Thyrotroph embryonic factor is downregulated in bladder cancer and suppresses proliferation and tumorigenesis via the AKT/FOXOs signalling pathway

Abstract Objectives Thyrotroph embryonic factor (TEF) plays an important role in several different processes in normal human cells; however, its function in malignant cells has not been fully elucidated. Materials and methods The mRNA levels of TEF in 408 bladder cancer (BC) samples from the Cancer Genome Atlas (TCGA) database were analysed in depth. Next, the expression of TEF in 7 BC cell lines was compared to that in normal bladder epithelial cells. The cell count, colony formation and anchorage‐independent growth assays as well as a nude mouse xenograft model were utilized to examine the effects of TEF on proliferation and tumorigenesis. Immunofluorescence staining, flow cytometry analysis and treatment with an AKT inhibitor were performed to explore the molecular regulation mechanisms of TEF in BC. Results Analysis of TCGA data indicated that TEF mRNA was decreased in BC samples compared to that in normal bladder epithelial cells and correlated with the poor survival of BC patients. Additional experiments verified that the mRNA and protein expression of TEF were significantly decreased in BC cells compared to that in normal bladder epithelial cells. Upregulation of TEF expression significantly retarded BC cell growth by inhibiting the G1/S transition via regulating AKT/FOXOs signalling. Conclusion Our results suggest that TEF might play an important role in suppressing BC cells proliferation and tumorigenesis.

conditions of the tumour to confirm the pathological diagnosis. 8,9 Unfortunately, cystoscopy is an invasive diagnostic procedure, and its use is limited by the lack of suitable biomarkers for pathological staining. Currently, the identification of new markers is of substantial importance for enabling an early diagnosis of BC. 10,11 Although a number of molecules related to BC have been described in the literature, such as forkhead box M1 (FOXM1), collagen type V alpha 2 chain (COL5A2) and N-myc, there are still no effective molecular targets for clinical diagnosis and therapy. 12,13 Thus, more specific and sensitive biomarkers and therapeutic targets need to be identified to improve the early diagnosis and prognosis of BC patients.
It is well known that the biological and molecular expression profiles of tumour cells are similar to those of embryonic cells but differ greatly from those of normal adult cells. 15 Therefore, an increasing number of studies have been devoted to studying the molecular differences between normal adult and embryonic cells to identify targets, such as alpha fetoprotein (AFP), for the diagnosis and treatment of malignant cells. 16 As a member of the proline and acidic amino acid-rich (PAR) subfamily of basic region/leucine zipper (bZIP) transcription factors, which also includes albumin D box-binding protein (DBP), human hepatic leucemia factor (HLF) and chicken vitellogenin gene-binding protein (VBP), Thyrotroph embryonic factor (TEF) is expressed in a broad range of cells and tissues in adult animals. 17,18 TEF expression appears to be restricted to the developing anterior pituitary gland during embryonic development. 20 In normal cells, TEF is an apoptotic regulator of hematopoietic progenitors and controls hematopoietic cell proliferation. 21 In mouse fibroblasts, TEF has been reported to control actin distribution and cell shape. 22 Furthermore, TEF has been reported to be involved in the development of various human diseases. For example, a study including 408 subjects using a linear regression model showed that TEF is associated with the Hamilton Rating Scale for Depression (HAMD) scores in patients with Parkinson's disease after adjusting for clinical variables. 23 Using multivariate generalized linear models, the association between TEF and depressive symptoms further confirmed that TEF is closely related to the development of human disease. 24 However, the role of TEF in malignant processes is still unknown.
In the present study, we demonstrate that TEF mRNA and protein expression are downregulated in BC cells and human BC tissues.
Overexpressed TEF significantly inhibits the G1/S transition in the cell cycle as well as tumorigenesis via regulating the AKT/FOXOs signalling pathway. In addition, opposite results were observed in TEF-silenced cells. These results suggest a potential role for TEF as a diagnostic marker and a valuable therapeutic target in BC.

| Microarray data processing and visualization
The RNA sequencing V2 profile dataset was downloaded on November 29th, 2014 from the Cancer Genome Atlas (TCGA) and contained 408 BC tissue and 19 adjacent normal bladder tissue samples. Profile data extractions were performed using Excel and MeV 4.9 (http://www.tm4.org/mev), and Gene Set Enrichment Analysis (GSEA) was performed using gsea 2.2.1 (http://www.broadinstitute. org/gsea).

| Statistical analysis
spss version 13.0 (SPSS Inc, Chicago, IL) was used in all statistical analyses. 25 The associations between the TEF RNA level and the clinicopathological characteristics of the TCGA data were analysed using the Chi-squared test. Survival curves were plotted using the Kaplan-Meier method and compared using the log-rank test. A twotailed P-value of <0.05 was considered statistically significant in all tests.

| Western blot analysis
Western blot analysis was performed according to standard methods as previously described. 27,28 Total protein was extracted from cell pellets or fresh tissue after harvest. The blots were immunostained with primary and secondary antibodies. Anti-TEF (1:500) was purchased from Abcam (Cambridge, MA); anti- Abcam), which served as a loading control.

| Vectors and retroviral infection
The TEF expression construct was generated by subcloning PCR-amplified full-length human TEF cDNA into the pMSCV- for quantitatively detecting the transcriptional activity of FOXOs was generated using the pGL3-enhancer plasmid (Promega, Madison, WI) as described previously. 29 According to the manufacturer's instructions, perifosine (20 µmol/L; Abcam), a novel AKT inhibitor, was used to verify that AKT is involved in regulating the effects of TEF on proliferation. Water, the storage solution and diluent of perifosine, was used as experimental control. After 6 days of continuous treatment of perifosine, RT4-TEF cells and HT-1376-TEF cells were used for other tests.

| Flow cytometry analysis
Flow cytometry analysis was used to measure the DNA distribution and to identify the cell cycle of the tested cells. Cells were seeded at an initial density of 50 000 cells in 100-mm dishes in the culture medium and allowed to attach for 24 hour. Then, the medium was

| Colony formation assay
Colony formation assays reflect cell proliferation according to the number of clones formed by adherent cells. As previously described, 1000 cells/well were incubated in 6-well plates. All stable cell lines were included in the colony formation assay. 26 Ten days later, the colonies were fixed with 10% formaldehyde for 5 minute and then stained with 1.0% crystal violet for 30 seconds. All experiments were performed in triplicate for each cell line.

| Bromodeoxyuridine labelling and immunofluorescence assay
Bromodeoxyuridine (BrdU) can be incorporated into the newly syn-

| TEF is downregulated in BC and correlates with poor patient survival
Analysing 408 BC cases from TCGA database showed that the mRNA expression of TEF was downregulated in BC samples compared to that in adjacent normal tissue samples (19 cases; P < 0.001, Figure 1A). To account for individual differences, the  Figure 1C). Meanwhile, the mRNA level of TEF was also downregulated in five fresh BC tissues (T) compared to that in the paired adjacent normal tissue (ANT; Figure 1E). Western blot analysis revealed that TEF protein expression was also downregulated in malignant BC cells (or tissues) compared to that in non-malignant cells (or tissues, Figure 1D,F). Taken together, these results strongly indicate that TEF is downregulated in human BC.
To investigate the significance of TEF downregulation in BC, data from the TCGA were categorized and analysed in depth. Using the mean value as a cut-off, all 400 patients with clinical staging data were categorized into one of two groups based on the TEF RNA level in their tumours: a high TEF RNA group and a low TEF RNA group. As shown in Figure 1G, the RNA expression of TEF was higher in early-stage tumours (stages I-II) and was lower in advanced-stage tumours (stages III-IV, P < 0.01). Kaplan-Meier survival curves were plotted for the 294 patients with available survival data, and the results showed that the overall survival of patients with low TEF RNA level was significantly shorter than that of patients with high TEF RNA level ( Figure 1H, P = 0.016).
Collectively, these results indicate that downregulation of TEF in BC patients correlates with poor survival.

| TEF regulates the proliferation of BC cells
Gene set enrichment analysis was used to explore the role of TEF

| TEF modulates the tumorigenesis of BC
As shown in Figure 3A,B, the colony number was significantly de-  Figure 3G, the protein level of Ki67 was low in samples  Figure 3G).
Taken together, these results suggest that TEF plays an important role in the tumorigenicity of BC cells.

| TEF arrests the cell cycle G1/S transition in BC cells
Gene set enrichment analysis also revealed that the mechanism responsible for the anti-proliferative effects of TEF involves the G1/S increased the percentage of cells in the G0/G1 phase but decreased that in the S phase. Conversely, TEF silencing significantly decreased the percentage of cells in the G0/G1 phase and increased that in the S phase ( Figure 4B). As shown in the representative images of Figure 4C and the statistical diagram in Figure 4D, the percentage of BrdU-incorporated cells dramatically decreased upon overexpressing TEF but significantly increased upon silencing TEF. These results clearly show that TEF blocks the G1/S transition of the cell cycle in BC cells.
To further investigate the biological function of TEF in arresting the cell cycle, genes closely related to the G1/S transition were examined by Western blot. As shown in Figure 4E, the protein expression levels of the cell cycle promoter cyclin D1, CDK4 and CDK6 were downregulated in TEF-overexpressing cells and upregulated in TEF-silenced cells. Conversely, the p21 Cip1 and p27 Kip1 protein expression levels were increased in TEF-overexpressing cells but decreased in TEF-silenced cells ( Figure 4E). Moreover, RT-qPCR analysis verified that the mRNA expression of p21 Cip1 and p27 Kip1 was increased in TEF-overexpressing cells but decreased in TEF-silenced cells ( Figure 4F).

| TEF regulates AKT/FOXOs signalling
Gene set enrichment analysis was also used to explore which intracellular signalling pathway(s) might be involved in TEF-mediated biological processes. As shown in Figure 5A, the TEF RNA level negatively correlated with AKT-activated gene signatures and positively correlated with FOXO1/FOXO4-activated gene signatures, suggesting that AKT/FOXOs signalling may be involved in the regulation of TEF. The phosphorylation of AKT, FOXO1 and FOXO4 was decreased in TEFoverexpressing cells but increased in TEF-silenced cells ( Figure 5B).
As shown in Figure 6C, the luciferase reporter assay demonstrated that the transcriptional activity of FOXOs was indeed increased in

TEF-overexpressing cells and decreased in TEF-silenced cells. Taken
together, these results suggest that the observed cell cycle arrest induced by TEF is associated with AKT kinase activity, which subsequently modulates the transactivation activities of FOXO factors.

| AKT/FOXOs pathway is involved in the antiproliferative activity of TEF
To further explore the relationship between AKT and the anti-pro- were treated with the AKT inhibitor perifosine to suppress the kinase activity of AKT. As shown in Figure 6A, the growth rate of perifosine-treated cells overexpressing TEF was significantly increased, but the growth rate was essentially unchanged when the same cells were treated with a control. The images and histogram obtained from a soft agar assay also gave similar results ( Figure 6B,C). In terms of cellular mechanisms, flow cytometry analysis showed that perifosine significantly abrogated the TEF-mediated reduction of cells in the S phase ( Figure 6D). BrdU labelling experiments also demonstrated that perifosine antagonizes the inhibitory effect of TEF protein on cell proliferation ( Figure 6E). Moreover, perifosine abrogated the effects of TEF on the expression of p21 Cip1 and p27 Kip1 ( Figure 6F). Our results suggest that AKT/FOXOs pathway mediates the anti-proliferative effect of TEF in BC cells.

| D ISCUSS I ON
Many transcription factors play key roles in cellular differentiation and the delineation of cell phenotypes. 32 Hunger et al 33   protein is related to stem cell-like properties and resistance to treatment. 41 In clear-cell renal carcinoma cells, overexpression of FOXO4 induces cellular apoptosis. 42 In BC, FOXO1 protein was reported to mediate cell apoptosis. 43 Other biological events, such as autophagic flux, oxidative stress, self-renewal, migration and invasion, were also demonstrated to be regulated by FOXOs. 44,45 However, mechanistic studies on FOXOs are rare and tend to focus on several key biological molecules, among which the most well-characterized and important is the Ser and Thr kinase AKT (also known as protein kinase B, PKB). 48 One direct piece of mechanistic evidence is that FOXOs protein phosphorylation is enhanced in the growth plates of AKT transgenic mice. 49 FOXO transcriptional effectiveness is largely determined by phosphorylation-dependent nucleocytoplasmic shuttling. 50 AKT phosphorylates FOXOs, which leads to their translocation to the cytoplasm, thereby promoting the proliferation of malignant cells. 51 This research shows that the phosphorylation of AKT, FOXO4 and FOXO1 is decreased in TEF-overexpressing cells but increased in TEF-silenced cells. Furthermore, a luciferase reporter assay demonstrated that the transcriptional activity of

FOXO was increased in TEF-overexpressing cells and decreased in
TEF-silenced cells. These results suggest that TEF protein regulates AKT/FOXOs signalling; however, the mechanism underlying these effects remains unclear. In glioma, HLF protein is directly bound to the miR-132 promoter to enhance the expression of miR-132. PTEN, a powerful regulator of AKT, is a target of miR-132. The binding partners of TEF are similar to those of HLF, indicating that TEF protein might regulate the activity of AKT by controlling the expression of micro-RNAs. This hypothesis is a topic worthy of further research.
Taken together, this study shows that TEF is downregulated in BC cells and plays an important role in suppressing proliferation and tumorigenesis by deactivating AKT/FOXOs signalling and suggests a potential role of TEF as a diagnostic marker and valuable therapeutic target in BC.

CO N FLI C T O F I NTE R E S T
The authors have no conflicts of interest to declare.