody was from Invitrogen Life Technologies (Carlsbad, CA); purified rabbit anti-phospho-PTEN (Ser380/Thr382/Thr383) antibody was from Cell Signaling Technology (Boston, MA); Affinity-isolated rabbit anti-actin antibody was from Sigma-Aldrich (St. Louis, MO). Hoechst33342 was from Dojindo (Kumamoto, Japan). The Hypoxyprobe-1 plus kit was purchased from Hypoxyprobe, Inc. (Burlington, MA). Horseradish peroxidase (HRP)-conjugated anti-mouse antibody, 3, 3′-diaminobenzidine (DAB) Substrate Kit, and Hematoxylin QS were from Vector Laboratories (Burlingame, CA). PhosSTOP was from Roche Applied Science (Mannheim, Germany). Can Get Signal was also purchased from Toyobo Co. (Tokyo, Japan). Doxycycline (Dox) was from Clontech (Mountain View, CA). pTRE-Tight vector and pTet-On Advanced were from Clontech (Mountain View, CA).
Vectors and gene transfection
A Dox-dependent gene expression system was applied to H358 cells, a human lung cancer cell line, carrying pTet-On Advanced (H358ON) . The H358ON cells that were generated contained GFP in the pTRE-Tight vector (GFP), GFP-PTEN in the pTRE-Tight vector (GFP-WildPTEN), or GFP-PTEN with a four-Ala substitution (S380A, T382A, T383A, and S385A) on the C-terminal tail in PTEN (PTEN4A) in the pTRE-Tight vector (GFP-PTEN4A).
Human lung cell lines H441, H358, A549, H157 and H1299, were maintained in RPMI medium as previously described . To examine the effect of hypoxia on phenotypic alterations in lung alveolar cell lines in vitro, the cells were cultured under hypoxia (1 % concentration of oxygen; 1 % O2) for the indicated periods using a hypoxic chamber (Wakenyaku Co. Ltd., Tokyo, Japan) for the indicated periods. For immunocytochemistry, some cells were cultured in an 8-well Lab-Tek Chamber Slide System (Nalge Nunc International, Naperville, IL). Pimonidazole hydrochloride at final concentration of 150 μM was added 90 min before immunostaining for pimonidazole .
We performed real-time PCR, by using a TaqMan ABI 7300 Sequence Detection System (PE Applied Biosystems, Foster City, CA). The following oligonucleotide primers and probe were used: twist (twist1: NM_000474) sense (5′-CCAGCTATGTGGCTCACGAG-3′) and antisense (5′-CTAGTGGGACGCGGACATGG-3′), internal fluorescence-labeled probe (FAM) (5′-CCTCCATCCTCCAGACCGAGAAGGCG-3′) . The twist mRNA expression levels were normalized to the glyceraldehydes-3-phophate dehydrogenase (GAPDH) mRNAs signal .
Western blotting analysis
Protein immunodetection was performed by electrophoretic transfer of SDS-PAGE, separation of proteins on nitrocellulose, incubation with primary antiboby, followed by the appropriate HRP-conjugate second antibody, and chemiluminescent second-step detection using ECL Plus (GE Healthcare Bio-Sciences Corp; Piscataway, NJ) . The amount of protein on the immunoblots was quantified by the image software Quantity One (BioRad Laboratories, Philadelphia, PA) .
Immunocytochemistry and immunehistochemistry
Immunocytochemistry and immunohistochemistry was done as previously described [9, 42, 43]. Immunofluorescent staining for PTEN in vitro was performed for H358 cells and H1299 cells and SAv-488-conjugated anti mouse antibody was utilized to visualize antibody binding. Nuclear staining was performed by Hoechst33342. To determine the impact of PTEN4A transduction on hypoxia-induced β-catenin translocation, double immunostaining for β-catenin and E-cadherin was done for H358 cells and H358ON cells with Dox-dependent GFP, GFP-WildPTEN, or GFP-PTEN4A according the previous procedure . We determined distribution of β-catenin by using confocal laser scanning microscopy (TiEA1R; Nikon Instech Co., Tokyo, Japan). Imaging software (NIS-Elements AR; Nikon Instech Co., Tokyo, Japan) was utilized for the intensity of β-catenin fluorescence. To evaluate the fluorescence, intensities over a random cross section of the cells were plotted . Measurement of fluorescence intensity in the nucleus and the cytoplasm was performed by evaluation of a minimum of five randomly selected high-power fields per sample . Immunostaining for pimonidazole in vivo was performed by following the manufacturer’s instructions with minor modifications . To evaluate PTEN expression in the tumor specimens in vivo, a DAB Substrate Kit was used to visualize antibody binding . Sections were counterstained with Hematoxylin QS.
Mouse tumor implantation in vivo
H358ON cells (3 × 106) expressing Dox-dependent GFP were inoculated subcutaneously into the flank of 6-week-old female nude mice. The treated mice were maintained on water containing Dox at a final concentration of 2 mg/ml .
Non-normally distributed results were analyzed by using the Mann–Whitney test for comparison between any two groups. Furthermore, they were analyzed by non-parametric equivalents of analysis of variance (ANOVA) for multiple comparisons. A p value of less than 0.05 was considered statistically significant.
Conception and design: NH. Development of methodology: TK, MK, DA, NH. Acquisition of data: TK, SM, MK, AA, NO, NH. Analysis and interpretation of data: KS, NH writing, review, and/or revision of the manuscript: TK, NH. Study supervision: KI, TK, YH. All authors read and approved the final manuscript.
Our study was supported by Kowa Life Science Foundation and Grant-in-Aid for Scientific Research (C) (21590987 and 24591162). This study is partly funded by a Grant to the Diffuse Lung Diseases Research Group from the Ministry of Health, Labour and Welfare, Japan.
The authors declared that they have no competing interests.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.