Abstract
Androgen receptor (AR) and estrogen receptor-β (ERβ) have been implicated in urothelial tumor outgrowth as promoters, while underlying mechanisms remain poorly understood. Our transcription factor profiling previously performed identified FOXO1 as a potential downstream target of AR in bladder cancer cells. We here investigated the functional role of FOXO1 in the development and progression of urothelial cancer in relation to AR and ERβ signals. In non-neoplastic urothelial SVHUC cells or bladder cancer lines, AR/ERβ expression or dihydrotestosterone/estradiol treatment reduced the expression levels of FOXO1 gene and induced those of a phosphorylated inactive form of FOXO1 (p-FOXO1). In chemical carcinogen-induced models, FOXO1 knockdown via shRNA or inhibitor treatment resulted in considerable induction of the neoplastic transformation of urothelial cells or bladder cancer development in mice. Similarly, FOXO1 inhibition considerably induced the viability, migration, and invasion of bladder cancer cells. Importantly, in FOXO1 knockdown sublines, an anti-androgen hydroxyflutamide or an anti-estrogen tamoxifen did not significantly inhibit the neoplastic transformation of urothelial cells, while dihydrotestosterone or estradiol did not significantly promote the proliferation or migration of urothelial cancer cells. In addition, immunohistochemistry in surgical specimens showed that FOXO1 and p-FOXO1 expression was down-regulated and up-regulated, respectively, in bladder tumor tissues, which was further associated with worse patient outcomes. AR or ERβ activation is thus found to correlate with inactivation of FOXO1 which appears to be their key downstream effector. Moreover, FOXO1, as a tumor suppressor, is likely inactivated in bladder cancer, which contributes in turn to inducing urothelial carcinogenesis and cancer growth.
Introduction
Urinary bladder cancer, mostly urothelial carcinoma, is one of the most commonly diagnosed neoplasms, with nearly 550,000 new cases and 200,000 deaths estimated in 2018 throughout the world, and age-standardized risks for the incidence and mortality are approximately four times higher in males (Bray et al. 2018). More than two-thirds of patients initially present with non-muscle-invasive disease and suffer from recurrence following transurethral surgery and currently available intravesical pharmacotherapy, while those with muscle-invasive tumors are at a high risk of disease progression and metastasis even after more aggressive treatment such as radical cystectomy with neoadjuvant/adjuvant chemotherapy (Chang et al. 2016, Milowsky et al. 2016, Flaig et al. 2018). Therefore, identification of key molecules or pathways responsible for urothelial cancer outgrowth is urgently required, which may successively provide targeted therapy that improves patient outcomes.
Emerging evidence suggests a vital role of sex hormone receptors, such as androgen receptor (AR) and estrogen receptor (ER)-β, in the modulation of both of two distinct events, urothelial tumorigenesis and tumor progression (Miyamoto et al. 2007, Hsu et al. 2013, 2014, Ide & Miyamoto 2015, Inoue et al. 2018a ). Specifically, androgens and estrogens have been shown to promote tumor outgrowth via the AR and ERβ pathways, respectively. Accordingly, AR and/or ERβ may represent key signal(s) that offer(s) a novel treatment strategy against urothelial cancer. Nonetheless, precise underlying mechanisms of how AR, ERβ, and their related signals function in urothelial cells remain far from being fully understood. We anticipated that AR or ERβ activation led to the modulation of key downstream effectors (i.e. activation of oncogenic molecules and inactivation of tumor suppressors) in urothelial cells. Meanwhile, another class of ER, ERα, has been found to be expressed only in a subset of urothelial tumors (Miyamoto et al. 2012, Kashiwagi et al. 2016, Ide et al. 2017), and its role in bladder cancer may thus be limited.
As described later, our transcription factor profiling and subsequent analysis in bladder cancer cells identified forkhead box O1 (FOXO1) as a potential downstream target of AR and ERβ signals. FOXO1 belongs to the forkhead transcription factor family known to involve some of the cellular functions, including cell growth and differentiation (Wang et al. 2014). More specifically, FOXO1 is targeted for phosphorylation by several protein kinases, including PI3K/AKT signaling, on specific sites (e.g. Ser256 by AKT), introducing inhibition of its interactions with DNA. Inactivation of FOXO1 by phosphorylation thereby leads to inhibition of apoptosis and promotion of cell-cycle progression and cell invasion (Medema et al. 2000, Alvarez et al. 2001, Ramaswamy et al. 2002). It has thus been suggested that FOXO1 generally functions as a tumor suppressor (Jiang et al. 2018). Indeed, down-regulation of FOXO1 expression has been documented in several types of malignancies, including bladder cancer (Kim et al. 2009, Lloreta et al. 2017, Zhang et al. 2017). Meanwhile, the functional interplay between FOXO1 and AR (Huang et al. 2004) or ER (Mazumdar & Kumar 2003, Nakajima et al. 2011) signals in prostate or breast cancer cells has been suggested.
Using preclinical models for bladder cancer, FOXO1 has also been studied, primarily as a downstream target of other molecules (Mao et al. 2015, Jiang et al. 2017, Zhu et al. 2017). However, its own functions in the development and growth of urothelial cancer have not been precisely determined. In the present study, we comprehensively assessed the role of FOXO1 in the neoplastic transformation or tumor initiation of urothelial cells and progression of bladder cancer, as well as the modulation of FOXO1 activity by AR and ERβ signals in urothelial cells.
Materials and methods
Antibodies and chemicals
Supplementary Table 1 (see section on supplementary materials given at the end of this article) lists all antibodies used for immunohistochemistry and/or Western blot in this study. We obtained AS1842856 from Millipore and dihydrotestosterone (DHT), hydroxyflutamide (HF), 17β-estradiol (E2), and tamoxifen (TAM) from Sigma-Aldrich.
Tissue microarray (TMA) and immunohistochemical staining
A set of TMA consisting of retrieved bladder tissue specimens obtained by transurethral surgery performed at The Johns Hopkins Hospital was constructed previously upon appropriate approval from the Institutional Review Board (Miyamoto et al. 2012, Kawahara et al. 2015). None of the patients had received therapy with radiation or anti-cancer drugs prior to the collection of the tissues. The TMA in which AR, ERα, and ERβ were immunohistochemically stained in our previous study (Miyamoto et al. 2012) consisted of 129 cases of urothelial neoplasm and corresponding benign-appearing urothelial tissues. All 51 patients with high-grade muscle-invasive tumor ultimately underwent radical cystectomy. Tumor progression was evaluated separately in 78 cases with non-muscle-invasive tumor (i.e. development of high-grade (from low-grade), invasive (from non-invasive), and/or muscle-invasive tumors) and 51 cases with muscle-invasive tumor (i.e. development of recurrent or metastatic tumors after cystectomy).
Immunohistochemistry was performed on the 5-µm sections, using a primary antibody to FOXO1 (clone C29H4; dilution 1:200) or p-FOXO1 (Ser256; dilution 1:200), as we recently described (Ide et al. 2019). All stains were manually quantified by a single pathologist (HM) who was blinded to sample identity. The German Immunoreactive Score, calculated by multiplying the percentage of immunoreactive cells (0% = 0; 1–10% = 1; 11–50% = 2; 51–80% = 3; 81–100% = 4) by staining intensity (negative = 0; weak = 1; moderate = 2; strong = 3), was considered negative (0; 0–1), weakly positive (1+; 2–4), moderately positive (2+; 6–8), or strongly positive (3+; 9–12).
Cell culture
An immortalized human normal urothelial cell line (SVHUC) and human urothelial carcinoma cell lines (UMUC3/5637/647V) were originally obtained from the American Type Culture Collection. All these lines were recently authenticated, using GenePrint 10 System (Promega), by the institutional core facility. Stable sublines, including SVHUC-vector/SVHUC-AR (Izumi et al. 2013), UMUC3-control-shRNA/UMUC3-AR-shRNA (Zheng et al. 2011), 5637-vector/5637-AR (Zheng et al. 2011), and 647V-vector/647V-AR (Li et al. 2014), were established in our previous studies. Similarly, ERβ-shRNA (sc-35325-V, Santa Cruz Biotechnology) or FOXO1-shRNA (sc-35382-V, Santa Cruz Biotechnology) was stably expressed in SVHUC/UMUC3/5637/647V and SVHUC/SVHUC-AR/UMUC3/5637-AR/647V-AR lines, respectively. A scrambled shRNA (sc-108080, Santa Cruz Biotechnology) was also expressed in each line. Multiple frozen aliquots were made upon the acquisition, and all experiments were performed with cells undergoing fewer than 20 passages. SVHUC (or its sublines) and UMUC3/5637/647V (or their sublines) were maintained in Kaighn’s modification of Ham’s F-12K and DMEM (Mediatech, Corning), respectively, supplemented with 10% fetal bovine serum (FBS), and cultured in phenol red-free medium supplemented with either 5% FBS or 5% charcoal-stripped FBS (primarily for experiments with DHT/E2 treatment) at least 24 h before experimental treatment.
Western blot
Equal amounts of proteins (30–50 µg) obtained from cell extracts were separated in 10% sodium dodecyl sulfate-PAGE, transferred to polyvinylidene difluoride membrane electronically, blocked with 0.3–3% milk, and incubated with a specific antibody and a secondary antibody (anti-mouse IgG HRP-linked antibody or anti-rabbit IgG HRP-linked antibody; Cell Signaling Technology) followed by scanning with an imaging system (ChemiDOC™ MP, Bio-Rad).
Reverse transcription (RT) and real-time polymerase chain reaction (PCR)
Total RNA isolated from cultured cells by TRIzol (Invitrogen) was reverse transcribed using oligo-dT primers and Omniscript reverse transcriptase (Qiagen). Real-time PCR was then performed using iQ™ SYBR® Green Supermix (BioRad) with C1000 Thermal Cycler (BioRad). The primer sequences are given in Supplementary Table 2.
Reporter gene assay
Cells at a density of 50–70% confluence in 24-well plates were co-transfected with 250 ng of a FOXO1 reporter plasmid DNA (Addgene) and 2.5 ng of a control reporter plasmid (pRL-CMV) using Lipofectamine® 3000 transfection reagent (Life Technologies). After transfection, the cells were cultured in the presence or absence of ligands for 24 h. Cell lysates were then assayed for luciferase activity measured using a Dual-Luciferase Reporter Assay kit (Promega).
Chromatin immunoprecipitation (ChIP)
ChIP assay was performed with a Magna ChIP kit (Millipore) according to the manufacturer’s recommended protocol. Briefly, confluent UMUC3 cells were fixed with 1% formaldehyde at room temperature for 10 min to crosslink histones and DNA. Then, cell lysates were sonicated and immunoprecipitated with an anti-AR antibody (N-20), an anti-ERβ antibody (B-3, Santa Cruz Biotechnology), or mouse IgG. After reverse-crosslinking, purified and immunoprecipitated DNA was PCR amplified with a FOXO1 promoter-specific primer set.
In vitro transformation
We used a method for the neoplastic transformation of SVHUC with exposure to a carcinogen, 3-methylcholanthrene (MCA), described in a previous study (Reznikoff et al. 1988), with minor modifications. Briefly, cells (2 × 106/10-cm culture dish incubated for 48 h) were cultured in F-12K containing 5 μg/mL MCA (Sigma-Aldrich). After the first 24 h of MCA exposure, FBS (1%) was added to the medium. After an additional 24 h, the cells were cultured in F-12K containing 5% FBS (without MCA) until near confluence. Subcultured cells (1:3 split ratio) were again cultured with MCA for two 48-h exposure periods using the previously mentioned protocol. These MCA-exposed cells were subcultured for 6 weeks and thereafter utilized for subsequent assays.
Cell proliferation
We used the methylthiazolyldiphenyl-tetrazolium bromide (MTT) assay to assess cell viability. Cells (5 × 103/well) seeded in 96-well tissue culture plates were cultured for up to 120 h and then incubated with 0.5 mg/mL of MTT (Sigma-Aldrich) for 3 h at 37°C. MTT was dissolved by dimethyl sulfoxide, and the absorbance was measured at a wavelength of 570 nm with background subtraction at 630 nm.
Plate colony formation
Cells (500/well) seeded in 12-well tissue culture plates were allowed to grow until colonies in the control well were easily distinguishable. The cells were then fixed with methanol and stained with 0.1% crystal violet. The number of colonies in photographed images was quantitated using ImageJ software (National Institutes of Health).
Mouse models for bladder cancer
The animal protocol used in this study was approved by the Johns Hopkins University Animal Care and Use Committee (#MO15M453).
SVHUC-derived cells (5 × 105/100 μL) exposed to MCA were resuspended in Matrigel (BD Biosciences) and subcutaneously injected into the flank of 6-week-old male immunocompromised NOD-SCID mice (Johns Hopkins University Research Animal Resources). Tumor volume was then estimated by serial caliper measurements of perpendicular diameters using the following formula: ((short diameter)2 × (longest diameter) × 0.5).
C57BL/6 mice (Johns Hopkins University Research Animal Resources) were supplied ad libitum with tap water containing 0.1% N-butyl-N-(4-hydroxybutyl)nitrosamine (BBN) (Sigma-Aldrich) at 6 weeks of age for 12 weeks and thereafter with tap water without BBN, as described previously (Miyamoto et al. 2007, Inoue et al. 2018b ). These mice also received injections of AS1842856 or vehicle only. Starting at 18 weeks of age, urine was assessed for the presence of hematuria twice a week using Chemostrip® 5 OB (Roche) urine test strips. When more than trace amount of blood was detected in three consecutive assessments, the animal was killed for macroscopic/microscopic analyses of the bladder and other major organs.
Cell migration
Scratch wound healing assay was adapted to evaluate the ability of cell migration. Cells at a density of ≥90% confluence in 12-well tissue culture plates were scratched manually with a sterile 200-μL plastic pipette tip. The wounded monolayers of the cells were allowed to heal for 24 h, and the width of the wound area was monitored with an inverted microscope. The normalized cell-free area in photographed images (24 h/0 h) was quantitated using ImageJ software.
Cell invasion
Cell invasiveness was determined using a Matrigel-coated transwell chamber (5.0-μm pore size polycarbonate filter with 6.5 mm in diameter; Corning). Cells (1 × 105) in 100 μL of serum-free medium were added to the upper chamber of the transwell, whereas 600 μL of medium containing 5% FBS was added to the lower chamber. After incubation for 24 h, invaded cells were fixed, stained with 0.1% crystal violet, and counted.
Statistical analysis
Fisher’s exact test or chi-square test was used to evaluate the associations between categorized variables. The numerical data were compared by Student’s t-test. Survival rates in patients and tumor development rates in mice were calculated by the Kaplan–Meier method and comparison was made by log-rank test. The Cox proportional hazards model was used to determine statistical significance of prognosticators in a multivariate setting. P values of less than 0.05 were considered to be statistically significant.
Results
Transcription factors down-regulated by androgen in bladder cancer cells
We anticipated that there were transcription factors whose activity was regulated via AR and/or ER activation in urothelial cells. Indeed, in our previous study (Kawahara et al. 2015), we performed a profiling assay, using the TF Activation Profiling Plate Array II (Signosis), in AR-positive bladder cancer UMUC3 cells with mock vs synthetic androgen methyltrienolone (R1881) treatment and identified several transcription factors, out of 96, whose activity was considerably induced by R1881. The transcription factor array data also displayed that 14 were down-regulated (i.e. >10% decrease) (Supplementary Fig. 1, unpublished data). We newly performed a quantitative RT-PCR analysis in UMUC3 incubated with vs without 10 nM DHT for 24 h to confirm the effect of androgen on the expression of three of the transcription factors in bladder cancer cells. DHT treatment was found to significantly reduce the expression of FOXO1 (58% decrease, P = 0.010), but not that of STAT4 (22% decrease, P > 0.1) or FOXC1 (10% decrease, P > 0.1). In addition, treatment with 100 nM E2 for 24 h in UMUC3 resulted in decreases in the expression of FOXO1 (28%, P = 0.039), STAT4 (2%, P > 0.1), or FOXC1 (6%, P > 0.1). We thus decided to investigate FOXO1 as a potential target of both AR and ER signals in bladder cancer in the present study.
Expression of FOXO1/p-FOXO1 in surgical specimens and its prognostic significance
We stained immunohistochemically for FOXO1 and its inactivated form, p-FOXO1, in a set of bladder TMA consisting of 129 cases of urothelial neoplasms and corresponding 80 non-neoplastic urothelial tissues. Positive signals of FOXO1 and p-FOXO1 were detected predominantly in the nucleus of benign or malignant urothelial cells (Fig. 1A).
Immunohistochemistry of FOXO1 and p-FOXO1 in surgical specimens. (A) Representative images of FOXO1(1+) or p-FOXO1(1+) expression (original magnification: ×400), as well as FOXO1(2+) or p-FOXO1(2+) expression (original magnification: ×200), in bladder cancer tissues. Kaplan–Meier analyses for recurrence-free survival in 78 patients with non-muscle-invasive tumor according to FOXO1 positivity (B; FOXO1-negative, n = 62 vs FOXO1-positive, n = 16) and progression-free survival in 51 patients with muscle-invasive tumor according to p-FOXO1 positivity (C; p-FOXO1-negative, n = 21 vs p-FOXO1-positive, n = 30). A full color version of this figure is available at https://doi.org/10.1530/ERC-20-0004.
Citation: Endocrine-Related Cancer 27, 4; 10.1530/ERC-20-0004
Immunohistochemistry of FOXO1 and p-FOXO1 in surgical specimens. (A) Representative images of FOXO1(1+) or p-FOXO1(1+) expression (original magnification: ×400), as well as FOXO1(2+) or p-FOXO1(2+) expression (original magnification: ×200), in bladder cancer tissues. Kaplan–Meier analyses for recurrence-free survival in 78 patients with non-muscle-invasive tumor according to FOXO1 positivity (B; FOXO1-negative, n = 62 vs FOXO1-positive, n = 16) and progression-free survival in 51 patients with muscle-invasive tumor according to p-FOXO1 positivity (C; p-FOXO1-negative, n = 21 vs p-FOXO1-positive, n = 30). A full color version of this figure is available at https://doi.org/10.1530/ERC-20-0004.
Citation: Endocrine-Related Cancer 27, 4; 10.1530/ERC-20-0004
Immunohistochemistry of FOXO1 and p-FOXO1 in surgical specimens. (A) Representative images of FOXO1(1+) or p-FOXO1(1+) expression (original magnification: ×400), as well as FOXO1(2+) or p-FOXO1(2+) expression (original magnification: ×200), in bladder cancer tissues. Kaplan–Meier analyses for recurrence-free survival in 78 patients with non-muscle-invasive tumor according to FOXO1 positivity (B; FOXO1-negative, n = 62 vs FOXO1-positive, n = 16) and progression-free survival in 51 patients with muscle-invasive tumor according to p-FOXO1 positivity (C; p-FOXO1-negative, n = 21 vs p-FOXO1-positive, n = 30). A full color version of this figure is available at https://doi.org/10.1530/ERC-20-0004.
Citation: Endocrine-Related Cancer 27, 4; 10.1530/ERC-20-0004
Table 1 summarizes the status of FOXO1 and p-FOXO1 expression and its associations with clinicopathological features of these 129 cases. Overall, FOXO1 or p-FOXO1 was positive in 40.0% (37.5% 1+, 2.5% 2+) or 23.8% (all 1+) of non-neoplastic tissues and in 13.2% (12.4% 1+, 0.8% 2+) or 55.1% (44.2% 1+, 10.9% 2+) of bladder tumors, respectively. Thus, the rates of FOXO1 and p-FOXO1 positivity (0 vs 1+/2+) were significantly lower and higher, respectively, in tumors than in benign tissues, suggesting inactivation of FOXO1 in urothelial neoplasms. In addition, FOXO1 positivity was significantly lower in high-grade (6.4%) or muscle-invasive (2.0%) tumors than in low grade (24.0%) or non-muscle-invasive (20.5%) tumors, while p-FOXO1 expression was significantly elevated in high-grade or pN+ cases. Meanwhile, in our previous study (Miyamoto et al. 2012), AR/ERα/ERβ was found to be immunoreactive in 58 (45.0%)/34 (26.4%)/64 (49.6%) of the present 129 tumors, respectively. Interestingly, there were positive or negative correlations between p-FOXO1 and AR expression (positive; P = 0.031), FOXO1 and ERα expression (positive; P = 0.001), and FOXO1 and ERβ expression (negative; P = 0.021).
Correlation between FOXO1/p-FOXO1 expression and clinicopathological profile of the patients.
| n | FOXO1 | p-FOXO1 | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Expression levels | P value | Expression levels | P value | |||||||
| 0 | 1+ | 2+ | 0 vs 1+/2+ | 0 | 1+ | 2+ | 0 vs 1+/2+ | 0/1+ vs 2+ | ||
| Non-neoplastic urothelium | 80 | 48 (60.0%) | 30 (37.5%) | 2 (2.5%) | <0.001 | 61 (76.3%) | 19 (23.8%) | 0 (0%) | <0.001 | 0.002 |
| Urothelial neoplasm | 129 | 112 (86.8%) | 16 (12.4%) | 1 (0.8%) | 58 (45.0%) | 57 (44.2%) | 14 (10.9%) | |||
| Sex | 0.959 | 0.393 | 0.279 | |||||||
| Male | 98 | 85 (86.7%) | 12 (12.2%) | 1 (1.0%) | 42 (42.9%) | 47 (48.0%) | 9 (9.2%) | |||
| Female | 31 | 27 (87.1%) | 4 (12.9%) | 0 (0%) | 16 (51.6%) | 10 (32.3%) | 5 (16.1%) | |||
| Tumor Grade | 0.004a | 0.101b | 0.010a | |||||||
| PUNLMP | 11 | 7 (63.6%) | 4 (36.4%) | 0 (0%) | 6 (54.5%) | 4 (36.4%) | 1 (9.1%) | |||
| Low-grade carcinoma | 39 | 31 (79.5%) | 8 (20.5%) | 0 (0%) | 21 (53.8%) | 18 (46.2%) | 0 (0%) | |||
| PUNLMP + Low-grade | 50 | 38 (76.0%) | 12 (24.0%) | 0 (0%) | 27 (54.0%) | 22 (44.0%) | 1 (2.0%) | |||
| High-grade carcinoma | 79 | 74 (93.7%) | 4 (5.1%) | 1 (1.3%) | 31 (39.2%) | 35 (44.3%) | 13 (16.5%) | |||
| Pathologic Stage | 0.002b | 0.485b | 0.154b | |||||||
| pTa | 75 | 61 (81.3%) | 13 (17.3%) | 1 (1.3%) | 35 (46.7%) | 35 (46.7%) | 5 (6.7%) | |||
| pT1 | 3 | 1 (33.3%) | 2 (66.7%) | 0 (0%) | 2 (66.7%) | 0 (0%) | 1 (33.3%) | |||
| Non-muscle-invasive | 78 | 62 (79.5%) | 15 (19.2%) | 1 (1.3%) | 37 (47.4%) | 35 (44.9%) | 6 (7.7%) | |||
| pT2 | 19 | 18 (94.7%) | 1 (5.3%) | 0 (0%) | 10 (52.6%) | 6 (31.6%) | 3 (15.8%) | |||
| pT3 | 24 | 24 (100%) | 0 (0%) | 0 (0%) | 10 (41.7%) | 12 (50.0%) | 2 (8.3%) | |||
| pT4 | 8 | 8 (100%) | 0 (0%) | 0 (0%) | 1 (12.5%) | 4 (50.0%) | 3 (37.5%) | |||
| Muscle-invasive | 51 | 50 (98.0%) | 1 (2.0%) | 0 (0%) | 21 (41.2%) | 22 (43.1%) | 8 (15.7%) | |||
| Lymph Node Involvement | 0.544 | 0.020 | 0.029 | |||||||
| pN0 | 36 | 35 (97.2%) | 1 (2.8%) | 0 (0%) | 19 (52.8%) | 13 (36.1%) | 4 (11.1%) | |||
| pN1-3 | 13 | 13 (100%) | 0 (0%) | 0 (0%) | 2 (15.4%) | 6 (46.2%) | 5 (38.5%) | |||
| AR expression | 0.079 | 0.031 | 0.461 | |||||||
| Negative | 71 | 65 (91.5%) | 6 (8.5%) | 0 (0%) | 38 (53.5%) | 24 (33.8%) | 9 (12.7%) | |||
| Positive | 58 | 47 (81.0%) | 10 (17.2%) | 1 (1.7%) | 20 (34.5%) | 33 (56.9%) | 5 (8.6%) | |||
| ERα expression | 0.001 | 0.605 | 0.084 | |||||||
| Negative | 95 | 88 (92.6%) | 7 (7.4%) | 0 (0%) | 44 (46.3%) | 38 (40.0%) | 13 (13.7%) | |||
| Positive | 34 | 24 (70.6%) | 9 (26.5%) | 1 (2.9%) | 14 (41.2%) | 19 (55.9%) | 1 (2.9%) | |||
| ERβ expression | 0.021 | 0.530 | 0.022 | |||||||
| Negative | 65 | 52 (80.0%) | 13 (20.0%) | 0 (0%) | 31 (47.7%) | 31 (47.7%) | 3 (4.6%) | |||
| Positive | 64 | 60 (93.8%) | 3 (4.7%) | 1 (1.6%) | 27 (42.2%) | 26 (40.6%) | 11 (17.2%) | |||
aPUNLMP + low-grade carcinoma vs high-grade carcinoma. bNon-muscle-invasive vs muscle-invasive.
PUNLMP: papillary urothelial neoplasm of low malignant potential.
We then performed Kaplan–Meier analysis coupled with the log-rank test to assess possible associations between FOXO1 or p-FOXO1 expression and patient outcomes. Patients with FOXO1-negative tumor had a significantly higher risk of recurrence of non-muscle-invasive disease (hazard ratio (HR) = 0.398; P = 0.016; Fig. 1B), but not progression of non-muscle-invasive (P = 0.471) or muscle-invasive (P = 0.291) disease or cancer-specific mortality of muscle-invasive disease (P = 0.289). p-FOXO1 positivity was also significantly associated with progression of muscle-invasive tumors (HR = 2.124; P = 0.041; Fig. 1C), but not recurrence (P = 0.487) or progression (P = 0.717) of non-muscle-invasive tumors or cancer-specific survival of muscle-invasive tumors (P = 0.108). Multivariate analysis with Cox model further revealed FOXO1 expression as an independent negative predictor for recurrence of non-muscle-invasive tumors (HR = 0.128; 95% CI = 0.017–0.940; P = 0.043), but not p-FOXO1 for progression of muscle-invasive tumors (P > 0.1).
Regulation of FOXO1 expression and activity by androgen/AR and estrogen/ERβ
We have previously demonstrated that UMUC3 possesses a functional AR and that SVHUC, 5637, and 647V are negative for AR (Miyamoto et al. 2007, Izumi et al. 2013, Ide et al. 2018). Similarly, ERβ has been shown to be positive in these four cell lines, whereas ERα is all negative (Izumi et al. 2014) (also see Supplementary Fig. 2). We have then established stable sublines expressing or silencing AR or ERβ, as well as their controls (i.e. SVUHC-vector/SVHUC-AR, SVHUC-control-shRNA/SVHUC-ERβ-shRNA, UMUC3-control-shRNA/UMUC3-AR-shRNA/UMUC3-ERβ-shRNA, 5637-vector/5637-AR, 5637-control-shRNA/5637-ERβ-shRNA, 647V-vector/647V-AR, 647V-control-shRNA/647V-ERβ-shRNA) (Fig. 2B and D). Using the parental or stable cell lines, we assessed the impact of AR/ERβ overexpression or knockdown, or androgen/estrogen treatment on the expression and activity of FOXO1.
Relationship between FOXO1/p-FOXO1 and AR/ERβ expression in urothelial cells. Quantitative real-time RT-PCR of FOXO1 (A) and Western blot of FOXO1 and p-FOXO1 (B) in AR-positive vs AR-negative cells. Quantitative real-time RT-PCR of FOXO1 (C) and Western blot of FOXO1 and p-FOXO1 (D) in ERβ-positive vs ERβ-negative cells. In RT-PCR analysis, the expression of FOXO1 gene was normalized to that of GAPDH, and transcription amount is presented relative to that in AR-negative or ERβ-positive cells. Each value represents the mean (+s.d.) from three independent experiments. *P < 0.05 (vs control lines). In Western blot, GAPDH served as a loading control.
Citation: Endocrine-Related Cancer 27, 4; 10.1530/ERC-20-0004
Relationship between FOXO1/p-FOXO1 and AR/ERβ expression in urothelial cells. Quantitative real-time RT-PCR of FOXO1 (A) and Western blot of FOXO1 and p-FOXO1 (B) in AR-positive vs AR-negative cells. Quantitative real-time RT-PCR of FOXO1 (C) and Western blot of FOXO1 and p-FOXO1 (D) in ERβ-positive vs ERβ-negative cells. In RT-PCR analysis, the expression of FOXO1 gene was normalized to that of GAPDH, and transcription amount is presented relative to that in AR-negative or ERβ-positive cells. Each value represents the mean (+s.d.) from three independent experiments. *P < 0.05 (vs control lines). In Western blot, GAPDH served as a loading control.
Citation: Endocrine-Related Cancer 27, 4; 10.1530/ERC-20-0004
Relationship between FOXO1/p-FOXO1 and AR/ERβ expression in urothelial cells. Quantitative real-time RT-PCR of FOXO1 (A) and Western blot of FOXO1 and p-FOXO1 (B) in AR-positive vs AR-negative cells. Quantitative real-time RT-PCR of FOXO1 (C) and Western blot of FOXO1 and p-FOXO1 (D) in ERβ-positive vs ERβ-negative cells. In RT-PCR analysis, the expression of FOXO1 gene was normalized to that of GAPDH, and transcription amount is presented relative to that in AR-negative or ERβ-positive cells. Each value represents the mean (+s.d.) from three independent experiments. *P < 0.05 (vs control lines). In Western blot, GAPDH served as a loading control.
Citation: Endocrine-Related Cancer 27, 4; 10.1530/ERC-20-0004
AR-positive urothelial cell lines showed considerably lower and higher expression levels of FOXO1 gene (Fig. 2A) and p-FOXO1 protein (Fig. 2B), respectively, compared with corresponding AR-negative lines. ERβ expression was also associated with reduced levels of FOXO1 gene (Fig. 2C) and elevated levels of p-FOXO1 protein (Fig. 2D). Consistent with these findings, DHT treatment reduced the levels of FOXO1 expression in AR-positive lines, but not in AR-negative cells, in a dose-dependent manner (Supplementary Fig. 3A), and the inhibitory effects of DHT on FOXO1 expression were at least partially restored by an anti-androgen HF (Fig. 3A). Additionally, in AR-positive bladder cancer lines, DHT induced p-FOXO1 expression, which was antagonized by HF (Fig. 3B). Similarly, E2 treatment resulted in decreases in the expression of FOXO1 gene in ERα-negative/ERβ-positive urothelial cell lines, but not in ERα-negative/ERβ-negative lines (Supplementary Fig. 3B), which was at least partially restored by an anti-estrogen TAM (Fig. 3C). In ERβ-positive lines, E2 increased the levels of p-FOXO1 (Fig. 3D). In these Western blots, DHT or E2 was treated for 24 h because time-dependent (up to 24 h) induction in p-FOXO1 expression was observed (Supplementary Fig. 4). Also, the ratio of phosphorylated to total FOXO1 expression was measured in each Western blot (Supplementary Table 3).
Effects of androgen or estrogen on FOXO1 and p-FOXO1 expression in urothelial cells. Quantitative real-time RT-PCR of FOXO1 in AR-positive cells treated with ethanol (mock), 10 nM DHT, and/or 1 μM HF (A) and ERβ-positive cells treated with ethanol (mock), 100 nM E2, and/or 1 μM TAM (C) for 24 h. The expression of FOXO1 gene was normalized to that of GAPDH, and transcription amount is presented relative to that of mock treatment. Each value represents the mean (+s.d.) from three independent experiments. *P < 0.05 (vs mock treatment). #P < 0.05 (vs DHT or E2 treatment). Western blot of p-FOXO1 in AR-positive cells treated with ethanol (mock), 10 nM DHT, and/or 1 μM HF (B) and ERβ-positive cells treated with ethanol (mock), 10 nM E2, and/or 1 μM TAM (D) for 24 h. GAPDH served as a loading control.
Citation: Endocrine-Related Cancer 27, 4; 10.1530/ERC-20-0004
Effects of androgen or estrogen on FOXO1 and p-FOXO1 expression in urothelial cells. Quantitative real-time RT-PCR of FOXO1 in AR-positive cells treated with ethanol (mock), 10 nM DHT, and/or 1 μM HF (A) and ERβ-positive cells treated with ethanol (mock), 100 nM E2, and/or 1 μM TAM (C) for 24 h. The expression of FOXO1 gene was normalized to that of GAPDH, and transcription amount is presented relative to that of mock treatment. Each value represents the mean (+s.d.) from three independent experiments. *P < 0.05 (vs mock treatment). #P < 0.05 (vs DHT or E2 treatment). Western blot of p-FOXO1 in AR-positive cells treated with ethanol (mock), 10 nM DHT, and/or 1 μM HF (B) and ERβ-positive cells treated with ethanol (mock), 10 nM E2, and/or 1 μM TAM (D) for 24 h. GAPDH served as a loading control.
Citation: Endocrine-Related Cancer 27, 4; 10.1530/ERC-20-0004
Effects of androgen or estrogen on FOXO1 and p-FOXO1 expression in urothelial cells. Quantitative real-time RT-PCR of FOXO1 in AR-positive cells treated with ethanol (mock), 10 nM DHT, and/or 1 μM HF (A) and ERβ-positive cells treated with ethanol (mock), 100 nM E2, and/or 1 μM TAM (C) for 24 h. The expression of FOXO1 gene was normalized to that of GAPDH, and transcription amount is presented relative to that of mock treatment. Each value represents the mean (+s.d.) from three independent experiments. *P < 0.05 (vs mock treatment). #P < 0.05 (vs DHT or E2 treatment). Western blot of p-FOXO1 in AR-positive cells treated with ethanol (mock), 10 nM DHT, and/or 1 μM HF (B) and ERβ-positive cells treated with ethanol (mock), 10 nM E2, and/or 1 μM TAM (D) for 24 h. GAPDH served as a loading control.
Citation: Endocrine-Related Cancer 27, 4; 10.1530/ERC-20-0004
The effects of AR/ERβ expression and androgen/estrogen treatment on the transcriptional activity of FOXO1 were then determined in the cell extracts with transfection of a luciferase reporter plasmid. AR expression or DHT treatment (Fig. 4A), as well as ERβ expression or E2 treatment (Fig. 4B), in three urothelial cell lines was all associated with reduced FOXO1 luciferase activity. In addition, DHT and E2 effects on FOXO1 transcription were abolished by HF and TAM, respectively.
Effects of androgen (A) and estrogen (B) on FOXO1 transcriptional activity in urothelial cells. Luciferase reporter activity in SVHUC-vector/SVHUC-AR/SVHUC-ERβ-shRNA, UMUC3-control-shRNA/UMUC3-AR-shRNA/UMUC3-ERβ-shRNA, and 5637-vector/5637-AR/5637-control-shRNA/5637-ERβ-shRNA transfected with FOXO1-Luc and subsequently cultured with or without ethanol (mock), DHT (1 nM), HF (1 µM), E2 (10 nM), and/or TAM (1 µM) for 24 h. Each value presented relative to that of control (vector only) or control-shRNA line or mock treatment represents the mean (+s.d.) from three independent experiments. *P < 0.05 (vs control line, AR-shRNA/ERβ-shRNA line, or mock treatment). #P < 0.05 (vs DHT or E2 treatment).
Citation: Endocrine-Related Cancer 27, 4; 10.1530/ERC-20-0004
Effects of androgen (A) and estrogen (B) on FOXO1 transcriptional activity in urothelial cells. Luciferase reporter activity in SVHUC-vector/SVHUC-AR/SVHUC-ERβ-shRNA, UMUC3-control-shRNA/UMUC3-AR-shRNA/UMUC3-ERβ-shRNA, and 5637-vector/5637-AR/5637-control-shRNA/5637-ERβ-shRNA transfected with FOXO1-Luc and subsequently cultured with or without ethanol (mock), DHT (1 nM), HF (1 µM), E2 (10 nM), and/or TAM (1 µM) for 24 h. Each value presented relative to that of control (vector only) or control-shRNA line or mock treatment represents the mean (+s.d.) from three independent experiments. *P < 0.05 (vs control line, AR-shRNA/ERβ-shRNA line, or mock treatment). #P < 0.05 (vs DHT or E2 treatment).
Citation: Endocrine-Related Cancer 27, 4; 10.1530/ERC-20-0004
Effects of androgen (A) and estrogen (B) on FOXO1 transcriptional activity in urothelial cells. Luciferase reporter activity in SVHUC-vector/SVHUC-AR/SVHUC-ERβ-shRNA, UMUC3-control-shRNA/UMUC3-AR-shRNA/UMUC3-ERβ-shRNA, and 5637-vector/5637-AR/5637-control-shRNA/5637-ERβ-shRNA transfected with FOXO1-Luc and subsequently cultured with or without ethanol (mock), DHT (1 nM), HF (1 µM), E2 (10 nM), and/or TAM (1 µM) for 24 h. Each value presented relative to that of control (vector only) or control-shRNA line or mock treatment represents the mean (+s.d.) from three independent experiments. *P < 0.05 (vs control line, AR-shRNA/ERβ-shRNA line, or mock treatment). #P < 0.05 (vs DHT or E2 treatment).
Citation: Endocrine-Related Cancer 27, 4; 10.1530/ERC-20-0004
Next, ChIP was performed to determine whether AR and ERβ regulate the expression of FOXO1. DNA fragments from UMUC3 cells immunoprecipitated by an AR or ERβ antibody or normal IgG were amplified by PCR with FOXO1 promoter-specific primers (Supplementary Fig. 5A). A PCR product was visualized from those precipitated by the ERβ antibody, but not control precipitations (Supplementary Fig. 5B), indicating interactions of ERβ with the FOXO1 promoter region. By contrast, no positive results were obtained in those precipitated by the AR antibody (figure not shown).
Role of FOXO1 in tumorigenesis
To investigate the impact of FOXO1 on urothelial tumorigenesis, we first used an in vitro model where non-neoplastic SVHUC cells could undergo stepwise neoplastic transformation upon exposure to a chemical carcinogen MCA (Reznikoff et al. 1988). SVHUC-derived cells were exposed to MCA and subsequently cultured for 6 weeks during the process of transformation with or without treatment of a FOXO1 inhibitor AS1842856, HF, or TAM. Oncogenic activity of MCA-SVHUC sublines was then monitored by cell viability (via MTT assay; Fig. 5A), colony formation (via clonogenic assay; Fig. 5B), tumor formation (via mouse xenograft model; Fig. 5C), and the expression level of tumor suppressors (via RT-PCR; Fig. 5D) without further drug treatment during these assays. We thus compared the degree of malignant transformation in urothelial cells with the carcinogen challenge, but not the efficacy of AS1842856, HF, or TAM in the growth of SVHUC-derived cells.
Effects of FOXO1 inactivation on neoplastic transformation of urothelial cells. SVHUC and its sublines exposed to MCA and subsequently cultured for 6 weeks in the absence or presence of AS1842856 (100 nM), HF (1 μM), or TAM (1 μM) were seeded for MTT assay (A; cultured for 7 days) or clonogenic assay (B; cultured for 2 weeks) without AS1842856/HF/TAM treatment. Cell viability or colony number (≥20 cells) is presented relative to that of mock-treated or control-shRNA lines. Each value represents the mean (+s.d.) from at least three independent experiments. *P < 0.05 (vs mock treatment or control-shRNA cells). SVHUC-control-shRNA and SVHUC-FOXO1-shRNA cells exposed to MCA and subsequently cultured for 6 weeks were implanted subcutaneously into the flank of male NOD-SCID mice (n = 10 in each group; Kaplan–Meier analysis for the formation of tumor exceeding 30 mm3 in its volume estimated (by the formula: (short diameter)2 × (longest diameter) × 0.5) or 5 mm in greatest dimension) (C) and subjected to RNA extraction and real-time RT-PCR of p21, p27, and UGT1A (D). In RT-PCR analysis, the expression of each specific gene was normalized to that of GAPDH, and transcription amount is presented relative to that of control-shRNA cells. Each value represents the mean (+s.d.) from three independent experiments. *P < 0.05 (vs control-shRNA cells). (E) Male and female C57BL/6 mice (n = 8/group) were treated with BBN in drinking water (for 12 weeks) as well as ethanol (mock; 1/2000 in 0.2 mL sterile distilled water) or AS1842856 (30 mg/kg) via s.c. injections three times a week, starting at 6 weeks of age. Kaplan–Meier curves for the detection of hematuria via three consecutive positive urine tests are shown.
Citation: Endocrine-Related Cancer 27, 4; 10.1530/ERC-20-0004
Effects of FOXO1 inactivation on neoplastic transformation of urothelial cells. SVHUC and its sublines exposed to MCA and subsequently cultured for 6 weeks in the absence or presence of AS1842856 (100 nM), HF (1 μM), or TAM (1 μM) were seeded for MTT assay (A; cultured for 7 days) or clonogenic assay (B; cultured for 2 weeks) without AS1842856/HF/TAM treatment. Cell viability or colony number (≥20 cells) is presented relative to that of mock-treated or control-shRNA lines. Each value represents the mean (+s.d.) from at least three independent experiments. *P < 0.05 (vs mock treatment or control-shRNA cells). SVHUC-control-shRNA and SVHUC-FOXO1-shRNA cells exposed to MCA and subsequently cultured for 6 weeks were implanted subcutaneously into the flank of male NOD-SCID mice (n = 10 in each group; Kaplan–Meier analysis for the formation of tumor exceeding 30 mm3 in its volume estimated (by the formula: (short diameter)2 × (longest diameter) × 0.5) or 5 mm in greatest dimension) (C) and subjected to RNA extraction and real-time RT-PCR of p21, p27, and UGT1A (D). In RT-PCR analysis, the expression of each specific gene was normalized to that of GAPDH, and transcription amount is presented relative to that of control-shRNA cells. Each value represents the mean (+s.d.) from three independent experiments. *P < 0.05 (vs control-shRNA cells). (E) Male and female C57BL/6 mice (n = 8/group) were treated with BBN in drinking water (for 12 weeks) as well as ethanol (mock; 1/2000 in 0.2 mL sterile distilled water) or AS1842856 (30 mg/kg) via s.c. injections three times a week, starting at 6 weeks of age. Kaplan–Meier curves for the detection of hematuria via three consecutive positive urine tests are shown.
Citation: Endocrine-Related Cancer 27, 4; 10.1530/ERC-20-0004
Effects of FOXO1 inactivation on neoplastic transformation of urothelial cells. SVHUC and its sublines exposed to MCA and subsequently cultured for 6 weeks in the absence or presence of AS1842856 (100 nM), HF (1 μM), or TAM (1 μM) were seeded for MTT assay (A; cultured for 7 days) or clonogenic assay (B; cultured for 2 weeks) without AS1842856/HF/TAM treatment. Cell viability or colony number (≥20 cells) is presented relative to that of mock-treated or control-shRNA lines. Each value represents the mean (+s.d.) from at least three independent experiments. *P < 0.05 (vs mock treatment or control-shRNA cells). SVHUC-control-shRNA and SVHUC-FOXO1-shRNA cells exposed to MCA and subsequently cultured for 6 weeks were implanted subcutaneously into the flank of male NOD-SCID mice (n = 10 in each group; Kaplan–Meier analysis for the formation of tumor exceeding 30 mm3 in its volume estimated (by the formula: (short diameter)2 × (longest diameter) × 0.5) or 5 mm in greatest dimension) (C) and subjected to RNA extraction and real-time RT-PCR of p21, p27, and UGT1A (D). In RT-PCR analysis, the expression of each specific gene was normalized to that of GAPDH, and transcription amount is presented relative to that of control-shRNA cells. Each value represents the mean (+s.d.) from three independent experiments. *P < 0.05 (vs control-shRNA cells). (E) Male and female C57BL/6 mice (n = 8/group) were treated with BBN in drinking water (for 12 weeks) as well as ethanol (mock; 1/2000 in 0.2 mL sterile distilled water) or AS1842856 (30 mg/kg) via s.c. injections three times a week, starting at 6 weeks of age. Kaplan–Meier curves for the detection of hematuria via three consecutive positive urine tests are shown.
Citation: Endocrine-Related Cancer 27, 4; 10.1530/ERC-20-0004
MTT assay showed that AS1842856 enhanced MCA-induced neoplastic transformation of SVHUC (AR-negative/ERβ-positive) or SVHUC-ERβ-shRNA (AR-negative/ERβ-negative) cells (Fig. 5A). Similarly, FOXO1 knockdown in AR-negative/ERβ-positive MCA-SVHUC cells (Supplementary Fig. 6) resulted in the induction of their neoplastic transformation (Fig. 5A and B). The effect of FOXO1 knockdown on neoplastic transformation was further confirmed by an animal experiment demonstrating significant delay in the formation of control xenograft tumors compared with MCA-SVHUC-FOXO1-shRNA xenografts (Fig. 5C). Additionally, significant down-regulation of the expression of three genes known to function as suppressors for bladder tumorigenesis, including cyclin-dependent kinase inhibitors p21 and p27, and UDP-glucuronosyltransferase-1A (UGT1A), a major phase II drug metabolism enzyme responsible for detoxifying various bladder carcinogens (Izumi et al. 2013, 2014), was seen in MCA-SVHUC-FOXO1-shRNA cells compared with control cells (Fig. 5D). However, although HF and TAM inhibited their neoplastic transformation, AS1842856 treatment or FOXO1-shRNA expression did not significantly induce that of AR-positive/ERβ-positive MCA-SVHUC-AR cells (Fig. 5A and B). Additional MTT assay in MCA-SVHUC-AR-FOXO1-shRNA showed that HF and TAM failed to significantly inhibit the transformation (Fig. 5A).
We also utilized an animal carcinogenesis model where a chemical carcinogen BBN could reliably induce the development of bladder tumor, especially in male rodents, to further assess the effects of the FOXO1 inhibitor on urothelial tumorigenesis. To detect bladder tumors at an early stage, hematuria was monitored via urine test strips. We then confirmed macroscopically and microscopically the development of high-grade carcinoma in the bladders from all the mice with three consecutive positive urine tests. In both male and female mice, AS1842856 treatment significantly facilitated tumor development (Fig. 5E). None of the animals developed upper urinary tract tumors or metastatic tumors when euthanized.
Role of FOXO1 in tumor progression
We next investigated whether FOXO1 signals contributed to regulating bladder cancer progression using cell line models. We compared cell viability (via MTT assay), migration (via wound-healing assay), and invasion (via transwell assay), as well as the expression of oncogenic or tumor suppressor genes (via RT-PCR), between FOXO1-positive lines vs their knockdown lines and/or FOXO1-positive lines with mock vs FOXO1 inhibitor treatments.
In AR-positive/ERβ-positive lines, FOXO1 silencing (Supplementary Fig. 6) or AS1842856 treatment resulted in considerable induction of cell proliferation (Fig. 6A and B) and migration (Fig. 6C and D). FOXO1 silencing also demonstrated marked increases in the invasion ability of AR-positive/ERβ-positive cells compared with that of control cells (Fig. 6E). In addition, there were significant increases in the expression levels of matrix metalloproteinase-2 (MMP2) and vascular endothelial growth factor (VEGF), and significant decreases in those of p21 and p27 (Fig. 6F).
Effects of FOXO1 inactivation on bladder cancer cell growth. MTT assay in UMUC3-control-shRNA/UMUC3-FOXO1-shRNA or 647V-AR-control-shRNA/647V-AR-FOXO1-shRNA cultured for 1–5 days (A), UMUC3, 647V-AR, or 647V-ERβ-shRNA cultured with either ethanol (mock) or 100 nM AS1842856 for 4 days (B), and UMUC3-control-shRNA/UMUC3-FOXO1-shRNA or 647V-AR-control-shRNA/647V-AR-FOXO1-shRNA cultured with either ethanol (mock), 1 nM DHT, or 10 nM E2 for 4 days (G). Cell viability is presented relative to that of each control line at day 1 (A) or mock treatment (B, G). Wound-healing assay in UMUC3-control-shRNA/UMUC3-FOXO1-shRNA or 647V-AR-control-shRNA/647V-AR-FOXO1-shRNA (C), UMUC3 or 647V-AR cultured with either ethanol (mock) or 100 nM AS1842856 (D), and UMUC3-control-shRNA/UMUC3-FOXO1-shRNA or 647V-AR-control-shRNA/647V-AR-FOXO1-shRNA cultured with either ethanol (mock), 1 nM DHT, or 10 nM E2 (H). The cells grown to confluence were gently scratched and the wound area was measured after 24 h culture in serum-free medium. The migration determined by the rate of cells filling the wound area is presented relative to that of each control line (C) or mock treatment (D, H). (E) Transwell invasion assay in UMUC3-control-shRNA/UMUC3-FOXO1-shRNA and 647V-AR-control-shRNA/647V-AR-FOXO1-shRNA cultured in the Matrigel-coated transwell chamber for 24 h. The number of invaded cells present in the lower chamber was counted under a light microscope (100× objective in five random fields). Cell invasion is presented relative to that of each control line. (F) Quantitative real-time RT-PCR of MMP-2, VEGF, p21, and p27 in UMUC3-control-shRNA/UMUC3-FOXO1-shRNA and 647V-AR-control-shRNA/647V-AR-FOXO1-shRNA. The expression of each specific gene was normalized to that of GAPDH, and transcription amount is presented relative to that in each control line. Each value represents the mean (+s.d.) from from independent experiments. *P < 0.05 (vs control-shRNA cells or mock treatment).
Citation: Endocrine-Related Cancer 27, 4; 10.1530/ERC-20-0004
Effects of FOXO1 inactivation on bladder cancer cell growth. MTT assay in UMUC3-control-shRNA/UMUC3-FOXO1-shRNA or 647V-AR-control-shRNA/647V-AR-FOXO1-shRNA cultured for 1–5 days (A), UMUC3, 647V-AR, or 647V-ERβ-shRNA cultured with either ethanol (mock) or 100 nM AS1842856 for 4 days (B), and UMUC3-control-shRNA/UMUC3-FOXO1-shRNA or 647V-AR-control-shRNA/647V-AR-FOXO1-shRNA cultured with either ethanol (mock), 1 nM DHT, or 10 nM E2 for 4 days (G). Cell viability is presented relative to that of each control line at day 1 (A) or mock treatment (B, G). Wound-healing assay in UMUC3-control-shRNA/UMUC3-FOXO1-shRNA or 647V-AR-control-shRNA/647V-AR-FOXO1-shRNA (C), UMUC3 or 647V-AR cultured with either ethanol (mock) or 100 nM AS1842856 (D), and UMUC3-control-shRNA/UMUC3-FOXO1-shRNA or 647V-AR-control-shRNA/647V-AR-FOXO1-shRNA cultured with either ethanol (mock), 1 nM DHT, or 10 nM E2 (H). The cells grown to confluence were gently scratched and the wound area was measured after 24 h culture in serum-free medium. The migration determined by the rate of cells filling the wound area is presented relative to that of each control line (C) or mock treatment (D, H). (E) Transwell invasion assay in UMUC3-control-shRNA/UMUC3-FOXO1-shRNA and 647V-AR-control-shRNA/647V-AR-FOXO1-shRNA cultured in the Matrigel-coated transwell chamber for 24 h. The number of invaded cells present in the lower chamber was counted under a light microscope (100× objective in five random fields). Cell invasion is presented relative to that of each control line. (F) Quantitative real-time RT-PCR of MMP-2, VEGF, p21, and p27 in UMUC3-control-shRNA/UMUC3-FOXO1-shRNA and 647V-AR-control-shRNA/647V-AR-FOXO1-shRNA. The expression of each specific gene was normalized to that of GAPDH, and transcription amount is presented relative to that in each control line. Each value represents the mean (+s.d.) from from independent experiments. *P < 0.05 (vs control-shRNA cells or mock treatment).
Citation: Endocrine-Related Cancer 27, 4; 10.1530/ERC-20-0004
Effects of FOXO1 inactivation on bladder cancer cell growth. MTT assay in UMUC3-control-shRNA/UMUC3-FOXO1-shRNA or 647V-AR-control-shRNA/647V-AR-FOXO1-shRNA cultured for 1–5 days (A), UMUC3, 647V-AR, or 647V-ERβ-shRNA cultured with either ethanol (mock) or 100 nM AS1842856 for 4 days (B), and UMUC3-control-shRNA/UMUC3-FOXO1-shRNA or 647V-AR-control-shRNA/647V-AR-FOXO1-shRNA cultured with either ethanol (mock), 1 nM DHT, or 10 nM E2 for 4 days (G). Cell viability is presented relative to that of each control line at day 1 (A) or mock treatment (B, G). Wound-healing assay in UMUC3-control-shRNA/UMUC3-FOXO1-shRNA or 647V-AR-control-shRNA/647V-AR-FOXO1-shRNA (C), UMUC3 or 647V-AR cultured with either ethanol (mock) or 100 nM AS1842856 (D), and UMUC3-control-shRNA/UMUC3-FOXO1-shRNA or 647V-AR-control-shRNA/647V-AR-FOXO1-shRNA cultured with either ethanol (mock), 1 nM DHT, or 10 nM E2 (H). The cells grown to confluence were gently scratched and the wound area was measured after 24 h culture in serum-free medium. The migration determined by the rate of cells filling the wound area is presented relative to that of each control line (C) or mock treatment (D, H). (E) Transwell invasion assay in UMUC3-control-shRNA/UMUC3-FOXO1-shRNA and 647V-AR-control-shRNA/647V-AR-FOXO1-shRNA cultured in the Matrigel-coated transwell chamber for 24 h. The number of invaded cells present in the lower chamber was counted under a light microscope (100× objective in five random fields). Cell invasion is presented relative to that of each control line. (F) Quantitative real-time RT-PCR of MMP-2, VEGF, p21, and p27 in UMUC3-control-shRNA/UMUC3-FOXO1-shRNA and 647V-AR-control-shRNA/647V-AR-FOXO1-shRNA. The expression of each specific gene was normalized to that of GAPDH, and transcription amount is presented relative to that in each control line. Each value represents the mean (+s.d.) from from independent experiments. *P < 0.05 (vs control-shRNA cells or mock treatment).
Citation: Endocrine-Related Cancer 27, 4; 10.1530/ERC-20-0004
We also performed MTT (Fig. 6G) and wound-healing (Fig. 6H) assays to compare the effects of androgen/estrogen on the growth of FOXO1-positive and FOXO1-negative cells. In accordance with previous observations (Miyamoto et al. 2007, Zheng et al. 2011, Huang et al. 2015, Inoue et al. 2018b , Ou et al. 2018), DHT and E2 induced the proliferation and migration of UMUC3/647V-AR cells, presumably via the AR and ERβ pathways, respectively. However, the stimulatory effects of DHT or E2 on cell proliferation and migration were considerably diminished in FOXO1 knockdown lines.
Discussion
Based on the results of our transcription factor profiling, we further investigated if FOXO1 activity was modulated by AR and ERβ signals in non-neoplastic urothelial cells and bladder cancer cells. We also assessed the impact of FOXO1 inactivation separately on urothelial tumorigenesis and bladder cancer progression. In addition, immunohistochemistry for FOXO1 and p-FOXO1 was performed in a set of TMA to determine if their expression could serve as prognostic markers in patients with bladder cancer.
FOXO1 has been implicated in tumorigenesis of a few types of malignancies, including osteosarcoma and prostate cancer (Guan et al. 2015, Yang et al. 2017). In particular, FOXO1 knockdown was shown to induce neoplastic formation (e.g. prostatic intraepithelial neoplasia) in the mouse prostate, presumably via activating ERG, a critical oncogenic molecule for prostate carcinogenesis (Yang et al. 2017). Indeed, earlier studies demonstrated that FOXO1 inactivation via its phosphorylation was a result of PI3K/AKT activation, leading to apoptosis suppression as well as cell-cycle progression involving a signaling cascade of cyclins and cyclin-dependent kinases such as cyclin D and p27 (Medema et al. 2000, Alvarez et al. 2001, Ramaswamy et al. 2002). We here demonstrated that FOXO1 inactivation was associated with the promotion of malignant transformation of urothelial cells or urothelial tumorigenesis, using two preclinical carcinogen-induced models, indicating its function as a tumor suppressor, while the functional role of AR and ERβ themselves was not assessed, particularly using in vivo models, in the present study. Specifically, in an in vitro system, FOXO1 knockdown or inhibitor treatment was found to induce MCA-mediated neoplastic transformation of AR-negative/ERβ-positive or AR-negative/ERβ-negative urothelial cells via monitoring cell viability, colony formation, and tumor formation in mice (knockdown only). Correspondingly, the expression levels of tumor suppressor genes were significantly reduced in these knockdown cells exposed to MCA. Furthermore, the preventive effect of the FOXO1 inhibitor on bladder tumor development was confirmed in both male and female mice treated with BBN. However, no significant effects of inhibitor or knockdown on neoplastic transformation were seen in AR-positive/ERβ-positive cells, presumably due to low basal expression of FOXO1 (Fig. 2A and C).
As aforementioned, FOXO1 is known to play an important role in modulating cell proliferation and invasion (Medema et al. 2000, Alvarez et al. 2001, Ramaswamy et al. 2002). Indeed, loss of FOXO1 in AR-positive or AR-negative prostate cancer lines has been shown to correlate with the induction of cell growth (Zhang et al. 2011, Yang et al. 2017). In bladder cancer cells, however, FOXO1 has been investigated, primarily as a downstream effector of, for instance, miRNA and p27 (Mao et al. 2015, Jiang et al. 2017, Zhu et al. 2017). Accordingly, the functional role of FOXO1 in the growth of urothelial cancer cells has not been well documented. We here demonstrated that FOXO1 knockdown or inhibitor treatment resulted in the induction of bladder cancer cell proliferation, migration, and invasion, as well as up-regulation of MMP-2 and VEGF or down-regulation of p21 and p27. Thus, FOXO1 inactivation was found to associate with the promotion of urothelial cancer progression. It still needs to be determined precisely how FOXO1 signaling regulates urothelial cancer outgrowth.
Activation of AR or ERβ signals has been correlated with the promotion of urothelial tumorigenesis and tumor progression (Miyamoto et al. 2007, Hsu et al. 2013, 2014, Ide & Miyamoto 2015, Inoue et al. 2018b ). In particular, early phase clinical trials (e.g. NCT02605863/NCT02788201/NCT02300610) have been conducted to assess the efficacy of anti-AR treatment in patients with urothelial cancer. However, it should be mentioned that three of four urothelial cell lines used in the present study lack the endogenous expression of AR. Transcription factors often work in cooperation with their co-regulatory proteins in a cell-specific manner, and these AR-negative cell lines may thus not express some co-regulators essential for modulating AR functions. Here the functional interplay between FOXO1 and AR/ER signals, as documented in prostate or breast cancer cells (Mazumdar & Kumar 2003, Huang et al. 2004), was confirmed in urothelial cells. Specifically, AR/ERβ overexpression or androgen/estrogen treatment reduced FOXO1 expression and more apparently induced p-FOXO1 expression in non-neoplastic and cancerous urothelial cells, suggesting an association between AR or ERβ activation and FOXO1 inactivation. Luciferase assay further showed that AR or ERβ activation resulted in inhibition of FOXO1 transcriptional activity. By contrast, FOXO1 knockdown or inhibitor treatment in bladder cancer cells did not significantly change the levels of AR and ERβ expression (data not shown). In addition, significant correlations between AR or ERβ expression vs FOXO1 or p-FOXO1 levels in bladder cancer tissue samples were observed. Thus, inactivation of a tumor suppressor FOXO1 is likely contributory to previously documented induction of urothelial tumorigenesis and tumor progression by androgens/AR and estrogens/ERβ. More strikingly, anti-androgen/anti-estrogen and androgen/estrogen failed to significantly inhibit the neoplastic transformation and promote the growth, respectively, of FOXO1-negative cells, suggesting that FOXO1 represents a critical downstream effector of AR and ERβ in urothelial cells. Further studies might be required to determine exactly how AR and ERβ signals modulate FOXO1 activity. Meanwhile, a ChIP assay revealed that at least ERβ could bind to the FOXO1 promoter in bladder cancer cells.
The expression of FOXO1 mRNA and/or protein in urothelial tumors has been assessed in surgical specimens (Kim et al. 2009, Lloreta et al. 2017, Zhang et al. 2017, Ide et al. 2019). Studies in bladder tissues demonstrated down-regulation of FOXO1 expression in tumors (vs corresponding non-cancerous tissues), particularly in higher grade/stage tumors, as well as its association with worse patient outcomes (Kim et al. 2009, Lloreta et al. 2017, Zhang et al. 2017). In one of the studies (Lloreta et al. 2017), FOXO1 expression was found to provide more precise prognostic information when p53 was highly expressed. We here confirmed these findings on FOXO1 expression in urothelial cancer and further demonstrated that loss of FOXO1 expression was an independent predictor of disease recurrence in patients with non-muscle-invasive bladder tumor following transurethral surgery. Additionally, in agreement with our previous immunohistochemistry data in upper urinary tract urothelial carcinomas (Ide et al. 2019), p-FOXO1 levels were higher in bladder tumors than in non-neoplastic bladder urothelial tissues, as well as in high-grade or lymph node-positive tumors, and p-FOXO1 positivity was associated with the risk of disease progression in patients with muscle-invasive tumor. FOXO1 expression in non-muscle-invasive tumors and p-FOXO1 expression in muscle-invasive tumors may thus serve as prognosticators of disease recurrence and progression, respectively.
In conclusion, the present findings indicate that FOXO1, as a tumor suppressor, is inactivated in bladder cancer, which is associated with tumor development and outgrowth. Moreover, activation of the AR or ERβ signaling pathway is found to correlate with inactivation of FOXO1 which appears to be its key downstream effector in urothelial cells. Accordingly, FOXO1 activation, together with AR and/or ERβ inactivation especially in AR/ERβ-positive cases, has the potential of being an effective chemopreventive and therapeutic approach for urothelial carcinoma.
Supplementary materials
This is linked to the online version of the paper at https://doi.org/10.1530/ERC-20-0004.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.
Funding
The research did not receive any specific grant from any funding agency in the public, commercial, or not-for-profit sector.
Author contribution statement
H I and H M performed concept and design of experiments. Acquisition of data was done by H I, T M, G J, T G, Y N, Y T, S I, Y L, E K, A S B, G J N, and T K. H I, T M, G J, T G, and H M performed analysis and interpretation of data. H I, T M, G J, T G, and H M wrote the manuscript. HM supervised the work.
References
Alvarez B, Martinez C, Burgering BM & Carrera AC 2001 Forkhead transcription factors contribute to execution of the mitotic programme in mammals. Nature 413 744–747. (https://doi.org/10.1038/35099574)
Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA & Jemal A 2018 Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA: A Cancer Journal for Clinicians 68 394–424. (https://doi.org/10.3322/caac.21492)
Chang SS, Boorjian SA, Chou R, Clark PE, Daneshmand S, Konety BR, Pruthi R, Quale DZ, Ritch CR, Seigne JD, et al. 2016 Diagnosis and treatment of non-muscle-invasive bladder cancer: AUA/SUO guideline. Journal of Urology 196 1021–1029. (https://doi.org/10.1016/j.juro.2016.06.049)
Flaig TW, Spiess PE, Agarwal N, Bangs R, Boorjian SA, Buyyounouski MK, Downs TM, Efstathiou JA, Friedlander T, Greenberg RE, et al. 2018 NCCN Guidelines Insights: bladder cancer, version 5.2018. Journal of the National Comprehensive Cancer Network 16 1041–1053. (https://doi.org/10.6004/jnccn.2018.0072)
Guan H, Tan P, Xie L, Mi B, Fang Z, Li J, Yue J, Liao H & Li F 2015 FOXO1 inhibits osteosarcoma oncogenesis via Wnt/β-catenin pathway suppression. Oncogenesis 4 e166. (https://doi.org/10.1038/oncsis.2015.25)
Hsu I, Vitkus S, Da J & Yeh S 2013 Role of oestrogen receptors in bladder cancer development. Nature Reviews: Urology 10 317–326. (https://doi.org/10.1038/nrurol.2013.53)
Hsu I, Chuang KL, Slavin S, Da J, Lim WX, Pang ST, O’Brien JH & Yeh S 2014 Suppression of ERβ signaling via ERβ knockout or antagonist protects against bladder cancer development. Carcinogenesis 35 651–661. (https://doi.org/10.1093/carcin/bgt348)
Huang H, Muddiman DC & Tindall DJ 2004 Androgens negatively regulate forkhead transcription factor FKHR (FOXO1) through a proteolytic mechanism in prostate cancer cells. Journal of Biological Chemistry 279 13866–13877. (https://doi.org/10.1074/jbc.M314143200)
Huang W, Chen Y, Liu Y, Zhang Q, Yu Z, Mou L, Wu H, Zhao L, Long T, Qin D, et al. 2015 Roles of ERβ and GRP30 in proliferative response of human bladder cancer cell to estrogen. BioMed Research International 2015 251780. (https://doi.org/10.1155/2015/251780)
Ide H & Miyamoto H 2015 Steroid hormone receptor signals as prognosticators for urothelial tumor. Disease Markers 2015 840640. (https://doi.org/10.1155/2015/840640)
Ide H, Inoue S & Miyamoto H 2017 Histopathological and prognostic significance of the expression of sex hormone receptors in bladder cancer: a meta-analysis of immunohistochemical studies. PLoS ONE 12 e0174746. (https://doi.org/10.1371/journal.pone.0174746)
Ide H, Inoue S, Mizushima T, Jiang G, Chuang KH, Oya M & Miyamoto H 2018 Androgen receptor signaling reduces radiosensitivity in bladder cancer. Molecular Cancer Therapeutics 17 1566–1574. (https://doi.org/10.1158/1535-7163.MCT-17-1061)
Ide H, Jiang G, Mizushima T, Fujita K, Inoue S, Yamaguchi S, Fushimi H, Nonomura N & Miyamoto H 2019 Forkhead box O1 as an indicator of prognosis is inactivated in urothelial carcinoma of the upper urinary tract. Oncology Letters 17 482–487. (https://doi.org/10.3892/ol.2018.9510)
Inoue S, Mizushima T & Miyamoto H 2018a Role of the androgen receptor in urothelial cancer. Molecular and Cellular Endocrinology 465 73–81. (https://doi.org/10.1016/j.mce.2017.06.021)
Inoue S, Ide H, Mizushima T, Jiang G, Netto GJ, Gotoh M & Miyamoto H 2018b Nuclear factor-κB promotes urothelial tumorigenesis and cancer progression via cooperation with androgen receptor signaling. Molecular Cancer Therapeutics 17 1303–1314. (https://doi.org/10.1158/1535-7163.MCT-17-0786)
Izumi K, Zheng Y, Hsu JW, Chang C & Miyamoto H 2013 Androgen receptor signals regulate UDP-glucuronosyltransferases in the urinary bladder: a potential mechanism of androgen-induced bladder carcinogenesis. Molecular Carcinogenesis 52 94–102. (https://doi.org/10.1002/mc.21833)
Izumi K, Li Y, Ishiguro H, Zheng Y, Yao JL, Netto GJ & Miyamoto H 2014 Expression of UDP-glucuronosyltransferase 1A in bladder cancer: association with prognosis and regulation by estrogen. Molecular Carcinogenesis 53 314–324. (https://doi.org/10.1002/mc.21978)
Jiang G, Huang C, Li J, Huang H, Jin H, Zhu J, Wu XR & Huang C 2017 Role of STAT3 and FOXO1 in the divergent therapeutic responses of non-metastatic and metastatic bladder cancer cells to miR-145. Molecular Cancer Therapeutics 16 924–935. (https://doi.org/10.1158/1535-7163.MCT-16-0631)
Jiang S, Li T, Yang Z, Hu W & Yang Y 2018 Deciphering the roles of FOXO1 in human neoplasms. International Journal of Cancer 143 1560–1568. (https://doi.org/10.1002/ijc.31338)
Kashiwagi E, Fujita K, Yamaguchi S, Fushimi H, Ide H, Inoue S, Mizushima T, Reis LO, Sharma R, Netto GJ, et al. 2016 Expression of steroid hormone receptors and its prognostic significance in urothelial carcinoma of the upper urinary tract. Cancer Biology and Therapy 17 1188–1196. (https://doi.org/10.1080/15384047.2016.1235667)
Kawahara T, Shareef HK, Aljarah AK, Ide H, Li Y, Kashiwagi E, Netto GJ, Zheng Y & Miyamoto H 2015 ELK1 is up-regulated by androgen in bladder cancer cells and promotes tumor progression. Oncotarget 6 29860–29876. (https://doi.org/10.18632/oncotarget.5007)
Kim TH, Jo SW, Lee YS, Kim YJ, Lee SC, Kim WJ & Yun SJ 2009 Forkhead box O-class 1 and forkhead box G1 as prognostic markers for bladder cancer. Journal of Korean Medical Science 24 468–473. (https://doi.org/10.3346/jkms.2009.24.3.468)
Li Y, Ishiguro H, Kawahara T, Miyamoto Y, Izumi K & Miyamoto H 2014 GATA3 in the urinary bladder: suppression of neoplastic transformation and down-regulation by androgens. American Journal of Cancer Research 4 461–473.
Lloreta J, Font-Tello A, Juanpere N, Frances A, Lorenzo M, Nonell L, de Muga S, Vázquez I, Cecchini L & Hernández-Llodrà S 2017 FOXO1 down-regulation is associated with worse outcome in bladder cancer and adds significant prognostic information to p53 overexpression. Human Pathology 62 222–231. (https://doi.org/10.1016/j.humpath.2016.12.022)
Mao XP, Zhang LS, Huang B, Zhou SY, Liao J, Chen LW, Qiu SP & Chen JX 2015 Mir-135a enhances cellular proliferation through post-transcriptionally regulating PHLPP2 and FOXO1 in human bladder cancer. Journal of Translational Medicine 13 86. (https://doi.org/10.1186/s12967-015-0438-8)
Mazumdar A & Kumar R 2003 Estrogen regulation of Pak1 and FKHR pathways in breast cancer cells. FEBS Letters 535 6–10. (https://doi.org/10.1016/s0014-5793(02)03846-2)
Medema RH, Kops GJ, Bos JL & Burgering BM 2000 AFX-like Forkhead transcription factors mediate cell-cycle regulation by Ras and PKB through p27kip1. Nature 404 782–787. (https://doi.org/10.1038/35008115)
Milowsky MI, Rumble RB, Booth CM, Gilligan T, Eapen LJ, Hauke RJ, Boumansour P & Lee CT 2016 Guideline on muscle-invasive and metastatic bladder cancer (European Association of Urology guideline): American Society of Clinical Oncology clinical practice guideline endorsement. Journal of Clinical Oncology 34 1945–1952. (https://doi.org/10.1200/JCO.2015.65.9797)
Miyamoto H, Yang Z, Chen YT, Ishiguro H, Uemura H, Kubota Y, Nagashima Y, Chang YJ, Hu YC, Tsai MY, et al. 2007 Promotion of bladder cancer development and progression by androgen receptor signals. Journal of the National Cancer Institute 99 558–568. (https://doi.org/10.1093/jnci/djk113)
Miyamoto H, Yao JL, Chaux A, Zheng Y, Hsu I, Izumi K, Chang C, Messing EM, Netto GJ & Yeh S 2012 Expression of androgen and oestrogen receptors and its prognostic significance in urothelial neoplasm of the urinary bladder. BJU International 109 1716–1726. (https://doi.org/10.1111/j.1464-410X.2011.10706.x)
Nakajima Y, Akaogi K, Suzuki T, Osakabe A, Yamaguchi C, Sunahara N, Ishida J, Kako K, Ogawa S, Fujimura T, et al. 2011 Estrogen regulates tumor growth through a nonclassical pathway that includes the transcription factors ERβ and KLF5. Science Signaling 4 ra22. (https://doi.org/10.1126/scisignal.2001551)
Ou Z, Wang Y, Chen J, Tao L, Zuo L, Sahasrabudhe D, Joseph J, Wang L & Yeh S 2018 Estrogen receptor β promotes bladder cancer growth and invasion via alteration of miR-92a/DAB2IP signals. Experimental and Molecular Medicine 50 152. (https://doi.org/10.1038/s12276-018-0155-5)
Ramaswamy S, Nakamura N, Sansal I, Bergeron L & Sellers WR 2002 A novel mechanism of gene regulation and tumor suppression by the transcription factor FKHR. Cancer Cell 2 81–91. (https://doi.org/10.1016/s1535-6108(02)00086-7)
Reznikoff CA, Loretz LJ, Christian BJ, Wu SQ & Meisner LF 1988 Neoplastic transformation of SV40-immortalized human urinary tract epithelial cells by in vitro exposure to 3-methylcholanthrene. Carcinogenesis 9 1427–1436. (https://doi.org/10.1093/carcin/9.8.1427)
Wang Y, Zhou Y & Graves DT 2014 FOXO1 transcription factors: their clinical significance and regulation. BioMed Research International 2014 925350. (https://doi.org/10.1155/2014/925350)
Yang Y, Blee AM, Wang D, An J, Pan Y, Yan Y, Ma T, He Y, Dugdale J, Hou X, et al. 2017 Loss of FOXO1 cooperates with TMPRSS2-ERG overexpression to promote prostate tumorigenesis and cell invasion. Cancer Research 77 6524–6537. (https://doi.org/10.1158/0008-5472.CAN-17-0686)
Zhang H, Pan Y, Zheng L, Choe C, Lindgren B, Jensen ED, Westendorf JJ, Cheng L & Huang H 2011 FOXO1 inhibits Runx2 transcriptional activity and prostate cancer cell migration and invasion. Cancer Research 71 3257–3267. (https://doi.org/10.1158/0008-5472.CAN-10-2603)
Zhang Y, Jia L, Zhang Y, Ji W & Li H 2017 Higher expression of FOXOs correlates to better prognosis of bladder cancer. Oncotarget 8 96313–96322. (https://doi.org/10.18632/oncotarget.22029)
Zheng Y, Izumi K, Yao JL & Miyamoto H 2011 Dihydrotestosterone upregulates the expression of epidermal growth factor receptor and ERBB2 in androgen receptor-positive bladder cancer cells. Endocrine-Related Cancer 18 451–464. (https://doi.org/10.1530/ERC-11-0010)
Zhu J, Li Y, Tian Z, Hua X, Gu J, Li J, Liu C, Jin H, Wang Y, Jiang G, et al. 2017 ATG7 overexpression is crucial for tumorigenic growth of bladder cancer in vitro and in vivo by targeting the ETS2/miRNA196b/FOXO1/p27 axis. Molecular Therapy: Nucleic Acids 7 299–313. (https://doi.org/10.1016/j.omtn.2017.04.012)
