Cultured cell lines were obtained from the Japanese Collection of Research Bioresources Cell Bank. Recombinant human TGF-β1 was purchased from PeproTech. Protease inhibitor cocktail, for use with mammalian cell and tissue extracts, and phosphatase inhibitor cocktails 1 and 2 were purchased from Sigma-Aldrich. The Protein kinase inhibitor (PKA), H-89 was obtained from Santa Cruz Biotechnology Inc. and okadaic acid (OA) was procured from Merck (Calbiochem, KGaA). All other purchased reagents were of analytical grade.

All cell lines were grown at 37°C and 5% CO₂. Human HSC-4 SCC cells (JCRB0624) were cultured in Eagle's minimum essential medium (MEM; Sigma-Aldrich) supplemented with 10% fetal bovine serum (FBS; Gibco BRL). SAS cells (JCRB0260) were cultured in PRIM1640 medium (Gibco BRL) supplemented with 10% FBS. HO-1-N1 cells (JCRB0831) were cultured in Dulbecco's modified Eagle's medium (DMEM) and Ham's F-12 medium (1:1; Gibco BRL) with 10% FBS. The culture medium was removed and replaced with serum-free medium 24 h prior to the TGF-β1-stimulated experiments. For time-course experiments, 2.0×10⁵ hOSCC cells were cultured in 500 µl of medium without serum containing 10 ng/ml TGF-β1, for 1 to 48 h in 12 or 24-well tissue culture plates.

For total RNA preparation, 2.0×10⁵ cells were cultured in 24-well tissue culture plates. Total RNA was isolated using the ISOGEN reagent (Nippon Gene), according to manufacturer's instructions. RNA was reverse transcribed into first-strand cDNA using a RT-PCR System kit (Takara Bio Inc.). qPCR was performed on a Thermal Cycler Dice Real Time System (Takara Bio) using SYBR Premix Ex Taq II (Takara Bio) with human gene-specific primers (Table I). Target gene expression was normalized to an internal β-actin reference and expressed in terms of fold-change relative to the control sample (27).

The sense sequences of human Slug siRNA (MISSION siRNA, Hs_SNAIS_9785, Sigma-Aldrich), and Sox9 siRNA (siRNA, Life Technologies) are 5′-GCAUUUGCAGACAGGUCAATT-3′ and 5′-UGAAGAAGGAGAGCGAGGAGGACAA-3′, respectively. Logarithmically growing cells were seeded at a density of 1×10⁵ cells in 24-well tissue culture plates and transfected with 10 nM of a specific siRNA using Lipofectamine RNAiMAX (Life Technologies), according to manufacturer's instructions. Forty-eight hours after transfection, cells were stimulated using 10 ng/ml TGF-β1 and then were used for RT-qPCR analysis to analyze vimentin gene expression or for wound healing assay, as described below. Stealth™ RNAi Negative Control High GC Duplex (Life Technologies), which does not possess significant homology to vertebrate gene sequences, was used as a negative control. Suppression of gene expression by siRNA was evaluated by RT-qPCR and western blot analyses were performed for targeted molecules.

For western blot experiments, 3.0×10⁶ cells were lysed in RIPA buffer (Sigma-Aldrich) containing a protease and phosphatase inhibitor cocktail (Sigma-Aldrich). The protein content of the samples was measured using BCA reagent (Thermo Fisher Scientific, Inc.). For the preparation of cell lysates to examine marker proteins, 1.0×10⁶ cells were cultured in a 6-well plate in serum-free MEM with or without 10 ng/ml TGF-β1 for the indicated times. Cells were dissolved in SDS sample buffer containing a protease and phosphatase inhibitor cocktail (Sigma-Aldrich). Acrylamide gels of 12.5% (ATTO Co.) for SDS-PAGE were used for protein separation, and the proteins were subsequently transferred onto PVDF membranes (Merck). Membranes were probed with primary antibodies, including mouse anti-N-cadherin (1:250, H-2; Santa Cruz Biotechnology) and rabbit anti-Sox9 (1:1,000, AB5535; Chemicon International Inc.) antibodies, while a mouse anti-β-actin antibody (1:1,000, clone C4; Santa Cruz) was used as a loading control in siRNA experiments. The blots were then incubated with alkaline phosphatase-conjugated secondary antibody, and subsequently, signals were detected using an alkaline phosphatase substrate kit (BCIP/NBT Substrate kit; Vector Laboratories Inc.).

The Boyden chamber-based cell migration assays were performed as follows. First, the cells were transfected with Slug siRNA as described above. Then, they were treated with 10 ng/ml TGF-β1 under serum-free conditions for 48 h. Subsequently, the cells were plated at a density of 1.0×10⁵ cells in the upper chamber of a Boyden chamber apparatus in serum-free media and were allowed to migrate into a medium containing 10% FBS in the lower chamber for 24 h at 37°C. Following the 24 h incubation period, the filter was fixed in 4% paraformaldehyde and stained with DAPI for 10 min. The cells that migrated to the underside of the membrane were counted in nine random fields under a fluorescence microscope. Data are the average of triplicate experiments. The values indicate the mean number of migrating cells compared with control. The level of significance was determined using the Tukey's multiple comparison test.

Cells plated on 8-well chamber slides were incubated at 37°C for 24 h and then stimulated with 10 ng/ml TGF-β1 for an additional time period of 48 h. Slides were fixed with 4% paraformaldehyde at room temperature for 15 min. Cells were then incubated with specific antibodies with 1:200 dilutions of mouse anti-Slug (A-7; Santa Cruz), rabbit anti-Sox9 (H-90; Santa Cruz) and rabbit anti-phospho-Sox9 (Ser181) (CSB-PA050120; Cusbio Technology) antibodies for 16 h at 4°C. After rinsing with phosphate-buffered saline (PBS), cells were incubated with secondary antibodies Alexa Fluor^® 488 goat anti-mouse or anti-rabbit antibodies (1:1,000; Life Technologies) for 1 h at room temperature and then stained with DAPI (1:500, Sigma-Aldrich) for 10 min. Slides were then washed and imaged using a fluorescence microscope (IX70; Olympus).

All experiments were performed at least in triplicate. Results are expressed as mean ± standard deviation (SD). Differences between two groups (control and TGF-β1-treated cells) for the time course of Sox9 expression, and the expressions of Slug, Sox9 and N-cadherin in hOSCC cells were analyzed using unpaired two-tailed Student's t-test. Differences among multiple samples for the siRNA- or inhibitor-treated experiments were compared using Tukey's multiple comparison test following ANOVA with IBM SPSS Statistics 24 software. P<0.05 was considered to indicate a statistically significant difference.

We reported that Slug plays an important role in the TGF-β-induced downregulation of E-cadherin expression in HSC-4 cells (16,17). We also found that Slug increased the expression of the mesenchymal marker, vimentin, but not N-cadherin. We previously demonstrated that the same Slug siRNA significantly downregulated mRNA expression of Slug in HSC-4 cells (16). Therefore, we further examined whether Slug increased the expression of mesenchymal markers other than vimentin to verify the extent to which Slug affected the TGF-β-induced EMT in HSC-4 cells. RT-qPCR analysis revealed that the TGF-β-induced expression of mesenchymal markers fibronectin (Fig. 1A) and thrombospondin-1 (Fig. 1B) were significantly downregulated following the administration of Slug siRNA. The status of Laminin α3 was not affected by Slug siRNA (Fig. 1C), suggesting that transcription factors other than Slug must also be involved in the TGF-β-induced EMT in HSC-4 cells.

The mRNA expression of Sox9 was found to be significantly upregulated at 3–48 h after TGF-β1 stimulation (10 ng/ml) in HSC-4 cells (Fig. 2). The expression of Sox9 mRNA continuously increased between 3 and 24 h following TGF-β1 stimulation, peaking at 24 h and then decreased at 48 h after stimulation. We previously reported that TGF-β1 (10 ng/ml) increased the mRNA expression of EMT-related transcription factor, Slug, at 1.5 h following stimulation, in a Smad signal transduction mechanism-dependent manner, in HSC-4 cells (17). These results indicate that Sox9 is not a direct target of Smad signaling.

As shown in Fig. 3A, we confirmed that control siRNA did not affect Sox9 mRNA expression, and that Sox9 siRNA significantly suppressed Sox9 mRNA expression in HSC-4 cells. Notably, Sox9 siRNA significantly abrogated the TGF-β1-induced upregulation of N-cadherin mRNA expression (Fig. 3B). In addition, we confirmed that Sox9 siRNA suppressed the TGF-β1-induced upregulation of N-cadherin expression at protein level (Fig. 3C). TGF-β1-induced cell migration was significantly and incompletely decreased by Sox9 siRNA in HSC-4 cells (Fig. 3D). These results suggest that N-cadherin promotes the migration activity of HSC-4 cells partially through Sox9-dependent pathway.

We previously reported that SAS and HO-1-N1 were TGF-β1-responsive hOSCC cells: TGF-β1 upregulated expression of fibronectin and plasminogen activator inhibitor-1 (16). Here, we examined whether TGF-β1 upregulated Slug, Sox9, and N-cadherin in SAS cells and HO-1-N1 cells as in HSC-4 cells. We found that TGF-β1 (10 ng/ml) significantly upregulated the mRNA expressions of Slug, and N-cadherin at 24 h following stimulation with TGF-β1 (Fig. 4A, B and C). In contrast, TGF-β1 (10 ng/ml) significantly upregulated the mRNA expression of Sox9 in both HSC-4 cells and SAS cells, but not in HO-1-N1 cells (Fig. 4A, B and C), suggesting that the TGF-β1-induced upregulation of N-cadherin expression is not always dependent on the upregulation of Sox9 expression in hOSCC cells.

The transcription factor Sox9 is known to translocate from cytoplasm into the nucleus in response to TGF-β1 stimulation in interstitial cells (28). In addition, cyclic AMP-dependent protein kinase (PKA)-mediated phosphorylation of Sox9 plays an important role in its transcriptional activity (29). We confirmed that TGF-β1 (10 ng/ml) induced nuclear translocation of Slug, as previously reported (30) (Fig. 5A). Similarly, we found that TGF-β1 (10 ng/ml) promoted nuclear translocation of total Sox9 and Ser-181-phosphorylated, which is known to activate the transcriptional activity of Sox9-target genes in chondrocytes (31) (Fig. 5A and B). Furthermore, we found that TGF-β1 stimulation resulted in an increase in total Sox9 and pSox9 levels (Fig. 5A and B). Interestingly, the pretreatment of HSC-4 cells with the PKA inhibitor, H-89 (15 µM), before TGF-β1 stimulation inhibited the TGF-β1-induced phosphorylation and nuclear translocation of pSox9 (Fig. 5B). Further, we observed that the PKA inhibitor significantly abrogated the TGF-β1 (10 ng/ml)-induced upregulation of N-cadherin mRNA expression (Fig. 5C). It was previously reported that protein phosphatase A2 (PP2A) negatively regulates PKA activity (32,33). We confirmed that the PP2A inhibitor okadaic acid (50 nM) clearly and significantly enhanced TGF-β1 (10 ng/ml)-induced upregulation of N-cadherin mRNA expression (Fig. 5D). These results suggest that the TGF-β1 promoted phosphorylation and nuclear translocation of Sox9 occurs in a PKA-dependent manner, possibly resulting in the upregulation of N-cadherin expression in HSC-4 cells.