RNA-binding motif protein 28 (RBM28) plays a critical regulatory role in various malignancies, yet, its expression pattern and functional role in gastric cancer remain unclear. This study aims to investigate the effect and molecular mechanism of RBM28 on the development and progression of gastric cancer (GC).

① Study materials and clinical sample grouping: Patients who visited Department of Gastroenterology of Daping Hospital between December 2021 and December 2023 and from whom tissue samples were collected in this study. Among them, 14 normal gastric mucosal tissues were assigned to Normal group, and 17 gastric cancer tissue samples were assigned to GC group. By integrating the data from The Cancer Genome Atlas (TCGA) database and our in-house RNA-seq data, differentially expressed members of the RBM family were screened, and RBM28 was identified as a significantly altered candidate gene. ② Tissue-level validation and prognostic analysis: Immunohistochemistry was performed to detect RBM28 protein expression in the Normal and GC groups. In addition, transcriptomics, and survival data of GC patients from the Kaplan-Meier Plotter database (n=876) were used to analyze the association between RBM28 expression and patient prognosis. ③ Cell model construction and functional assays: GC cell lines HGC-27 and AGS were employed as experimental models. Stable RBM28-knockdown cell lines and their corresponding control cells were established via lentiviral infection. Cell proliferation was assessed using CCK-8 assay at days 1, 2, 3, and 5 after seeding, with 6 replicate wells per group. Colony formation assay was also performed to evaluate clonogenic proliferative capacity. ④ Cell cycle analysis: Propidium iodide (PI) staining combined with flow cytometry was utilized to observe cell cycle distribution of the control and RBM28-knockdown groups, and to determine the effect of RBM28 on G1/S phase progression. ⑤ Mechanistic investigation and rescue validation: Pathway enrichment analysis was conducted to identify key genes associated with RBM28-related cell cycle and DNA replication pathways. A DONSON overexpression rescue model was further constructed, including the Control, shRBM28, and shRBM28+DONSON groups. DONSON expression was verified by RT-qPCR. Subsequently, CCK-8 assay, colony formation assay, and flow cytometry were performed to evaluate the changes in cell proliferation and cell cycle distribution after DONSON rescue, thereby determining whether RBM28 regulates G1/S phase progression and malignant proliferation of GC cells through DONSON.

① Integrated analysis of the TCGA database and our in-house dataset, as well as immunohistochemical validation, demonstrated that the protein level of RBM28 was significantly upregulated in GC tissues (P<0.0001), and high RBM28 expression was significantly associated with poor prognosis (HR=2.36, 95%CI: 1.45 to 3.85, P=0.0004). ② Knockdown of RBM28 significantly inhibited the proliferative capacity of GC cells (P<0.01). ③ Mechanistically, we found that high RBM28 expression was closely associated with the cell cycle DNA replication pathways. RBM28 knockdown induced G1 phase arrest in GC cells (P<0.05), and DONSON was identified as a key regulatory factor involved in this process. ④ Furthermore, overexpression of DONSON following RBM28 knockdown significantly reversed G1/S phase arrest and restored the suppressed proliferative capacity of GC cells (P<0.05).

RBM28 is highly expressed in GC tissues, and promotes G1/S phase progression by upregulating DONSON expression, thereby enhancing the proliferative capacity of GC cells and driving tumor progression. The RBM28/DONSON axis may serve as a potential novel target for GC molecularly targeted therapy.

To determine the expression level of TG-interacting factor 1 (TGIF1) in gastric cancer (GC) and its correlation with prognosis, and investigate the characteristics of immune microenvironment of TGIF1-overexpressing GCs and its clinical significance.

Integrated analysis was performed using transcriptomic data from The Cancer Genome Atlas (TCGA), Asian Cancer Research Group (ACRG), and our in-house GC transcriptome database. Immunohistochemistry (IHC) assay was employed to compare TGIF1 expression between GC and normal gastric mucosa tissues. Based on Kaplan-Meier Plotter online database, the correlation between TGIF1 expression and clinical prognosis was evaluated. Transcriptomic data were analyzed to identify functional enrichment features of GC with high TGIF1 expression. The GENIE3 toolkit and STRING database were utilized to predict TGIF1-regulated target genes and protein-protein interaction (PPI) networks, respectively to explore the potential immune-related signaling pathways regulated by TGIF1. Single-cell RNA sequencing (scRNA-seq) was applied to analyze the association between TGIF1 expression and tumor microenvironment (TME) disorder.

Transcriptomic analysis revealed significantly higher TGIF1 expression in GC tissues compared to normal tissues (P<0. 01). Patients with high TGIF1 expression exhibited poorer clinical prognosis (P<0. 05). IHC assay confirmed elevated TGIF1 expression in GC tissues than normal tissues (P<0. 01). GC with high TGIF1 expression was enriched in immune regulatory pathways, including immunosuppressive cytokines such as chemokine C-C motif ligand 20 (CCL20). These tumors displayed an immunosuppressive TME, characterized by abundant immunosuppressive NK cells, mast cells, regulatory T cells (Tregs), myeloid cells, and exhausted CD8⁺ T cells. Cell interaction analysis suggested that TGIF1-overexpressing GC cells may engage with Tregs via the CCL20-CCR6 axis. Co-high expression of signature gene sets from these interacting cells was associated with significantly shorter survival in the patients (P<0. 05).

TGIF1 is highly expressed in GC tissues and may serve as a biomarker for immunosuppressive TME and poor prognosis. TGIF1-overexpressing GC cells may interact with multiple immune cells, particularly Tregs, to induce an immunosuppressive microenvironment.

To explore the genome-wide distribution of histone H3K27ac in intestinal type gastric cancer, analyze remodeling features of enhancers and regulome and construct a prediction model for prognosis.

H3K27ac CUT&Tag sequencing and RNA sequencing were performed in intestinal type gastric cancer tissues from 15 patients and normal gastric mucosa tissues from 18 healthy volunteers. Bioinformatics analysis was performed to identify the differences in genome distribution of H3K27ac modifications. Based on the distribution characteristics of H3K27ac, the enhancer elements were identified and the remodeling characteristics of enhancer and related regulome were explored. The prediction model for prognosis based on enhancer related target genes was constructed by univariate Cox and multivariate Cox regression analyses.

The histone H3K27ac modification was mainly distributed in the enhancer region and displayed no significant differences in the genomic distribution patterns between normal and cancer tissues. Compared with normal gastric mucosa, the level of enhancer H3K27ac modification was higher in intestinal type gastric cancer. A total of 8 847 enhancers with increased activity in intestinal type gastric cancer were identified, accounting for 8.3% of all enhancers, which might promote malignant behaviors such as proliferation and adhesion of gastric cancer cells. A prognosis-predicting model established based on a panel of 6 genes that upregulated by the acquired enhancer in cancers, which was able to predict the overall survival of patients.

Enhancer remodeling is one of the significant epigenetic features of intestinal type gastric cancer. These enhancers may drive malignant growth and adhesion of cancer cells by upregulating the expression of MYC, E2F3 and other genes. A prognosis model based on enhancer target genes is constructed.

To characterize the silencers in genomic loci that were labeled by H3K37me3 modification in gastric high-grade intraepithelial neoplasia (HGIN) tissues, and identify their key downstream genes.

Gastric mucosa were collected from 48 donors, including 24 patients with HGIN and 24 healthy controls who underwent dndoscopic biopsy or endoscopic gastric mucosal dissection in our hospital. Genomic distribution of histone H3K27me3 modification was retrieved by targeted chromosome Cleavage Under Target & Tagmentation (CUT&Tag) sequencing. Bioinformatics analysis tools were used to compare the characteristics and differences of the silencer signals between the 2 types of tissues. The RNA sequencing data and the public high-through chromosome conformation capture (Hi-C) data were integrated to analyze the target genes regulated by silencer remodeling and their potential biological processes.

Compared with normal gastric mucosal tissue, the number of H3K27me3 modifications and signal intensity in HGIN tissues were significantly reduced, which was manifested by global remodeling of H3K27me3 signals. High-throughput RNA sequencing and bioinformatics analysis showed that there were a total of 8 887 differentially expressed genes for HGIN tissues, including 4 335 up-regulated genes and 4 552 down-regulated ones, and among them, CTNNB1 and other oncogenes were significantly up-regulated. Integrated analysis of epigenomics and transcriptomics data reveled that extensive remodeling of silencers in HGIN. Loss of silencer may regulate the expression of metabolic genes related to amino acid biosynthesis, arginine and proline metabolism and glycolysis at the transcriptional level, such as phosphoglycerate kinase 1 (PGK1), which may promote development of gastric precancerous lesions.

Global remodeling of silencers, namely loss of H3K27me3 modification, is an epigenomic characteristics of HGIN tissue, which may lead to expression dysregulation of glycolysis and amino acid metabolism regulators, such as PGK1.

To systemically characterize the genome-wide distribution of H3K27me3-marked silencers in gastric intestinal metaplasia（IM）tissues and to elucidate a novel epigenetic regulatory mechanism of IM.

Gastric antrum mucosa samples were collected from 22 healthy individuals and 39 patients with IM. The modification of H3K27me3 and the transcriptome characteristics were detected by Cleavage Under Targets and Tagmentation（CUT&Tag）sequencing and RNA sequencing, respectively. Various bioinformatics tools, such as Ngsplot, ChIPseeker, MAnorm2, GGBiPlot, edgeR, Homer and others were applied for analyzing the silencer signal and its regulation on gene expression of IM tissues.

CUT&Tag sequencing and RNA sequencing data were of great quality, and there was no significance in the whole-genome distribution of H3K27me3 modification between normal gastric mucosa and IM tissues. Genome-scale distribution of H3K27me3-marked silencers was also portrayed, which revealed weakened silencer signal intensity in IM tissues（P<0.05）. Principal component analysis（PCA）showed that the silencer characteristics in IM were greatly different from those in normal tissues, indicating extensive silencer remodeling in the IM tissues. Combined with RNA sequencing analysis, silencer remodeling was closely associated with the increased expression of CDX1 and other IM-related genes, as well as with the possible activation of IM-related metabolic pathways. In addition, loss of silencer signal recruited various transcription factors such as ATOH1（P<0.01）and ONECUT2（P<0.01）to form a regulatory network, which were highly expressed in IM tissues, regulating the expression of CDX1 and other key genes of IM, and affecting cell biological processes like coll fate commitment.

Integration analysis of epigenetic and transcriptomic sequencing data discovers silencer remodeling as a hallmark epigenetic feature in gastric IM tissues. Loss-of-function in silencer loci may recruit IM-regulatory transcription factors such as ATOH1 and ONECUT2, which may generate a network to impact on the expression of key genes and the metabolism of IM cells.

To profile the genome-wide distribution of H3K27ac-marked enhancers in gastric intestinal metaplasia（IM）tissues and to investigate the potential regulatory mechanism of enhancers on IM.

IM tissues from 41 patients and normal gastric mucosa from 21 healthy individuals were collected for analysis using Cleavage Under Targets and Tagmentation（CUT&Tag）sequencing and RNA sequencing. The number and signal intensity of H3K27ac modifications in the 2 tissues were compared to reveal the reprogramming characteristics of the enhancers in the IM tissues. Meanwhile, the underlying regulatory mechanism of enhancer-associated transcription factors on the expression of IM-related genes were investigated.

As compared with normal gastric mucosa, the number and signal intensity of H3K27ac modifications were significantly elevated in the IM tissues, and the reprogramming areas were mainly located in the enhancers. The enhancers with greatly increased activity accounted for 92.4% of all variant enhancer loci in IM, which improved the expression of many intestinal epithelium-related genes, leading to altered phenotypes and biological properties of metaplastic mucous cells. The binding motifs of 14 transcription factors including CDX2 were identified by enrichment analysis in the enhancer loci. These transcription factors were up-regulated in the IM tissues and thus constituted a transcriptional regulatory network, and may play an important role in enhancer activating IM-related genes.

Through integration analysis of epigenetic and transcriptomic sequencing, enhancer reprogramming is identified as a critical epigenetic modification feature in gastric IM. Multiple transcription factors like CDX2 might be recruited to the enhancer loci, and then generate a regulatory network to synergistically up-regulate the expression of IM-related genes.