Deficiency of Thioredoxin Binding Protein-2 (TBP-2) Enhances TGF-β Signaling and Promotes Epithelial to Mesenchymal Transition

Background Transforming growth factor beta (TGF-β) has critical roles in regulating cell growth, differentiation, apoptosis, invasion and epithelial-mesenchymal transition (EMT) of various cancer cells. TGF-β-induced EMT is an important step during carcinoma progression to invasion state. Thioredoxin binding protein-2 (TBP-2, also called Txnip or VDUP1) is downregulated in various types of human cancer, and its deficiency results in the earlier onset of cancer. However, it remains unclear how TBP-2 suppresses the invasion and metastasis of cancer. Principal Findings In this study, we demonstrated that TBP-2 deficiency increases the transcriptional activity in response to TGF-β and also enhances TGF-β-induced Smad2 phosphorylation levels. Knockdown of TBP-2 augmented the TGF-β-responsive expression of Snail and Slug, transcriptional factors related to TGF-β-mediated induction of EMT, and promoted TGF-β-induced spindle-like morphology consistent with the depletion of E-Cadherin in A549 cells. Conclusions/Significance Our results indicate that TBP-2 deficiency enhances TGF-β signaling and promotes TGF-β-induced EMT. The control of TGF-β-induced EMT is critical for the inhibition of the invasion and metastasis. Thus TBP-2, as a novel regulatory molecule of TGF-β signaling, is likely to be a prognostic indicator or a potential therapeutic target for preventing tumor progression.


Introduction
Transforming growth factor-b (TGF-b) has dual functions in cancer [1]. TGF-b acts as a tumor suppressor in the early stage of tumor development, and contradictorily, promotes the invasion and metastasis of tumor cells in the late stage. Recently, many studies have shown that TGF-b promotes cancer progression by inducing Epithelial-mesenchymal transition (EMT), which is a crucial process to acquire the ability to execute the invasionmetastasis steps of cancer [2,3]. TGF-b induces the expression of several transcription factors driven to EMT [4], including Snail/ SNAI1 [5] and Slug/SNAI2 [6], which act directly or indirectly as a repressor of E-Cadherin. The loss of E-Cadherin is a fundamental event in EMT [7,8].
There is the growing evidence that TBP-2 plays as a suppressor of cancer. TBP-2 is downregulated in various human cancer cells [25,26]. TBP-2 overexpression inhibits proliferation via cell cycle arrest [12,27,28,29] and promotes apoptosis [30]. In human T cell lymphocyte virus type 1 (HTLV-I)-infected T cells, TBP-2 regulates cell growth and its expression is associated with responsiveness to IL-2-dependent growth [31], and plays a key role in glucocorticoid-induced cell death [32]. In vivo studies, TBP-2 overexpression suppressed tumor growth and metastasis of the transplanted tumor. Point mutation or knock out of TBP-2 gene in mice show the higher incidence of hepatocellular carcinoma [33,34]. TBP-2 knock out mice also shows the earlier onset of Nbutyl-N-(4-hydroxybutyl) nitrosamine (BBN)-induced bladder carcinoma [35].
These results collectively support that TBP-2 deficiency contributes to the progression and metastasis of cancer, however, detail mechanisms of TBP-2 in this process has not been sufficiently elucidated. In the late stage of cancer cells, TBP-2 expression is downregulated and TGF-b elicits cancer malignancy driving EMT. This correlation provides the hypothesis that TBP-2 regulates TGF-b-associated cancer development in the late stage.
In the present study, we examined the role of TBP-2 in TGF-b signaling. TBP-2 deficiency increased TGF-b signaling by enhancing Smad2 phosphorylation levels, and upregulated TGFb-induced expression of Snail or Slug, resulting in acceleration of TGF-b-driven EMT. These findings show a novel function of TBP-2, as a regulator of TGF-b signaling, and provide new insights to the mechanisms of TGF-b-induced EMT.

TBP-2 Deficiency Enhances Transcriptional Activity of TGF-b Signaling
To investigate the function of TBP-2 in TGF-b signaling, we performed promoter assay using 96CAGA-Luc (TGF-b-responsive promoter-reporter), which is the most frequently used reporter system for TGF-b/Smad signal transduction, in WT (Wild Type: TBP-2 +/+ ) mouse embryonic fibroblasts (MEFs) and TBP-2 2/2 MEFs. The results showed that transcriptional activity in response to TGF-b is enhanced in TBP-2 2/2 MEFs compared with WT MEFs (Fig. 1A). The efficiency of TBP-2 knockdown in A549 and MDA-MB-231 cells was confirmed by real-time RT-PCR (

TBP-2 Deficiency Increases TGF-b-mediated Phosphorylation of Smad2
Next, we analyzed the level of TGF-b-mediated phosphorylation of Smad2 in WT and TBP-2 2/2 MEFs by the western blot analyses. The phospho-Smad2 protein level was declined at 20 hour-TGF-b stimulation in WT MEFs, but was continuously elevated in TBP-2 2/2 MEFs (Fig. 3A). Similarly, phospho-Smad2 levels were enhanced with TGF-b stimulation for 12, 24 and 36 hours in TBP-2 knockdown-A549 cells (Fig. 3B). In addition, total Smad2 protein levels went down for 4 hours, responding to TGF-b stimulation, but were unchanged between 4 to 20 hours in WT MEFs, whereas no significant differences from 0 to 20 hours with TGF-b stimulation in TBP-2 2/2 MEFs (Fig. 3A).

TBP-2 Deficiency Enhances the Induction of Snail and Slug by TGF-b
TGF-b induces the expression of transcriptional factors involved in EMT, including Snail and Slug. As the induction of Snail or Slug is a crucial step for EMT, the effect of TBP-2 knockdown on

TBP-2 Deficiency Promotes TGF-b-induced EMT
Then, we evaluated the effects of TBP-2 knockdown in TGF-binduced EMT. Knockdown of TBP-2 promoted TGF-b-induced morphological changes in A549 (Fig. 5) and 253J cells (data not shown). In the presence of 2.5 ng/ml TGF-b for 24 or 36 hours, TGF-b-driven spindle-like morphology was significantly observed in TBP-2 knockdown-A549 cells. To quantify the morphological changes, we measured the length of the longest diagonal line of each cell. TBP-2 knockdown-cells with TGF-b stimulation significantly lengthened more than control cells (Fig. S1). Consistently, the depletion of E-Cadherin, an epithelial marker, was quickened, and similarly the induction of vimentin, a mesenchymal marker, was elevated in TBP-2 knockdown-A549 cells (Fig. 6). These results indicate that TBP-2 deficiency accelerates the TGF-b-driven EMT phenotype.

Discussion
In this study, we demonstrated that deficiency of TBP-2 increases TGF-b-responsive transcriptional activity and upregulates Smad2 phosphorylation levels, resulting in the acceleration of TGF-b-induced EMT.
TBP-2 deficiency contributes to upregulate transcriptional activities for several stimuli or ligands. We or other groups reported that peroxisome proliferator activated receptor (PPAR) or insulin target genes are upregulated in TBP-2 2/2 mice, and that TBP-2 negatively regulates PPAR transcriptional activity in vitro [23]. TBP-2 deficiency might maintain the level of transcriptional activities with the imperfection of biological feedback.
TBP-2 deficiency also results in the enhancement of phosphorylation of signal transducers. Regarding the relationship between TBP-2 and cell signaling, it was reported that phosphorylation of ERK is enhanced in TBP-2-KO mice [36] bladders during BBNinduced bladder carcinogenesis [35]. Our previous study showed that TBP-2 is a negative regulator of TRX [11], and other group reported that overexpression of TRX elevates the ERK1/2 phosphorylation levels [37]. These reports suggest that TBP-2 deficiency facilitates TRX activity, resulting in enhancement of the phosphorylation levels of signal transducer, such as ERK1/2. However, TBP-2 deficiency did not change the protein levels of TRX in the presence or absence of TGF-b (data not shown), so   The re-expression of TBP-2 using expression vector in TBP-2 2/2 MEFs failed to rescue the knock out effects of TBP-2 on the CAGA promoter. We also performed the experiments on the gain-of-function of TBP-2 using expression vector in A549 and MDA-MB-231 cell lines. The results unexpectedly showed that the overexpression of TBP-2 did not lead to the opposite of the loss-of-function results (data not shown). These results might be caused by the difficulty in controlling the expression level of TBP-2 within the physiological range. Since TBP-2 is a multifunctional protein targeting several molecules, the superabundant expression of TBP-2 might cause unexpected effects, which should be dissected in our future study.
It has been also reported that TBP-2 deficiency promotes TNFa-induced NF-kB activity [34], that TBP-2 inhibits mTOR activity by binding REDD1 protein [38], and that TBP-2 deficiency enhances the phosphorylation of Akt in response to insulin [16,24]. The present study shows that TBP-2 deficiency enhances TGF-b-mediated Smad2 phosphorylation level. These findings suggest that TBP-2 act as a crucial feedback regulator for various biological responses. TBP-2 might be essential for protein phosphatases or protein degradation systems.
TBP-2 deficiency enhanced TGF-b signaling and upregulated Smad7 expression ( Fig. 1 and 2). Smad7, one of inhibitory Smads, plays an essential role in the negative feedback regulation of TGFb signaling [39], however, TBP-2 deficiency enhanced TGF-bmediated Smad2 phosphorylation (Fig. 3) irrespective of increasing Smad7 expression. In the negative feedback of TGF-b signaling, Smad7 requires to bind to Smad ubiquitin regulatory factor 2 (Smurf2), HECT type E3 ligases containing WW domain [39,40]. Smad7-Smurf2 complex binds to the activated TGF-b receptors, and induces their degradation [41,42]. In addition, Smurf2 also decreases the protein levels of Smad2 in response to TGF-b stimulation. Our results showed that total Smad2 protein levels went down for 4 hours, responding to TGF-b stimulation in WT MEFs, but no significant differences in TBP-2 2/2 MEFs. TBP-2 contains two PPxY motifs, which are reported to interact with WW domain. TBP-2 interacts with Smurf2 in co-immnoprecipita-tion assay (data not shown), providing the hypothesis that TBP-2 is required for functions of Smurf2 in the negative feedback of TGFb signaling. The significance of TBP-2-Smurf2 interaction has been entirely unclear and will be examined in detail.
In conclusion, we demonstrated that TBP-2 deficiency enhances Smad2 phosphorylation level, resulting in acceleration of TGF-bdriven EMT. Our findings show a novel mechanisms of cancer suppression associated with TBP-2 and provide new insights into TGF-b-mediated EMT. TBP-2 is likely to be a prognosis indicator by monitoring TBP-2 expression in tumor, and a potential therapeutic target in the inhibition of EMT.

Reagents and Antibodies
TGF-b1 was purchased from R&D systems. Stealth small interfering RNA (siRNA) for TBP-2 (UCAAUUCGAGCAGA-GACAGACACCC) and a negative control were purchased from Invitrogen. The antibodies used were as follows: anti-phospho-Smad2 (Ser465/467) (138D4) and anti-Smad2 (L16D3) antibodies were purchased from Cell Signaling. Anti-Txnip antibody and Anti-Vimentin were from MBL. Anti-E-Cadherin antibody was from Transduction Laboratories. Anti-b-actin antibody was from Santa Cruz. Anti-a-tubulin antibody was from Sigma.

Cell Culture
Primary wild-type and TBP-2 2/2 mouse embryonic fibroblasts (MEFs) were generated as previously described [24]. Human lung adenocarcinoma cell line A549 was obtained from Health Science Research Bank. Human breast cancer cell line MDA-MB-231 was from DS Pharma Biomedical. MEFs, A549 and MDA-MB-231 cells were cultured in Dulbecco's modified Eagle's medium (DMEM) with 10% fetal bovine serum (FBS), 1% penicillin/ streptomycin antibiotics, and 2 mM L-glutamine. The culture was maintained at 37uC with 5% CO 2 .

RNA Interference
All knockdown assay using siRNAs were performed with Lopofectamine 2000 (Invitrogen) according to the manufacturer's instruction. The cells were used after 36 hours from transfection.

Transient Transfection and Luciferase Reporter Assay
Cells were transiently transfected with pGL3 96CAGA-MLP-Luc and pRL-TK (Promega) using TransIT-LT1 (Takara) according to the manufacturer's instruction. pRL-TK was used as a control of the efficiency of transfection. At the same time of transfection, cells were under the condition of serum deprivation. After 20 hours of transfection, cells were stimulated with TGF-b for 20 hours. Luciferase activity was measured with the Dual-Luciferase reporter system (Promega).

RNA Isolation, RT-PCR and Real-time Quantitative PCR
Total RNAs were extracted using TRIzol (Invitrogen), and were reverse-transcribed using High-Capacity cDNA Reverse Transcription Kits (Applied Biosystems) according to the manufacturer's instruction. Real-time PCR was performed with Power STBR Green PCR Master Mix (Applied Biosystems), using b-actin as an internal control for normalization. Fluorescent detection and data analyses were performed using ABI 7500 Sequence Detection System (Applied Biosystems). Primers for PCR analyses were listed in Table S1.

Immunoblotting Analysis
For western blotting, the cells were lysed in CelLytic M Cell Lysis Reagent (Sigma-Aldrich) containing a protease inhibitor cocktail (Roche) and phosphatase inhibitor (Nacalai Tesque). The lysate were boiled with Laemmli Smaple Buffer (BIO-RAD) at 95uC for 3 minutes. The samples were subjected to SDS-PAGE, transferred to PVDF membranes, and incubated with primary antibodies. The membranes were washed and incubated with horseradish peroxidase-conjugated secondary anti-mouse-or antirabbit-immunoglobulin G (GE Lifesciences). Finally, chemiluminescence was detected using Chemi-Lumi One L kit (Nacalai Tesque), and luminescence images were analyzed by LAS 3000 or LAS 4000 (GE Lifesciences). Figure S1 The length of the longest diagonal line of TBP-2 siRNA-A549 and control siRNA-A549 cells in the presence or absence of TGF-b (2.5 ng/ml for 36 hours).

Supporting Information
The length of each cell was calculated from expanded photos (200 cells). (TIF)