Single-Cell Genetic Architecture Identifies GSTM1 as a Novel Therapeutic Target for Reversing T-Cell Ferroptosis in NSCLC

Single-Cell Genetic Architecture Identifies GSTM1 as a Novel Therapeutic Target for Reversing T-Cell Ferroptosis in NSCLC | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Single-Cell Genetic Architecture Identifies GSTM1 as a Novel Therapeutic Target for Reversing T-Cell Ferroptosis in NSCLC Shouyong Xiao, Siyun Wu, Xianfeng Shao, Ming Chao, Quibo Huang, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9010841/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 12 You are reading this latest preprint version Abstract Background Ferroptosis has emerged as a pivotal mechanism in cancer surveillance, particularly in determining the efficacy of immunotherapy. However, the precise genetic switches that regulate ferroptosis sensitivity within specific immune cell lineages remain poorly characterized due to the resolution limits of bulk tissue analysis. Methods To address this, we constructed a high-resolution genetic map of ferroptosis in non-small cell lung cancer (NSCLC). We applied a multi-stage integrative framework, harmonizing single-cell expression quantitative trait loci (sc-eQTL) from 14 immune cell populations with large-scale genomic data from the FinnGen cohort (N = 385,195). Results Our analysis identified 40 putative causal genes driving NSCLC susceptibility. A key finding is that GSTM1 expression in CD8 + naive T cells confers a robust protective effect (OR = 0.60, P = 2.17 x 10^-4). Multi-omic validation indicates that GSTM1 functions as a "metabolic shield," enhancing T-cell functional quality rather than simple infiltration density. Furthermore, we reveal a striking cell-state-dependent antagonism for MAPK3, which acts as a protective factor in the naive state (OR = 0.67) but switches to a risk-promoting factor in effector CD4 + T cells (OR = 1.39). Conclusions This study uncovers a lineage-specific genetic architecture of ferroptosis, highlighting GSTM1 and MAPK3 as critical regulators of T-cell metabolic fitness. These findings suggest that stratifying patients based on GSTM1 status could serve as a novel strategy to guide precision immunotherapy. Non-small cell lung cancer Ferroptosis Mendelian randomization Single-cell eQTL CD8 + T cells GSTM1 Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 1. Background Non-small cell lung cancer (NSCLC) remains the leading cause of cancer-related mortality globally. While immune checkpoint blockade (ICB) targeting PD-1/PD-L1 has revolutionized clinical care, primary and acquired resistance remains a major bottleneck, with only ~ 20% of unselected patients deriving durable benefit[ 1 , 2 ]. Emerging evidence suggests that the metabolic fitness of tumor-infiltrating lymphocytes (TILs) is a critical determinant of ICB efficacy. Specifically, ferroptosis—an iron-dependent form of non-apoptotic cell death driven by lipid peroxidation—has been identified as a "metabolic checkpoint" that limits the survival and cytotoxicity of CD8 + T cells in the oxidative tumor microenvironment (TME) [ 3 , 4 ]. Seminal work by Wang et al. (Nature, 2019) demonstrated that CD8 + T cells suppress tumor growth by inducing ferroptosis in cancer cells; however, the T cells themselves are also vulnerable to ferroptosis, which can lead to metabolic exhaustion and immune failure[ 4 , 5 ]. Therefore, identifying the specific genetic switches that regulate ferroptosis sensitivity within immune cells is crucial for developing next-generation immunotherapies. Current understanding of ferroptosis regulation largely relies on bulk tissue transcriptomics or preclinical mouse models, which fail to capture the complex, cell-state-dependent genetic architecture of the human immune system. For instance, a gene that protects tumor cells might paradoxically impair T-cell function, or have divergent effects in naive versus effector states[ 6 ]. To address this resolution gap, we performed a systematic genetic dissection of ferroptosis in NSCLC. Moving beyond traditional association studies, we integrated single-cell expression quantitative trait loci (sc-eQTL) from 14 immune cell lineages (OneK1K cohort) with large-scale genomic data from the FinnGen study (N = 385,195). Our goal was not merely to catalog genes, but to identify lineage-specific therapeutic targets. Here, we report that GSTM1 and MAPK3 act as cell-state-dependent regulators of ferroptosis, providing genetic evidence for a "Quality over Quantity" paradigm in anti-tumor immunity. 2. Methods 2.1 Study Design and Data Sources The overall study design is illustrated in Fig. 1 . We implemented a three-stage integrative analysis: (1) identification of causal genes using Two-Sample Mendelian Randomization (MR); (2) fine-mapping of causal variants via Bayesian colocalization; and (3) clinical validation using single-cell tumor transcriptomics. Detailed information on all data sources is provided in Table 1 . 2.2 Ferroptosis-Related Gene Selection A comprehensive panel of 483 ferroptosis-related genes was curated from FerrDb [ 7 ], a manually annotated database comprising drivers, suppressors, and markers of ferroptosis. 2.3 Single-Cell eQTL and GWAS Datasets Cell-type-specific instrumental variables were derived from the OneK1K cohort [ 8 ], which profiled sc-eQTLs across 14 immune cell types (including CD4 + subsets, CD8 + subsets, and NK cells) from 982 donors. Genetic associations for NSCLC were retrieved from the FinnGen R12 release [ 9 ], comprising 2,564 cases and 382,631 controls ( Table 1 ). 2.4 Mendelian Randomization Analysis For each gene-cell pair, independent significant cis-eQTLs (P < 5 × 10^-8, r^2 < 0.001) were selected as instrumental variables. The distributions of SNP counts, P-values, and effect sizes for these instrumental variables are detailed in Supplementary Figs. 2–4, verifying the robustness of our selection process. The Inverse Variance Weighted (IVW) method was used as the primary test[ 10 ]. Sensitivity analyses, including MR-Egger and Weighted Median, were performed to assess horizontal pleiotropy ( Supplementary Table 3 ). Correction for multiple testing was applied using the Benjamini-Hochberg method. 2.5 Bayesian Colocalization To distinguish valid causal instruments from linkage disequilibrium (LD) artifacts, we performed colocalization analysis using the coloc R package [ 11 ]. We tested five hypotheses (H0 to H4), where a posterior probability of H4 (PP.H4 > 0.8) indicates a shared causal variant between the sc-eQTL and GWAS signals. 2.6 Single-Cell Validation (TISCH2) The expression landscape of prioritized genes was validated using the TISCH2 database[ 12 ]. We analyzed the NSCLC_GSE131907 dataset to compare gene expression levels between tumor-infiltrating lymphocytes (TILs) and peripheral blood cells. 3. Results 3.1 Systematic Identification of Ferroptosis Drivers in NSCLC We first established the baseline genetic landscape by screening 483 ferroptosis-related genes using summary-level data from the FinnGen R12 cohort (N = 385,195 individuals, Table 1 ). As detailed in Table 2 , this systematic screen identified 40 genes with significant causal associations with NSCLC risk (P < 0.05). Among the protective factors shown in Fig. 2 , MAPK3 exhibited the most robust evidence (OR = 0.910, 95% CI: 0.876–0.946, P = 1.58 × 10^-6), followed by MAP3K11 (OR = 0.913, P = 3.37 × 10^-5) and TMBIM4 (OR = 0.864, P = 2.80 × 10^-4). On the risk side, HDDC3 was identified as the strongest risk factor (OR = 1.267, 95% CI: 1.123–1.430, P = 1.27 × 10^-4), along with CHP1 (OR = 1.233, P = 1.10 × 10^-3). Supplementary Table 1 provides the complete summary statistics for all screened genes. Importantly, sensitivity analyses using MR-Egger and Weighted Median methods showed consistent effect directions ( Supplementary Table 3 ), ruling out significant horizontal pleiotropy. 3.2 Cell-Type Specificity of Genetic Regulation While whole-blood analysis provided a broad overview, integrating single-cell eQTL data (OneK1K, N = 982 donors) revealed that these genetic effects are highly cell-type specific. Table 3 lists the 8 significant gene-cell type pairs identified after strict filtering. The heatmap in Fig. 3 illustrates this heterogeneity. For instance, while GSTM1 showed a moderate protective effect in whole blood (OR = 0.949, Table 2 ), single-cell analysis pinpointed its primary site of action to T cells. Specifically, GSTM1 expression was significantly protective in CD8 + Naive T cells (OR = 0.595, P = 2.17 × 10^-4) and CD4 + Naive T cells (OR = 0.764, P = 7.70 × 10^-4), but showed no significant effect in B cells or Monocytes ( Table 3 ). Similarly, HRAS was identified as a protective factor specifically in CD8 + Effector T cells (OR = 0.814, P = 0.017), a granularity that was masked in the bulk analysis. 3.3 Deep Dive into GSTM1: A T-Cell Guardian We focused on GSTM1 due to its strong effect size and lineage specificity. As shown in Fig. 4 , the protective effect of GSTM1 forms a gradient across T cell differentiation states. The strongest protection was observed in the naive CD8 + compartment (OR = 0.595), which slightly attenuated in the effector compartment (OR = 0.755, P = 0.045, Table 3 ). To validate the genetic architecture, we performed Bayesian colocalization. Figure 5 A shows that the probability of a shared causal variant (PP.H4) varies across lineages, peaking in CD8 + Effector Memory T cells (PP.H4 = 0.12). The regional association plot (Fig. 5 B) and the SNP distribution data in Supplementary Fig. 2 confirm that the instruments used for GSTM1 are robust (Median N SNPs = 17) and distinct from potential confounders. 3.4 MAPK3 Exhibits Antagonistic Pleiotropy A key finding of this study is the cell-state-dependent antagonism of MAPK3 . As detailed in Table 3 and visualized in Fig. 6 , MAPK3 expression confers significant protection in naive lymphocytes: Naive CD4 + T cells : OR = 0.667 (95% CI: 0.537–0.829, P = 2.56 × 10^-4) Naive B cells : OR = 0.677 (95% CI: 0.544–0.844, P = 5.29 × 10^-4) However, this effect is strikingly reversed in the effector state. In Effector CD4 + T cells , higher MAPK3 expression is associated with increased lung cancer risk (OR = 1.388, 95% CI: 1.089–1.769, P = 8.11 × 10^-3). This bidirectional effect explains why the whole-blood signal (OR = 0.910, Table 2 ) appeared weaker than the specific naive cell signals—the bulk tissue average likely diluted the opposing effects from different cell subsets. 3.5 Clinical Validation and Mechanism To bridge these genetic findings with clinical reality, we analyzed tumor transcriptomics (Fig. 7 ). Consistent with the genetic "Quality over Quantity" hypothesis, GSTM1 expression in the tumor microenvironment correlated positively with functional T-cell signatures rather than simple infiltration density. Based on these multi-omic lines of evidence, we propose the mechanism illustrated in Fig. 8 : GSTM1 upregulation enhances glutathione (GSH) production and GPX4 activity, thereby shielding CD8 + T cells from lipid-ROS-induced ferroptosis and preserving their anti-tumor efficacy. 4. Discussion In this study, we leveraged single-cell genetics to dismantle the complexity of ferroptosis regulation in lung cancer. Our central finding is the identification of GSTM1 and MAPK3 as potent, lineage-dependent regulators. Unlike traditional bulk-tissue analyses, our sc-eQTL approach revealed that the genetic control of ferroptosis varies not just between lineages but also between differentiation states. 4.1 GSTM1: A Metabolic Checkpoint for T-Cell Fitness Our results align with the emerging paradigm establishing ferroptosis as a metabolic vulnerability in CD8 + T cells. Wang et al. previously demonstrated that maintaining redox homeostasis is essential for CD8 + T cell anti-tumor function[ 4 ]. Extending this, we identified GSTM1 as a critical "metabolic shield." GSTM1 encodes a cytosolic enzyme that detoxifies lipid peroxidation products. Our single-cell validation (Fig. 7 ) indicates that GSTM1 expression is positively correlated with T-cell cytotoxicity scores but not necessarily with the absolute number of infiltrating cells. This supports a "Quality over Quantity" model: high GSTM1 levels may not recruit more T cells, but ensure that infiltrating T cells survive the lipid-ROS-rich tumor microenvironment, preventing them from undergoing ferroptosis and sustaining their killing capacity [ 13 ]. 4.2 The "Decoupling" Phenomenon: Unmasking Gene-Environment Interactions A key observation in our study was the "decoupling" between the strong causal effect of GSTM1 in MR analysis (OR = 0.60) and the moderate colocalization probability (PP.H4 = 0.12) at the GWAS locus. Rather than a technical artifact, we propose this reflects a profound Gene-Environment Interaction (GxE) . As demonstrated by Goto et al., the prognostic impact of the GSTM1 null genotype is latent and profoundly stratified by environmental triggers—specifically, smoking status[ 14 ]. In non-smokers, the lack of GSTM1 may be tolerated; however, under the oxidative stress of cigarette smoke, the defect becomes lethal. Since standard GWAS cohorts mix smokers and non-smokers, the "environmental noise" dilutes the genetic signal, leading to lower colocalization scores. Our sc-eQTL MR, by isolating the specific genetic effect on gene expression, successfully unmasked this causal link. 4.3 Antagonistic Pleiotropy of MAPK3 The cell-state-dependent antagonism of MAPK3 is a novel finding. We observed that MAPK3 expression is protective in naive T cells but risk-promoting in effector T cells. This dichotomy can be explained by the dual role of MAPK signaling. In the naive state, robust MAPK signaling is required for initial antigen priming and activation[ 15 ]. However, in the effector phase, chronic hyperactivation of the MAPK pathway is a hallmark of T-cell exhaustion [ 16 ]. Xu et al. recently reported that excessive lipid uptake promotes exhaustion via metabolic stress[ 17 ]. Our data suggests MAPK3 follows a "Goldilocks" principle: beneficial for priming naive cells, but detrimental if overactive in effector cells, likely by accelerating exhaustion-associated ferroptosis. This insight challenges the use of broad-spectrum MAPK inhibitors, which might inadvertently impair naive T-cell priming. 4.4 Clinical Implications : From Genetics to Therapy Our findings have immediate translational relevance. First, GSTM1 status could serve as a precision biomarker. Ritambhara et al. previously linked GSTM1 polymorphisms to chemotherapy toxicity[ 18 ]. ; our data extends this to immunotherapy. We propose that NSCLC patients with GSTM1-null genotypes or low intratumoral GSTM1 expression may possess T cells that are hypersensitive to ferroptosis. Second, this "metabolic defect" is potentially druggable. Since GSTM1 functions by replenishing glutathione (GSH) to fuel GPX4 activity, patients with low GSTM1 might benefit from combinatorial therapies that include ferroptosis inhibitors (e.g., liproxstatin-1) or clinically available antioxidants such as N-acetylcysteine (NAC). By artificially boosting the antioxidative capacity of T cells, we may be able to rescue the "ferroptosis-prone" phenotype caused by GSTM1 deficiency, thereby sensitizing resistant tumors to PD-1 blockade. This hypothesis aligns with recent preclinical success in manipulating ferroptosis to enhance ICB efficacy [ 19 ]. 4.5 Limitations Our study has limitations. First, the GWAS data (FinnGen) is predominantly of European ancestry. Given that GSTM1 deletion frequencies vary significantly between Asian and European populations[ 20 , 21 ], validation in Asian cohorts is warranted. Second, while we utilized large-scale single-cell datasets, direct experimental validation (e.g., GSTM1 knockdown in T-cell co-culture models) would further strengthen the mechanistic conclusions. Declarations Availability of data and materials The data supporting the findings of this study are publicly available. 1.NSCLC GWAS summary statistics are available from the FinnGen study (Release R12) at https://www.finngen.fi/en/access_results. 2.Single-cell eQTL summary statistics were obtained from the OneK1K cohort at https://onek1k.org/. 4.Ferroptosis-related genes were retrieved from the FerrDb V2 database at http://www.zhounan.org/ferrdb/. 5.Single-cell tumor transcriptomic data for validation are accessible via the TISCH2 database at http://tisch.comp-genomics.org/. All other data generated or analyzed during this study are included in this published article and its supplementary information files. Conflict of Interest Disclosures The authors declare that they have no competing interests. Ethics approval and consent to participate This study is a secondary analysis of publicly available, de-identified summary-level data. The original studies (including FinnGen and OneK1K) obtained relevant ethical approvals and informed consent from all participants. According to institutional policies, further specific ethical approval for this meta-analysis of summary statistics was not required. Funding This work was supported by: The Joint Basic Research Program of Yunnan Provincial Department of Science and Technology and Kunming Medical University (Grant No. 202401AY07001-369) The Yunnan Provincial Innovative Research Team for Thoracic Tumor Prevention and Control (Grant No. 202405AS350015) The Development of a Precision Prevention and Full-Cycle Intelligent Management System for Regionally Prevalent Lung Cancer in Yunnan (Grant No. 202303AC100203) The Expert Workstation of Dr. Li Yin (Grant No. 202405AF140055) Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication. Author Contributions Dr. Xiao had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Concept and design: Xiao, Ye. Acquisition, analysis, or interpretation of data: Wu, Shao, Chao, Huang, Li. Drafting of the manuscript: Xiao. Critical revision of the manuscript for important intellectual content: All authors. Statistical and bioinformatics analysis: Xiao, Shao. Obtained funding: Ye. Administrative, technical, or material support: Wu, Chao, Huang, Li. Supervision: Ye. Acknowledgements We gratefully acknowledge the participants and investigators of the FinnGen study, the OneK1K cohort, and the eQTLGen Consortium for sharing the GWAS and sc-eQTL summary statistics used in this research. We also thank the developers and maintainers of FerrDb and TISCH2 for providing valuable resources for ferroptosis and single-cell research. Artificial Intelligence (AI) Use Statement The authors declare that no artificial intelligence tools were used in the generation of study data, statistical analyses, or scientific interpretation. Language editing support was limited to standard grammar and clarity checks, with full responsibility for the content retained by the authors. References Siegel RL, Giaquinto AN, Jemal A. Cancer statistics, 2024. CA Cancer J Clin. 2024;74:12–49. Chen DS, Mellman I. Elements of cancer immunity and the cancer-immune set point. Nature. 2017;541:321–30. Dixon SJ, Lemberg KM, Lamprecht MR, Skouta R, Zaitsev EM, Gleason CE, Patel DN, Bauer AJ, Cantley AM, Yang WS, et al. Ferroptosis: an iron-dependent form of nonapoptotic cell death. Cell. 2012;149:1060–72. Wang W, Green M, Choi JE, Gijon M, Kennedy PD, Johnson JK, Liao P, Lang X, Kryczek I, Sell A, et al. CD8(+) T cells regulate tumour ferroptosis during cancer immunotherapy. Nature. 2019;569:270–4. Tang R, Xu J, Zhang B, Liu J, Liang C, Hua J, Meng Q, Yu X, Shi S. Ferroptosis, necroptosis, and pyroptosis in anticancer immunity. J Hematol Oncol. 2020;13:110. Hong Y. Single-Cell Transcriptome-Wide Mendelian Randomization and Colocalization Analyses Uncover Cell-Specific Mechanisms in Epigenetic Age Acceleration. FASEB J. 2025;39:e71310. Zhou N, Yuan X, Du Q, Zhang Z, Shi X, Bao J, Ning Y, Peng L. FerrDb V2: update of the manually curated database of ferroptosis regulators and ferroptosis-disease associations. Nucleic Acids Res. 2023;51:D571–82. Yazar S, Alquicira-Hernandez J, Wing K, Senabouth A, Gordon MG, Andersen S, Lu Q, Rowson A, Taylor TRP, Clarke L, et al. Single-cell eQTL mapping identifies cell type-specific genetic control of autoimmune disease. Science. 2022;376:eabf3041. Kurki MI, Karjalainen J, Palta P, Sipila TP, Kristiansson K, Donner KM, Reeve MP, Laivuori H, Aavikko M, Kaunisto MA, et al. FinnGen provides genetic insights from a well-phenotyped isolated population. Nature. 2023;613:508–18. Hemani G, Zheng J, Elsworth B, Wade KH, Haberland V, Baird D, Laurin C, Burgess S, Bowden J, Langdon R et al. The MR-Base platform supports systematic causal inference across the human phenome. Elife 2018, 7. Giambartolomei C, Vukcevic D, Schadt EE, Franke L, Hingorani AD, Wallace C, Plagnol V. Bayesian test for colocalisation between pairs of genetic association studies using summary statistics. PLoS Genet. 2014;10:e1004383. Han Y, Wang Y, Dong X, Sun D, Liu Z, Yue J, Wang H, Li T, Wang C. TISCH2: expanded datasets and new tools for single-cell transcriptome analyses of the tumor microenvironment. Nucleic Acids Res. 2022;51:D1425–31. Board PG, Menon D. Glutathione transferases, regulators of cellular metabolism and physiology. Biochim Biophys Acta. 2013;1830:3267–88. Goto I, Yoneda S, Yamamoto M, Kawajiri K. Prognostic significance of germ line polymorphisms of the CYP1A1 and glutathione S-transferase genes in patients with non-small cell lung cancer. Cancer Res. 1996;56:3725–30. Jiang X, Stockwell BR, Conrad M. Ferroptosis: mechanisms, biology and role in disease. Nat Rev Mol Cell Biol. 2021;22:266–82. Wherry EJ, Kurachi M. Molecular and cellular insights into T cell exhaustion. Nat Rev Immunol. 2015;15:486–99. Xu S, Chaudhary O, Rodriguez-Morales P, Sun X, Chen D, Zappasodi R, Xu Z, Pinto AFM, Williams A, Schulze I, et al. Uptake of oxidized lipids by the scavenger receptor CD36 promotes lipid peroxidation and dysfunction in CD8(+) T cells in tumors. Immunity. 2021;54:1561–e15771567. Ritambhara TS, Vijayaraghavalu S, Siddiqui MA, Al-Khedhairy AA, Kumar M. Clinical response of carboplatin-based chemotherapy and its association to genetic polymorphism in lung cancer patients from North India - A clinical pharmacogenomics study. J Cancer Res Ther. 2022;18:109–18. Zhou J, Zhang H, Pan WW, Peng C, Tang H, Yuan G, Peng F. Isocucurbitacin B targets STAT3 to induce ferroptosis and promote anti-PD1 immunotherapy responses in breast cancer. Int J Surg 2026. Chen J, Yang H, Teo ASM, Amer LB, Sherbaf FG, Tan CQ, Alvarez JJS, Lu B, Lim JQ, Takano A, et al. Genomic landscape of lung adenocarcinoma in East Asians. Nat Genet. 2020;52:177–86. Hayes JD, Strange RC. Glutathione S-transferase polymorphisms and their biological consequences. Pharmacology. 2000;61:154–66. Tables Tables 1 to 3 are available in the Supplementary Files section. Additional Declarations No competing interests reported. Supplementary Files Table1.xlsx Table 1. Data sources and characteristics of the GWAS and sc-eQTL datasets used in this study. Table2.xlsx Table 2. Summary of the 40 ferroptosis-related genes causally associated with NSCLC risk identified by MR. Table3.xlsx Table 3. Significant cell-type-specific causal associations (P < 0.05) identified between ferroptosis genes and NSCLC risk. SupplementaryTable3.xlsx SupplementaryTable2.xlsx SupplementaryTable1.xlsx SupplementaryFigure2.pdf SupplementaryFigure4.pdf SupplementaryFigure1.pdf SupplementaryFigure3.pdf Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 30 Apr, 2026 Reviews received at journal 28 Apr, 2026 Reviewers agreed at journal 16 Apr, 2026 Reviews received at journal 13 Apr, 2026 Reviewers agreed at journal 05 Apr, 2026 Reviewers agreed at journal 04 Apr, 2026 Reviewers agreed at journal 30 Mar, 2026 Reviewers invited by journal 30 Mar, 2026 Editor invited by journal 10 Mar, 2026 Editor assigned by journal 04 Mar, 2026 Submission checks completed at journal 04 Mar, 2026 First submitted to journal 02 Mar, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9010841","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":614973023,"identity":"3db55922-0fcf-44eb-b303-cc0fd8334a40","order_by":0,"name":"Shouyong Xiao","email":"","orcid":"","institution":"The Third Affiliated Hospital of Kunming Medical University (Yunnan Cancer Hospital, Yunnan Hospital of Peking University Cancer Hospital)","correspondingAuthor":false,"prefix":"","firstName":"Shouyong","middleName":"","lastName":"Xiao","suffix":""},{"id":614973025,"identity":"bee6c5d3-498d-4c21-9caf-799c1ea5ebe2","order_by":1,"name":"Siyun Wu","email":"","orcid":"","institution":"The Third Affiliated Hospital of Kunming Medical University, Yunnan Cancer Hospital, Yunnan Hospital of Peking University Cancer Hospital)","correspondingAuthor":false,"prefix":"","firstName":"Siyun","middleName":"","lastName":"Wu","suffix":""},{"id":614973030,"identity":"e5ac7668-cc93-4b37-a87a-ca350c69bede","order_by":2,"name":"Xianfeng Shao","email":"","orcid":"","institution":"The Third Affiliated Hospital of Kunming Medical University (Yunnan Cancer Hospital, Yunnan Hospital of Peking University Cancer Hospital)","correspondingAuthor":false,"prefix":"","firstName":"Xianfeng","middleName":"","lastName":"Shao","suffix":""},{"id":614973031,"identity":"0aefa6ca-9a65-47a9-bcb8-8134bfa96aeb","order_by":3,"name":"Ming Chao","email":"","orcid":"","institution":"The Third Affiliated Hospital of Kunming Medical University (Yunnan Cancer Hospital, Yunnan Hospital of Peking University Cancer Hospital)","correspondingAuthor":false,"prefix":"","firstName":"Ming","middleName":"","lastName":"Chao","suffix":""},{"id":614973032,"identity":"1da80ed4-d050-4a91-8370-67d26bb09e09","order_by":4,"name":"Quibo Huang","email":"","orcid":"","institution":"The Third Affiliated Hospital of Kunming Medical University (Yunnan Cancer Hospital, Yunnan Hospital of Peking University Cancer Hospital)","correspondingAuthor":false,"prefix":"","firstName":"Quibo","middleName":"","lastName":"Huang","suffix":""},{"id":614973033,"identity":"ddb5f80c-7def-4887-93d4-a22139299aaf","order_by":5,"name":"Guangjian Li","email":"","orcid":"","institution":"The Third Affiliated Hospital of Kunming Medical University (Yunnan Cancer Hospital, Yunnan Hospital of Peking University Cancer Hospital)","correspondingAuthor":false,"prefix":"","firstName":"Guangjian","middleName":"","lastName":"Li","suffix":""},{"id":614973034,"identity":"fbc9b13f-11dd-49f4-9349-1e904418dd49","order_by":6,"name":"Lianhua Ye","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA5UlEQVRIiWNgGAWjYBACPgkgkVAhwcwvAeZLyBDUwgZS+eCMDbvkDAbGBqAWHqK0MD5sS+M3uAHWwkCEFukewwcJbIeljW83H390o8aCh4H98NENeLXInDE2SOA5bGx251hic84xoMN40tJu4HdYjplEgsThZLMbOYbNOWxALRI8ZoS0mP9IMDhcv3kGSMs/4rSYMSQkpDEbSAC15LYRpSWtWCLhgA2zxI20xNm5fRI8bIT8wi+RvPHjz3/AqJyRfOBzzrc6OX72w8fwamFg4DBAsxe/chBgf0BYzSgYBaNgFIxsAAAv9kPWusCOEAAAAABJRU5ErkJggg==","orcid":"","institution":"The Third Affiliated Hospital of Kunming Medical University (Yunnan Cancer Hospital, Yunnan Hospital of Peking University Cancer Hospital)","correspondingAuthor":true,"prefix":"","firstName":"Lianhua","middleName":"","lastName":"Ye","suffix":""}],"badges":[],"createdAt":"2026-03-02 13:39:53","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9010841/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9010841/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":105984370,"identity":"ea1b7ac6-2bb2-467d-9dbe-4d2bf27c8371","added_by":"auto","created_at":"2026-04-02 07:13:48","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":1340982,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eStudy Design.\u003c/strong\u003e Overview of the integrative framework combining Ferroptosis Gene Set (N=483), Single-Cell eQTL data (OneK1K), and NSCLC GWAS (FinnGen). The workflow proceeds from MR screening to cell-state specificity analysis and multi-stage validation.\u003c/p\u003e","description":"","filename":"Figure1studydesign.png","url":"https://assets-eu.researchsquare.com/files/rs-9010841/v1/d7964dd04e4682da1f90dd52.png"},{"id":105984335,"identity":"fd7f1f1f-3ad8-41d9-af1a-3be11d155d77","added_by":"auto","created_at":"2026-04-02 07:13:42","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":76721,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eIdentification of NSCLC-Associated Ferroptosis Genes.\u003c/strong\u003e Forest plot displaying the odds ratios (OR) of 40 significant ferroptosis-related genes. Genes are ranked by p-value. Notable protective genes include \u003cstrong\u003eMAPK3\u003c/strong\u003e (OR=0.91) and \u003cstrong\u003eGSTM1\u003c/strong\u003e (OR=0.95 in whole blood), while \u003cstrong\u003eHDDC3\u003c/strong\u003e (OR=1.27) and \u003cstrong\u003eCHP1\u003c/strong\u003e (OR=1.23) are top risk factors.\u003c/p\u003e","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-9010841/v1/736225fb19086380d0a068c8.png"},{"id":105984355,"identity":"f5f4cdef-1667-47be-b1d2-5d574ab05bd9","added_by":"auto","created_at":"2026-04-02 07:13:45","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":165581,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCell-Type Specificity of Genetic Regulation.\u003c/strong\u003e (A) Heatmap and (B) Dot plot illustrating the heterogeneous causal effects of key ferroptosis genes across immune cell types. The size of dots represents significance, and color indicates effect direction (Blue=Protective, Red=Risk).\u003c/p\u003e","description":"","filename":"Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-9010841/v1/2587d82629aec7f5aa1cf16a.png"},{"id":105984338,"identity":"884d9ea1-044a-40a5-902f-d266eb4ca325","added_by":"auto","created_at":"2026-04-02 07:13:43","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":220467,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eGSTM1 Effects Across Cell Types.\u003c/strong\u003eDetailed forest plot of \u003cem\u003eGSTM1\u003c/em\u003e causal effects in T cell subsets. The strongest protection is seen in \u003cstrong\u003eCD8+ Naive T cells\u003c/strong\u003e (OR=0.60), followed by \u003cstrong\u003eCD4+ Naive T cells\u003c/strong\u003e (OR=0.76) and \u003cstrong\u003eCD8+ Effector T cells\u003c/strong\u003e(OR=0.76).\u003c/p\u003e","description":"","filename":"Figure4.png","url":"https://assets-eu.researchsquare.com/files/rs-9010841/v1/289901435af9be9a40dfb5a4.png"},{"id":105984367,"identity":"6ef373ee-a2fa-4677-8e52-9e47e8f28f83","added_by":"auto","created_at":"2026-04-02 07:13:48","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":111780,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eColocalization Landscape of GSTM1.\u003c/strong\u003e (A) Lollipop plot showing the Posterior Probability of Hypothesis 4 (PP.H4) across immune lineages, peaking in \u003cstrong\u003eCD8+ Effector Memory T cells\u003c/strong\u003e (PP.H4 = 0.12). (B) Regional mirror plot comparing \u003cem\u003eGSTM1\u003c/em\u003e eQTL signals (top, blue) and NSCLC GWAS signals (bottom, red) at Chromosome 1 (~110 Mb).\u003c/p\u003e","description":"","filename":"Figure5.png","url":"https://assets-eu.researchsquare.com/files/rs-9010841/v1/3c40a906ca58660316de7c7c.png"},{"id":105984344,"identity":"b67db39c-108e-4e02-bf52-8e9828f4610e","added_by":"auto","created_at":"2026-04-02 07:13:43","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":110634,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMAPK3 Shows Cell-State-Dependent Effects.\u003c/strong\u003eComparison of \u003cem\u003eMAPK3\u003c/em\u003e causal estimates in naive versus effector populations. \u003cem\u003eMAPK3\u003c/em\u003e is protective in \u003cstrong\u003eNaive B cells\u003c/strong\u003e (OR=0.68) and \u003cstrong\u003eNaive CD4+ T cells\u003c/strong\u003e (OR=0.67) but shifts to a risk factor in \u003cstrong\u003eEffector CD4+ T cells\u003c/strong\u003e (OR=1.39).\u003c/p\u003e","description":"","filename":"Figure6.png","url":"https://assets-eu.researchsquare.com/files/rs-9010841/v1/3c64e94b2dd60920ec56f850.png"},{"id":105984350,"identity":"f25c5de0-5085-4384-9e38-b52be10fb683","added_by":"auto","created_at":"2026-04-02 07:13:44","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":169680,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eClinical Validation: Infiltration vs. Prognosis.\u003c/strong\u003e (A) Forest plot of survival associations (Z-scores) for various immune cell infiltration estimates (TIMER, CIBERSORT, etc.) in LUAD/LUSC, showing a lack of significant prognostic value for infiltration density alone. (B-D) Correlation analysis supporting the functional role of specific T cell states.\u003c/p\u003e","description":"","filename":"Figure7.png","url":"https://assets-eu.researchsquare.com/files/rs-9010841/v1/9c5ef8682a44ac38a39962dd.png"},{"id":105984357,"identity":"8ce0b45b-d97b-4cf0-9311-0c7184a642a3","added_by":"auto","created_at":"2026-04-02 07:13:45","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":63933,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eProposed Mechanism.\u003c/strong\u003eSchematic diagram illustrating the \u003cstrong\u003eGSTM1-GSH-GPX4\u003c/strong\u003e axis. \u003cem\u003eGSTM1\u003c/em\u003eupregulation leads to increased GSH and GPX4 activity, reducing Lipid ROS and inhibiting ferroptosis, ultimately enhancing CD8+ T cell survival and anti-tumor immunity.\u003c/p\u003e","description":"","filename":"Figure8.png","url":"https://assets-eu.researchsquare.com/files/rs-9010841/v1/f03a4069d705dddb6b0ab629.png"},{"id":106095787,"identity":"76a68f5d-3e8e-48cd-b3d6-a891ce1f97c9","added_by":"auto","created_at":"2026-04-03 11:51:08","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2854942,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9010841/v1/715b0e11-1b58-4a56-81bd-7fa00609374f.pdf"},{"id":106093896,"identity":"789ad974-928c-443a-8789-2f532b9f7c19","added_by":"auto","created_at":"2026-04-03 11:39:54","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":5793,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTable 1.\u003c/strong\u003e Data sources and characteristics of the GWAS and sc-eQTL datasets used in this study.\u003c/p\u003e","description":"","filename":"Table1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-9010841/v1/65b88f382a160a74dab617d0.xlsx"},{"id":105984337,"identity":"395a5b20-3700-463f-8782-49e4e10a0c48","added_by":"auto","created_at":"2026-04-02 07:13:43","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":7175,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTable 2.\u003c/strong\u003e Summary of the 40 ferroptosis-related genes causally associated with NSCLC risk identified by MR.\u003c/p\u003e","description":"","filename":"Table2.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-9010841/v1/95e07a46d0b0b1dafeac78e5.xlsx"},{"id":105984354,"identity":"878a27de-dcb1-4fb5-8749-d1015063553b","added_by":"auto","created_at":"2026-04-02 07:13:45","extension":"xlsx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":5588,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTable 3.\u003c/strong\u003e Significant cell-type-specific causal associations (P \u0026lt; 0.05) identified between ferroptosis genes and NSCLC risk.\u003c/p\u003e","description":"","filename":"Table3.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-9010841/v1/fa2ef00252453db059af6ff6.xlsx"},{"id":105984339,"identity":"55c2f4b7-e6fb-4268-92ee-2d8d8c4d49d9","added_by":"auto","created_at":"2026-04-02 07:13:43","extension":"xlsx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":14317,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTable3.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-9010841/v1/d5f1d15bdce615143ef4cee0.xlsx"},{"id":106093519,"identity":"9cc54943-1637-4039-ad78-456bc978f34b","added_by":"auto","created_at":"2026-04-03 11:37:44","extension":"xlsx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":6368,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTable2.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-9010841/v1/0c1c3406da26edba8ee608d4.xlsx"},{"id":105984346,"identity":"cd1ecf7b-6bf2-49db-815c-5a8b9c6c8d24","added_by":"auto","created_at":"2026-04-02 07:13:44","extension":"xlsx","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":7178,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTable1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-9010841/v1/a9f209a35b3ef7a10870f7c1.xlsx"},{"id":105984351,"identity":"58600a18-022c-4745-9f27-1e0c3ba03150","added_by":"auto","created_at":"2026-04-02 07:13:44","extension":"pdf","order_by":7,"title":"","display":"","copyAsset":false,"role":"supplement","size":5293,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryFigure2.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9010841/v1/6c1126df2f06a5bb1893fc83.pdf"},{"id":105984372,"identity":"ba6fa444-461d-49f6-b053-cb0261b7230d","added_by":"auto","created_at":"2026-04-02 07:13:49","extension":"pdf","order_by":8,"title":"","display":"","copyAsset":false,"role":"supplement","size":10056,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryFigure4.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9010841/v1/1e0ed050eca4eebe50f7a31b.pdf"},{"id":105984369,"identity":"fe465714-31ff-447b-8a67-2fe0d8020030","added_by":"auto","created_at":"2026-04-02 07:13:48","extension":"pdf","order_by":9,"title":"","display":"","copyAsset":false,"role":"supplement","size":10928,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryFigure1.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9010841/v1/b783056d62c4082eef677577.pdf"},{"id":105984347,"identity":"6e4daa86-3a10-4617-8688-5e3b7e7a802e","added_by":"auto","created_at":"2026-04-02 07:13:44","extension":"pdf","order_by":10,"title":"","display":"","copyAsset":false,"role":"supplement","size":5384,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryFigure3.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9010841/v1/9705e2e802a9363097859b8f.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Single-Cell Genetic Architecture Identifies GSTM1 as a Novel Therapeutic Target for Reversing T-Cell Ferroptosis in NSCLC","fulltext":[{"header":"1. Background","content":"\u003cp\u003eNon-small cell lung cancer (NSCLC) remains the leading cause of cancer-related mortality globally. While immune checkpoint blockade (ICB) targeting PD-1/PD-L1 has revolutionized clinical care, primary and acquired resistance remains a major bottleneck, with only\u0026thinsp;~\u0026thinsp;20% of unselected patients deriving durable benefit[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. Emerging evidence suggests that the metabolic fitness of tumor-infiltrating lymphocytes (TILs) is a critical determinant of ICB efficacy. Specifically, ferroptosis\u0026mdash;an iron-dependent form of non-apoptotic cell death driven by lipid peroxidation\u0026mdash;has been identified as a \"metabolic checkpoint\" that limits the survival and cytotoxicity of CD8\u0026thinsp;+\u0026thinsp;T cells in the oxidative tumor microenvironment (TME) [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eSeminal work by Wang et al. (Nature, 2019) demonstrated that CD8\u0026thinsp;+\u0026thinsp;T cells suppress tumor growth by inducing ferroptosis in cancer cells; however, the T cells themselves are also vulnerable to ferroptosis, which can lead to metabolic exhaustion and immune failure[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. Therefore, identifying the specific genetic switches that regulate ferroptosis sensitivity within immune cells is crucial for developing next-generation immunotherapies.\u003c/p\u003e \u003cp\u003eCurrent understanding of ferroptosis regulation largely relies on bulk tissue transcriptomics or preclinical mouse models, which fail to capture the complex, cell-state-dependent genetic architecture of the human immune system. For instance, a gene that protects tumor cells might paradoxically impair T-cell function, or have divergent effects in naive versus effector states[\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eTo address this resolution gap, we performed a systematic genetic dissection of ferroptosis in NSCLC. Moving beyond traditional association studies, we integrated single-cell expression quantitative trait loci (sc-eQTL) from 14 immune cell lineages (OneK1K cohort) with large-scale genomic data from the FinnGen study (N\u0026thinsp;=\u0026thinsp;385,195). Our goal was not merely to catalog genes, but to identify lineage-specific therapeutic targets. Here, we report that GSTM1 and MAPK3 act as cell-state-dependent regulators of ferroptosis, providing genetic evidence for a \"Quality over Quantity\" paradigm in anti-tumor immunity.\u003c/p\u003e"},{"header":"2. Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Study Design and Data Sources\u003c/h2\u003e \u003cp\u003eThe overall study design is illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. We implemented a three-stage integrative analysis: (1) identification of causal genes using Two-Sample Mendelian Randomization (MR); (2) fine-mapping of causal variants via Bayesian colocalization; and (3) clinical validation using single-cell tumor transcriptomics. Detailed information on all data sources is provided in \u003cb\u003eTable\u0026nbsp;1\u003c/b\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Ferroptosis-Related Gene Selection\u003c/h2\u003e \u003cp\u003eA comprehensive panel of 483 ferroptosis-related genes was curated from FerrDb [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e], a manually annotated database comprising drivers, suppressors, and markers of ferroptosis.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Single-Cell eQTL and GWAS Datasets\u003c/h2\u003e \u003cp\u003eCell-type-specific instrumental variables were derived from the \u003cb\u003eOneK1K cohort\u003c/b\u003e [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e], which profiled sc-eQTLs across 14 immune cell types (including CD4\u0026thinsp;+\u0026thinsp;subsets, CD8\u0026thinsp;+\u0026thinsp;subsets, and NK cells) from 982 donors. Genetic associations for NSCLC were retrieved from the \u003cb\u003eFinnGen R12 release\u003c/b\u003e [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e], comprising 2,564 cases and 382,631 controls (\u003cb\u003eTable\u0026nbsp;1\u003c/b\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.4 Mendelian Randomization Analysis\u003c/h2\u003e \u003cp\u003eFor each gene-cell pair, independent significant cis-eQTLs (P\u0026thinsp;\u0026lt;\u0026thinsp;5 \u0026times; 10^-8, r^2\u0026thinsp;\u0026lt;\u0026thinsp;0.001) were selected as instrumental variables. \u003cb\u003eThe distributions of SNP counts, P-values, and effect sizes for these instrumental variables are detailed in Supplementary Figs.\u0026nbsp;2\u0026ndash;4, verifying the robustness of our selection process.\u003c/b\u003e The Inverse Variance Weighted (IVW) method was used as the primary test[\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. Sensitivity analyses, including MR-Egger and Weighted Median, were performed to assess horizontal pleiotropy (\u003cb\u003eSupplementary Table\u0026nbsp;3\u003c/b\u003e). Correction for multiple testing was applied using the Benjamini-Hochberg method.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.5 Bayesian Colocalization\u003c/h2\u003e \u003cp\u003eTo distinguish valid causal instruments from linkage disequilibrium (LD) artifacts, we performed colocalization analysis using the coloc R package [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. We tested five hypotheses (H0 to H4), where a posterior probability of H4 (PP.H4\u0026thinsp;\u0026gt;\u0026thinsp;0.8) indicates a shared causal variant between the sc-eQTL and GWAS signals.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e2.6 Single-Cell Validation (TISCH2)\u003c/h2\u003e \u003cp\u003eThe expression landscape of prioritized genes was validated using the TISCH2 database[\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. We analyzed the NSCLC_GSE131907 dataset to compare gene expression levels between tumor-infiltrating lymphocytes (TILs) and peripheral blood cells.\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Results","content":"\u003cp\u003e \u003cb\u003e3.1 Systematic Identification of Ferroptosis Drivers in NSCLC\u003c/b\u003e We first established the baseline genetic landscape by screening 483 ferroptosis-related genes using summary-level data from the FinnGen R12 cohort (N\u0026thinsp;=\u0026thinsp;385,195 individuals, \u003cb\u003eTable\u0026nbsp;1\u003c/b\u003e). As detailed in \u003cb\u003eTable\u0026nbsp;2\u003c/b\u003e, this systematic screen identified 40 genes with significant causal associations with NSCLC risk (P\u0026thinsp;\u0026lt;\u0026thinsp;0.05).\u003c/p\u003e \u003cp\u003eAmong the protective factors shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, \u003cb\u003eMAPK3\u003c/b\u003e exhibited the most robust evidence (OR\u0026thinsp;=\u0026thinsp;0.910, 95% CI: 0.876\u0026ndash;0.946, P\u0026thinsp;=\u0026thinsp;1.58 \u0026times; 10^-6), followed by \u003cb\u003eMAP3K11\u003c/b\u003e (OR\u0026thinsp;=\u0026thinsp;0.913, P\u0026thinsp;=\u0026thinsp;3.37 \u0026times; 10^-5) and \u003cb\u003eTMBIM4\u003c/b\u003e (OR\u0026thinsp;=\u0026thinsp;0.864, P\u0026thinsp;=\u0026thinsp;2.80 \u0026times; 10^-4). On the risk side, \u003cb\u003eHDDC3\u003c/b\u003e was identified as the strongest risk factor (OR\u0026thinsp;=\u0026thinsp;1.267, 95% CI: 1.123\u0026ndash;1.430, P\u0026thinsp;=\u0026thinsp;1.27 \u0026times; 10^-4), along with \u003cb\u003eCHP1\u003c/b\u003e (OR\u0026thinsp;=\u0026thinsp;1.233, P\u0026thinsp;=\u0026thinsp;1.10 \u0026times; 10^-3). \u003cb\u003eSupplementary Table\u0026nbsp;1\u003c/b\u003e provides the complete summary statistics for all screened genes. Importantly, sensitivity analyses using MR-Egger and Weighted Median methods showed consistent effect directions (\u003cb\u003eSupplementary Table\u0026nbsp;3\u003c/b\u003e), ruling out significant horizontal pleiotropy.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003e3.2 Cell-Type Specificity of Genetic Regulation\u003c/b\u003e While whole-blood analysis provided a broad overview, integrating single-cell eQTL data (OneK1K, N\u0026thinsp;=\u0026thinsp;982 donors) revealed that these genetic effects are highly cell-type specific. \u003cb\u003eTable\u0026nbsp;3\u003c/b\u003e lists the 8 significant gene-cell type pairs identified after strict filtering.\u003c/p\u003e \u003cp\u003eThe heatmap in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e illustrates this heterogeneity. For instance, while \u003cb\u003eGSTM1\u003c/b\u003e showed a moderate protective effect in whole blood (OR\u0026thinsp;=\u0026thinsp;0.949, \u003cb\u003eTable\u0026nbsp;2\u003c/b\u003e), single-cell analysis pinpointed its primary site of action to T cells. Specifically, \u003cb\u003eGSTM1\u003c/b\u003e expression was significantly protective in \u003cb\u003eCD8\u0026thinsp;+\u0026thinsp;Naive T cells\u003c/b\u003e (OR\u0026thinsp;=\u0026thinsp;0.595, P\u0026thinsp;=\u0026thinsp;2.17 \u0026times; 10^-4) and \u003cb\u003eCD4\u0026thinsp;+\u0026thinsp;Naive T cells\u003c/b\u003e (OR\u0026thinsp;=\u0026thinsp;0.764, P\u0026thinsp;=\u0026thinsp;7.70 \u0026times; 10^-4), but showed no significant effect in B cells or Monocytes (\u003cb\u003eTable\u0026nbsp;3\u003c/b\u003e). Similarly, \u003cb\u003eHRAS\u003c/b\u003e was identified as a protective factor specifically in \u003cb\u003eCD8\u0026thinsp;+\u0026thinsp;Effector T cells\u003c/b\u003e (OR\u0026thinsp;=\u0026thinsp;0.814, P\u0026thinsp;=\u0026thinsp;0.017), a granularity that was masked in the bulk analysis.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003e3.3 Deep Dive into GSTM1: A T-Cell Guardian\u003c/b\u003e We focused on \u003cb\u003eGSTM1\u003c/b\u003e due to its strong effect size and lineage specificity. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, the protective effect of \u003cem\u003eGSTM1\u003c/em\u003e forms a gradient across T cell differentiation states. The strongest protection was observed in the naive CD8\u0026thinsp;+\u0026thinsp;compartment (OR\u0026thinsp;=\u0026thinsp;0.595), which slightly attenuated in the effector compartment (OR\u0026thinsp;=\u0026thinsp;0.755, P\u0026thinsp;=\u0026thinsp;0.045, \u003cb\u003eTable\u0026nbsp;3\u003c/b\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo validate the genetic architecture, we performed Bayesian colocalization. Figure\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA shows that the probability of a shared causal variant (PP.H4) varies across lineages, peaking in CD8\u0026thinsp;+\u0026thinsp;Effector Memory T cells (PP.H4\u0026thinsp;=\u0026thinsp;0.12). The regional association plot (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eB) and the SNP distribution data in \u003cb\u003eSupplementary Fig.\u0026nbsp;2\u003c/b\u003e confirm that the instruments used for \u003cem\u003eGSTM1\u003c/em\u003e are robust (Median N SNPs\u0026thinsp;=\u0026thinsp;17) and distinct from potential confounders.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003e3.4 MAPK3 Exhibits Antagonistic Pleiotropy\u003c/b\u003e A key finding of this study is the cell-state-dependent antagonism of \u003cb\u003eMAPK3\u003c/b\u003e. As detailed in \u003cb\u003eTable\u0026nbsp;3\u003c/b\u003e and visualized in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e, \u003cb\u003eMAPK3\u003c/b\u003e expression confers significant protection in naive lymphocytes:\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eNaive CD4\u0026thinsp;+\u0026thinsp;T cells\u003c/b\u003e: OR\u0026thinsp;=\u0026thinsp;0.667 (95% CI: 0.537\u0026ndash;0.829, P\u0026thinsp;=\u0026thinsp;2.56 \u0026times; 10^-4)\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eNaive B cells\u003c/b\u003e: OR\u0026thinsp;=\u0026thinsp;0.677 (95% CI: 0.544\u0026ndash;0.844, P\u0026thinsp;=\u0026thinsp;5.29 \u0026times; 10^-4)\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003eHowever, this effect is strikingly reversed in the effector state. In \u003cb\u003eEffector CD4\u0026thinsp;+\u0026thinsp;T cells\u003c/b\u003e, higher \u003cem\u003eMAPK3\u003c/em\u003e expression is associated with increased lung cancer risk (OR\u0026thinsp;=\u0026thinsp;1.388, 95% CI: 1.089\u0026ndash;1.769, P\u0026thinsp;=\u0026thinsp;8.11 \u0026times; 10^-3). This bidirectional effect explains why the whole-blood signal (OR\u0026thinsp;=\u0026thinsp;0.910, \u003cb\u003eTable\u0026nbsp;2\u003c/b\u003e) appeared weaker than the specific naive cell signals\u0026mdash;the bulk tissue average likely diluted the opposing effects from different cell subsets.\u003c/p\u003e \u003cp\u003e \u003cb\u003e3.5 Clinical Validation and Mechanism\u003c/b\u003e To bridge these genetic findings with clinical reality, we analyzed tumor transcriptomics (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e). Consistent with the genetic \"Quality over Quantity\" hypothesis, \u003cem\u003eGSTM1\u003c/em\u003e expression in the tumor microenvironment correlated positively with functional T-cell signatures rather than simple infiltration density. Based on these multi-omic lines of evidence, we propose the mechanism illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e: \u003cem\u003eGSTM1\u003c/em\u003e upregulation enhances glutathione (GSH) production and GPX4 activity, thereby shielding CD8\u0026thinsp;+\u0026thinsp;T cells from lipid-ROS-induced ferroptosis and preserving their anti-tumor efficacy.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eIn this study, we leveraged single-cell genetics to dismantle the complexity of ferroptosis regulation in lung cancer. Our central finding is the identification of \u003cb\u003eGSTM1\u003c/b\u003e and \u003cb\u003eMAPK3\u003c/b\u003e as potent, lineage-dependent regulators. Unlike traditional bulk-tissue analyses, our sc-eQTL approach revealed that the genetic control of ferroptosis varies not just between lineages but also between differentiation states.\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e4.1 GSTM1: A Metabolic Checkpoint for T-Cell Fitness\u003c/h2\u003e \u003cp\u003eOur results align with the emerging paradigm establishing ferroptosis as a metabolic vulnerability in CD8\u0026thinsp;+\u0026thinsp;T cells. Wang et al. previously demonstrated that maintaining redox homeostasis is essential for CD8\u0026thinsp;+\u0026thinsp;T cell anti-tumor function[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. Extending this, we identified \u003cb\u003eGSTM1\u003c/b\u003e as a critical \"metabolic shield.\" \u003cb\u003eGSTM1\u003c/b\u003e encodes a cytosolic enzyme that detoxifies lipid peroxidation products. Our single-cell validation (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e) indicates that \u003cb\u003eGSTM1\u003c/b\u003e expression is positively correlated with T-cell cytotoxicity scores but not necessarily with the absolute number of infiltrating cells. This supports a \u003cb\u003e\"Quality over Quantity\"\u003c/b\u003e model: high \u003cb\u003eGSTM1\u003c/b\u003e levels may not recruit more T cells, but ensure that infiltrating T cells survive the lipid-ROS-rich tumor microenvironment, preventing them from undergoing ferroptosis and sustaining their killing capacity [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e4.2 The \"Decoupling\" Phenomenon: Unmasking Gene-Environment Interactions\u003c/h2\u003e \u003cp\u003eA key observation in our study was the \"decoupling\" between the strong causal effect of \u003cb\u003eGSTM1\u003c/b\u003e in MR analysis (OR\u0026thinsp;=\u0026thinsp;0.60) and the moderate colocalization probability (PP.H4\u0026thinsp;=\u0026thinsp;0.12) at the GWAS locus. Rather than a technical artifact, we propose this reflects a profound \u003cb\u003eGene-Environment Interaction (GxE)\u003c/b\u003e. As demonstrated by Goto et al., the prognostic impact of the \u003cb\u003eGSTM1\u003c/b\u003e null genotype is latent and profoundly stratified by environmental triggers\u0026mdash;specifically, smoking status[\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. In non-smokers, the lack of \u003cb\u003eGSTM1\u003c/b\u003e may be tolerated; however, under the oxidative stress of cigarette smoke, the defect becomes lethal. Since standard GWAS cohorts mix smokers and non-smokers, the \"environmental noise\" dilutes the genetic signal, leading to lower colocalization scores. Our sc-eQTL MR, by isolating the specific genetic effect on gene expression, successfully unmasked this causal link.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e4.3 Antagonistic Pleiotropy of MAPK3\u003c/h2\u003e \u003cp\u003eThe cell-state-dependent antagonism of \u003cb\u003eMAPK3\u003c/b\u003e is a novel finding. We observed that \u003cb\u003eMAPK3\u003c/b\u003e expression is protective in naive T cells but risk-promoting in effector T cells. This dichotomy can be explained by the dual role of MAPK signaling. In the naive state, robust MAPK signaling is required for initial antigen priming and activation[\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. However, in the effector phase, chronic hyperactivation of the MAPK pathway is a hallmark of \u003cb\u003eT-cell exhaustion\u003c/b\u003e [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. Xu et al. recently reported that excessive lipid uptake promotes exhaustion via metabolic stress[\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. Our data suggests \u003cb\u003eMAPK3\u003c/b\u003e follows a \"Goldilocks\" principle: beneficial for priming naive cells, but detrimental if overactive in effector cells, likely by accelerating exhaustion-associated ferroptosis. This insight challenges the use of broad-spectrum MAPK inhibitors, which might inadvertently impair naive T-cell priming.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003e4.4 Clinical Implications : From Genetics to Therapy\u003c/h2\u003e \u003cp\u003eOur findings have immediate translational relevance. First, GSTM1 status could serve as a precision biomarker. Ritambhara et al. previously linked GSTM1 polymorphisms to chemotherapy toxicity[\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. ; our data extends this to immunotherapy. We propose that NSCLC patients with GSTM1-null genotypes or low intratumoral GSTM1 expression may possess T cells that are hypersensitive to ferroptosis. Second, this \"metabolic defect\" is potentially druggable. Since GSTM1 functions by replenishing glutathione (GSH) to fuel GPX4 activity, patients with low GSTM1 might benefit from combinatorial therapies that include ferroptosis inhibitors (e.g., liproxstatin-1) or clinically available antioxidants such as N-acetylcysteine (NAC). By artificially boosting the antioxidative capacity of T cells, we may be able to rescue the \"ferroptosis-prone\" phenotype caused by GSTM1 deficiency, thereby sensitizing resistant tumors to PD-1 blockade. This hypothesis aligns with recent preclinical success in manipulating ferroptosis to enhance ICB efficacy [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e4.5 Limitations\u003c/h2\u003e \u003cp\u003eOur study has limitations. First, the GWAS data (FinnGen) is predominantly of European ancestry. Given that \u003cb\u003eGSTM1\u003c/b\u003e deletion frequencies vary significantly between Asian and European populations[\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e], validation in Asian cohorts is warranted. Second, while we utilized large-scale single-cell datasets, direct experimental validation (e.g., \u003cb\u003eGSTM1\u003c/b\u003e knockdown in T-cell co-culture models) would further strengthen the mechanistic conclusions.\u003c/p\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u0026nbsp;\u003c/strong\u003eThe data supporting the findings of this study are publicly available.\u003c/p\u003e\n\u003cp\u003e1.NSCLC GWAS summary statistics are available from the FinnGen study (Release R12) at https://www.finngen.fi/en/access_results.\u003c/p\u003e\n\u003cp\u003e2.Single-cell eQTL summary statistics were obtained from the OneK1K cohort at https://onek1k.org/.\u003c/p\u003e\n\u003cp\u003e4.Ferroptosis-related genes were retrieved from the FerrDb V2 database at http://www.zhounan.org/ferrdb/.\u003c/p\u003e\n\u003cp\u003e5.Single-cell tumor transcriptomic data for validation are accessible via the TISCH2 database at http://tisch.comp-genomics.org/. All other data generated or analyzed during this study are included in this published article and its supplementary information files.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflict of Interest Disclosures\u003c/strong\u003e The authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e This study is a secondary analysis of publicly available, de-identified summary-level data. The original studies (including FinnGen and OneK1K) obtained relevant ethical approvals and informed consent from all participants. According to institutional policies, further specific ethical approval for this meta-analysis of summary statistics was not required.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u0026nbsp;\u003c/strong\u003eThis work was supported by:\u003c/p\u003e\n\u003cp\u003eThe Joint Basic Research Program of Yunnan Provincial Department of Science and Technology and Kunming Medical University (Grant No. 202401AY07001-369)\u003c/p\u003e\n\u003cp\u003eThe Yunnan Provincial Innovative Research Team for Thoracic Tumor Prevention and Control (Grant No. 202405AS350015)\u003c/p\u003e\n\u003cp\u003eThe Development of a Precision Prevention and Full-Cycle Intelligent Management System for Regionally Prevalent Lung Cancer in Yunnan (Grant No. 202303AC100203)\u003c/p\u003e\n\u003cp\u003eThe Expert Workstation of Dr. Li Yin (Grant No. 202405AF140055)\u003c/p\u003e\n\u003cp\u003eRole of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contributions Dr. Xiao\u003c/strong\u003e had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConcept and design:\u0026nbsp;\u003c/strong\u003eXiao, Ye.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcquisition, analysis, or interpretation of data:\u003c/strong\u003e Wu, Shao, Chao, Huang, Li.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDrafting of the manuscript:\u003c/strong\u003e Xiao.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCritical revision of the manuscript for important intellectual content:\u0026nbsp;\u003c/strong\u003eAll authors.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStatistical and bioinformatics analysis:\u0026nbsp;\u003c/strong\u003eXiao, Shao.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eObtained funding:\u0026nbsp;\u003c/strong\u003eYe.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAdministrative, technical, or material support:\u0026nbsp;\u003c/strong\u003eWu, Chao, Huang, Li.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSupervision:\u003c/strong\u003e Ye.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u0026nbsp;\u003c/strong\u003eWe gratefully acknowledge the participants and investigators of the FinnGen study, the OneK1K cohort, and the eQTLGen Consortium for sharing the GWAS and sc-eQTL summary statistics used in this research. We also thank the developers and maintainers of FerrDb and TISCH2 for providing valuable resources for ferroptosis and single-cell research.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eArtificial Intelligence (AI) Use Statement\u003c/strong\u003e The authors declare that no artificial intelligence tools were used in the generation of study data, statistical analyses, or scientific interpretation. Language editing support was limited to standard grammar and clarity checks, with full responsibility for the content retained by the authors.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eSiegel RL, Giaquinto AN, Jemal A. Cancer statistics, 2024. CA Cancer J Clin. 2024;74:12\u0026ndash;49.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen DS, Mellman I. Elements of cancer immunity and the cancer-immune set point. Nature. 2017;541:321\u0026ndash;30.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDixon SJ, Lemberg KM, Lamprecht MR, Skouta R, Zaitsev EM, Gleason CE, Patel DN, Bauer AJ, Cantley AM, Yang WS, et al. Ferroptosis: an iron-dependent form of nonapoptotic cell death. Cell. 2012;149:1060\u0026ndash;72.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang W, Green M, Choi JE, Gijon M, Kennedy PD, Johnson JK, Liao P, Lang X, Kryczek I, Sell A, et al. CD8(+) T cells regulate tumour ferroptosis during cancer immunotherapy. Nature. 2019;569:270\u0026ndash;4.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTang R, Xu J, Zhang B, Liu J, Liang C, Hua J, Meng Q, Yu X, Shi S. Ferroptosis, necroptosis, and pyroptosis in anticancer immunity. J Hematol Oncol. 2020;13:110.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHong Y. Single-Cell Transcriptome-Wide Mendelian Randomization and Colocalization Analyses Uncover Cell-Specific Mechanisms in Epigenetic Age Acceleration. FASEB J. 2025;39:e71310.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhou N, Yuan X, Du Q, Zhang Z, Shi X, Bao J, Ning Y, Peng L. FerrDb V2: update of the manually curated database of ferroptosis regulators and ferroptosis-disease associations. Nucleic Acids Res. 2023;51:D571\u0026ndash;82.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYazar S, Alquicira-Hernandez J, Wing K, Senabouth A, Gordon MG, Andersen S, Lu Q, Rowson A, Taylor TRP, Clarke L, et al. Single-cell eQTL mapping identifies cell type-specific genetic control of autoimmune disease. Science. 2022;376:eabf3041.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKurki MI, Karjalainen J, Palta P, Sipila TP, Kristiansson K, Donner KM, Reeve MP, Laivuori H, Aavikko M, Kaunisto MA, et al. FinnGen provides genetic insights from a well-phenotyped isolated population. Nature. 2023;613:508\u0026ndash;18.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHemani G, Zheng J, Elsworth B, Wade KH, Haberland V, Baird D, Laurin C, Burgess S, Bowden J, Langdon R et al. The MR-Base platform supports systematic causal inference across the human phenome. Elife 2018, 7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGiambartolomei C, Vukcevic D, Schadt EE, Franke L, Hingorani AD, Wallace C, Plagnol V. Bayesian test for colocalisation between pairs of genetic association studies using summary statistics. PLoS Genet. 2014;10:e1004383.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHan Y, Wang Y, Dong X, Sun D, Liu Z, Yue J, Wang H, Li T, Wang C. TISCH2: expanded datasets and new tools for single-cell transcriptome analyses of the tumor microenvironment. Nucleic Acids Res. 2022;51:D1425\u0026ndash;31.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBoard PG, Menon D. Glutathione transferases, regulators of cellular metabolism and physiology. Biochim Biophys Acta. 2013;1830:3267\u0026ndash;88.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGoto I, Yoneda S, Yamamoto M, Kawajiri K. Prognostic significance of germ line polymorphisms of the CYP1A1 and glutathione S-transferase genes in patients with non-small cell lung cancer. Cancer Res. 1996;56:3725\u0026ndash;30.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJiang X, Stockwell BR, Conrad M. Ferroptosis: mechanisms, biology and role in disease. Nat Rev Mol Cell Biol. 2021;22:266\u0026ndash;82.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWherry EJ, Kurachi M. Molecular and cellular insights into T cell exhaustion. Nat Rev Immunol. 2015;15:486\u0026ndash;99.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu S, Chaudhary O, Rodriguez-Morales P, Sun X, Chen D, Zappasodi R, Xu Z, Pinto AFM, Williams A, Schulze I, et al. Uptake of oxidized lipids by the scavenger receptor CD36 promotes lipid peroxidation and dysfunction in CD8(+) T cells in tumors. Immunity. 2021;54:1561\u0026ndash;e15771567.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRitambhara TS, Vijayaraghavalu S, Siddiqui MA, Al-Khedhairy AA, Kumar M. Clinical response of carboplatin-based chemotherapy and its association to genetic polymorphism in lung cancer patients from North India - A clinical pharmacogenomics study. J Cancer Res Ther. 2022;18:109\u0026ndash;18.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhou J, Zhang H, Pan WW, Peng C, Tang H, Yuan G, Peng F. Isocucurbitacin B targets STAT3 to induce ferroptosis and promote anti-PD1 immunotherapy responses in breast cancer. Int J Surg 2026.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen J, Yang H, Teo ASM, Amer LB, Sherbaf FG, Tan CQ, Alvarez JJS, Lu B, Lim JQ, Takano A, et al. Genomic landscape of lung adenocarcinoma in East Asians. Nat Genet. 2020;52:177\u0026ndash;86.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHayes JD, Strange RC. Glutathione S-transferase polymorphisms and their biological consequences. Pharmacology. 2000;61:154\u0026ndash;66.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"},{"header":"Tables","content":"\n\u003cp\u003eTables 1 to 3 are available in the Supplementary Files section.\u003c/p\u003e\n"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"discover-oncology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"dion","sideBox":"Learn more about [Discover Oncology](https://www.springer.com/12672)","snPcode":"","submissionUrl":"","title":"Discover Oncology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Discover Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Non-small cell lung cancer, Ferroptosis, Mendelian randomization, Single-cell eQTL, CD8 + T cells, GSTM1","lastPublishedDoi":"10.21203/rs.3.rs-9010841/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9010841/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eFerroptosis has emerged as a pivotal mechanism in cancer surveillance, particularly in determining the efficacy of immunotherapy. However, the precise genetic switches that regulate ferroptosis sensitivity within specific immune cell lineages remain poorly characterized due to the resolution limits of bulk tissue analysis.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eTo address this, we constructed a high-resolution genetic map of ferroptosis in non-small cell lung cancer (NSCLC). We applied a multi-stage integrative framework, harmonizing single-cell expression quantitative trait loci (sc-eQTL) from 14 immune cell populations with large-scale genomic data from the FinnGen cohort (N\u0026thinsp;=\u0026thinsp;385,195).\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eOur analysis identified 40 putative causal genes driving NSCLC susceptibility. A key finding is that GSTM1 expression in CD8\u0026thinsp;+\u0026thinsp;naive T cells confers a robust protective effect (OR\u0026thinsp;=\u0026thinsp;0.60, P\u0026thinsp;=\u0026thinsp;2.17 x 10^-4). Multi-omic validation indicates that GSTM1 functions as a \"metabolic shield,\" enhancing T-cell functional quality rather than simple infiltration density. Furthermore, we reveal a striking cell-state-dependent antagonism for MAPK3, which acts as a protective factor in the naive state (OR\u0026thinsp;=\u0026thinsp;0.67) but switches to a risk-promoting factor in effector CD4\u0026thinsp;+\u0026thinsp;T cells (OR\u0026thinsp;=\u0026thinsp;1.39).\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e \u003cp\u003eThis study uncovers a lineage-specific genetic architecture of ferroptosis, highlighting GSTM1 and MAPK3 as critical regulators of T-cell metabolic fitness. These findings suggest that stratifying patients based on GSTM1 status could serve as a novel strategy to guide precision immunotherapy.\u003c/p\u003e","manuscriptTitle":"Single-Cell Genetic Architecture Identifies GSTM1 as a Novel Therapeutic Target for Reversing T-Cell Ferroptosis in NSCLC","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-02 07:13:32","doi":"10.21203/rs.3.rs-9010841/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-04-30T18:43:09+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-28T13:49:23+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"193121163575182526902707899605008295892","date":"2026-04-16T12:35:02+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-13T17:05:12+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"66460137332111145089507433393899471620","date":"2026-04-05T05:36:58+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"34124732458288998544688644894951232520","date":"2026-04-04T17:11:08+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"333804333376006420128928885260593838630","date":"2026-03-30T14:41:58+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-03-30T13:41:30+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-03-10T16:56:02+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-03-04T05:40:55+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-03-04T05:37:39+00:00","index":"","fulltext":""},{"type":"submitted","content":"Discover Oncology","date":"2026-03-02T13:29:52+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"discover-oncology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"dion","sideBox":"Learn more about [Discover Oncology](https://www.springer.com/12672)","snPcode":"","submissionUrl":"","title":"Discover Oncology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Discover Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"5a0146c6-50ce-4c00-8f0c-d3ebb4bfe8ee","owner":[],"postedDate":"April 2nd, 2026","published":true,"recentEditorialEvents":[{"type":"decision","content":"Revision requested","date":"2026-04-30T18:43:09+00:00","index":"","fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-05-18T07:53:21+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-02 07:13:32","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9010841","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9010841","identity":"rs-9010841","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}