Thalidomide Inhibits TGF-β1-induced Epithelial to Mesenchymal Transition in Alveolar Epithelial Cells via Smad-Dependent and Smad-Independent Signaling Pathways

Recent evidence indicates that the epithelial to mesenchymal transition (EMT) in primary alveolar cells (AECs) plays an important role in idiopathic pulmonary fibrosis (IPF). In vivo models have suggested that thalidomide (THL) has anti-fibrotic effects against pulmonary fibrosis, but the underlying mechanism of this effect is not clear. This study investigated whether THL regulates alveolar EMT and the possible mechanisms underlying this process. CCL-149 cells were treated with TGF-β1 in the presence of THL at the indicated concentrations. EMT was assessed by changes in cell morphology and in phenotypic markers. Signaling pathways involved in EMT were characterized by western blot analysis. THL inhibited the TGF-β1 induction of α-SMA, vimentin, MMP-2/-9 and collagen type IV expression and restored the morphological changes in primary alveolar epithelial cells caused by TGF-β1. TGF-β1 induction of α-SMA expression was partially dependent on the activation of p38, JNK, ERK, Akt, Smad 2 and Smad3. Moreover, THL inhibited TGF-β1-induced phosphorylation of p38, JNK, ERK, Akt, GSK3β, Smad 2 and Smad3 without altering the total expression levels of those proteins. These findings indicate that TGF-β1-induced EMT in alveolar epithelial cells is inhibited by THL via both Smad-dependent and non-Smad-dependent signaling pathways and suggests therapeutic approaches for targeting this process in pulmonary fibrosis.

Thalidomide (THL) has been recognized as a therapeutic agent for multiple myeloma (MM). Recent studies have shown that thalidomide has anti-fibrotic effects in animal models of bleomycin-induced 12 and paraquat-induced pulmonary fibrosis 13 . Although the anti-fibrotic effects of thalidomide in pulmonary fibrosis have been demonstrated, the exact mechanism of this effect remains poorly understood. The results of previous studies have suggested that the anti-fibrotic effects of thalidomide on pulmonary fibrosis might be related to suppression of the ERK1/2 signaling pathway 14 . Liu et al. reported that thalidomide is able to inhibit JNK signaling in an animal model of pulmonary fibrosis 15 . In addition, a study by Liang et al. suggested that thalidomide has an anti-fibrotic effect that is caused by suppression of TGF-β1-induced p38 and Smad3 signaling 16 . Considering the anti-fibrotic effect of thalidomide in pulmonary fibrosis, the possible mechanisms of that effect and the characteristics of EMT in pulmonary fibrosis, we hypothesized that thalidomide inhibits TGF-β1-induced EMT in AECs. In the present study, we induced EMT in vitro to investigate the potential effects of thalidomide on alveolar EMT and the possible mechanisms of this effect.

Results
Thalidomide reversed TGF-β1-induced EMT in AECs. Stimulation with 5 ng/mL TGF-β1 for 48 h resulted in apparent changes in cellular morphology and changes in the expression of cellular markers. As seen in Fig. 1A, treatment with thalidomide restored the morphological changes observed in cells after TGF-β1 stimulation. The majority of TGF-β1-treated cells in the thalidomide-treated groups showed an epithelial morphology instead of a fibroblast-like morphology. Western-blot assays showed that thalidomide significantly inhibited the TGF-β1-induced increases in α-SMA and vimentin in a dose-dependent manner (p < 0.05 for both α-SMA and vimentin). Treatment with thalidomide significantly prevented the TGF-β1-induced decrease in TGF-β1 (5 ng/mL) were treated with THL (1, 10, or 100 μg/mL) or SB431542 (5 μM) for 48 hours, and the lysates were analyzed for ZO-1, E-Cadherin, α-SMA and vimentin. β-Actin was used as the loading control. Data are expressed as the means ± SD, and n = 3 in each group. THL: thalidomide; SB: SB431542, TGF-β1R-Kinase inhibitor; ZO-1: zonula occludens-1; α-SMA: α-smooth muscle actin; # p < 0.05, ## p < 0.01 compared to the TGF-β1 group (untreated); * p < 0.05, ** p < 0.01 compared to all other groups. the epithelial markers ZO-1 and E-Cadherin (p < 0.05 for both ZO-1 and E-Cadherin) (Fig. 1B). In addition, ELISAs showed that pretreatment with TGF-β1 significantly increased the secretion of MMP-2, MMP-9 and collagen type IV in CCL-149 cells. However, thalidomide abrogated these changes in a dose-dependent manner (p < 0.05) (Fig. 2). Moreover, immunofluorescence microscopy demonstrated localized expression of the epithelial marker E-Cadherin and the mesenchymal marker α-SMA in CCL-149 cells (Fig. 3). Cells in the control group expressed significant amounts of E-Cadherin and low levels of α-SMA on their cell membranes. TGF-β1 markedly reduced the expression of E-Cadherin and increased the expression of α-SMA. However, cells concurrently Figure 2. ELISA analysis. CCL-149 cells stimulated with TGF-β1 (5 ng/mL) were treated with THL (1, 10, or 100 μg/mL) or SB431542 (5 μM) for 48 hours, and then the cell culture media were collected for ELISA analysis of MMP-2 (A), MMP-9 (B) and collagen type IV (C) secretions. THL: thalidomide; SB431542: TGF-β1R-Kinase inhibitor; MMP-2, and -9: matrix metalloproteinase-2, and -9; # p < 0.05, ## p < 0.01 compared to the TGF-β1 group (untreated); *p < 0.05, **p < 0.01 compared to all other groups. . Thalidomide (THL) reverts TGF-β1-induced EMT as detected by confocal immunofluorescence microscopy. CCL-149 cells were grown on coverslips and treated with TGF-β1 in the presence or absence of THL (100 μg/mL) for 48 h. Cells were fixed and stained with α-SMA and E-Cadherin antibodies followed by incubation with fluorescently-tagged secondary antibodies. DAPI was used to stain nuclei. α-SMA (red), E-Cadherin (green), and nuclei (blue) were visualized by confocal fluorescence microscopy (400×). treated with TGF-β1 and thalidomide retained high levels of expression of the epithelial marker E-Cadherin without a significant increase in the expression of the mesenchymal marker α-SMA.
the role of p38, JNK, and ERK in TGF-β1-induced EMT. When cells were treated with inhibitors specific for these proteins, including SB203580 (p38 inhibitor), PD098059 (JNK inhibitor) and SP600125 (ERK inhibitor), the activation of p38, JNK and ERK by TGF-β1 and the expression of the mesenchymal marker α-SMA was significantly inhibited (p < 0.05) (Fig. 4A). Importantly, we observed that thalidomide inhibited the activation of these MAPKs after TGF-β1 treatment in a dose-dependent manner (p < 0.05) without altering the total expression levels of those proteins (Fig. 4B). These results indicate that thalidomide might inhibit TGF-β1-induced EMT in pulmonary epithelial cells through the inhibition of MAPK signaling pathways.

Thalidomide inhibited TGF-β1-induced phosphorylation of Akt and GSK-3β. A previous study
confirmed that TGF-β1 induces the phosphorylation of Akt and its downstream protein GSK-3β in primary AECs 17 . In this study, we also found that TGF-β1 induced phosphorylation of Akt in CCL-149 cells in vitro. In addition, treatment with the PI3-K/Akt specific inhibitor LY294002 significantly inhibited the TGF-β1-induced phosphorylation of Akt and the subsequent induction of α-SMA expression (Fig. 5A). Furthermore, TGF-β1 treatment also induced the phosphorylation of GSK-3β in AECs. These results suggest that the Akt/GSK-3β pathway is involved in TGF-β1-induced EMT in CCL-149 cells. However, treatment with thalidomide inhibited TGF-β1-induced Akt and GSK-3β phosphorylation ( Fig. 5B and C) in a dose-dependent manner (p < 0.05). In addition, as shown above, thalidomide also inhibited TGF-β1-induced production of α-SMA and vimentin and restored the morphological changes caused by TGF-β1. Thus, we conclude that the Akt/GSK-3β pathway is involved in thalidomide-mediated anti-EMT effects. Thalidomide inhibited TGF-β1-induced phosphorylation of Smad 2/3. Smad transcription factors are well-known mediators of TGF-β1 signaling. A previous study showed that Smads are activated by TGF-β1 and regulate EMT-related changes in pulmonary epithelial cells 18 . To confirm the role of Smad2/3 in TGF-β1-induced EMT, Smad2 and Smad3 were knocked down using siRNAs. As seen in Fig. 6A, the knockdown of Smad2 and Smad3 proteins was comparable in cells with or without TGF-β1 treatment. We also investigated whether co-treatment with Smad2 and Smad3 siRNAs enhanced the knockdown of target genes in this study. However, no significant changes were observed in cells co-treated with Smad2/3 siRNAs compared to cells treated with single siRNAs (data not shown). Moreover, we investigated the effects of the knockdown of Smad2 and Smad3 on TGF-β1-induced α-SMA expression. The TGF-β1-induced increase in α-SMA expression was significantly attenuated by knockdown of both Smad2 and Smad3. Furthermore, the TGF-β1-induced increase in α-SMA expression was completely inhibited in cells transfected with both Smad2 and Smad3 siRNAs (Fig. 6B). These results confirm that both Smad2 and Smad3 are involved in the TGF-β1-induced increase in α-SMA expression in CCL-149 cells. Importantly, cells concurrently treated with TGF-β1 and thalidomide showed significantly lower levels of α-SMA expression and decreased phosphorylation of Smad2/3 compared to cells treated with Cells transfected with Smad2 siRNA and/or Smad3 siRNA were treated with or without TGF-β1 (5 ng/mL) for 48 h. Cells were lysed and α-SMA levels were evaluated by western blot. (C) Cells were stimulated with TGF-β1 (5 ng/mL) in the presence of vehicle, SB431542 (5 μM) or increasing doses of THL (1, 10, and 100 μg/ mL) for 48 h. Then, cells were lysed and the phosphorylation of Smad2 and Smad3 was evaluated by western blot. β-Actin was used as a loading control. Data are expressed as the means ± SD, and n = 3 in each group. THL: thalidomide; SB431542: TGF-β1R-Kinase inhibitor; α-SMA: α-smooth muscle actin; # p < 0.05, ## p < 0.01 compared to the TGF-β1 group (untreated); * p < 0.05, ** p < 0.01 compared to all other groups.

Discussion
This study is the first to demonstrate that thalidomide inhibits TGF-β1-induced alveolar epithelial to mesenchymal transition (EMT) in CCL-149 cells. Thalidomide significantly reduced the expression of mesenchymal markers, increased the expression of epithelial markers after TGF-β1 stimulation, and reverted the TGF-β1-induced morphological changes in alveolar epithelial cells. We also found that the inhibitory effects of thalidomide on TGF-β1-induced EMT may be related to inhibition of the TGF-β1-mediated signaling pathways Smad2/3, Akt/ GSK-3β and MAPK.
Currently, increasing attention has been given to the process of EMT and its possible role in the pathogenesis of pulmonary fibrosis. Numerous studies have reported EMT as a response to TGF-β1 treatment in vitro in alveolar epithelial cells 19,20 . The mechanism of TGF-β1-induced EMT is complicated because TGF-β1 signals are transmitted through multiple pathways, including the Smad, MAPK and PI3K pathways. The TGF-β1 signal itself is predominantly transduced by the Smad proteins, including Smad2, Smad3 and Smad4 21 . Binding of TGF-β1 to its receptors results in the phosphorylation of Smad proteins, causing translocation of the Smad complex into the nucleus where it regulates the expression of target genes. Although both Smad2 and Smad3 are important for the regulation of TGF-β1 signals, the mechanisms of action and the functions of these two Smads are different. Experimental Smad3-deficient mouse models of EMT and fibrosis demonstrated that Smad3 signaling is important in fibrosis 22 ; however, the role of Smad2 in TGF-β1-induced EMT remains unclear. In the present study, we observed that the TGF-β1-induced phosphorylation of Smad2 and Smad3 and the TGF-β1-induced increase in α-SMA production were partially dependent on Smad2 and Smad3 and were reversed by knockdown of both Smad2 and Smad3. These results are consistent with those of previous reports 23 . Moreover, the results of our study suggest that thalidomide inhibits TGF-β1-induced EMT in pulmonary epithelial cells by blocking the phosphorylation of both Smad2 and Smad3. However, previous studies demonstrated that Smad3, but not Smad2, is required for the fibrotic response to TGF-β1 24,25 . In addition, a study of alterations in gene expression during EMT in IHAEo − cells showed that only collagen I gene expression was Smad2/3 dependent; all other alterations in gene expression during EMT appeared to be Smad independent 18 . The discrepancies between the findings of these studies may be due to the cell types used. However, the Smad dependency of EMT changes in CCL-149 cells (other than those studied here) requires further investigation.
In addition to the Smad pathways, TGF-β utilizes several non-Smad pathways, such as the MAP kinase and phosphatidylinositol-3-kinase/AKT pathways, to regulate a wide array of cellular functions. To date, the best-characterized non-Smad pathway utilized by TGF-β is the p38 MAPK and JNK signaling cascade. It is well known that TGF-β can rapidly activate p38 MAPK and JNK through MKK3/6 26 and MKK4 27 , respectively. In addition, the p38 MAPK and JNK pathways play very important roles in TGF-β-induced EMT. Inhibiting p38 or JNK activity using specific inhibitors or gene silencing can inhibit TGF-β-induced changes in cell morphology and expression of phenotypic markers 18,28 . Similar to p38 and JNK, the ERK pathway is another non-Smad signaling pathway necessary for TGF-β-induced EMT 29 . Liang et al. reported that thalidomide treatment significantly inhibited TGF-β-induced p38 phosphorylation but did not have a significant effect on ERK or JNK phosphorylation 24 . An experimental in vivo pulmonary fibrosis rat model demonstrated that thalidomide reduced the degree of bleomycin-induced pulmonary fibrosis by down-regulating the expression of p-JNK and α-SMA 15 . Moreover, Choe et al. reported that the anti-fibrotic effect of thalidomide on pulmonary fibrosis might be related to its inhibition of TGF-β1-induced activation of the ERK1/2 signaling pathway 14 .
In the present study, we found that TGF-β1 activated these MAPK cascades and that the TGF-β1-induced increase in the expression of mesenchymal markers, including α-SMA and vimentin, was inhibited by p38-, JNK-and ERK-specific inhibitors. This suggests that the TGF-β1-induced increase in the expression of α-SMA and vimentin is mediated by the activation of these MAPKs. We also observed that thalidomide treatment resulted in a significant reduction in the levels of phosphorylated p38, JNK and ERK in CCL-149 cells, indicating that the inhibitory effect of thalidomide on TGF-β1-induced EMT is associated with the suppression of MAPK signaling pathways.
Furthermore, several studies have suggested that the PI3K/Akt pathway is another non-Smad pathway that contributes to TGF-β1-induced EMT. Inhibition of Akt activity attenuates TGF-β1-induced EMT in kidney epithelial cells 30 . In this study, we observed that thalidomide inhibited TGF-β1-induced phosphorylation of Akt and that a specific PI3K inhibitor inhibited Akt activation and α-SMA production after TGF-β1 stimulation. This suggests that the attenuation of TGF-β1-induced EMT by thalidomide is associated with the PI3K/Akt pathway. In addition, the activation of GSK-3β, a downstream effector of Akt, was suppressed by thalidomide treatment. Previous studies have shown that inactivation of GSK-3β could lead to the stabilization of key mediators of EMT, including SNAI1 and β-catenin 31,32 . These findings lead us to hypothesize that the inhibitory effect of thalidomide on TGF-β1-induced EMT may be partly mediated by inhibition of the PI3-K/ Akt/GSK-3β pathway.
In a bleomycin model of pulmonary fibrosis, approximately one third of fibroblasts are of epithelial origin 33 . EMT may also contribute to fibroblastic foci and fibrotic development in idiopathic pulmonary fibrosis (IPF) 34,35 . However, Rock et al. suggested that epithelial cells do not generate myofibroblasts in the context of pulmonary fibrosis 9 . These contradictory results may be due to differences in the study methods, the pulmonary fibrosis model, or the cell lines. Collectively, the role of EMT in the development of pulmonary fibrosis still needs to be confirmed in future studies. Our study provides some insight into the mechanism by which thalidomide is able to prevent EMT in a pulmonary epithelial cell line. Finally, there were three main limitations in this study. First, although we identified that both Smad-dependent and non-Smad-dependent pathways were involved in Scientific REPORTS | 7: 14727 | DOI:10.1038/s41598-017-15239-2 TGF-β1-induced EMT in pulmonary epithelial cell lines, we failed to determine which pathway plays the main role in this pathological process. Second, further studies are required to investigate the relationship between the Smad-dependent and non-Smad-dependent pathways during TGF-β1-induced EMT in pulmonary epithelial cells. Third, we used a fetal rat pulmonary epithelial cell line (CCL-149) instead of a human pulmonary epithelial cell line (such as A549) in this study. Thus, the different characteristics between species should be considered.

Conclusions
The results of this study demonstrate that treatment with thalidomide inhibits TGF-β1-induced α-SMA and vimentin production and TGF-β1-induced cell morphological changes in pulmonary epithelial cells by suppressing both Smad-dependent and non-Smad-dependent pathways. Thus, thalidomide may be valuable as a therapeutic agent for the inhibition of lung fibrosis.

Cell culture and treatments. Rat lung epithelial cells (CCL-149) were purchased from Wuhan Biofavor
Biotech Service CO., LTD (Wuhan, Hubei, China). Cells were cultured in DMEM/F12 supplemented with 10% fetal bovine serum (Invitrogen, San Diego, CA) and 1% streptomycin-penicillin solution. Cultures were maintained in a humidified 5% CO 2 incubator at 37°C. After 24 hours of serum starvation, cells were maintained in growth media supplemented with 5 ng/mL TGF-β1 (dissolved in 10 mM citric acid, pH = 3.0) for 48 h to induce EMT 36 . In the majority of experiments, cells were maintained in media supplemented with TGF-β1 with or without thalidomide (1, 10, or 100 μg/mL) for 48 h to investigate the effects of thalidomide on EMT in CCL-149 epithelial cells. To investigate pathways possibly involved in the effect of thalidomide on EMT, the cells were preincubated for 1 h with SB431542 (5 μM), LY294002 (3 μM), SB203580 (20 μM), SP600125 (10 μM) and PD098059 (40 μM) before treatment with exogenous TGF-β1 with or without thalidomide. The concentrations of thalidomide and inhibitors were determined according to the results of our pilot study. Cells were harvested at the indicated time points for protein extraction and western blotting and the cell culture media was collected for ELISA analysis.
Smad2 and Smad3 siRNA transfection. Pre-designed Smad2 (Ambion, Austin, TX) and Smad3 (Thermo, PA, USA) siRNAs were transfected into cells using siPORT ™ NeoFX ™ Transfection Agent (Ambion) at a concentration of 20 nM. The siRNA-transfection reagent complexes were prepared as described by the manufacturer. Cells were treated with or without 5 ng/mL TGF-β1 in the presence or absence of different concentrations of thalidomide for 48 h. The efficiency of siRNA gene silencing was evaluated with western blots for Smad2 and Smad3. The effect of siRNA transfection on TGF-β1-induced α-SMA production was analyzed by western blotting as described below.
ELISA. As described above, the cell culture media was collected for the detection of MMP-2, MMP-9 and collagen type IV secreted from CCL-149 cells in response to pretreatment with TGF-β1 and thalidomide. Culture media was centrifuged at 3000 rpm for 20 min. Quantification of MMP-2, MMP-9 and collagen type IV was performed using commercial ELISA kits according to the manufacturer's instructions.
Briefly, cells cultured on coverslips were washed thrice with PBS and fixed with paraformaldehyde (4%) for 15 min. Following three washes with PBS, the cells were blocked with 0.1% Triton X-100 and 2% normal serum for 30 min at room temperature (RT) and then incubated with specific primary antibodies, including those against α-SMA (BM0002, Boster, Wuhan, China) and E-Cadherin (sc-7870, Santa Cruz, CA, USA), at 4°C overnight. The cells were then stained with the appropriate secondary antibodies (Boster, Wuhan, China). The cells were also stained with 4,6-diamidino-2-phenylindole (DAPI) to visualize the nuclei and were washed with PBST four times for 5 min each time. The slides were viewed with an epi-fluorescence microscope (Nikon Eclipse 80i, Nikon, Tokyo, Japan) equipped with a digital camera (DS-Ri1, Nikon, Tokyo, Japan).
Statistical analysis. Data are expressed as the means ± SD. Data were analyzed using the statistical analysis software SPSS version 20.0 (SPSS Inc., Chicago, IL, USA). Comparisons of several means were performed using one-way and repeated measure two-way analysis of variance followed by the Tukey-Kramer test to identify significant differences between groups. All P values were two-tailed, and a P value of less than 0.05 was considered significant.