Introduction
Biomarkers associated with the progression from gastric intestinal metaplasia (GIM) to gastric adenocarcinoma (GA) could provide valuable insights into identifying patients with increased risk for GA. Currently, the discovery of effective GA biomarkers poses a challenge due to the complexity of predictive and actionable value. Rarely can tumorigenesis be attributed solely to an individual molecule, as GA arises from the complex interplay of various intrinsic and environmental factors and collective interactions among multiple molecules.
Aims & Methods
The aim of this study was to identify potential biomarkers for the GIM-to-GA progression and potential drug repurposing for GIM. Sixty-four surgical biopsies were collected from 16 patients (4 tissue biopsies per patient) at St. Olav’s Hospital, Norway. Transcriptome datasets of GIM, GA, and normal tissues were compared within the same stomachs. Computational platform ScType was used for cell-type identification based on given scRNA-seq data and a comprehensive cell marker database. Ingenuity pathway analysis, Pagerank centrality, GO and MSigDB enrichments, CopyKat, aneuploid analysis, docking analysis, pathology and immunohistochemistry (IHC) were used for data and pathway mining, protein-protein interaction (PPI), and molecular structure-based drug repurposing.
Results
GIM displayed 110 canonical pathways that exhibited a negative correlation with GA. Wnt/β-catenin, ERK/MAPK, mTOR, neuregulin, and BRCA1 pathways were activated in both GIM and GA. Conversely, pathways such as HIPPO, p53, toll-like receptor, aryl hydrocarbon receptor, LPS/IL-1-mediated inhibition of RXR function, and eIF2 were inactivated in GIM but activated in GA. GIM exhibited elevated gene expression compared to GA in tumor microenvironment-related cells. Particularly, RBP2 expression and goblet cell-related gene expression were found to be restricted in GIM, but not in GA. Chromosome instability occurred mainly in GA cells with high rates of gain or loss of chromosomes 1, 7, 12, 21 (gain), and 5, 7, 8, X (loss). CD44 IHC expression was expanded from the basal part of mucosa in normal and GIM stomachs to the upper part of the mucosa in GA stomachs. PPI analysis of 865 GIM-related proteins showed seven network nodes (i.e., drug targets) including epidermal growth factor receptor (EGFR), SRC (proto-oncogene tyrosine-protein kinase), paxillin (PXN), Jun proto-oncogene (JUN), breast cancer type 1 susceptibility protein (BRCA1), tumor protein P53 (TP53 or p53), mouse double minute 2 homolog (MDM2), and CD44. These network node proteins were also shown in The Human Protein Atlas and IHC, with the exception of SRC in GA. Computational drug repurposing of approved and investigational drugs/compounds (e.g., as found on www.clinicaltrials.gov) identified the following drugs as effective for GIM: dasatinib with or without erlotinib, nilotinib, afatinib on SRC, SM1-71 on SRC, JNJ-26483327 on EGFR, Afatinib on EGFR, ingenol mebutate on protein kinase C (PKC) family, amoxapine, benoxaprofen, avobenzone, amlexanox, alprazolam, praziquantel, ibuprofen piconol, tranilast, and restoril.
Conclusion
The druggable protein network associated with GIM progression toward GA was identified. The distinct phenotypic patterns between GIM and GA might challenge the notion that GIM serves as the primary precancerous lesion in GA tumorigenesis rather than shares a common cause with tissue-resident stem cells. Further validation of the druggable protein network is needed for the development of a cost-effective surveillance program and preventive treatment of the GIM-to-GA progression.
