RETRACTED ARTICLE: K-ras-ERK1/2 down-regulates H2A.X^Y142ph through WSTF to promote the progress of gastric cancer

Human GC cell line SNU-16 and MKN1 cells were purchased from Shanghai Institute for Biological Science (Shanghai, China). Cells were maintained at 37 °C, 5% CO₂ in RPMI-1640 (Gibco Laborato-ties, Grand Island, NY) with 100 units/ml penicillin, 100 μg/ml streptomycin and 10% fetal bovine serum (FBS, Life Science, UT, USA).

Empty-pEGFP-N1 vector, pEGFP-K-Ras^WT, pEGFP-K-Ras^G12V/T35S plasmids were transfected in cells. The transfection with specific gene with HA-tag was used for screening out the target factor through western blot. pEGFP-K-Ras^G12V/T35S plasmids were obtained by site-directed mutagenesis. SiRNAs (Shanghai GenePharma, Shanghai, China) refers to using interference RNA to silence the goal RNA (Mouse double minute 2 homolog (MDM2) or EYA3). The pEGFP-H2A.X^Y142A construct was constructed using the TaKaRa MutanBEST Kit (#D401) (TaKaRa, Shiga, Japan) as recommend by the manufacturer.

Cells at the density of 5 × 10⁵ per well were cultured in 6-well plates for 12 h in the darkness. Then the cells were transfected with plasmids or siRNA using Lipofectamine 2000 (Invitrogen, Carlsbad, CA). Then qRT-PCR and western blot were used to determine the transfection efficiency after 48 h of transfection.

MTT (Sigma-Aldrich, St Louis, MO, USA) was used for the detecting cell viability. After cells were cultured for 48 h, 20 μl 5 mg/mL MTT was administrated to each well. Cells were cultured for 4 h. Afterward, we used 100 μl dimethyl sulfoxide (Sigma-Aldrich, St Louis, MO, USA) to lyse formazan crystal. The value was obtained at 570 nm by a multiwell spectrophotometer (Emax; Molecular Devices, Sunnyvale, CA).

RT-PCR method was referred to what described in the report [17]. Total RNA was obtained from SNU-16 cells using TRIzol (Invitrogen) reagent. DNase-I-treated total RNA was supplied for first-strand cDNA synthesis by M-MuLV reverse transcriptase (Fermentas, York, UK) and oligo-dT primers (Invitrogen). QuantiTect SYBR Green PCR Kit (Qiagen, Hilden, Germany) was used to amplify the target sequence. GAPDH was an internal control for detecting RNA expression based on triplicate experiments.

Soft-agar assay was performed to measure the cell colonies ability [18]. The cells suspended in full culture medium with 0.35% low-melting agarose, then cells were transferred into solidified 0.6% agarose in six-well culture plates (1 × 10³ cells/well). The number of the colonies was counted 3 weeks later using microscopically (40 ×).

Cell migration was evaluated by using a modified two-chamber migration Transwell (Corning Costa, NY, US) with a pore size of 8 μm. 100 μl (around 2 × 10⁵ cells/ml) cell suspension without serum was added to upper Transwell. 600 μl complete medium was added in the lower compartment. Cells were maintained for 24 h at 37 °C, 5% CO₂. After incubation, cells at the upper surface of the filter were removed by a cotton swab, and the filter was fixed with methanol for 5 min. Cells at the lower surface of the filter were stained by 0.1% Giemsa (Sigma-Aldrich) for 15 min. Cells were counted by 100 × microscope.

Protein was obtained using RIPA lysis buffer (Cat. No:R0010, Solarbio, Beijing, China) supplemented with protease inhibitors (Thermo Fisher Scientific, Rockford, IL). The BCA™ Protein Assay Kit (Pierce, Appleton, WI, USA) was used for determining protein concentrations. Western blot system was established using a Bio-Rad Bis-Tris Gel system following the manufacturer’s instructions. Primary antibodies were prepared in 5% blocking buffer and diluted according to the product instruction. These primary antibodies were incubated in membrane and maintained at 4 °C overnight at recommended concentration. Then secondary antibody incubated with horseradish peroxidase (HRP) conjugated secondary antibody. Captured the signals, and Image Lab™ Software (Bio-Rad, Shanghai, China) quantified the intensity of the bands.

SNU-16 cells were cultured until reach 75–80% confluence, and then cells were washed by PBS to remove the non-adherent cells. Collected cells were all adherent cells and fixed with cold 70% ethanol. Cells were washed with PBS again. After that, cells were stained with 4, 6-diamidino-2-phenylindole (DAPI) (Partec, Germany) and cultured in the darkness for 30 min. Flow cytometry was used for detecting cell cycle distribution. The percentage of cells in different cell cycle stages was calculated.

Cells were fixed in 1% formaldehyde, then lysed, and sonicated cells. Then 5 mg chromatin was immunoprecipitated with Dynabeads. After that, purified DNA was used for PCR amplification at the CYR61, IGFBP3, WNT16B, NT5E, GDF15, CARD16 promoter. The detailed process could refer to the literature [19].

Data was analyzed by Graphpad 6.0 statistical software (GraphPad, San Diego, CA, USA). Data were present as mean + SD. The statistical analyses were performed using the Student’s t-test or one way ANOVA followed by Duncan post-hoc of multiple comparisons. A P value of < 0.05 was considered significant (* P < 0.05, ** P < 0.01 or ***P < 0.001).

Ras/ERK signal pathway was often found to be closely related with GC [20]. In our study, SNU-16 and MKN1 cells were respectivly transfected with empty-pEGFP-N1 vector, pEGFP-K-Ras^WT and pEGFP-K-Ras^G12V/T35S plasmids. Results showed that Ras^G12V/T35S decreased the expression of H2A.X^Y142ph level (P < 0.01, Fig. 1a). In addition, we measured the effects of Ras^G12V/T35S activated phosphorylation of ERK1/2 (Fig. 1b). To confirm this suggestion, another cell line MKN1 was used and similar results were also observed in this cell line as what we describe in SNU-16 cells (P < 0.01, Fig. 1c-d), which suggested that Ras^G12V/T35S played a role as a switch for the phosphorylation of ERK.

Histone modification influenced cell growth and cell metastasis [21]. In the current study, we constructed H2A.X^Y142A plasmids to mimic the situation of the phosphorylation of H2A.X^Y142. The mimicked H2A.X^Y142A plasmids were co-transfected with Ras^G12V/T35S plasmids into SNU-16 cells and MKN1 cells. Hence, we examined the effects of phosphorylation of histone H2A in GC cell viability, cell colony ability and cell migration. Firstly, SNU-16 cell was transfected with H2A.X^Y142A.With the increasing concentration, the expression of phosphorylation of H2A.X^Y142 was inhibited in a dose-dependent manner (Fig. 2a). Moreover, results showed that Ras/ERK pathway significantly increased cell viability (P < 0.001) while H2A.X^Y142A decreased cell viability to some extent in SNU-16 cells (P < 0.001, Fig. 2b). This result suggested that Ras/ERK has the ability to increase cell viability while H2A.X^Y142A decreased cell viability, which indicating H2A.XY142A suppressed cell growth in GC. In addition, the number of colonies (P < 0.01, Fig. 2c) and cell migration (P < 0.001, Fig. 2d) revealed the similar trend by Ras^G12V/T35S pathway in SNU-16 cells. Similar results were also observed in MKN1 cell line (P < 0.05 or P < 0.01, Fig. 2e-h). Taken together, we inferred that the phosphorylation of H2A.X^Y142 was involved in the progression of GC cells.

Ras/ERK pathway was a complex and exact regulation pathway which was modulated by diverse downstream factors [22]. We explored the expression of these important downstream factors CYR61 [23], IGFBP3 [24], WNT16B [25], NT5E [26], GDF15 [27], CARD16 [28], which are involved in the tumor cell growth and cell metastasis. Then, we found that Ras^G12V/T35S increased the expression of WNT16B (P < 0.05), NT5E (P < 0.01) while decreased the expression of IGFBP3 (P < 0.01), GDF15 (P < 0.01) and CARD16 (P < 0.01). Meanwhile, we found that co-transfection with Ras^G12V/T35S and H2A.X^Y142A decreased the expression of CYR61 (P < 0.05), IGFBP (P < 0.001), WNT16B (P < 0.05) and increased the expression of NT5E (P < 0.05), GDF15 (P < 0.05) and CARD16 (P < 0.05) as relative to Ras^G12V/T35S group, which suggested that H2A.X^Y142A was down-regulated these downstream genes of Ras^G12V/T35S pathway. In addition, the ChIP assay results showed Ras^G12V/T35S decreased expression of these genes (P < 0.05, P < 0.01 or P < 0.001, Fig. 3b), this suggested that H2A.X^Y142ph could down-regulate gene expression via directly binding to these gene’s promoters. In conclusion, H2A.X^Y142ph could down-regulate the downstream transcription progression of the downstream factors.

Previous study reported that the effects of dephosphorylation of histone H2A by the EYA1/3 also address significant influence in human cancers [16]. Thereafter, knock-down EYA3 expression to investigate whether upregulation of phosphorylation of H2A.X^Y142 could modulate cell phenotype was performed. qRT-PCR and western blot was used to determine the transfection efficiency. We used two interference miRNAs to silence the expression of EYA3, which named si-EYA3–1 and EYA3–2, respectively. Result in Fig. 4a showed that si-EYA3–1 and si-EYA3–2 both decreased EYA3 expression in mRNA level and adding si-EYA3–1 and si-EYA3–2 both up-regulated the phosphorylation of H2A.X^Y142. After that, we detected the effects of si-EYA3–1/2 on cell viability, cell migration, cell cycle stage and cell relative mRNA levels. Results showed that si-EYA3–1/2 decreased cell viability (P < 0.05), cell migration (P < 0.05) and cell percent in S stage as compared with Ras^G12V/T35S alone (Fig. 4b-d). Furthermore, EYA3–1/2 also altered the expression of the downstream factors of Ras^G12D/T35V pathway compared with Ras^G12V/T35S alone in GC cells (P < 0.05, P < 0.01 or P < 0.001, Fig. 4e), which suggesting that EYA3 also participated in the regulation of the downstream factors of Ras^G12V/T35S pathway.

WSTF makes mutant H2A.X phosphorylated in Tyr 142, and the activity of WSTF exert indispensable functions in regulating multiple events [29]. Further experiments were performed to explore the exact mechanism how WSTF affects the phosphorylation of H2A.X^Y142. Result from Fig. 5a showed that no obviously difference was observed in the expression of WSTF and EYA3 in the group pEGFP-N1 and Ras^G12V/T35S. This suggested that WSTF was not played part in the transcription level. Afterwards, we found that the expression of EYA3 in protein level was no difference (Fig. 5b) while WSTF was significantly decreased (Fig. 5c) in the group Ras^G12V/T35S. Combined the result of Fig. 5b and c, we inferred that the effects of WSTF induced by Ras^G12V/T35S was in translation level. Interestingly, the result in Fig. 5d confirmed our inference, WSTF and H2A.X^Y142ph revealed the same trend reacted in group Ras^G12V/T35S. Further result showed that Ras^G12V/T35S could bind to the promoters of genes to reduce the input levels of WSTF in downstream factors of Ras^G12V/T35S pathway (Fig. 5e). Moreover, the proteasome inhibitor MG132 administration made the changes of WSTF expression by Ras^G12V/T35S disappeared, which indicated high inhibition for WSTF (Fig. 5f). Afterwards, we found that without MG132 administration, the phosphorylation level of H2A.X^Y142ph was decreased with the delaying of the transfection time (24, 48, 51, 54 and 60 h) in SNU-16 cells (Fig. 5g). Similar, under the same treatment, we found that the accumulated levels of WSTF was decreasing with the increasing of the transfection time (Fig. 5h). On the opposite, with the MG132 supplement, the decreased phosphorylation level of H2A.X^Y142ph by Ras^G12V/T35S was vanished. Instead the phosphorylation level of H2A.X^Y142ph was increased after administration of MG132 in different time treatment (0, 3, 6 and 12 h) (Fig. 5i). Taken together, these findings indicated that Ras^G12V/T35S pathway affected the phosphorylation of H2A.X^Y142ph was through WSTF.

It is well validated that histone modification was mediated by MDM2 in various cells [30]. Further experiments were performed to explore the mechanism about whether MDM2 was involved in the regulation of WSTF on H2A.X^Y142ph. We co-transfected Ras^G12V/T35S and WSTF with tagged HA into cells. The result of western blot showed that WSTF expression was decreased with the increasing of MDM2 (Fig. 6a). Further result showed without transfecting WSTF, the expression of WSTF was still decreased with the increasing of MDM2, which indicated that there might be a negative relationship between MDM2 and WSTF (Fig. 6b). Furthermore, we co-transfected Ras^G12V/T35S and WSTF with tag HA, the expression of WSTF was inhibited when transfected MDM2-His while WSTF expression was enhanced when transfected MDM2-MU (Fig. 6c), which strongly suggested that there was closely negative relationship between MDM2 and WSTF expression. Moreover, As shown in Fig. 6d, Ras^G12V/T35S and detected the expression of WSTF got the similar result as Fig. 6c, which confirmed the strong negative association between MDM2 and WSTF. We thereafter determined the relationship between MDM2 and H2A.X^{Y142 ph} and result in Fig. 6e. The result showed that Ras^G12V/T35S inhibited H2AX.^Y142ph while up-regulated MDM2 expression. Then si-MDM2 to silence MDM2 (Fig. 6f) and we found that si-MDM2 could enhance the expression of H2A.X^Y142ph, which confirmed the result in Fig. 6e. Then through these experiments, we concluded the cascade reaction might be Ras^G12V/T35S positively regulated MDM2 expression, and then MDM2 negatively regulated WSTF, and WSTF positively regulate H2A.X^{Y142 ph}.