Transcriptional Regulation of ING5 and its Suppressive Effects on Gastric Cancer

Abstract

ING5 targets histone acetyltransferase or histone deacetylase complexes for local chromatin remodeling. Its transcriptional regulation and suppressive effects on gastric cancer remain elusive. Luciferase assay, EMSA, and ChIP were used to identify the cis-acting elements and trans-acting factors of the ING5 gene. We analyzed the effects of SAHA on the aggressive phenotypes of ING5 transfectants, and the effects of different ING5 mutants on aggressive phenotypes in SGC-7901 cells. Finally, we observed the effects of ING5 abrogation on gastric carcinogenesis. EMSA and ChIP showed that both SRF (-717 to -678 bp) and YY1 (-48 to 25bp) interacted with the promoter of ING5 and up-regulated ING5 expression in gastric cancer via SRF-YY1-ING5-p53 complex formation. ING5, SRF, and YY1 were overexpressed in gastric cancer, (P<0.05), and associated with worse prognosis of gastric cancer patients (P<0.05). ING5 had positive relationships with SRF and YY1 expression in gastric cancer (P<0.05). SAHA treatment caused early arrest at S phase in ING5 transfectants of SGC-7901 (P<0.05), and either 0.5 or 1.0 μM SAHA enhanced their migration and invasion (P<0.05). The wild-type and mutant ING5 transfectants showed lower viability and invasion than the control (P<0.05) with low CDC25, VEGF, and MMP-9 expression. Gastric spontaneous adenocarcinoma was observed in Atp4b-cre; ING5^f/f, Pdx1-cre; ING5^f/f, and K19-cre; ING5^f/f mice. ING5 deletion increased the sensitivity of MNU-induced gastric carcinogenesis. ING5 mRNA might be a good marker of gastric carcinogenesis, and poor prognosis. ING5 expression was positively regulated by the interaction of SRF-YY1-ING5-p53 complex within the ING5 promoter from -50 bp upstream to the transcription start site. ING5 deletion might contribute to the tumorigenesis and histogenesis of gastric cancer.

Keywords: ING5; gastric cancer; transcriptional regulation; tumor suppressor; tumorigenesis.

Figure 1

The promoter activity and trans-acting factors of ING5 gene Different upstream wild-type (WT) and mutant (MUT) DNA fragments of ING5 were inserted into the pGL-3 basic vector, as shown schematically (A). These plasmids and pRL-TK were co-transfected into HEK293 cells and the cell lysates were subjected to luciferase reporter assay (B). We also predicted trans-acting factors that bind to the promoter of ING5 (C). Among them, His-tagged Sp-1 (Sp), WT-1 (W1), PPAR-γ1 (PPAR), SRF, YY1, CTCF, and Pax5 expression plasmids were constructed using pET-28 vector, induced by IPTG, and purified by Ni-IDA (D). In EMSA, the recombinant proteins were incubated with probes and subjected to electrophoresis (E). To confirm the EMSA results, we carried out ChIP in SGC-7901 and HEK293 cells using either anti-SRF or anti-YY1 antibody (F). M, marker; pel, pellet after sonication; sup, supernatant after sonication; Ni-IDA, supernatant throughout Ni-IDA column; Was, washed target protein; Rem, remnant protein in column; NC, negative control; PC, positive control; cold com, cold competitor; Mut com, mutant competitor; ChIP, chromatin immunoprecipitation; Ctr, control; ND, no DNA; NC in ChIP, negative control using normal mouse IgG ChIP; PC in ChIP, positive control using anti-Pol II ChIP; INPUT, 2% input.

Figure 2

The promoting effects of both SRF and YY1 proteins on ING5 expression The results of RT-PCR and Western blot revealed the successful silencing of SRF and YY1 expression in AGS and SGC-7901 cells, which showed low ING5 mRNA and protein expression (A). Co-IP was performed to explore whether SRF bind to YY1, p53 and ING5 or YY1 to SRF, p53 and ING5 in both gastric cancer cells, treated with SFR or YY1 siRNA (B). Double immunofluorescence was carried out to observe the co-localization of YY1, SRF or ING5, and p53 in gastric cancer cells (C). Ctr, control; RT-PCR, reverse-transcriptional polymerase chain reaction; WB, western blot; Co-IP, co-immunoprecipitation.

Figure 3

The clinicopathological significance of ING5, SRF and YY1 mRNA expression in gastric cancer Oncomine (A), Xiantao (B), and/or UALCAN (C) datasets were employed to analyze ING5 mRNA expression in gastric cancer, and the correlations of its expression with pathological parameters of cancers were analyzed using TCGA (D). Kaplan–Meier curves were used to analyze the relationships of ING5 mRNA expression with overall (OS), progression-free (PFS), and post-progression (PPS) survival, according to Kaplan-Meier plotter (E). The relationship between SFR and ING5 mRNA expression was analyzed using TCGA data (F). SRF mRNA expression was explored in gastric cancer using UALCAN (G). The prognostic significance of SRF mRNA expression was investigated using Kaplan-Meier plotter (H). We also investigated the relationship between YY1 and ING5 mRNA expression using TCGA data (I). YY1 mRNA expression was explored in gastric cancer using Xiantao (J) and UALCAN (K). The prognostic significance of SRF mRNA expression was studied using Xiantao platform (L). N, normal mucosa; T, tumor; IT, intestinal type; DT, diffuse type; MT, mixed-type; HR, hazard ratio. ; **p<0.01.

Figure 4

The anti-tumor effects of SAHA on ING5-overexpressing gastric cancer cells SGC-7901 and its ING5 transfectants (Clones 4 and 8) were treated with SAHA at 0.5 and 1.0 μM. After exposure, they were subjected to PI staining and flow cytometry (A). Among them, clone 8 was stained using IdU and CIdU to differentiate the early and late phases of DNA synthesis (B). Transwell (C) and wound healing (D) assays were employed to analyze the effects of SAHA on migration and invasion of these cells. Note: *, compared with 0μm, p<0.05.

Figure 5

The effects of ING5 domains on phenotypes and related proteins of gastric cancer cells FLAG-tagged wild-type (WT) and mutant (M1–M4) ING5 fragments were inserted into pcDNA3.1 vector (A). These WT and mutant ING5-expressing plasmids were transfected into SGC-7901 cells by Western blot (B), and all transfectants were subjected to the examination of cell viability (C). After treatment with cisplatin (10 μmol/L, 36 h), we also employed PI-Annexin V-FITC staining to examine apoptosis (D). Transwell assay was utilized to observe migration and invasion (E). Finally, phenotype-related proteins were screened by western blot (F). Note: *, compared with mutant, p<0.05.

Figure 6

The effects of conditional ING5 knockout on gastric carcinogenesis PCR primers were designed (A) and used for PCR of tail (B) DNA. We performed tube (C) and in situ (D) PCR amplification targeting mutant and deleted ING5 using the stomach samples of Capn8-cre/ING5^f/f (CapnI5), Atp4b-cre/ING5^f/f (AtpI5), PGC-cre/ING^f/f (PGCI5), Pdx1-cre/ING5^f/f (PdxI5), and K19-cre/ING^f/f (K19I5) and. MNU was orally administered into these wild-type and conditional knockout mice in accordance with the schedule (E). The stomach of wild-type (WT) and knockout (KO) mice was grossly (F) and histologically (G) observed with or without exposure to MNU (240 mg/L) until 68 weeks. The histological findings on the gastric lesions are summarized in (H).P, primer; NC, negative control.