Interferon signaling and ferroptosis in tumor immunology and therapy

The study investigates how activation of the interferon (IFN) signaling pathway influences tumor cell ferroptosis and its relationship with CD8⁺ T cell exhaustion within the tumor microenvironment. Interferons regulate proliferation, apoptosis, immune responses, and antiviral defense, and are pivotal for CD8⁺ T cell activation and function in cancer therapy. Ferroptosis, an iron-dependent lipid peroxidation-driven cell death, has emerged as a promising anti-cancer strategy given iron’s essential role in tumor metabolism and survival. Persistent antigenic stimulation in tumors leads to CD8⁺ T cell exhaustion, diminishing anti-tumor immunity. The authors hypothesize that interferon signaling modulates ferroptosis in tumor cells and promotes CD8⁺ T cell exhaustion via specific molecular mediators, thereby impacting patient prognosis and responses to immunotherapy. The purpose is to delineate key genes and downstream pathways connecting IFN signaling, T cell exhaustion, and ferroptosis to inform novel therapeutic targets and optimization of immunotherapy.

Background literature highlights: (1) IFN signaling has dual roles in antitumor immunity, affecting proliferation, apoptosis, and immune regulation; IFN-γ is central to anti-tumor responses and can induce ferroptosis by altering tumor antioxidant and iron metabolism pathways. (2) Ferroptosis is implicated as a therapeutic approach across cancers, with CD8⁺ T cells and fatty acid metabolism (e.g., ACSL4) orchestrating tumor ferroptosis and immunity. (3) Chronic type I IFN (IFN-I) signaling can drive terminal CD8⁺ T cell exhaustion and reduce efficacy of PD-1 blockade; transcriptional regulators (e.g., IRF2) mediate IFN-driven exhaustion. (4) The HSP70 family, including HSPA6, is associated with stress responses, cancer progression, and immune modulation; HSP70/TLR2 activation may suppress CD8⁺ T cell responses. These studies provide context for examining how IFN-α/γ, CD8⁺ T cell functional states, and ferroptotic pathways intersect in cancer.

- Study design: Integrated single-cell transcriptomics, bioinformatics (pan-cancer TCGA/UCSC Xena and FerrDb), in vitro functional assays, and in vivo mouse xenograft models to dissect the IFN–CD8⁺ T cell exhaustion–ferroptosis axis. - Single-cell CD8⁺ T cell profiling: CD8⁺ T cells were isolated from tumors of nine cancer types using a mouse in situ transplantation model. NOG mice received intratumoral injections of human cancer cell lines (TE354.T basal cell carcinoma, BJAB B-cell lymphoma, TE-1 esophageal, KMS-11 multiple myeloma, OVCAR-3 ovarian, PANC-1 pancreatic, A-498 renal, ACT-1 thyroid, KI.F. endometrial) and IV injections of normal human CD8⁺ T cells. After 30 days, tumor and adjacent (normal) tissues and CD8⁺ T cells were collected. CD8⁺ T cells were processed on the 10x Genomics Chromium platform; libraries sequenced on Illumina HiSeq4000 PE125. Data processed with Cell Ranger (v2.2.0), STAR alignment, Seurat QC (nFeature_RNA > 500; nCount_RNA > 1000 and < 20000; percent.mt < 25), PCA/t-SNE for clustering. PCs with p < 0.05 were used for clustering. - Identification of exhausted and IFN-activated CD8⁺ T cells: Exhaustion markers PDCD1, CXCL13, LAYN used to label exhausted subsets. IFN-activated cells labeled by STAT1, IFIT1, ISG15, CCR1 expression. Pseudotime/trajectory analyses assessed temporal relationships between IFN activation and exhaustion. - Differential expression and enrichment: Merged expression matrices from six cancer types’ normal vs tumor CD8⁺ T cells to identify DEGs (|log2FC| > 1, p < 0.05). Eight DEGs identified: HSPA6, CXCL13, HSPA1A, HSPA1B, HSPB1, BAG3, IGLV3-1, DNAJB1. GO/KEGG via ClusterProfiler identified enrichment in stress-related processes and pathways including protein processing in ER, MAPK signaling, estrogen signaling. GEPIA2 used for pan-cancer expression and co-expression with exhaustion signatures, highlighting HSPA6. - Downstream factor prediction and interaction validation: GeneMANIA and Coexpedia predicted HSPA6 interactors (DNAJB1/DNAJB4 etc.), nominating DNAJB1 as a key downstream factor. Co-immunoprecipitation (Co-IP) using HSPA6 antibody confirmed HSPA6–DNAJB1 interaction. - Pan-cancer IRG/ferroptosis analyses: FerrDb-derived ferroptosis-related interferon genes (IRGs; 14 genes). TCGA pan-cancer expression (n ≈ 10,327 tumors; 730 normals) via UCSC Xena analyzed for differential expression of IFNA1, IFNA5, IFNA21, IFNG, etc., correlations among IRGs, and survival associations (survival package in R). GEPIA validated selected survival findings (e.g., THYM). - In vitro CD8⁺ T cell exhaustion model: Human CD8⁺ T cells (purity >95–98%) cultured 8 days with IL-2 and Dynabeads CD3/CD28; recombinant human IFN-α added throughout to induce exhaustion. Lentiviral transduction generated sh-HSPA6, sh-DNAJB1, or sh-NC CD8⁺ T cells (MOI 10; puromycin selection). Flow cytometry assessed exhaustion markers (PD-1, TIM3, LAG-3, CTLA-4), activation/proliferation (Ki67, CD69), and cytokines (IFN-γ, IL-2, TNF-α). RT-qPCR and Western blot measured HSPA6 and DNAJB1 expression. ELISA quantified IFN-γ. - Co-culture and ferroptosis assays: Transwell co-culture (0.4 µm) of CD8⁺ T cells with nine tumor cell lines (HT-29, HuH-7, A-549, HeLa, MCF-7, MKN-45, SK-MEL-1, etc.). IFN-α-stimulated CD8⁺ T cells assessed for effects on tumor cell viability (MTT), apoptosis (flow cytometry), ferroptosis-related mRNAs (COX2, ACSL4, NOX1, GPX4 by RT-qPCR), and lipid peroxidation products MDA and LPO (ELISA). Separate in vitro stimulation of tumor cells with IFN-γ tested direct effects; anti-IFN-γ neutralizing antibody vs IgG control assessed specificity. - In vivo xenograft and PD-1 blockade: NOG mice (n=6/group) received subcutaneous SK-MEL-1 injections (4×10⁶ cells). Groups: control (normal CD8⁺ T cells + IgG), anti-PD-1 alone, sh-NC+anti-PD-1, sh-HSPA6+anti-PD-1; additional reporting includes sh-DNAJB1 groups. Anti-PD-1 (2 mg/kg IP twice weekly). Tumor volume tracked for 30 days; endpoint tumor weight recorded. Tumor-infiltrating CD8⁺ T cell function by flow cytometry; tumor IFN-γ (ELISA); ferroptosis-related gene expression (RT-qPCR) and MDA/LPO (ELISA). - Statistics: Data as mean ± SD. Two-group comparisons by t-test; multi-group by one-way ANOVA; repeated measures ANOVA for longitudinal tumor volumes. Significance at p < 0.05, p < 0.01, p < 0.001. - Data availability: scRNA-seq data deposited in SRA (BioProject PRJNA1140668; listed SRR accessions).

- CD8⁺ T cell exhaustion is widespread in tumors: Single-cell analysis across nine cancer types showed increased diversity of CD8⁺ T cell subsets in tumors vs normal tissues and significant expansion of clusters expressing exhaustion markers PDCD1, CXCL13, LAYN in multiple tumor types (e.g., BCL, ESCA, MM, OV, PACA, RC, UCEC). - IFN activation correlates with exhaustion: Tumor samples had increased clusters expressing STAT1, IFIT1, ISG15, CCR1 (interferon-activated CD8⁺ T cells). Trajectory analyses in ESCA, MM, OV, PACA, RC, UCEC indicated exhausted CD8⁺ T cells arise coincident with or following IFN-activated states, supporting IFN-driven exhaustion in tumors. - HSPA6 identified as a key exhaustion mediator: Among eight DEGs between normal and tumor CD8⁺ T cells, HSPA6, HSPA1A, HSPA1B, HSPB1 were enriched in MAPK signaling. GEPIA2 pan-cancer analyses showed upregulation of these genes and positive correlation with T cell exhaustion signatures, with HSPA6 showing the strongest association. - IFN-α induces CD8⁺ T cell exhaustion in vitro: IFN-α increased PD-1, TIM-3, LAG-3, CTLA-4 and reduced Ki67, CD69, IFN-γ, IL-2, TNF-α in CD8⁺ T cells vs control (n=3 independent experiments; p < 0.05/0.01). IFN-α upregulated HSPA6 mRNA/protein; shRNA silencing of HSPA6 (sh-HSPA6-2) reversed IFN-α–induced exhaustion phenotypes. - DNAJB1 functions downstream of HSPA6: Bioinformatic predictions (GeneMANIA, Coexpedia) and Co-IP supported HSPA6–DNAJB1 interaction. IFN-α increased DNAJB1 mRNA/protein; HSPA6 silencing reduced DNAJB1 expression, while DNAJB1 silencing did not alter HSPA6 protein, placing DNAJB1 downstream of HSPA6 in exhaustion regulation. - Pan-cancer IRG/ferroptosis landscape: Among 14 ferroptosis-related IRGs, IFNA1/5/21 and IFNG were relatively highly expressed in cancers; IFNG was significantly upregulated in several tumor types (e.g., UCEC, STAD, READ). Correlations among IRGs were observed (e.g., IFNA13 with IFNA6/IFNA2). Survival analyses linked IRG expression to prognosis; IFNG associated with better OS in BRCA, HNSC, SKCM, UCEC, but worse in KIRC and THYM. - IFN-α-stimulated CD8⁺ T cells attenuate tumor ferroptosis in co-culture: Continuous IFN-α stimulation increased tumor cell viability and decreased apoptosis across nine cell lines; tumor cells showed decreased COX2, ACSL4, NOX1 and increased GPX4 mRNA, with reduced MDA and LPO (p < 0.05), indicating reduced ferroptosis. - IFN-γ promotes tumor ferroptosis directly: IFN-γ treatment decreased tumor cell viability, increased apoptosis, elevated COX2, ACSL4, NOX1 and reduced GPX4 mRNA, with increased MDA/LPO; anti-IFN-γ neutralization reversed these effects, confirming IFN-γ–driven ferroptosis. - Silencing HSPA6/DNAJB1 enhances PD-1 blockade in vivo: In SK-MEL-1 xenografts (n=6/group), sh-HSPA6 or sh-DNAJB1 reduced tumor volume/weight; anti-PD-1 alone had limited effect, whereas sh-HSPA6+anti-PD-1 or sh-DNAJB1+anti-PD-1 significantly suppressed tumor growth. Tumor-infiltrating CD8⁺ T cells from sh-HSPA6/DNAJB1 groups showed reduced exhaustion markers, increased Ki67/CD69, elevated effector cytokines, higher tumor IFN-γ, increased ferroptosis marker expression (COX2, ACSL4, NOX1 up; GPX4 down), and increased MDA/LPO (p < 0.05). - Overall mechanism: Chronic IFN-α drives CD8⁺ T cell exhaustion via HSPA6–DNAJB1, reducing IFN-γ secretion, thereby diminishing IFN-γ–induced tumor ferroptosis. Silencing HSPA6/DNAJB1 restores CD8⁺ function, enhances IFN-γ, increases tumor ferroptosis, and improves PD-1 therapy efficacy.

The study addresses whether and how interferon signaling links CD8⁺ T cell exhaustion to tumor ferroptosis and impacts immunotherapy outcomes. Single-cell and trajectory analyses demonstrate that interferon-activated states precede or accompany expansion of exhausted CD8⁺ T cells in tumors. Mechanistically, HSPA6 emerges as a central mediator: IFN-α induces HSPA6, which via downstream DNAJB1 promotes an exhaustion program, reducing CD8⁺ effector functions and IFN-γ production. This, in turn, blunts IFN-γ–driven ferroptotic death in tumor cells, allowing greater tumor cell viability and survival. Conversely, direct IFN-γ exposure induces ferroptosis in tumor cells, with hallmark increases in lipid peroxidation (MDA/LPO) and ferroptosis-associated genes (COX2, ACSL4, NOX1) and decreased GPX4. Pan-cancer analyses connect IFNG expression with patient outcomes in a cancer-type–specific manner, consistent with variable roles for IFN-γ and the tumor microenvironment. Functionally, silencing HSPA6 or DNAJB1 reinvigorates CD8⁺ T cells, increases IFN-γ, elevates ferroptosis within tumors, and potentiates PD-1 blockade in vivo, directly linking the mechanistic axis to therapeutic benefit. These findings highlight the HSPA6–DNAJB1 pathway as a modulator of the IFN–exhaustion–ferroptosis triad and suggest that mitigating IFN-α–driven T cell exhaustion can restore ferroptotic vulnerability of tumors and enhance checkpoint inhibition.

This work defines an IFN-α–HSPA6–DNAJB1 axis that promotes CD8⁺ T cell exhaustion, suppresses IFN-γ production, and attenuates ferroptosis in tumor cells. HSPA6 is identified as a key regulator of exhaustion, with DNAJB1 acting downstream. Restoring CD8⁺ T cell function by silencing HSPA6/DNAJB1 increases IFN-γ, enhances tumor ferroptosis, and significantly improves anti-PD-1 efficacy in a melanoma xenograft model. Pan-cancer analyses underscore the prognostic relevance of IFN-related ferroptosis genes, particularly IFNG. These findings provide mechanistic insight and nominate HSPA6/DNAJB1 as potential therapeutic targets to optimize immunotherapy and ferroptosis-based strategies. Future research should: (1) validate these mechanisms in clinical samples and prospective trials; (2) delineate detailed signaling events and upstream regulators of HSPA6/DNAJB1 in T cells; (3) refine therapeutic regimens combining interferons, ferroptosis inducers, and checkpoint inhibitors with optimized dosing and timing; and (4) explore biomarkers predicting response to HSPA6/DNAJB1 targeting.

- Translational scope: Findings rely heavily on mouse models, xenografts, and in vitro systems; clinical validation with patient samples is needed. - Mechanistic depth: While HSPA6–DNAJB1 is implicated, full molecular signaling and network interactions governing exhaustion and ferroptosis are not fully elucidated. - Potential confounders: HSPA6, as a heat shock protein, may be induced by infections or stress; its expression could reflect infection status rather than tumor-specific processes in some contexts. - Tumor heterogeneity and microenvironment: Variability of HSPA6 expression among tumor cells and the influential role of the tumor microenvironment limit generalizability; correlations with tumor stage were inconsistent across cancers. - Cancer-type specificity: Prognostic impact of IFNG and IRGs varies by cancer, suggesting context-dependent effects requiring tailored strategies.