Integrative single-cell, spatial, and bulk transcriptomics reveal an FMR1–FTO axis linked to the immune-excluded phenotype in gastric cancer

Chenxi Mao, M Zhang, Kangjie Zhou, Yidong Hong, Yiqian Han, Li Zhao, Meirong Liang, Jingzhou Zhang, Nan Hu, Fenglei WuFrontiers in Immunology 2026 · DOI

TL;DR

This study reveals a novel molecular axis, FMR1-FTO, that drives an "immune-excluded" phenotype in gastric cancer. This axis creates a physical barrier of cancer-associated fibroblasts (CAFs) that prevents T cells from reaching the tumor, hindering immunotherapy effectiveness. Targeting this FMR1-FTO axis could be a new strategy to improve cancer treatment.

What Problem Does This Paper Address?

Immune checkpoint inhibitors (ICIs) have revolutionized cancer treatment, but their effectiveness is often limited by the tumor microenvironment (TME). A major challenge is the "immune-excluded" phenotype, where T cells are present but physically blocked from entering the tumor core by a dense stromal barrier. While we know this barrier is often formed by cancer-associated fibroblasts (CAFs) and extracellular matrix components, the specific molecular mechanisms that drive this exclusion, particularly those involving RNA modifications like N6-methyladenosine (m6A), remain unclear. This paper aims to unravel how m6A regulators contribute to this immune exclusion in gastric cancer and identify potential therapeutic targets.

Methods

The researchers employed a multi-omics approach, integrating several cutting-edge techniques:

  • Single-cell RNA sequencing (scRNA-seq): Used to map the cellular landscape and identify different cell types and their gene expression profiles within the TME.
  • Spatial transcriptomics (ST): Allowed them to visualize gene expression patterns within the physical context of the tumor tissue, revealing the spatial relationships between different cell types and their molecular states.
  • Bulk transcriptomics: Analyzed larger datasets to identify molecular subtypes and correlate them with clinical outcomes and drug responses.
  • CellChat: A computational tool used to infer intercellular communication networks between different cell types based on ligand-receptor interactions.
  • m6A regulator scoring: A novel score was developed to quantify the spatial distribution of m6A-related gene expression patterns.
  • Histopathology (IHC and multiplex immunofluorescence): Used to validate findings at the protein level, visualizing the spatial distribution of key proteins like FMR1 and CD8+ T cells in tumor tissues.
  • Functional assays (Western blotting, CHX chase, MG132 treatment, co-immunoprecipitation): Employed to investigate the molecular mechanisms, specifically the post-translational regulation of FTO by FMR1.
  • Molecular docking and molecular dynamics simulations: Used to predict and assess the structural compatibility and stability of protein interactions.

Datasets

The study utilized several publicly available datasets:

  • scRNA-seq data: From the Gene Expression Omnibus (GEO) database (GSE183904), comprising 26 gastric cancer (GC) and 10 normal tissue samples.
  • Spatial transcriptomics data: From GEO (GSE251950), using the 10x Visium platform.
  • Bulk transcriptomic datasets: Five independent GC cohorts from GEO (GSE15459, GSE34942, GSE57303, GSE62254, GSE84437) with clinical follow-up data.
  • Genomic Drug Sensitivity Consortium (GDSC) database: Used for predicting chemotherapy drug responses.

The paper does not introduce a new benchmark dataset but rather integrates and analyzes existing ones to uncover novel biological insights.

Key Insights and Ideas

  • CAF-driven stromal barrier is key to immune exclusion: The study strongly reinforces that CAFs create a physical barrier, primarily through collagen deposition, that effectively prevents CD8+ T cells from infiltrating the tumor core.
  • m6A regulators are spatially coupled with immune exclusion: A novel "m6A regulator score" revealed that elevated m6A activity is specifically concentrated within tumor nests, not the surrounding stroma, suggesting a tumor-intrinsic program driving exclusion.
  • FMR1-FTO axis as a novel regulator of immune exclusion: The research identifies a critical link between the RNA-binding protein FMR1 and the m6A demethylase FTO. This axis appears to be a central player in establishing the immune-excluded phenotype.
  • FMR1 stabilizes FTO protein via post-translational modification: A significant finding is that FMR1, an RNA-binding protein, directly influences FTO protein stability, protecting it from proteasomal degradation. This is a non-canonical function for FMR1 and a novel mechanism for regulating FTO levels.
  • m6A subtyping predicts clinical outcomes and immunotherapy response: The study successfully classified gastric cancer into three distinct m6A-based subtypes, with the "immune-excluded" subtype (Cluster 2) showing significantly worse survival and predicted resistance to ICIs.

Key Results

  • Spatial transcriptomics confirmed immune exclusion: Visualizations clearly showed tumor nests surrounded by CAF-rich stroma, with T cells trapped in the periphery.
  • Elevated m6A regulator scores in tumor nests: Quantitative analysis showed significantly higher m6A regulator scores within tumor nests compared to the stroma across multiple samples.
  • Three distinct m6A molecular subtypes identified: These subtypes were characterized as "immune-desert," "immune-excluded," and "immune-inflamed," with distinct stromal and immune activation profiles.
  • Immune-excluded phenotype (Cluster 2) linked to poor prognosis: Patients in this cluster had significantly worse overall survival and were predicted to have a low response rate to ICIs.
  • FMR1 protein expression is high in immune-excluded tumors: Histopathological validation confirmed that FMR1 protein levels were significantly elevated in immune-excluded samples, correlating with CD8+ T cell arrest.
  • FMR1 knockdown reduces FTO protein levels and stability: Functional experiments demonstrated that reducing FMR1 expression led to decreased FTO protein abundance and accelerated FTO turnover, indicating FMR1's role in stabilizing FTO.
  • FMR1 and FTO physically interact: Co-immunoprecipitation assays provided evidence of a protein-protein interaction between FMR1 and FTO.
  • Immune-excluded tumors exhibit chemoresistance: Predictive analysis suggested that tumors with the immune-excluded phenotype are broadly resistant to common chemotherapeutic agents.

Limitations and Future Work

The authors acknowledge several limitations:

  • In vitro and correlative evidence: While strong, the findings are primarily based on correlative analyses of clinical samples and in vitro experiments. In vivo validation using immunocompetent mouse models is needed to establish causal relationships.
  • Resolution of FMR1/FTO interaction: The precise subcellular localization of FMR1 and FTO at the CAF-tumor cell interface could not be fully resolved due to technological limitations.
  • Indirect measurement of m6A abundance: The study inferred an activated m6A regulatory program rather than directly quantifying global or transcript-specific m6A levels. Future work should include dedicated epitranscriptomic assays.
  • Mechanistic gaps in FMR1-FTO interaction: While an association and MG132-sensitive regulation were shown, direct RNA-independent interaction confirmation (e.g., RNase-treated Co-IP) and FTO ubiquitination studies are needed. FTO mRNA levels under FMR1 perturbation were also not quantified.

Future work should focus on:

  • In vivo validation: Using animal models to confirm the role of the FMR1-FTO axis in driving immune exclusion and its therapeutic potential.
  • Targeting the FMR1-FTO axis: Developing strategies to disrupt this axis for improved immunotherapy response.
  • Investigating downstream targets of FTO: Elucidating the specific transcripts regulated by FTO in the context of stromal remodeling and immune evasion.
  • Exploring FMR1's non-canonical functions: Further understanding how FMR1 influences protein stability beyond its known RNA-binding roles.

Summary updated

Abstract

Background Immune exclusion is a major barrier to immune checkpoint inhibitor (ICI) efficacy in gastric cancer, yet the spatial mechanisms by which m 6 A regulators drive this phenotype remain unclear. Methods We integrated single-cell RNA-seq, spatial transcriptomics, and functional assays to map the gastric cancer microenvironment and derived an m 6 A regulator score to quantify the spatial coupling between m 6 A patterns and immune exclusion. Results CAFs emerged as a central hub that excluded T cells via collagen–integrin interactions and MIF signaling, forming a CAF-defined collagen barrier that impeded CD8 + T-cell entry into the tumor core. In immune-excluded samples, m 6 A regulator scores were selectively elevated in tumor nests compared with surrounding stroma. Across five bulk transcriptomic datasets, this phenotype was associated with reduced predicted ICI responsiveness, activation of chemotherapy resistance programs, and poor survival. Mechanistically, FTO was highly expressed in the immune-excluded phenotype and correlated with stromal activation and T-cell exclusion. Serial immunohistochemistry and multiplex immunofluorescence revealed high FMR1 protein expression in immune-excluded tumors, with CD8 + T cells largely confined to the stroma. Cycloheximide (CHX) chase and MG132 treatment showed that FMR1 depletion reduced FTO protein abundance and accelerated FTO turnover in an MG132-sensitive manner, consistent with a post-translational regulatory relationship. Conclusions Collectively, our data support an FMR1–FTO module associated with the immune-excluded phenotype and nominate this axis as a potential vulnerability for disrupting stromal immune barriers. The FMR1–FTO axis may represent a candidate target for strategies aimed at relieving immune exclusion and improving immunotherapy sensitivity.
BibTeX
@inproceedings{mao2026integrative,
  author = {Chenxi Mao and M Zhang and Kangjie Zhou and Yidong Hong and Yiqian Han and Li Zhao and Meirong Liang and Jingzhou Zhang and Nan Hu and Fenglei Wu},
  title = {{Integrative single-cell, spatial, and bulk transcriptomics reveal an FMR1–FTO axis linked to the immune-excluded phenotype in gastric cancer}},
  booktitle = {Frontiers in Immunology},
  year = {2026},
  doi = {10.3389/fimmu.2026.1713267},
}
Figure