Multi-Omics Integrative Analysis of the Aspirin-Gut-Brain-Glioma Axis: Transcriptomic, Proteomic, Epigenetic, Mendelian Randomization, and Single-Cell Transcriptomic Evidence Converges on NEO1/Hepcidin Iron Reprogramming and Ferroptosis Vulnerability
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Background
Despite epidemiological interest in aspirin’s chemopreventive potential against glioma, the underlying multi-layered molecular mechanisms — spanning COX-2/PGE2 signaling, iron metabolism, ferroptosis, epigenetic regulation, and the NEO1/hepcidin regulatory axis — have not been systematically characterized at the multi-omics level.
Methods
We conducted an integrative multi-omics analysis leveraging TCGA-GBM (n=172) and TCGA-LGG (n=534) transcriptomes, CPTAC GBM proteomics (n=99), TCGA HM450K DNA methylation data (GBM n=140, LGG n=516), GEO aspirin perturbation datasets, IEU OpenGWAS summary statistics, and independent single-cell RNA-seq data (GSE131928, 28 GBM patients). Eight analytical tracks were executed: (1) COX-2/PGE2 pathway profiling, (2) BBB tight junction characterization, (3) GEO-derived aspirin response signature projection, (4) gut-brain axis evaluation, (5) Mendelian randomization (MR) using PTGS2 cis-SNPs, (6) iron metabolism and ferroptosis pathway analysis, (7) NEO1/HFE2/BMP6/HAMP regulatory axis characterization with multi-omics validation, and (8) single-cell transcriptomic validation across GBM malignant cell states.
Results
Transcriptomic analysis revealed profound reprogramming of the NEO1/hepcidin iron regulatory axis in GBM: HAMP (hepcidin) was massively upregulated (log2FC=+2.92, P=5.0×10⁻³⁷), accompanied by TFRC upregulation (log2FC=+1.38, HR=2.30, P=3.6×10⁻⁴²) and NEO1 downregulation (log2FC=-0.57, HR=0.59, P=4.6×10⁻⁶). De novo HM450K methylation analysis revealed HAMP as the dominant epigenetic target in the iron network, exhibiting the strongest hypomethylation signal (Δβ=-0.265, P=1.4×10⁻⁴⁸), while NEO1 and TFRC showed constitutively low baseline methylation (β<0.05). Gene set enrichment analysis identified ferroptosis driver genes (NES=+1.861, P=0.030) and the iron deficiency response pathway (NES=+1.698, P=0.010) as the most significantly enriched pathways in GBM. Molecular subtype analysis revealed that the mesenchymal GBM subtype exhibits the highest iron metabolism gene expression. CPTAC protein-level estimation confirmed directionally concordant changes, and Mendelian randomization established a causal relationship between PTGS2 expression and glioma risk (IVW OR=1.31, P=1.1×10⁻⁴). The COX risk score demonstrated superior prognostic power (HR=1.93, P=5.3×10⁻⁵³, C-index=0.80). Single-cell RNA-seq analysis of GBM (GSE131928, 28 patients) validated that iron metabolism gene expression is heterogeneously distributed across malignant cell states, with the mesenchymal (MES) state exhibiting the highest HAMP expression (0.338 vs. 0.029–0.044 in other states, P<0.001) and elevated ferroptosis vulnerability. PTGS2 and ACSL4 were selectively enriched in the MES state, consistent with the GSEA-identified ferroptosis driver pathway activation. GPX4 was universally highly expressed across all cell states (mean 1.46–1.68), indicating pan-GBM dependence on GPX4-mediated ferroptosis suppression.
Conclusions
This multi-omics investigation reveals that the NEO1/hepcidin iron regulatory axis is epigenetically reprogrammed in glioma, driving iron-dependent vulnerability that bridges COX-2 signaling with ferroptosis susceptibility. The convergent evidence from transcriptomics, proteomics, epigenomics, and causal inference provides a comprehensive mechanistic framework for aspirin’s protective effects against glioma and identifies the NEO1/HAMP/TFRC axis as a promising therapeutic target.
