SOX2 mRNA expression in gastric cancer cell lines and primary gastric carcinoma tissues. (A) RT–PCR analysis of SOX2 mRNA levels in 10 gastric cancer cell lines and the normal stomach mucosae (normal st.). GAPDH expression was used as an internal loading control. RT (+ or −) indicates reverse transcriptase added or not, and H₂O indicates no RNA added. (B) Quantitative real-time RT–PCR analysis of SOX2 mRNA levels in primary gastric carcinoma samples and corresponding noncancerous gastric mucosae from the same patients. SOX2 expression levels were normalised by internal GAPDH expression. The assay was performed in triplicate, and the bars indicate s.d. (C) Representative results of the endpoint RT–PCR of SOX2 in primary gastric carcinomas (lanes Ca) and noncancerous gastric mucosae (lanes N).

Effects of SOX2 on cell cycle. (A) Cell-cycle analysis of parental, Ad-GFP- and Ad-SOX2-infected OUMS37 cells at 24 h after infection. DNA content was measured by propidium iodide staining on flow cytometry. The percentages of cell-cycle phases are shown in each panel. (B) Differential expression of the cell-cycle-regulatory proteins associated with SOX2 overexpression. Western blot analysis was performed to compare expression levels of cyclin D1, pRb, p27^Kip1 and p21^Cip1 proteins among parental, Ad-GFP- and Ad-SOX2-infected cells. Expression of α-tubulin was used as a protein loading control. (C) Western blot analysis of cell-cycle-regulatory proteins expression after transfection with SOX2 siRNA (S) or control siRNA (C) in MKN45 and TGBC11TKB cell lines. Ad-SOX2-infected NUGC3 cells (Ad-SOX2) were used as a positive control of the SOX2 protein band. Expression of α-tubulin was used as a protein loading control.

Methylation status of SOX2 in gastric cancer cell lines. (A) Schematic representation of the human SOX2 gene. Open and closed boxes indicate the untranslating and coding regions, respectively, and an arrow denotes the transcription start site (+1). Vertical bars show CpG sites. Arrows below the CpG sites indicate the regions subjected to MSP (MSP-A and MSP-B) and bisulphite sequencing (BS-A and BS-B). (B) Representative results of the demethylation analysis of SOX2. Gastric cancer cell lines were treated with (lanes A) or without (lanes M; mock) 5-aza-2′-deoxycytidine (5 μM), and SOX2 expression was examined by RT–PCR as described in Figure 1A. p27^Kip1 expression was also examined. (C) MSP analyses in gastric cancer cell lines and the normal stomach. The primer region is depicted in panel A. Bands in the ‘Mt' lanes, as shown by arrowheads, are PCR products obtained with methylation-specific primers; those in the ‘U' lanes are obtained with unmethylated-specific primers. (D) Sodium bisulphite DNA sequencing of SOX2 in gastric cancer cell lines (KATOIII, MKN74 and HSC59). The analysed regions are shown in panel A. Each horizontal row of circles represents analysis, in single clone of bisulphite-treated DNA, of 22 or 41 CpG sites (for BS-A or BS-B, respectively) contained in the regions. Solid and open circles represent methylated and unmethylated CpG sites, respectively.

Effects of SOX2 overexpression on the proliferation of gastric epithelial cell lines. (A) Morphological changes in OUMS37, NUGC3 and GCIY cell lines after infection with the Ad-SOX2 or Ad-GFP. The MOI of the adenovirus vectors were 80 for OUMS37 and 20 for NUGC3 and GCIY, respectively. The left and right panels show the phase contrast micrographs appearance and fluorescent micrographs for GFP expression in the same fields. (B) Western blot analysis of SOX2 protein expression after adenovirus-mediated overexpression in OUMS37, NUGC3 and GCIY cell lines. SOX2 protein expression levels were compared among parental cells (P), Ad-GFP-infected cells (MOI: G80 for OUMS37, and G20 for NUGC3 and GCIY, respectively) and Ad-SOX2-infected cells (MOI: S40 and S80 for OUMS37, and S10 and S20 for NUGC3 and GCIY, respectively) at 48 h after adenovirus infection. α-tubulin expression was used as a protein loading control. (C) In vitro cell proliferation assay after SOX2 overexpression in gastric cell lines. The number of viable cells was measured by WST-1 assay of 1, 3 and 5 days after adenovirus infection. The assay was performed in triplicate, and bars indicate s.d.

Analyses of apoptosis associated with SOX2 overexpression in OUMS37 cells. (A) Percentage of the sub-G₁ DNA fraction in OUMS37 cells at 48 and 72 h after infection. DNA was stained with propidium iodide and the sub-G₁ fraction was measured by flow cytometry. The percentages of the sub-G₁ fractions, indicated by the bars, are shown in each panel. (B) Apoptosis detection by APOPercentage Apoptosis assay. Top panels: phase contrast micrographs indicate representative areas of the control (Ad-GFP 40MOI) and SOX2-overexpressing (Ad-SOX2 40MOI) OUMS37 cells at 72 h after adenovirus infection. Bottom panel: purple-red-stained apoptotic cells were counted in multiple randomly selected fields and the percentage (per total cells) was determined (columns, mean of 10 fields; bars, s.d.). (C) DNA ladder assay by agarose gel electrophoresis. DNA extracted from the control (parental and Ad-GFP 40MOI for 72 h) and SOX2-overexpressing (Ad-SOX2 40MOI for 48 and 72 h) OUMS37 cells was electrophoresed on a 2% agarose gel and detected by ethidium bromide staining. Arrows indicate the molecular weights of DNA base pairs. (D) Detection of activated caspase-3 in SOX2-overexpressing OUMS37 cells on western blot analysis. The reduction of caspase-3 and accompanying appearance of cleaved-caspase-3 indicate activation of caspase-3. The reduction of a caspase-3 substrate, PARP, also reflects activation of caspase-3. Expression of α-tubulin was used as a protein loading control.

Methylation status of SOX2 in primary gastric carcinoma tissues, and its involvement in prognosis. (A) Representative results of MSP analysis of SOX2 in primary gastric carcinomas (lanes Ca) and corresponding noncancerous gastric mucosae (lanes N). All analyses were carried out by using MSP-B primers (shown in Figure 5A). (B) Sodium bisulphite DNA sequencing of SOX2 in noncancerous gastric mucosae (Case 8N) and primary gastric carcinoma (Case 8Ca) tissue samples. The analysed regions (BS-B) are shown in Figure 5A. Each horizontal row of circles represents analysis, in single clone of bisulphite-treated DNA, of 41 CpG sites contained in the regions. Solid and open circles represent methylated and unmethylated CpG sites, respectively. (C) Kaplan–Meier survival analysis of SOX2 methylation in 52 advanced gastric cancer patients. Patients with SOX2 methylation (Mt; dotted) had a significantly poorer outcome than those without methylation (U; solid) (P=0.0062).