🧬 biology

Systems Biology Analysis Identifies Immune Related Biomarkers and Therapeutic Targets in EBV Positive and EBV Negative Intestinal Type Gastric Cancer

This study employs an integrated systems biology approach to identify distinct immune-related hub genes, specifically CXCL10, VCAM1, and CD40, as potential biomarkers and therapeutic targets for Epstein–Barr virus-positive intestinal-type gastric cancer.

Original authors: Farnaz Taghinasab, Negar Mottaghi-Dastjerdi, Mohammad Soltany-Rezaee-Rad

Published 2026-07-02

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Farnaz Taghinasab, Negar Mottaghi-Dastjerdi, Mohammad Soltany-Rezaee-Rad

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ⚕️ This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer

The Big Picture: Two Types of the Same Crime

Imagine the stomach is a city, and "Gastric Cancer" is a crime wave happening there. For a long time, doctors treated all stomach cancers as if they were the same type of criminal gang. However, scientists recently realized there are actually two distinct gangs:

The "EBV-Negative" Gang: These criminals operate without a specific viral boss.
The "EBV-Positive" Gang: These criminals are led by a virus called the Epstein-Barr Virus (EBV).

This virus is like a notorious crime lord that takes over the city. While we know this "EBV gang" exists, we didn't fully understand how they operate differently from the other gang, specifically within the "Intestinal Type" of stomach cancer (a specific neighborhood of the city).

The Investigation: A Digital Detective Team

The researchers in this paper acted like a team of digital detectives. Instead of looking at one clue at a time, they used a "Systems Biology" approach. Think of this as using a massive, interconnected map of the city rather than just looking at a single street.

They gathered data from three different public databases (like collecting evidence from three different police precincts) containing genetic blueprints from stomach cancer patients. They compared the blueprints of the EBV-positive patients against the EBV-negative patients to see what was different.

The Process: Sifting Through the Noise

The Filter: They found over 7,500 genes (the city's instruction manuals) that were behaving differently between the two groups. That's too many to study individually, so they filtered the list down to the top 200 most chaotic genes.
The Network Map: They took these 200 genes and built a "relationship map" (a Protein-Protein Interaction network). Imagine a social network where every gene is a person, and lines connect people who talk to each other.
Finding the Leaders (Hub Genes): In any social network, some people are just "connectors" who know everyone and run the show. The researchers used a computer program to find the 12 most important "leaders" in this cancer network.

The Key Findings: Who Are the Leaders?

The study identified 12 key "leaders" (Hub Genes) that drive the behavior of the EBV-positive cancer. Here is what they found about them:

The Immune Callers (CXCL10, VCAM1, CD40): These genes act like loudspeakers or sirens. They are turned up (amplified) in the EBV-positive cancer. Their job is to call in the body's immune system (the police). This explains why EBV-positive tumors are full of immune cells—they are constantly ringing the alarm bell.
The Identity Losers (PGA4, LTF): These genes are turned down (silenced). They are responsible for the stomach's normal job of making digestive juices. When these shut off, the stomach cells lose their "identity" and become cancerous.
The Survival Predictor (RBP4): One gene, RBP4, was unique. It didn't act like a loudspeaker or a silencer in the same way. Instead, the study found that patients with high levels of RBP4 had a harder time surviving. It's like a "bad omen" marker that predicts a tougher battle, regardless of whether the virus is present.
The Drug Targets (DPP4, IDO1, CD40): The researchers checked a "drug catalog" to see if any existing medicines could target these leaders.
- DPP4 is a target for drugs already used for diabetes (like Sitagliptin). This suggests these drugs might be "repurposed" to fight this cancer.
- CD40 and IDO1 are targets for immunotherapy drugs (drugs that help the immune system fight cancer).

The Clusters: How the Gangs Organize

When the researchers looked at how these genes worked together, they found two main "squads":

The Immune Squad: A group of genes focused on calling in immune cells, fighting inflammation, and managing the body's defense systems.
The Metabolic Squad: A group focused on blood clotting, fat processing, and energy management.

The Conclusion: What Does This Mean?

The paper concludes that EBV-positive stomach cancer is a very specific type of disease. It is defined by:

A loud immune response: The body is constantly trying to fight the virus-driven cancer.
A loss of stomach function: The cells stop acting like normal stomach cells.
Specific targets: We now have a list of 12 specific "leaders" to watch.

Crucially, the paper states:

CXCL10, VCAM1, and CD40 are the best markers to identify this specific "EBV-positive" type of cancer.
PGA4 and LTF show that the stomach has lost its normal function.
RBP4 is a warning sign for patients who might have a poorer outcome.
DPP4 is the most promising target for using existing drugs (drug repurposing).

The researchers emphasize that these are computer-based predictions. They have identified the suspects and the potential weapons, but they state that these findings need to be tested in real-world labs and clinics to prove they work in actual patients. They are a roadmap for future doctors and scientists to follow.

Technical Summary: Systems Biology Analysis of EBV-Positive and EBV-Negative Intestinal-Type Gastric Cancer

Problem Statement
Epstein–Barr virus-associated gastric cancer (EBVaGC) is a distinct molecular subtype of gastric cancer (GC), characterized by an immune-rich tumor microenvironment. While EBVaGC is generally recognized as immunologically distinct, the specific transcriptomic alterations that differentiate EBV-positive from EBV-negative tumors specifically within the intestinal-type histological subtype remain incompletely defined. Current literature often treats EBVaGC broadly or focuses on diffuse types, leaving a gap in understanding the specific immune and molecular drivers of the intestinal subtype. This study aims to address this gap by identifying immune-related biomarkers and therapeutic targets specific to EBV-positive intestinal-type GC using an integrated systems biology approach.

Methodology
The study employed a multi-step computational workflow integrating public transcriptomic data with network analysis and clinical validation:

Data Acquisition and Integration: Three GEO microarray datasets (GSE51575, GSE62254, and GSE66229) were selected based on strict inclusion criteria: human gastric cancer tissue, defined EBV status, intestinal-type histology, and sufficient biological replicates.
Differential Expression Analysis (DEA): Using GEO2R, differentially expressed genes (DEGs) were identified by comparing EBV-positive versus EBV-negative intestinal-type samples within each dataset. Statistical significance was determined using Benjamini–Hochberg FDR correction ( $p < 0.05$ ).
Data Harmonization: Gene symbols were annotated, duplicates were collapsed by retaining the highest absolute log fold change ($|logFC|$), and lists were merged to create a non-redundant set of 7,576 unique DEGs (3,118 upregulated, 4,458 downregulated).
Network Construction and Hub Identification: The top 200 most dysregulated genes (ranked by $|logFC|$) were selected to construct a Protein-Protein Interaction (PPI) network using STRING (v12) and Cytoscape. Hub genes were identified using the CytoHubba plugin based on four algorithms: MCC, MNC, Degree, and DMNC.
Functional and Module Analysis: Gene Ontology (GO) and Reactome pathway enrichment were performed using Enrichr. Network modules were detected using the CytoCluster plugin (IPCA algorithm) to identify functional clusters.
Validation and Clinical Correlation:
- Expression Validation: Hub gene expression was validated against TCGA-STAD data via UALCAN, comparing tumor vs. normal tissues and analyzing stage-specific patterns.
- Survival Analysis: Kaplan-Meier curves and log-rank tests were used to assess the association between hub gene expression and overall survival (OS).
- Therapeutic Assessment: Gene-drug interactions were queried using DGIdb and DrugBank to identify potential repurposable drugs or therapeutic targets.

Key Results

Differential Expression: The integrated analysis identified 7,576 unique DEGs. The top 200 genes were prioritized for network analysis.
Hub Genes: Twelve hub genes were prioritized based on topological centrality: VCAM1, LTF, CD40, CXCL10, CXCL8, RBP4, ULBP2, LAX1, IGLL5, PGA4, IDO1, and DPP4.
Functional Enrichment: Enrichment analyses revealed that the DEGs are primarily involved in immune-cell trafficking, chemokine signaling, complement/Fc receptor pathways, TNF receptor signaling, and epithelial remodeling.
Network Modules: Cluster analysis identified two primary functional modules:
1. Hemostatic-Metabolic Module: Involving platelet activation, lipid/retinoid metabolism (e.g., RBP4, ORM1), and angiotensinogen metabolism.
2. Immune/Chemokine Module: Centered on cytokine signaling, IL-10 feedback, and TNFR2 non-canonical NF-κB signaling, driven by genes such as CXCL10, CXCL8, CD40, and IDO1.
Biomarker Applicability:
- EBV-Associated Immune Markers: CXCL10, VCAM1, and CD40 were identified as key EBV-associated immune biomarkers. CXCL10 and VCAM1 showed strong upregulation in tumors and early-stage elevation, with CXCL10 being specifically higher in EBV-positive intestinal GC.
- Loss-of-Expression Markers: PGA4 and LTF were downregulated in tumors, with PGA4 showing marked suppression across all stages, indicating a loss of gastric secretory identity.
- Prognostic Marker: RBP4 was the only hub gene significantly associated with overall survival (high expression correlated with poorer survival, $p=0.019$ ), though it showed lower expression in EBV-positive tumors compared to EBV-negative ones.
- Other Markers: ULBP2, IDO1, CXCL8, and DPP4 were generally lower in EBV-positive intestinal tumors, suggesting they may be more relevant to EBV-negative contexts, general progression, or druggability rather than EBV-specific activation.
Therapeutic Targets: The gene-drug interaction network highlighted several potential targets:
- DPP4: Identified as the strongest drug-repurposing axis, with existing inhibitors (e.g., Sitagliptin, Linagliptin).
- CD40: A strong immunotherapy target with associated antibodies (e.g., Dacetuzumab).
- IDO1: An immunometabolic target with inhibitors (e.g., Epacadostat).
- CXCL8/CXCL10: Linked to inflammatory and antiviral-associated compounds.

Significance and Claims
The authors claim that this study provides a network-based framework for prioritizing biomarkers and therapeutic targets specifically for EBV-positive intestinal-type GC, moving beyond general descriptions of EBVaGC as "immune-rich."

Specificity: The study distinguishes EBV-associated immune signals (e.g., CXCL10, VCAM1) from broader inflammatory programs (e.g., CXCL8) and progression markers (e.g., RBP4, ULBP2).
Multigene Potential: The authors suggest that no single gene captures all diagnostic and prognostic features; instead, a multigene framework combining immune markers (CXCL10, VCAM1, CD40) with differentiation-loss markers (PGA4, LTF) offers superior biological and clinical resolution.
Translational Value: The identification of DPP4 as a highly tractable target with existing clinical inhibitors, alongside CD40 and IDO1, offers immediate candidates for therapeutic exploration and drug repurposing.
Modesty and Limitations: The authors explicitly state that these findings are computationally prioritized signals. They acknowledge limitations, including the use of a low-confidence STRING threshold, the exclusion of moderately altered genes, and the lack of EBV-stratification in the TCGA survival analysis. Consequently, they emphasize that the identified biomarkers and targets require validation in independent, EBV-stratified cohorts, protein-level assays, and functional experiments before clinical application.