Long Noncoding RNA IRAIN Drives Immunometabolic Reprogramming in Glioma via the IGF1R-JAK-STAT-BIRC5 Axis

Summary

Using integrated machine learning across multiple cohorts, we built a robust IMRG prognostic signature for glioma. IRAIN is markedly downregulated, inversely linked to grade and survival, and suppresses IGF1R-JAK2-STAT3-BIRC5 signaling, restraining proliferation, migration, and angiogenesis. Findings nominate IRAIN and IMRGs as biomarkers and therapeutic targets.

Abstract

Glioma is an aggressive malignancy with limited therapeutic options and a poor prognosis. Its progression is closely linked to metabolic reprogramming and immune evasion, underscoring the need to identify molecular regulators that integrate these processes. Here, we established a robust immunometabolic-related gene (IMRG) prognostic signature and elucidated the role of the long noncoding RNA IRAIN in regulating glioma immunometabolism through the IGF1R-JAK-STAT-BIRC5 axis. Transcriptomic and clinical data from TCGA, CGGA (693/325), and GEO (GSE43378) cohorts were analyzed. Differential expression and weighted gene co-expression network analysis (β = 8) ensured scale-free topology, and a leave-one-out cross-validation framework combining ten machine-learning algorithms produced 101 prognostic models. The optimal 17-gene IMRG signature was validated across all cohorts and independently predicted overall survival. Functional assays demonstrated that IRAIN overexpression inhibited IGF1R, suppressed JAK2/STAT3 phosphorylation, and downregulated BIRC5, thereby promoting apoptosis and reducing angiogenesis. Low IRAIN expression correlated with immunosuppressive Tregs and M2 macrophages and with elevated tumor-microenvironment scores, suggesting an immune-evasive phenotype. These findings establish IRAIN as a key regulator of glioma immunometabolic reprogramming and validate the IMRG signature as a reliable prognostic tool, highlighting the IRAIN-IGF1R-JAK-STAT-BIRC5 pathway as a mechanistic bridge linking immune and metabolic dysregulation in glioma and offering potential targets for precision therapy.

Introduction

Glioma represents the most common primary malignancy of the central nervous system and remains among the deadliest human cancers despite progress in neurosurgery, radiotherapy, and chemotherapy. High-grade gliomas, particularly glioblastoma multiforme (GBM), are characterized by rapid proliferation, diffuse infiltration, and inevitable recurrence^1,2. Recently, a range of effective therapeutic approaches has arisen, such as surgical interventions, immunotherapy, and chemotherpy³. Median survival remains approximately 14-18 months, even with maximal therapy, underscoring the need for nove....

Protocol

All procedures involving human tissues complied with institutional guidelines and the Declaration of Helsinki and were approved by the Institutional Review Board of Fujian Medical University (Approval No. 2021KYB089). Written informed consent was obtained from all participants prior to tissue procurement.

Gene expression and survival analysis
RNA sequencing data and corresponding clinical information were obtained from multiple public databases. 1) TCGA cohort: RNA-seq (FPKM) data for 175 glioblastoma multiforme (GBM) and 534 low-grade glioma (LGG) samples were downloaded from The Cancer Genome Atlas (https://portal.g....

Results

IRAIN is downregulated in glioma and associated with adverse clinicopathologic features
The analysis process is illustrated in Figure 1. IRAIN expression was first assessed by qRT-PCR in primary astrocytes, glioma tissues and glioma cell lines. IRAIN levels were markedly reduced in both low-grade and high-grade glioma tissues and in all four glioma cell lines (SHG44, A172, U251 and T98G) compared with normal astrocytes (Figure 2A). Consist.......

Discussion

Glioma remains one of the most lethal malignancies of the central nervous system, characterized by profound intratumoral heterogeneity and resistance to conventional therapy. Despite surgical resection combined with chemoradiotherapy, recurrence and mortality remain high, and the median survival of patients, particularly those with glioblastoma, has shown limited improvement over recent decades^4,24. Increasing evidence indicates that metabolic reprogramming and i.......

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Annexin V-PI Apoptosis Kit	Southern Biotechnology, Birmingham, AL, USA	NA	Used for evaluating apoptosis
AnnexinV-PI Apoptosis Kit	Southern Biotechnology	10010-02	Used for apoptosis assessment by flow cytometry.
Anti-IGF1, Anti-IGF1R, Anti-JAK2, Anti-STAT3, Anti-Survivin, Anti-β-actin	Abcam	ab182408, ab108596, ab109085, ab76424, ab8227	Antibodies used for Western Blot analysis of signaling pathways.
BiocManager	CRAN	1.30.25	Install and manage Bioconductor packages in R ≥ 3.6.
Bio-Rad Cfx96 System	Bioconductor	1.5.1	Programmatic access to STRING protein–protein interactions.
Caret	CRAN	0.4	Time dependent ROC/AUC for survival models.
Chinese Glioma Genome Atlas ( CGGA)	NA	CGGA693 and CGGA325	Used for model consruction and validation
Clusterprofiler	Bioconductor	2.18.0	Parse MAF files; compute tumor mutational burden (TMB) and mutation landscape.
Corrplot	CRAN	3.5.0	General plotting (boxplots, violin plots, scatter, trend lines).
Coxboost	CRAN	7.3-60	stepAIC for stepwise Cox model selection.
Data.Table	CRAN	readr2.1.5	Fast reading of delimited files; robust UTF-8 handling.
Dose	Bioconductor	3.17.0	Human gene annotation (Entrez, Ensembl, SYMBOL mappings).
Dynamictreecut	CRAN	1.73	Weighted gene co expression network analysis; pickSoftThreshold, TOM, module detection.
Edger	Bioconductor	3.56.2	Differential expression (voom/linear models); Wilcoxon/empirical Bayes; volcano/heatmap inputs.
Enrichplot	Bioconductor	3.26.2	Disease Ontology enrichment and GSEA helpers (used with clusterProfiler).
Flashclust	CRAN	1.63-1	Adaptive branch cutting for WGCNA module detection.
Gbm	CRAN	1.12	Supervised principal components for survival analysis.
Gene Expression Omnibus (GEO)	NA	GSE43378	Used for model consruction and validation
GEOquery	Bioconductor	NA	Download and parse GEO datasets.
Ggplot2	CRAN	1.0.12	Publication ready heatmaps of expression signatures.
Glmnet	CRAN	3.3.1	Random survival forest for right censored outcomes.
Horseradish Peroxidase-labeled anti-rabbit secondary antibodies	Abcam	ab6721	Secondary antibodies for detection in Western Blot.
Human glioblastoma cell lines (SHG44, U251, A172, T98G)	American Tissue Culture Collection	NA	Human glioblastoma cell lines for glioma research. Cells were cultured in DMEM + 10% FBS, 2 mM L-glutamine.
Igraph	CRAN	10.0.1	Access MSigDB gene sets (e.g., metabolic MRGs); convenient tidy frames.
Image J software	Media Cybernetics, USA	152	Used for visualization
Lentiviral vector encoding lncRNA-IRAIN	GeneChem, Shanghai, China	NA	Lentiviral vectors used for transfection of glioma and HEB cells to establish stable clones.
Limma	Bioconductor	3.48.0	Batch correction (ComBat) and surrogate variable analysis.
Maftools	Bioconductor	1.52.0	Harrell’s concordance index (C index) and survival comparison utilities.
Mass	CRAN	4.1-8	Penalized Cox models: Lasso, Ridge, and Elastic Net.
Msigdbr	CRAN	0.0.5	Support vector machine methods for survival data.
MTT	Sigma-Aldrich	M2128	Used for cell proliferation assays.
Normal glial cells (HEB)	American Tissue Culture Collection	NA	Normal human glial cell line for comparison studies. Cultured in DMEM + 10% FBS.
Org.Hs.Eg.Db	Bioconductor	4.8.3	GO/KEGG enrichment and GSEA; supports multiple ID types.
Pheatmap	CRAN	2.0.0	Grammar of data science; includes dplyr, tidyr, purrr, ggplot2 for wrangling and plotting.
Plsrcox	CRAN	1.5	Likelihood based boosting for Cox models.
PVDF Membranes	Millipore	IPVH00010	Membranes used for protein transfer after SDS-PAGE.
Randomforestsrc	CRAN	6.0-94	Unified resampling (including LOOCV), tuning grids, and model pipelines.
Readr	CRAN	NA	Retrieve matrices/phenotypes from UCSC Xena (e.g., TCGA/GTEx hubs).
SDS-Page Gel (12%)	Bio-Rad	185-5096	System used for quantitative PCR.
Stringdb	Bioconductor	1.20.3	Visualization for enrichment results (dotplot, cnetplot, ridgeplot).
Superpc	CRAN	1.7.7	Partial least squares regression adapted to Cox models.
Survcomp	Bioconductor	3.42.4	Counts normalization and dispersion estimation when needed (optional, complements limma voom).
Survival	CRAN	1.01-2	Fast hierarchical clustering used by WGCNA (optional).
Survivalsvm	CRAN	2.2.2	Generalized boosted regression modeling (gradient boosting).
Survminer	CRAN	3.8-3	Cox proportional hazards models; Kaplan–Meier curves.
TCGAbiolinks	Bioconductor	T9039	Programmatic access to TCGA data; download/prepare expression and clinical data.
The Cancer Genome Atlas Program ( TCGA)	NA	https://www.cancer.gov/tcga; 175 GBM and 534 LGG	Used for model consruction and validation
Tidyverse	CRAN	1.17.6	High performance data manipulation for large matrices/tables.
Timeroc	CRAN	0.5.0	KM visualization and risk table rendering.
TRIzol	Invitrogen	15596026	Reagent for RNA extraction from cells.
UCSC Xena	NA	https://xenabrowser.net/datapages/	Used for model consruction and validation
Ucscxenatools	Bio-Rad	4561044	Used for protein separation in Western Blot analysis.
Wgcna	CRAN	0.95	Correlation matrices for checkpoint genes and feature relationships.