A subscription to JoVE is required to view this content. Sign in or start your free trial.

Research Article

Multistage Genetic, Transcriptomic, and Single-Cell Evidence Prioritizes MAP1LC3A among Ferroptosis-Related Genes in Glioblastoma

94 views

⸱

September 11th, 2026

 ,  ,  ,  ,  ,  ,  , 
1Department of Cerebrovascular Neurosurgery, The First People's Hospital of Zhaoqing, 2The National Key Clinical Specialty, Guangdong Provincial Key Laboratory on Brain Function Repair and Regeneration, Department of Neuro-oncological Surgery, Zhujiang Hospital, Southern Medical University

* These authors contributed equally

In This Article

Summary

A genetically anchored, multistage framework integrating Mendelian randomization, tumor transcriptomics, and single-cell analyses prioritized MAP1LC3A as a ferroptosis-related gene associated with glioblastoma susceptibility and a candidate for future experimental validation.

Abstract

Glioblastoma (GBM) remains a highly aggressive malignancy, and the contribution of ferroptosis-related genes to disease susceptibility remains incompletely understood. A genetically anchored, multistage framework was applied to prioritize ferroptosis-related genes associated with GBM. Among 483 genes curated from FerrDb V2, 315 had candidate cis-expression quantitative trait loci (cis-eQTLs) in eQTLGen, 250 retained at least three independent instruments after linkage disequilibrium clumping, and 226 yielded valid inverse-variance weighted (IVW) Mendelian randomization estimates using a GBM genome-wide association study comprising 6,183 cases and 18,169 controls. Thirty-four genes met the exploratory discovery criteria of P < 0.05 and a Benjamini–Hochberg false discovery rate (BH-FDR) < 0.20, with directionally concordant Bayesian weighted Mendelian randomization (BWMR) estimates. Replication-stage Mendelian randomization using GTEx V10 whole-blood cis-eQTLs supported four genes: ATG7, RPTOR, MAP1LC3A, and CHMP6. Evaluation across three independent tumor–control transcriptomic cohorts demonstrated that MAP1LC3A was consistently downregulated in tumor tissue and showed a significant random-effects pooled estimate (log₂ fold change, −1.273; 95% confidence interval, −1.625 to −0.920; false discovery rate = 0.016), whereas the other three genes lacked comparable cross-cohort statistical support. Single-cell virtual knockout analysis was subsequently performed in a patient-balanced subset of 2,400 malignant cells selected from 4,916 eligible cells across 20 adult IDH-wild-type GBM tumors. Across five independently seeded runs, 3, 15, 4, and 7 robust downstream genes were identified for ATG7, RPTOR, MAP1LC3A, and CHMP6, respectively. The resulting consensus sets comprised 17 unique genes, with RND3 shared across all four targets. Gene Ontology analysis indicated enrichment of cell-adhesion and cell-surface processes, whereas no KEGG or Reactome pathways remained significant after multiple-testing correction. Collectively, these findings prioritize MAP1LC3A for future experimental investigation while distinguishing genetic association, tumor-expression concordance, and computational perturbation from definitive evidence of causality or mechanism.

Introduction

Glioblastoma (GBM) remains a paradigmatic treatment-refractory malignancy. Despite increasingly refined molecular classification and multidisciplinary care, durable improvements in patient outcomes have been limited1. For medically fit patients, current management consists of maximal safe resection followed by radiotherapy with concomitant and adjuvant temozolomide, a regimen established in a landmark randomized trial and retained in contemporary clinical guidelines1,2. Nevertheless, diffuse infiltration and extensive cellular and molecular heterogeneity limit durable disease control, and most patients ultimately experience progression or recurrence, for which no universally effective standard treatment exists1,3. This persistent gap between advances in disease characterization and clinical outcomes underscores the need to identify biologically relevant molecular dependencies that may inform novel therapeutic strategies for GBM.

Ferroptosis is an iron-dependent form of regulated cell death characterized by uncontrolled phospholipid peroxidation and failure of cellular antioxidant defenses, distinguishing it mechanistically from apoptosis and other canonical cell death programs4,5. This process is particularly relevant to GBM, where genetic alterations and metabolic plasticity reshape iron homeostasis, redox balance, and lipid metabolism. Integrated genomic and lipidomic profiling has demonstrated that CDKN2A deletion redistributes oxidizable polyunsaturated fatty acids, thereby creating genotype-dependent ferroptosis susceptibility in GBM models6. Likewise, paired analyses of primary and recurrent tumors have identified relapse-associated alterations in GPX4, ACSL4, and other ferroptosis regulators7. Experimental modulation of ferroptosis-defense pathways has also been shown to influence temozolomide responsiveness in GBM cells and xenograft models8. Collectively, these findings identify ferroptosis as a biologically plausible therapeutic vulnerability in GBM. However, they primarily reflect tumor-state associations or context-dependent experimental observations and do not establish whether constitutive variation in ferroptosis-related gene expression contributes to inherited susceptibility to GBM.

Most human studies investigating ferroptosis in glioma have relied on differential expression, survival modeling, and molecular subtyping analyses using TCGA, CGGA, and GEO datasets9,10. Although these studies established the prognostic relevance of ferroptosis-related transcriptional programs, their observational design cannot determine whether altered gene expression contributes to GBM susceptibility or instead represents a consequence of tumor development. Transcriptome-wide Mendelian randomization has subsequently identified genetically regulated, tissue-dependent genes associated with glioma risk, whereas more recent expression quantitative trait locus (eQTL)- and protein quantitative trait locus (pQTL)-based studies have begun to prioritize potential therapeutic targets for GBM11,12. Nevertheless, previous investigations have generally adopted transcriptome-wide or drug-target-oriented approaches rather than evaluating a prespecified, comprehensive set of ferroptosis-related genes. Furthermore, few studies have integrated a large GBM genome-wide association study (GWAS) with replication-stage Mendelian randomization using an independent eQTL resource, followed by evaluation across multiple tumor–control transcriptomic cohorts. This distinction is important because the genetic regulation of gene expression varies substantially across tissues, and blood-derived eQTL associations cannot be assumed to reflect regulatory effects within brain tumors11,13. Therefore, an integrative framework that combines genetic association, replication-stage Mendelian randomization, cross-cohort tumor transcriptomics, and patient-derived single-cell functional prediction is needed to identify ferroptosis-related genes supported by convergent lines of evidence of their involvement in GBM.

This study investigated whether genetically regulated expression of ferroptosis-related genes is associated with susceptibility to GBM. Discovery-stage and replication-stage Mendelian randomization were combined with gene-expression analyses across three independent tumor–control transcriptomic cohorts. Genes supported by both Mendelian randomization stages were subsequently evaluated in patient-derived single-cell transcriptomic data using virtual perturbation to characterize predicted transcriptional responses in malignant cells. Rather than prioritizing genes solely based on tumor expression signatures, this multistage framework first leveraged inherited genetic variation and then assessed disease-relevant expression patterns alongside cell-resolved computational perturbation profiles. The resulting convergent evidence was used to prioritize ferroptosis-related genes for future experimental investigation in GBM.

Access restricted. Please log in or start a trial to view this content.

Protocol

This study was granted an exemption from ethical review by the Medical Ethics Committee of The First People's Hospital of Zhaoqing (Reference No. B2026-08-03). The study used retrospectively collected, de-identified summary-level genetic and transcriptomic data, including Data Access Committee-approved controlled-access data from EGAD00010001657 and datasets obtained from GEO, eQTLGen, and GTEx in accordance with their applicable access and use conditions. No new participants were recruited, no biospecimens were collected, and no identifiable participant-level data were accessed. Ethical approval and informed consent for the original studies were obtained by the respective data generators, and the controlled-access data were used in accordance with the applicable Data Access Agreement.

Study design

This study employed a multistage framework to prioritize ferroptosis-related genes associated with glioblastoma (GBM) susceptibility and to evaluate their disease-relevant transcriptional effects (Figure 1). First, ferroptosis-related genes curated from FerrDb V2 were evaluated using two-sample Mendelian randomization (MR) with cis-expression quantitative trait locus (cis-eQTL) data and a large GBM genome-wide association study (GWAS). Inverse-variance weighted (IVW) MR served as the primary screening method, Bayesian weighted Mendelian randomization (BWMR) provided a complementary assessment of robustness, and an independent eQTL dataset was used for replication-stage MR. Second, genes supported by the genetic analyses were evaluated across three independent tumor–control transcriptomic cohorts, followed by cross-cohort meta-analysis. Third, patient-derived single-cell transcriptomic data were used to perform virtual gene perturbation in malignant cells and identify reproducible downstream transcriptional responses. These responses were subsequently characterized through functional enrichment and shared-network analyses. Overall, the genetic analyses were designed to identify genes associated with GBM susceptibility, whereas the transcriptomic and single-cell analyses assessed biological concordance and generated hypotheses for subsequent experimental validation.

Data sources

Ferroptosis-related genes were obtained from FerrDb V2, yielding 483 unique human protein-coding genes after gene-symbol harmonization and removal of duplicate entries14. Whole-blood cis-expression quantitative trait locus (cis-eQTL) summary statistics from the eQTLGen Consortium were used as the exposure dataset for discovery-stage Mendelian randomization (MR), whereas GTEx release V10 whole-blood cis-eQTL data served as an independent exposure dataset for replication-stage MR11,15. GBM outcome associations were obtained from controlled-access genome-wide association study (GWAS) summary statistics available through the European Genome-phenome Archive, comprising 6,183 cases and 18,169 controls of European ancestry16. Tissue-level expression of genetically prioritized genes was evaluated across three independent Gene Expression Omnibus (GEO) cohorts: GSE196533, comprising 61 grade 4 glioma samples annotated as GBM in the deposited metadata and nine non-neoplastic brain samples; GSE4290, comprising GBM and epilepsy-derived non-tumor brain samples; and GSE116520, containing paired tumor-core and peritumoral specimens together with non-neoplastic controls17,18,19. Patient-derived Smart-seq2 data from GSE131928 were used for malignant-cell virtual perturbation analysis20. Dataset characteristics and their respective analytical roles are summarized in Table 1. All analyses were performed using previously collected, de-identified datasets for which ethical approval and informed consent had been obtained in the original studies.

Genetic instrument selection and data harmonization

Candidate instruments were restricted to cis-expression quantitative trait loci (cis-eQTLs) associated with gene expression at genome-wide significance (P < 5 × 10⁻8). Variants were clumped using the European reference panel from the 1000 Genomes Project with a linkage disequilibrium (LD) threshold of r2 < 0.001 within a 10,000-kb window. Genes retaining fewer than three independent instruments after clumping were excluded from the primary Mendelian randomization (MR) analysis. Given the exploratory screening objective, a minimum of three instruments was prespecified to retain genes with sparse but strong cis-eQTL support while permitting multi-instrument IVW estimation. This threshold was accompanied by stringent LD clumping and F-statistic filtering; estimates based on only three or four instruments were interpreted cautiously, and sensitivity analyses were performed only when methodologically applicable. Instrument strength was assessed for each variant using the F statistic (F = β2/SE2), where β and SE represent the cis-eQTL effect estimate and its standard error, respectively. Variants with F < 10 were excluded to minimize weak-instrument bias21,22. Exposure and outcome datasets were harmonized by aligning effect alleles and effect directions. Duplicate variants, variants unavailable in the GBM GWAS dataset, and variants with incompatible allele coding were excluded. Because effect-allele frequencies were unavailable for the GBM GWAS, palindromic variants with ambiguous strand orientation were removed rather than inferred. For the same reason, a formal Steiger directionality test was not performed.

Mendelian randomization analyses

The association between genetically predicted gene expression and GBM susceptibility was evaluated using two-sample Mendelian randomization (MR). In the discovery stage, only genes with at least three independent cis-expression quantitative trait locus (cis-eQTL) instruments were included, and the inverse-variance weighted (IVW) method served as the primary analytical approach. Effect estimates were reported as odds ratios (ORs) with 95% confidence intervals (CIs) per unit increase in genetically predicted gene expression. To account for multiple testing across the genes evaluated, IVW P values were adjusted using the Benjamini–Hochberg procedure23. Genes with P < 0.05 and a Benjamini–Hochberg false discovery rate (BH-FDR) < 0.20 were retained as exploratory candidates. This relatively permissive FDR threshold was selected to reduce premature exclusion of potentially relevant genes during the discovery stage; therefore, candidate status was interpreted in conjunction with subsequent analyses rather than as confirmatory evidence. Bayesian-weighted Mendelian randomization (BWMR) was subsequently applied to the discovery-stage candidates using the same harmonized instruments24. Concordance between the IVW and BWMR results was assessed based on both statistical significance and the direction of the effect. Where permitted by the number of available instruments, Cochran's Q test, the MR-Egger intercept test, MR-PRESSO, and leave-one-out analyses were performed to evaluate heterogeneity, horizontal pleiotropy, influential outlying variants, and the influence of individual single-nucleotide polymorphisms (SNPs)22.

Replication-stage MR was performed using GTEx V10 whole-blood cis-eQTL data. The Wald ratio method was applied when only a single instrument was available, whereas the IVW method was used for genes with two or more instruments. Replication was defined by P < 0.05, BH-FDR < 0.20, and an effect direction consistent with the corresponding discovery-stage estimate. Because several GTEx replication estimates were based on only one or two instruments, they were interpreted as supportive evidence for replication rather than as independent evidence of causality.

Cross-cohort transcriptomic evaluation

The four genes prioritized by the discovery- and replication-stage Mendelian randomization (MR) analyses were evaluated across three independent transcriptomic cohorts. For GSE196533, raw RNA-sequencing count data were analyzed using DESeq225. Genes with counts below 10 in all but two samples were excluded, whereas the four target genes were retained regardless of expression filtering. Differential expression was assessed between 61 grade 4 glioma samples annotated as GBM in the deposited metadata and nine non-neoplastic brain samples.

For GSE4290, four samples lacking an explicit histopathological diagnosis were excluded, leaving 77 GBM and 23 non-tumor brain samples. Processed microarray intensities were log2-transformed, quantile-normalized, and analyzed using robust empirical Bayes linear models implemented in limma26. When multiple probes mapped to the same gene, the probe with the highest mean expression across all included samples was selected independently of differential expression significance.

GSE116520 comprised paired tumor-core and peritumoral specimens from 17 patients together with eight non-neoplastic controls. The deposited log-transformed, quantile-normalized expression data were analyzed using limma. Within-patient correlations between tumor-core and peritumoral samples were accounted for using patient-level blocking and the duplicateCorrelation function. Tumor-core versus control was the prespecified comparison for the cross-cohort meta-analysis, whereas peritumoral comparisons and the ordered control–peritumoral–tumor-core trend were evaluated separately.

Study-specific log2 fold changes and standard errors were pooled using a restricted maximum-likelihood random-effects model with Hartung–Knapp inference, as implemented in metafor. Between-study heterogeneity was assessed using Cochran's Q statistic and I2. Pooled P values for the four target genes were adjusted using the Benjamini–Hochberg procedure. Strong transcriptomic support was defined by a meta-analysis false discovery rate (FDR) < 0.05, cohort-level FDR significance in at least two datasets, and concordant effect directions across all three cohorts.

Single-cell virtual knockout analysis

Patient-derived Smart-seq2 data from GSE131928 were used to evaluate the four MR-replicated genes in malignant cells. Adult malignant cells were identified according to the annotations provided in the original study, and equal numbers of cells were randomly sampled from each eligible patient to minimize imbalance in patient representation. Virtual knockout was performed using scTenifoldKnk and repeated across five independent runs. Genes with a Benjamini–Hochberg-adjusted P < 0.05 in an individual run were considered significant. Downstream genes reproduced in at least three of the five runs were defined as the primary consensus set, whereas a more stringent four-of-five criterion was used for sensitivity analysis. These results were interpreted as computational predictions of transcriptional perturbation rather than evidence of direct molecular regulation.

Functional enrichment and shared-network analysis

Functional enrichment analysis was performed using target-specific downstream genes reproducibly identified in at least three of five virtual knockout runs. Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG), and Reactome pathway enrichment were assessed using one-sided hypergeometric tests, with the 1,004 genes included in the network inference serving as the background gene set. P values were adjusted separately for each annotation database using the Benjamini–Hochberg procedure, and an adjusted P < 0.05 was considered statistically significant.

A bipartite network was constructed to represent the relationships between the four knockout targets and their consensus downstream genes. Genes associated with multiple targets were identified based on their shared degree, and overlap among target-specific gene sets was quantified using intersection counts and Jaccard indices. Network edges represent associations between reproducible computational perturbations and should not be interpreted as evidence of direct molecular interactions.

Statistical analysis and reproducibility

Unless otherwise specified, statistical tests were two-sided, and multiple comparisons were controlled using the Benjamini–Hochberg procedure. Analysis-specific significance criteria are described in the corresponding subsections. All analyses were performed using R or Python. Randomized procedures used prespecified seeds, and the analytical code, software versions, and detailed parameter settings were archived to support reproducibility. All datasets were previously collected and de-identified; ethical approval and informed consent were obtained in the original studies.

Access restricted. Please log in or start a trial to view this content.

Results

Selection of ferroptosis-related genes and genetic instruments

A total of 483 ferroptosis-related genes were obtained from FerrDb V2 (Figure 1). Among these, 315 genes were matched to the eQTLGen dataset and had at least one candidate cis-expression quantitative trait locus (cis-eQTL). After linkage disequilibrium clumping, 250 genes retained at least three independent candidate instruments. Following outcome-variant lookup, allele harmonization, ...

Access restricted. Please log in or start a trial to view this content.

Discussion

This study integrated genetic association, replication-stage Mendelian randomization (MR), tumor transcriptomics, and patient-derived single-cell modeling to prioritize ferroptosis-related genes associated with glioblastoma (GBM) susceptibility. Screening 483 curated genes against a GBM genome-wide association study (GWAS) comprising 6,183 cases and 18,169 controls identified 34 discovery-stage candidates, four of which—ATG7, RPTOR, MAP1LC3A, and CHMP6—were supported in the replication analysis using an indep...

Access restricted. Please log in or start a trial to view this content.

Disclosures

The authors declare no competing interests.

Acknowledgements

The authors acknowledge the Cancer Genomics team at The Institute of Cancer Research for providing access to the glioma GWAS summary statistics through the European Genome-phenome Archive (dataset EGAD00010001657). The original generation of these data was supported by Cancer Research UK, including the Bobby Moore Fund, the Wellcome Trust, and the DJ Fielding Medical Research Trust (C1298/A8362).

Access restricted. Please log in or start a trial to view this content.

Materials

List of materials used in this article
NameCompanyCatalog NumberComments
BWMRR packageBWMRBayesian weighted Mendelian randomization
DESeq2BioconductorVersion 1.46.0RNA-sequencing differential-expression analysis
FerrDb V2FerrDbVersion 2Source of 483 curated ferroptosis-related genes
Glioblastoma bulk microarrayNCBI Gene Expression OmnibusGSE4290Transcriptomic evaluation cohort
Glioblastoma GWAS summary statisticsEuropean Genome-phenome ArchiveEGAD00010001657Controlled-access outcome data; 6,183 cases and 18,169 controls
Glioblastoma Smart-seq2 single-cell RNA sequencingNCBI Gene Expression OmnibusGSE131928Malignant-cell virtual-knockout analysis
Grade 4 glioma bulk RNA sequencingNCBI Gene Expression OmnibusGSE196533Transcriptomic evaluation cohort
GTEx whole-blood cis-eQTL summary statisticsGenotype-Tissue Expression projectGTEx V10Replication-stage exposure data
limmaBioconductorVersion 3.62.2Microarray differential-expression analysis
metaforR packageVersion 4.8-0Random-effects meta-analysis
RR Foundation for Statistical ComputingVersion 4.4.2Statistical computing environment
scTenifoldKnkR packageVersion 1.0.3Single-cell virtual-knockout analysis
Tumour-core and peritumoural bulk microarrayNCBI Gene Expression OmnibusGSE116520Transcriptomic evaluation cohort
TwoSampleMRR packageVersion 0.6.29Two-sample Mendelian randomization
Whole-blood cis-eQTL summary statisticseQTLGen ConsortiumeQTLGenDiscovery-stage exposure data

References

  1. Weller M, et al. EANO guidelines on the diagnosis and treatment of diffuse gliomas of adulthood. Nat Rev Clin Oncol. 2021;18:170-186.
  2. Stupp R, et al. Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med. 2005;352:987-996.
  3. McBain C, et al. Treatment options for progression or recurrence of glioblastoma: a network meta-analysis. Cochrane Database Syst Rev. 2021;5:CD013579.
  4. Dixon SJ, et al. Ferroptosis: an iron-dependent form of nonapoptotic cell death. Cell. 2012;149:1060-1072.
  5. Stockwell BR, et al. Ferroptosis: A regulated cell death nexus linking metabolism, redox biology, and disease. Cell. 2017;171:273-285.
  6. Minami JK, et al. CDKN2A deletion remodels lipid metabolism, priming glioblastoma for ferroptosis. Cancer Cell. 2023;41:1048-1060.e1049.
  7. Kram H, et al. Glioblastoma relapses show increased markers of vulnerability to ferroptosis. Front Oncol. 2022;12:841418.
  8. Miao Z, et al. A targetable PRR11-DHODH axis drives ferroptosis- and temozolomide-resistance in glioblastoma. Redox Biol. 2024;73:103220.
  9. Liu HJ, et al. Ferroptosis-related gene signature predicts glioma cell death and glioma patient progression. Front Cell Dev Biol. 2020;8:538.
  10. Dong J, et al. Ferroptosis-related gene contributes to immunity, stemness, and predicts prognosis in glioblastoma multiforme. Front Neurol. 2022;13:829926.
  11. Robinson JW, et al. Transcriptome-wide Mendelian randomization study prioritizing novel tissue-dependent genes for glioma susceptibility. Sci Rep. 2021;11:2329.
  12. Zhang H, Wang Z, Qiao X, Wu J, Cheng C. Investigating potential drug targets for the treatment of glioblastoma: a Mendelian randomization study. BMC Cancer. 2025;25:654.
  13. GTEx Consortium. The GTEx Consortium atlas of genetic regulatory effects across human tissues. Science. 2020;369:1318-1330.
  14. Zhou N, et al. FerrDb V2: update of the manually curated database of ferroptosis regulators and ferroptosis-disease associations. Nucleic Acids Res. 2023;51:D571-D582.
  15. Võsa U, et al. Large-scale cis- and trans-eQTL analyses identify thousands of genetic loci and polygenic scores that regulate blood gene expression. Nat Genet. 2021;53:1300-1310.
  16. Melin BS, et al. Genome-wide association study of glioma subtypes identifies specific differences in genetic susceptibility to glioblastoma and non-glioblastoma tumors. Nat Genet. 2017;49:789-794.
  17. Zeng C, et al. Dissection of transcriptomic and epigenetic heterogeneity of grade 4 gliomas: implications for prognosis. Acta Neuropathol Commun. 2023;11:133.
  18. Sun L, et al. Neuronal and glioma-derived stem cell factor induces angiogenesis within the brain. Cancer Cell. 2006;9:287-300.
  19. Kruthika BS, et al. Transcriptome profiling reveals PDZ binding kinase as a novel biomarker in peritumoral brain zone of glioblastoma. J Neurooncol. 2019;141:315-325.
  20. Neftel C, et al. An integrative model of cellular states, plasticity, and genetics for glioblastoma. Cell. 2019;178:835-849.e821.
  21. Burgess S, Thompson SG. Avoiding bias from weak instruments in Mendelian randomization studies. Int J Epidemiol. 2011;40:755-764.
  22. Papadimitriou N, et al. Physical activity and risks of breast and colorectal cancer: a Mendelian randomisation analysis. Nat Commun. 2020;11:597.
  23. Song W, et al. Causal relationship between gut microbiota and lung squamous cell carcinoma: a bidirectional two-sample Mendelian randomization study. Postgrad Med J. 2025;101:526-534.
  24. Zhao J, et al. Bayesian weighted Mendelian randomization for causal inference based on summary statistics. Bioinformatics. 2020;36:1501-1508.
  25. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15:550.
  26. Ritchie ME, et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015;43:e47.
  27. Yun D, et al. A novel prognostic signature based on glioma essential ferroptosis-related genes predicts clinical outcomes and indicates treatment in glioma. Front Oncol. 2022;12:897702.
  28. Li R, et al. Identification of candidate genes associated with prognosis in glioblastoma. Front Mol Neurosci. 2022;15:913328.
  29. Bai H, Inoue J, Kawano T, Inazawa J. A transcriptional variant of the LC3A gene is involved in autophagy and frequently inactivated in human cancers. Oncogene. 2012;31:4397-4408.
  30. Hayashima K, Kimura I, Katoh H. Role of ferritinophagy in cystine deprivation-induced cell death in glioblastoma cells. Biochem Biophys Res Commun. 2021;539:56-63.
  31. Wu Y, et al. ALDH1-mediated autophagy sensitizes glioblastoma cells to ferroptosis. Cells. 2022;11:4015.
  32. Carriere A, et al. ERK1/2 phosphorylate Raptor to promote Ras-dependent activation of mTOR complex 1 (mTORC1). J Biol Chem. 2011;286:567-577.
  33. Dai E, Meng L, Kang R, Wang X, Tang D. ESCRT-III-dependent membrane repair blocks ferroptosis. Biochem Biophys Res Commun. 2020;522:415-421.
  34. Comincini S, et al. microRNA-17 regulates the expression of ATG7 and modulates the autophagy process, improving the sensitivity to temozolomide and low-dose ionizing radiation treatments in human glioblastoma cells. Cancer Biol Ther. 2013;14:574-586.
  35. Wang L, et al. Autophagy mediates glucose starvation-induced glioblastoma cell quiescence and chemoresistance through coordinating cell metabolism, cell cycle, and survival. Cell Death Dis. 2018;9:213.

Access restricted. Please log in or start a trial to view this content.

Reprints and Permissions

Explore More Articles

Ferroptosis GenesGlioblastoma SusceptibilityMendelian RandomizationSingle Cell AnalysisTranscriptomic CohortseQTL AnalysisGene OntologyTumor DownregulationGenetic Association
Contact Us
Recommend to Library

Privacy

Terms of Use

Policies

Copyright © 2026 MyJoVE Corporation. All rights reserved.