TRIM36 and CAMK2N2 regulate ferroptosis and antigen presentation in small cell lung cancer

Yifan Cai; Shuang Zhu; Honglin Wang; Fang Huang; Zhongyuan Yin; Jun Nie

doi:10.1016/j.isci.2026.115310

. 2026 Mar 11;29(4):115310. doi: 10.1016/j.isci.2026.115310

TRIM36 and CAMK2N2 regulate ferroptosis and antigen presentation in small cell lung cancer

Yifan Cai ^1,^2,^3,⁶, Shuang Zhu ^1,^2,^3,⁶, Honglin Wang ^4,⁶, Fang Huang ^1,^2,³, Zhongyuan Yin ^1,^2,^3,^∗, Jun Nie ^5,^7,^∗∗

PMCID: PMC13049528 PMID: 41940332

Summary

Small cell lung cancer (SCLC) is a highly aggressive tumor with poor prognosis. Ferroptosis is closely linked to tumor antigen presentation: it affects antigen presentation efficiency via immunostimulatory signals, while CD8⁺ T cell activation induced by antigen presentation promotes tumor cell ferroptosis by secreting IFNγ. This study used multi-omics analyses and machine learning to screen key genes, verified by in vitro/in vivo experiments. TRIM36 and CAMK2N2 were significantly upregulated in SCLC, negatively correlating with patient survival, effector memory CD8⁺ T cell infiltration, and tumor MHC I expression. They suppress SCLC antigen presentation via ferroptosis-dependent/independent mechanisms, limiting T cell function. TRIM36 and CAMK2N2 are promising SCLC biomarkers and therapeutic targets, providing clues to unravel ferroptosis-antigen presentation associations in tumor cells and optimize immunotherapeutic strategies.

Subject areas: Cell biology, Immunology

Graphical abstract

Highlights

•
TRIM36/CAMK2N2 upregulate in SCLC, low CD8⁺ T infiltration, poor survival
•
TRIM36/CAMK2N2 drive SCLC ferroptosis resistance via GPX4/FTH downregulation
•
TRIM36/CAMK2N2 inhibit SCLC antigen presentation via dual ferroptosis mechanisms
•
TRIM36/CAMK2N2 are SCLC biomarkers and therapeutic targets

Cell biology; Immunology

Introduction

Small cell lung cancer (SCLC) is a subtype of lung cancer with extremely high malignancy, characterized by rapid progression, early metastasis, and extremely poor prognosis.¹ Despite some advances in chemotherapy and immunotherapy for SCLC, the clinical efficacy remains less than desirable, largely owing to insufficient apprehension of its profoundly complex molecular mechanisms and the absence of effective targeted therapeutic avenues.²

Ferroptosis, a form of regulated cell death characterized by iron-dependent accumulation of lipid peroxides and depletion of glutathione (GSH), driven by inactivation of glutathione peroxidase 4 (GPX4), has emerged as a key factor influencing tumor progression and immune regulation.³ A burgeoning compendium of research underscores that ferroptosis, delineated as a singular form of non-apoptotic cell demise, may herald therapeutic prospects for addressing neoplastic entities exhibiting intractability to standard-of-care treatment modalities.⁴^,⁵

The tumor microenvironment is a complex ecosystem composed of various components such as tumor cells, immune cells, and stromal cells,⁶^,⁷ and there exists an intricate interaction between ferroptosis of tumor cells and the tumor microenvironment.⁸ On one hand, ferroptosis occurring in tumor cells can trigger immune responses within the tumor microenvironment.⁹ For instance, tumor cells treated with ferroptosis inducers release a variety of immunostimulatory signals in the early stage.¹⁰ These signals, as markers of immunogenic cell death, can promote the maturation of dendritic cells, enhance the phagocytic efficiency of macrophages toward ferroptotic tumor cells, facilitate the polarization of M1-type macrophages, and boost the activity of intratumoral T cells, thereby activating anti-tumor immune responses.²^,¹¹ On the other hand, antigen presentation between tumor cells and CD8⁺ T cells is the core link of adaptive anti-tumor immunity. CD8⁺ T cells exert their cytotoxic effects to kill tumor cells by recognizing tumor antigens presented by MHC class I (MHC I) on the surface of tumor cells. CD8⁺ T cells can promote ferroptosis in tumor cells through the secretion of IFNγ.¹² This process depends both on the restriction of cystine uptake and the enhancement of polyunsaturated fatty acid phospholipid (PUFA-PL) synthesis, and these metabolic changes may affect antigen presentation machinery (APM) of tumor cells.¹³ Meanwhile, products such as lipid peroxides released by ferroptotic tumor cells may alter the properties or presentation efficiency of tumor antigens, thereby influencing the recognition of tumor antigens by CD8⁺ T cells.¹⁴

However, current research on the key genes that simultaneously regulate tumor cell ferroptosis and APM remains relatively scarce. Identifying these genes with co-regulatory effects is helpful to reveal how ferroptosis in tumor cells influences the anti-tumor immune response of CD8⁺ T cells through regulating the APM genes. This study aims to explore the key genes that simultaneously affect APM and ferroptosis, conduct an in-depth analysis of their regulatory mechanisms in the interaction between tumor cells and CD8⁺ T cells, and provide a theoretical basis and potential targets for optimizing the immunotherapeutic strategies for SCLC.

Results

Study design and preliminary data analysis

This study established a comprehensive research workflow encompassing differential analysis, weighted gene co-expression network analysis (WGCNA), gene set enrichment analysis (GSEA), screening of key genes, immune cell infiltration analysis, and validation with external datasets (Figure 1A). Two datasets, GEO: GSE60052 ¹⁵ and GSE149507,¹⁶ were used in this study. GEO: GSE60052 ¹⁵ was applied for WGCNA, GSEA, differential expressed genes (DEGs) screening, key gene identification, immune cell infiltration analysis, expression differences, and survival correlation analysis; GEO: GSE149507 ¹⁶ was used for external validation of key gene expression patterns and diagnostic efficacy (Figure 1B). KEGG enrichment analysis revealed that pathways such as the PI3K-Akt signaling pathway and human papillomavirus infection were significantly enriched in SCLC (Figure 1C). The volcano plot of DEGs clearly showed a large number of genes with significant differential expression in the datasets (Figure 1D). GO enrichment analysis revealed that differentially expressed genes were significantly enriched in terms related to small GTPase-mediated signal transduction, wound healing, and cell-substrate adhesion in biological processes, collagen-containing extracellular matrix and focal adhesion in cellular components, and actin binding as well as GTPase/nucleoside-triphosphatase regulator activities in molecular functions, indicating their potential roles in cell signaling, extracellular matrix organization, and cytoskeletal regulation (Figure 1E). The heatmap of DEGs in samples indicated that the expression patterns of these genes could effectively distinguish tumor samples from normal samples (Figure 1F).

Biological processes including tube morphogenesis and positive regulation of cell migration were significantly enriched, suggesting that these biological processes are relatively active in SCLC (Figure 1G). To further characterize term relationships, a subset of enriched terms was selected for network plotting, with edges connecting terms of similarity >0.3. We prioritized terms with the most significant p values from each of 20 clusters (≤15 terms per cluster and ≤250 terms total). Networks were visualized in Cytoscape (http://metascape.org), where nodes (enriched terms) were colored by cluster ID (Figure 1H) and p value (Figure 1I).

Significant differences in biological processes and signaling pathways between SCLC and normal tissues

GEO: GSE60052 ¹⁵ was applied for GSEA. GSEA was performed to explore the functional terms enriched between tumor and normal tissues. It was found that processes related to vascular development and pathways such as adherens junction were enriched in tumor tissues (Figures 2A and 2C), while processes like epidermal cell differentiation and pathways including cytokine-cytokine receptor interaction were enriched in normal tissues (Figures 2B and 2D). These results confirm the presence of abnormal pathway reprogramming in SCLC, providing a direction for subsequent module screening via WGCNA.

Screening of key genes based on DEGs and WGCNA

Differential analysis of gene expression between tumor and normal groups and WGCNA were performed based on GEO: GSE60052 ¹⁵ RNA-seq data. Boxplots showed that both ferroptosis scores and APM scores were higher in tumor tissues than in normal tissues (Figures 2E and 2F). WGCNA was used to identify gene modules closely related to APM and ferroptosis. The results indicated no obvious outliers in sample clustering with good consistency, which is suitable for subsequent analyses (Figure 2G). An appropriate soft threshold was determined through scale independence and average connectivity analyses to construct the co-expression network (Figure 2H). Genes were clustered into multiple modules, laying a foundation for the correlation analysis between modules and phenotypes (Figure 2I). The brown module showed the strongest correlation with the tumor phenotype (Figure 2J) and was closely associated with APM and ferroptosis (Figures 2K and 2L). A Venn diagram revealed 186 genes that are both DEGs and co-expression module genes, which may be key genes (Figure 2M). Random forest (RF) model analysis showed that the model had the best predictive performance (lower root-mean-square error [RMSE]) when 4 variables were included (Figure 2N).

Correlation analysis between immune cells and key genes

The data used for genes’ correlation with immune cells is sourced from GEO: GSE60052.¹⁵ The correlation heatmap of immune cell subsets revealed a positive correlation between Tregs and M2 macrophages, indicating the synergistic effect of immunosuppressive cells in jointly inhibiting anti-tumor immunity; CD8⁺ T cells were negatively correlated with Tregs, suggesting the antagonism between effector cells and inhibitory cells. This map reveals a correlation matrix in the SCLC immune microenvironment (Figure 3A). The correlation matrix between key genes and immune cells showed that TRIM36 was positively correlated with Tregs and M2 macrophages, and negatively correlated with central memory CD8⁺ T cells, suggesting that this gene may promote the infiltration of immunosuppressive cells. GPX3 was positively correlated with CD8⁺ T cells, indicating that this gene may enhance anti-tumor immunity (Figure 3B). Scatterplots of key genes and immune cells showed that GPX3 expression was positively correlated with effector memory CD8⁺ T cells, while TRIM36, RND2, and CAMK2N2 were negatively correlated with effector memory CD8⁺ T cells (Figure 3C). The four key genes also showed significant correlations (p < 0.05) with various immune cells (e.g., activated CD4⁺ T cells and activated dendritic cells), indicating that these key genes may play roles in regulating the immune microenvironment (Figures 3D and 3E).

Correlation analysis based on immune cells and key genes

(A) Heatmap of correlations among immune cell subsets.

(B) Correlation matrix between key genes and immune cells.

(C) Scatterplot of the correlation between key genes and effector memory CD8⁺ T cells.

(D) Scatterplot of the correlation between key genes and activated CD4⁺ T cells.

(E) Scatterplot of the correlation between key genes and activated dendritic cells. Statistical significance: ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001. Correlations were analyzed using Spearman’s rank correlation test.

Expression differences and diagnostic efficacy of key genes

The expression and diagnostic efficacy of key genes were initially evaluated using GEO: GSE60052 ¹⁵ and further validated with GEO: GSE149507 ¹⁶ microarray data. The expression of key genes in tumor and normal samples showed significant differences (p < 0.001), among which GPX3 was more highly expressed in normal samples, and the other three were upregulated in tumor samples (Figure 4A). Preliminary analysis based on bulk RNA-seq data suggests that these genes hold promise as potential diagnostic markers for SCLC, and their specificity needs to be further confirmed by validating their cellular origins (Figures 4B–4E). Verification of expression differences of key genes in the GEO: GSE149507 ¹⁶ dataset showed no significant difference in RND2 expression (p > 0.05), while the changes in other genes were consistent with previous results (Figure 4F), and their AUC values were all relatively high (Figures 4G–4J). To further confirm the robustness of these genes as potential diagnostic markers and avoid overfitting, we performed 5-fold cross-validation on both the GEO: GSE60052 ¹⁵ and GSE149507 ¹⁶ datasets. The resulting receiver operating characteristic (ROC) curves for GPX3, TRIM36, CAMK2N2, and RND2 (Figures S1A–S1D) demonstrated stable diagnostic performance across cross-validation folds, with AUC values consistent with the initial analysis. These cross-validated results confirm that the diagnostic efficacy of the key genes is not driven by overfitting, reinforcing their potential as reliable diagnostic biomarkers for SCLC. Due to the limitations of accessing clinical SCLC specimens, we further validated the protein expression of these key genes using a mouse orthotopic lung cancer model established by intrapleural injection of RP1 cells (a murine SCLC cell line). Western blot (WB) analysis of orthotopic tumor tissues and adjacent normal lung tissues confirmed that TRIM36, CAMK2N2, and RND2 were significantly upregulated at the protein level in tumor tissues, while GPX3 was downregulated (Figures S1E and S1F).

To clarify the cellular specificity of these key genes, we performed single-cell transcriptome analysis. The data were sourced from Chan et al.¹⁷ A UMAP dimensionality reduction plot of the single-cell transcriptome data shows the distribution of various cell types in the SCLC microenvironment, including immune cells (T cells, B cells, macrophages, neutrophils, and dendritic cells) and different SCLC subtypes (SCLC-A, SCLC-N, and SCLC-P), presenting the spatial distribution pattern of these cells (Figure 4K). UMAP plots illustrate the expression distribution of key genes (GPX3, RND2, TRIM36, and CAMK2N2) in the SCLC single-cell transcriptome (Figure 4L). Then, we quantified the expression levels of CAMK2N2, TRIM36, RND2, and GPX3 in different cell clusters. All four key genes are mainly expressed in the subtypes of SCLC cells (SCLC-A, SCLC-N, and SCLC-P), while they are hardly expressed in immune cells (T cells, B cells, macrophages, neutrophils, and DCs) (Figures S1G–S1J).

Survival analysis of patients based on key genes and their correlation with APM genes

The data used for survival analysis is sourced from the George’s cohort.¹⁸ The survival was significantly longer in the GPX3 low-expression group and RND2 high-expression group (both p < 0.0001) (Figures 5A and 5C), while it was significantly shorter in the TRIM36 high-expression group and CAMK2N2 high-expression group (both p < 0.0001) (Figures 5B and 5D). We focused on TRIM36 and CAMK2N2, two genes that are more highly expressed in tumor tissues and whose high expression is negatively correlated with survival. Meanwhile, a scatterplot was made based on the expression of key genes and the average correlation between the expression of key genes and the expression of APM genes. The data used for genes’ correlation with APM genes is sourced from GEO: GSE60052.¹⁵ The results showed that TRIM36 and CAMK2N2 were located in the upper left corner, indicating their higher expression and a significant negative correlation with APM genes (Figure 5E). Among them, TRIM36 and CAMK2N2 were significantly negatively correlated with MHC I (Figures 5F and 5G).

TRIM36 and CAMK2N2 promote tumor growth in vivo and inhibit the expression of MHC I

Murine shRNA was constructed, and the knockdown efficiency of Camk2n2 and Trim36 was verified in RP1 cells, with both expressions significantly reduced (p < 0.001) (Figure 6A). ShCamk2n2 and ShTrim36 groups exhibited smaller tumor volumes (Figure 6B), lighter tumor weights (p < 0.001) (Figure 6C), and slower tumor growth rates (p < 0.001) compared to the NC group (Figure 6D). Moreover, the mean fluorescence intensity (MFI) of H2D/H2K and beta-2 microglobulin (B2M) on tumor cell surfaces (Figures 6E and 6F) in tumor tissues were also markedly higher in these two groups than in the NC group (p < 0.01). In addition, the number of CD3⁺CD8⁺T cells and IFNγ⁺CD8⁺T cells in the ShCamk2n2 group and ShTrim36 group was significantly increased compared with that in the NC group (Figures 6G–6I).

*TRIM36* and *CAMK2N2* promote tumor growth *in vivo*, inhibit the expression of MHC I, and suppress CD8⁺ T cell activation in SCLC

(A) qRT-PCR detection of the relative expression levels of *Camk2n2* gene in the NC group and ShCamk2n2 group, and *Trim36* gene in the NC group and ShTrim36 group of RP1 cells. qRT-PCR data were normalized to β-actin (ACTB) expression.

(B) Gross morphology of subcutaneous tumors in the NC group, ShCamk2n2 group, and ShTrim36 group after dissection on the 24th day when mice were injected with cells (n = 5 per group).

(C) Statistics of tumor weights in each group.

(D) Curves of tumor volume changes over time in each group.

(E and F) Flow cytometry detection and statistics of the expression of H2D/H2K and B2M in tumors.

(G–I) Flow cytometry analysis of the proportion of CD3⁺CD8⁺ T cells and IFNγ⁺CD8⁺ T cells. Cell experiments were independently repeated 3 times (n = 3). Data are presented as mean ± SD. Statistical significance: ∗∗p < 0.01, ∗∗∗p < 0.001. Statistical analyses were performed using GraphPad Prism software. Multiple group comparisons were conducted using one-way analysis of variance (one-way ANOVA) followed by Bonferroni correction for pairwise comparisons. NC, negative control; MFI, mean fluorescence intensity.

The number of CD45⁺ cells in the ShTrim36 group was significantly increased compared with that in the NC group, while there was no significant difference between the NC group and the ShCamk2n2 group (Figure S2A). Immunofluorescence showed that the relative expression levels of B2M in the ShCamk2n2 group and ShTrim36 group were significantly higher than those in the NC group (Figures S2B and S2C); TUNEL staining (green) showed that the proportion of TUNEL⁺ cells in the total cells in the ShCamk2n2 group and ShTrim36 group was significantly increased compared with that in the NC group (Figures S2D and S2E).

To further verify the effect of genes on tumor growth, we, respectively, subjected RP1 cells with stable knockdown of Camk2n2 and Trim36 to puromycin selection for 10 days. On the basis of the obtained cell lines, stable overexpression plasmids were administered. The five groups of cells (NC+Vector, ShCamk2n2+Vector, ShCamk2n2+rCamk2n2, ShTrim36+Vector, and ShTrim36+rTrim36) were injected subcutaneously in equal amounts, and the subcutaneous tumor tissues were detected by quantitative real-time PCR (qRT-PCR). The expression level of Camk2n2 gene in the ShCamk2n2 group was significantly lower than that in the NC+Vector group, while that in the ShCamk2n2+Camk2n2 group was increased compared with that in the ShCamk2n2+Vector group, and the results of ShTrim36 were similar (Figures S2F and S2G). The tumor mass and volume in the ShCamk2n2+Vector group and ShTrim36+Vector group were significantly lower than those in the NC+Vector group, while those in the ShCamk2n2+rCamk2n2 group and ShTrim36+rTrim36 group were significantly higher than those in the ShCamk2n2+Vector group and ShTrim36+Vector group (Figures S2H–S2J).

TRIM36 and CAMK2N2 promote ferroptosis resistance in SCLC cells

Human shRNA was constructed, and the knockdown efficiency of CAMK2N2 and TRIM36 was verified in DMS114 cells, with both expressions significantly reduced (p < 0.05) (Figure 7A). The cell viability curves of DMS114 cells treated with Ras-selective lethal 3 (RSL3) and Erastin, respectively, showed that the IC50 values were 1.912 and 11.57 μM (Figures S3A and S3B). Therefore, for subsequent propidium iodide (PI) staining, colony formation, and ROS detection of DMS114 cells, treatment with 1 μM RSL3 and 10 μM Erastin was used, respectively.

*TRIM36* and *CAMK2N2* promote resistance to RSL3-induced ferroptosis in DMS114

(A) qRT-PCR detection of the relative expression levels of *CAMK2N2* gene in the NC group and ShCAMK2N2 group, and *TRIM36* gene in the NC group and ShTRIM36 group of DMS114 cells.

(B and D) There are four groups of cells: untreated, NC, ShCAMK2N2, and ShTRIM36. Each group was subjected to three treatments: control, RSL3 (treated with 1 μmol/L for 6 h), and RSL3 + Fer-1 (co-treated with Fer-1 and 1 μmol/L RSL3 for 6 h). After adding 5 μg/mL PI working solution, fluorescence microscopy was used to observe the images and statistically analyze the proportion of PI-negative cells. Upper: bright field; lower: cell PI staining. Scale bars, 100 μm

(C) After the four groups of cells (untreated, NC, ShCAMK2N2, and ShTRIM36) were treated with 1 μmol/L RSL3, CCK-8 assay was used to analyze the relative cell viability of different groups at different time points.

(E) Each of the four groups of cells was subjected to three treatments: control, RSL3 (treated with 1 μmol/L for 6 h), and RSL3 + Fer-1 (co-treated with Fer-1 and 1 μmol/L RSL3 for 6 h). After 6 h, the medium was changed, and the cells were cultured for another 10 days, followed by images of colony formation plates stained with crystal violet.

(F) Statistics of the number of colony formations (colonies with >50 cells were counted as positive).

(G and H) Results of flow cytometry detection with Bodipy-FITC staining and MFI statistics. All cell experiments were independently repeated three times (n = 3). Data are presented as mean ± SD. Statistical significance: ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. Statistical analyses were performed using GraphPad Prism software. Multiple group comparisons were conducted using one-way ANOVA followed by Bonferroni correction. Fluorescence microscopy images and colony formation were quantified using ImageJ software.

For PI staining of cells, after RSL3 treatment, the ShCAMK2N2 group and ShTRIM36 group had more positive cells, and ferrostatin-1 (Fer-1) could alleviate this phenomenon (Figures 7B and 7D). The curves of cell viability changes over time showed that after RSL3 treatment, the cell viability of the ShCAMK2N2 group and ShTRIM36 group decreased more rapidly than that of the NC group (p < 0.001) (Figure 7C). Consistent results were observed following erastin treatment (Figures S3C–S3E). Colony formation assays showed that when ferroptosis was induced by RSL3, the colony formation ability of cells in the ShCAMK2N2 group and ShTRIM36 group was significantly lower than that in the NC group, and Fer-1 could partially reverse this effect (Figures 7E and 7F). Consistent results were observed following erastin treatment (Figures S3F and S3G). After RSL3 treatment at the same concentration and time, the level of lipid peroxidation in cells of the ShCAMK2N2 and ShTRIM36 groups was higher, indicating a more severe degree of ferroptosis (Figures 7G and 7H). These results collectively indicate that TRIM36 and CAMK2N2 can promote ferroptosis resistance in SCLC cells.

To investigate the role of CAMK2N2 and TRIM36 in H1688 cells, we first verified the knockdown efficiency: compared with the NC group, the expression of CAMK2N2 in the ShCAMK2N2 group and TRIM36 in the ShTRIM36 group was significantly reduced (Figures S4A and S4B). Cell viability curves showed that the IC50 values of H1688 cells treated with RSL3 and erastin were 8.157 and 17.11 μM, respectively (Figures S4C and S4D). For RSL3-induced ferroptosis, PI staining showed that the proportion of PI-positive cells in the ShCAMK2N2 and ShTRIM36 groups was higher following RSL3 treatment, and Fer-1 could alleviate this condition (Figures S4E and S4G). The time-dependent cell viability curve indicated that after RSL3 treatment, the cell viability of the ShCAMK2N2 and ShTRIM36 groups decreased more rapidly (Figure S4F). Consistent results were observed with erastin treatment (Figures S4H–S4J). Colony formation assays demonstrated that RSL3 treatment significantly reduced the colony-forming ability of the ShCAMK2N2 and ShTRIM36 groups, and Fer-1 could partially reverse this effect. Consistent results were observed with erastin treatment (Figures S4K–S4N). Collectively, these results indicate that CAMK2N2 and TRIM36 can promote ferroptosis resistance in H1688 cells, which is consistent with the findings in DMS114 cells.

Notably, in the absence of ferroptosis inducers, there were no significant differences in the growth rates of DMS114 and H1688 cells among the four groups (untreated, NC, ShCAMK2N2, and ShTRIM36). These findings further illustrate that the inhibitory effect of CAMK2N2 and TRIM36 on tumor growth is closely associated with increased ferroptosis sensitivity (Figures S5A–S5E).

Knockdown of CAMK2N2 or TRIM36 enhances ferroptosis susceptibility via downregulating GPX4/FTH1, with CAMK2N2 disrupting ER calcium homeostasis

To investigate the mechanism by which CAMK2N2 and TRIM36 affect ferroptosis sensitivity, we detected the common ferroptosis-mediating molecules GPX4 (a crucial enzyme for lipid peroxide detoxification) and FTH1 (involved in iron homeostasis) protein expression levels by WB. The results showed that in H1688, DMS114, and RP1 cells, the protein expression levels of GPX4 and FTH1 in the ShCAMK2N2 group and ShTRIM36 group were significantly lower than those in the NC group. This indicates that the depletion of CAMK2N2 and TRIM36 impairs the cellular defense against lipid peroxidation and disrupts iron metabolism, thereby enhancing the susceptibility to ferroptosis (Figures S5F–S5H).

CAMK2N2 is a specific inhibitor of calcium/calmodulin-dependent protein kinase II (CaMKII), which inhibits CaMKII activity by directly binding to its catalytic domain. To explore the molecular mechanism by which CAMK2N2 participates in ferroptosis, we performed co-localization experiments of calreticulin and calnexin. The results showed that the Pearson correlation coefficient of calreticulin and calnexin co-localization in the ShCAMK2N2 group was significantly lower than that in the NC group (Figure S5I). This indicates that the ShCAMK2N2 group may induce endoplasmic reticulum calcium homeostasis imbalance, thereby triggering lipid peroxidation and abnormal expression of ferroptosis-related proteins, ultimately promoting the ferroptosis process.

TRIM36 and CAMK2N2 inhibit the expression of APM genes in SCLC cells

The mRNA expressions of APM genes such as H2d, H2k, B2m, Tap1, Psmb8, Psmb9, and Nlrc5 in RP1 cells of the ShCamk2n2 and ShTrim36 groups were upregulated compared with those in the NC group, which was more pronounced following IFNγ stimulation (p < 0.05) (Figures 8A–8D). The protein expressions of human leukocyte antigen-ABC (HLA-ABC) and B2M in H1688 cells of the ShCAMK2N2 and ShTRIM36 groups were significantly higher than those in the NC group (Figures 8E, 8F, S6A, and S6B). The MFI of H2D/H2K and B2M on the surface of knockdown RP1 cells in the ShCamk2n2 and ShTRIM36 groups was significantly higher than that in the NC group (p < 0.05) (Figures 8G–8J), indicating that both TRIM36 and CAMK2N2 inhibit the expression of APM genes in SCLC cells.

*TRIM36* and *CAMK2N2* inhibit the expression of antigen presentation genes in SCLC cells

(A and B) The relative expression levels of *H2D*, *H2K*, *B2m*, and *Tap1* genes in the untreated group, NC group, ShCamk2n2 group, and ShTrim36 group of RP1 cells with or without IFNγ treatment detected by qRT-PCR. qRT-PCR data were normalized to ACTB expression.

(C and D) The relative expression levels of *Psmb8*, *Psmb9*, and *Nlrc5* genes in the untreated group, NC group, ShCamk2n2 group, and ShTrim36 group with or without IFNγ treatment detected by qRT-PCR. qRT-PCR data were normalized to ACTB expression.

(E and F) WB detection of the expression levels of HLA-ABC, B2M, and ACTB proteins in the untreated group, NC group, ShCAMK2N2 group, and ShTRIM36 group of H1688 cells with or without IFNγ treatment. WB band intensities were quantified using ImageJ software (normalized to ACTB).

(G–J) MFI of H2D/H2K and B2M in the untreated group, NC group, ShCamk2n2 group, and ShTrim36 group with or without IFNγ treatment in RP1 cells detected by flow cytometry. All cell experiments were independently repeated three times (n = 3). Data are presented as mean ± SD. Statistical significance definitions: n.s., no significance; ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001. Statistical analyses were performed using GraphPad Prism software. Multiple group comparisons were conducted using one-way ANOVA followed by Bonferroni correction. NC, negative control; HLA, human leukocyte antigen; B2M, beta-2 microglobulin; H2D/H2K, histocompatibility 2-D region antigen/H2-K region antigen; MFI, mean fluorescence intensity.

TRIM36 and CAMK2N2 suppress antigen presentation in SCLC through ferroptosis-dependent and independent mechanisms, limiting T cell effector function

To investigate the link between the effects of TRIM36 and CAMK2N2 on ferroptosis and antigen presentation, we established an mSCLC-OVA co-culture system (Figure 9A): OT-I T cells were isolated from OT-I mice, and RP1-OVA cells (SCLC cells stably expressing full-length OVA) were subjected to NC, ShCamk2n2, or ShTrim36 treatment, followed by co-culture with OT-I T cells at ratios of 2:1, 4:1, and 8:1. The secretion of IFNγ and TNFα was measured by enzyme-linked immunosorbent assay (ELISA). Flow cytometry analysis of the MFI of SIINFEKL-Kb (OVA peptide bound to H2-Kb) on the surface of tumor cells showed that, compared with the NC group, the expression of SIINFEKL-Kb was significantly higher in the ShCamk2n2 and ShTrim36 groups. This expression was further enhanced by RSL3 and reversed by Fer-1 (Figures 9B and 9C). Correspondingly, ELISA results demonstrated that OT-I T cells co-cultured with ShCamk2n2 or ShTrim36-treated RP1-OVA cells secreted remarkably higher levels of IFNγ (Figure 9D) and TNFα (Figure 9E) across all three co-culture ratios. We then selected the 4:1 co-culture ratio to observe the effect of ferroptosis induction on OT-I T cells. Under RSL3 stimulation, the secretion of IFNγ (Figure 9F) and TNFα (Figure 9G) was synergistically increased in the ShCamk2n2 and ShTRIM36 groups, while Fer-1 abrogated this effect. Collectively, these results indicate that depletion of Camk2n2 or Trim36 enhances antigen presentation and T cell effector function in SCLC through two aspects: on the one hand, it directly promotes antigen presentation independently of ferroptosis; on the other hand, it further potentiates this process by increasing ferroptosis sensitivity, ultimately synergistically enhancing T cell effector function.

Discussion

Ferroptosis, a specialized mode of regulated cell death dependent on iron, is driven by the toxic buildup of lipid peroxides within cellular membranes—and it has rapidly emerged as a highly promising area of exploration in oncology research.¹⁹ Beyond directly inhibiting tumor progression, triggering ferroptosis also offers substantial potential to enhance the efficacy of immunotherapeutic strategies and overcome acquired resistance to conventional cancer treatments.²⁰ The latest research has found that inducing ferroptosis can become a potential strategy for treating lung cancer, enhancing the efficacy of ferroptosis in inhibiting SCLC and strengthening the therapeutic effect of anti-PD-L1 immunotherapy.⁴ However, current research on SCLC mostly focuses on a single dimension—either concentrating on strategies to induce ferroptosis or exploring mechanisms to reverse immune escape independently—while rarely uncovering key molecules that simultaneously regulate two core malignant traits: ferroptosis resistance and antigen presentation deficiency. Building on this, the present study, through multi-omics screening and functional validation, identifies TRIM36 and CAMK2N2 as dual-function genes in SCLC that link ferroptosis resistance to antigen presentation suppression. We systematically clarify their regulatory mechanisms and clinical significance, thereby providing molecular targets and a theoretical foundation for addressing therapeutic challenges in SCLC.

The core role of WGCNA is to construct a co-expression network among genes, cluster genes with similar expression patterns into different modules, and then explore gene modules and hub genes that are significantly associated with specific biological phenotypes (APM and ferroptosis).²¹ In this study, WGCNA divided genes into multiple co-expression modules, among which the brown module showed a significant positive correlation with tumor ferroptosis and APM. Genes within this module not only exhibit high co-expression but also have a significant association with tumor phenotypes, confirming that it is not a random cluster but precisely linked to SCLC phenotypes. By taking the intersection of WGCNA module genes and DEGs, combined with machine learning validation, the study screened out key genes that simultaneously meet the criteria of “core of co-expression module” and “significantly differentially expressed.” These genes are core targets that simultaneously affect ferroptosis and APM in SCLC tumor cells. The finally identified key genes such as TRIM36 and CAMK2N2 show significant expression differences between tumor and normal tissues. ROC curves indicate that they have good diagnostic efficacy, and this was partially verified in the external GEO: GSE149507 ¹⁶ dataset, suggesting that they can serve as potential diagnostic markers for SCLC. Survival analysis further revealed that high expression of TRIM36 and CAMK2N2 is associated with poor prognosis in SCLC patients, highlighting their value in the diagnosis and treatment of SCLC.

In terms of ferroptosis regulatory mechanisms, in vitro and in vivo experiments confirmed that knockdown of TRIM36 and CAMK2N2 can significantly inhibit tumor growth and is closely related to ferroptosis resistance. After knockdown of TRIM36 and CAMK2N2, SCLC cells showed increased sensitivity to RSL3 and erastin, decreased cell viability, weakened colony-forming ability, elevated lipid peroxidation levels, and TUNEL staining indicated increased cell death, all supporting the role of these two genes in promoting ferroptosis resistance. TRIM36 is a member of the tripartite motif (TRIM) protein family and has been identified as an androgen-responsive gene. The role of TRIM36 varies in different tumors. It exerts a tumor-suppressive effect in various cancers including colorectal cancer,²² hepatocellular carcinoma,²³ human esophageal squamous cell carcinoma,²⁴ and prostate cancer.²⁵ However, it has a significant promoting effect on clear cell renal cell carcinoma, facilitating its proliferation, thereby participating in cell division and cell cycle progression, and influencing the tumor immune microenvironment, making it a promising biomarker.²⁶ As a crucial enzyme for lipid peroxide detoxification, reduced expression of GPX4 directly leads to the accumulation of intracellular lipid peroxidation products, triggering ferroptosis. In contrast, FTH1 is involved in iron homeostasis regulation, and its downregulation exacerbates intracellular iron overload, further promoting lipid peroxidation through the Fenton reaction. This study found that knockdown of TRIM36 and CAMK2N2 significantly decreased the protein expression of GPX4 and FTH1 in multiple SCLC cell lines, including H1688, DMS114, and RP1. This result reveals the core molecular mechanism by which these two genes regulate ferroptosis—by simultaneously downregulating GPX4 (impairing lipid peroxide detoxification) and FTH1 (exacerbating iron overload), TRIM36 and CAMK2N2 create a “lipid peroxidation + iron overload” synergistic effect, thereby enhancing ferroptosis sensitivity.

CAMK2N2, namely calmodulin kinase II inhibitor 2, is a potent and specific intracellular inhibitor of calcium/calmodulin-dependent protein kinase II (CaM-kinase II, CAMK2). It can bind to CaM-kinase II, sequester Ca²⁺/calmodulin thereon, and thereby inhibit the activity of CaM-kinase II.²⁷ Currently, there are very few reports on the relationship between CAMK2N2 and cancer. In this study, it was found that inhibiting CAMK2N2 can regulate ferroptosis in SCLC, enhance the expression of APM genes, and promote CD8⁺ T cell infiltration. This result echoes the findings of Zhao et al.²⁸ in head and neck squamous cell carcinoma—that after ferroptosis induction, the exposure of calreticulin on the surface of tumor cells increases, which acts as an “eat-me” signal to initiate an immunogenic response. Since the function of calreticulin depends on intracellular calcium homeostasis,²⁹ and CAMK2N2, as an endogenous inhibitor of CaMKII, primarily functions to regulate calcium/calmodulin-dependent signaling pathways and affect the efficiency of intracellular calcium signal transduction³⁰; this study found that knocking down CAMK2N2 significantly reduced the Pearson correlation coefficient of the co-localization of calreticulin and calnexin, suggesting that ShCAMK2N2 promotes ferroptosis and immune regulation by disrupting the calcium homeostasis in the endoplasmic reticulum. Both collectively point to the potential cross-regulatory network of “calcium signal-ferroptosis-immune response.”

The efficacy of immunotherapy in SCLC patients is closely associated with the immune microenvironment and APM of the tumor cells.³¹^,³²^,³³ Only 36% of SCLC tumors exhibit significant CD45⁺ immune cell infiltration, and high CD8⁺ T cell infiltration is associated with longer overall survival in limited-stage SCLC. More than two-thirds of SCLC cases present with “immune desert” or “low infiltration” phenotypes, indicating weak overall immunogenicity of this malignancy.³⁴ The process of tumor antigen presentation begins with the degradation of abnormal proteins (such as mutant oncoproteins) within tumor cells into short peptides by immunoproteasomes, which contain catalytic subunits encoded by PSMB8 and PSMB9 genes. These peptides are transported into the endoplasmic reticulum with the assistance of molecules like TAP1. Within the endoplasmic reticulum, the peptides bind to MHC I (H2D/H2K in mice and HLA-A, B, C in human), which then assemble with B2M to form a stable complex. This peptide-MHC I-B2M complex is transported to the surface of tumor cells via the secretory pathway, where it is recognized by the T cell receptor of CD8⁺ T cells. Meanwhile, the CD8 co-receptor enhances this interaction, ultimately activating cytotoxic T cells to eliminate tumor cells.³⁵ IFNγ secreted by CD8⁺ T cells can promote the entire process of antigen presentation, including the activation of the transcription factor NLRC5, thereby facilitating the transcription of other APM genes.³⁶ In the results of this study, immune cell infiltration analysis revealed that SCLC exhibits an immunosuppressive microenvironment, which is consistent with previous reports.³⁷ TRIM36 was positively correlated with immunosuppressive cells (such as Tregs and M2 macrophages) and negatively correlated with effector memory CD8⁺ T cells, and CAMK2N2 showed a similar correlation pattern, indicating that they may be involved in shaping the immunosuppressive microenvironment and inhibiting anti-tumor immunity. Meanwhile, both genes were significantly negatively correlated with APM genes, especially those associated with MHC I. Experiments further confirmed that after knockdown of TRIM36 and CAMK2N2, the expressions of APM genes such as H2D, H2K, and B2M in SCLC cells were significantly upregulated, and the mRNA level of NLRC5 also increased. These findings suggest that TRIM36 and CAMK2N2 may reduce the mRNA expression of NLRC5, thereby downregulating the expressions of MHC I and other APM genes. This process can inhibit the APM genes, reduce the display of antigens on the surface of tumor cells, and suppress the activation of CD8⁺ T cells, ultimately inhibiting anti-tumor immunity and promoting SCLC progression.

It is worth noting that this study has confirmed that TRIM36 and CAMK2N2 regulate the antigen presentation of SCLC through a dual pathway of “ferroptosis-dependent + ferroptosis-independent.” Without ferroptosis inducer (RSL3), knocking down these two genes can significantly enhance the expression of SIINFEKL-Kb and the secretion of IFNγ and TNFα of CD8⁺ T cells. Cytokine secretion is an early and specific indicator of T cell activation, which can directly reflect the efficiency of the “tumor cell antigen presentation-T cell activation” process.³⁸ The results indicated that knocking down TRIM36 and CAMK2N2 can directly relieve the inhibition of APM genes (such as H2D, H2K, and NLRC5), and this “ferroptosis-independent” effect is the basis for the regulation of immune escape by these two genes; while after RSL3 induces ferroptosis, this effect is significantly amplified and can be reversed by Fer-1, suggesting that ferroptosis enhances antigen presentation through exacerbating lipid peroxidation, promoting the exposure of calreticulin, and other “immunogenic ferroptosis” characteristics.

In summary, this study identifies TRIM36 and CAMK2N2 as key genes that simultaneously regulate ferroptosis and APM in SCLC. They promote SCLC progression through mechanisms such as enhancing ferroptosis resistance, inhibiting antigen presentation, and shaping an immunosuppressive microenvironment. These findings provide a theoretical foundation for the precise diagnosis, prognostic evaluation, and development of immunotherapeutic targets for SCLC. Subsequent explorations addressing the limitations of this study are expected to bring breakthroughs in SCLC treatment.

Limitations of the study

A limitation of this study is the lack of cell-type resolution in our bulk RNA-seq and microarray datasets. Since our samples contain mixed cell populations (tumor, stromal, and immune cells), we cannot definitively attribute gene expression profiles and pathway enrichments solely to SCLC cells (some signals may derive from non-tumor cells). Thus, the depicted pathways are not entirely SCLC-specific. In future work, integrating single-cell RNA-seq or cell-type deconvolution will be critical to dissect the cellular origin of these signatures and clarify their specificity. Second, there is insufficient research on the direct interaction between these two genes and CD8⁺ T cells, which can be further analyzed through experiments such as co-culture in the future. Third, the study focuses on the individual roles of TRIM36 and CAMK2N2, but whether there is a synergistic or antagonistic relationship between them in regulating ferroptosis and APM, as well as the interaction network with other key genes (such as GPX3, RND2), has not been systematically explored. The overall mechanism of multi-gene co-regulation still needs to be improved. Fourth, this study did not further investigate the association between the expression of TRIM36 and CAMK2N2 and neuroendocrine (NE) subtypes of SCLC, yet this dimension is likely critical for deciphering their immunoregulatory mechanisms. Accumulating evidence has established that NE subtypes of SCLC-namely NE-high and NE-low³⁹—not only dictate the pattern of immune cell infiltration but also exhibit strong correlations with the expression of antigen-presenting molecules (e.g., MHC II) and immune checkpoint molecules (e.g., IDO and TIM3).⁴⁰ Specifically, the NE-low subtype tends to exhibit greater CD8⁺ T cell infiltration but concurrently overexpresses immunosuppressive molecules such as IDO, whereas the NE-high subtype is predominantly characterized by an “immune desert” phenotype.⁴¹ Notably, several key questions remain unresolved in our study: first, do the inhibitory effects of TRIM36 and CAMK2N2 on MHC I expression and CD8⁺ T cell infiltration differ across distinct NE subtypes—for instance, are these effects only significant in the NE-high subtype? Second, is the ability of TRIM36 and CAMK2N2 to regulate ferroptosis resistance also modulated by NE subtype status? To address these knowledge gaps, future studies should recruit SCLC cohorts with larger sample sizes, integrate the detection of NE subtype-specific markers (e.g., ASCL1 and NEUROD1), and thereby clarify the functional specificity of TRIM36 and CAMK2N2 across different NE subtypes of SCLC.

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Jun Nie (jun_nie@aliyun.com).

Materials availability

This study did not generate new unique reagents.

Data and code availability

•
The datasets analyzed during the current study are available in the Gene Expression Omnibus (GEO) repository, with accession numbers GEO: GSE60052: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE60052) and GSE149507: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE149507). All datasets used in this study are fully publicly available with no access restrictions.
•
Custom code: Not applicable. No custom code was generated in this study.
•
This study did not generate new standardized datasets that require deposition in a Cell Press-approved repository.
•
Any additional information required to reanalyze the data reported in this article is available from the lead contact upon request.

Acknowledgments

We would like to express our gratitude to the Medical Subcenter of HUST Analytical & Testing Center for providing technical support during the data acquisition process of this study. This study was supported by the National Natural Science Foundation of China (grant no. 82472879) to F.H., the Natural Science Foundation of Hubei Province (grant no. 2022CFB306) to J.N., and the Natural Science Foundation of Hubei Province (grant no. 2024AFD436) to Z.Y.

Author contributions

Conceptualization, Y.C. and S.Z.; methodology, Y.C., S.Z., and H.W.; formal analysis, Y.C., S.Z., and F.H.; investigation, Y.C., S.Z., H.W., F.H., and Z.Y.; validation, Y.C., S.Z., and J.N.; resources, Y.C. and S.Z.; software, Y.C. and H.W.; data curation, Y.C., S.Z., and F.H.; writing – original draft, Y.C. and S.Z.; writing – review and editing, Y.C., Z.Y., and J.N.; visualization, Y.C., S.Z., and H.W.; funding acquisition, F.H. and J.N.

Declaration of interests

The authors declare no competing interests.

Declaration of generative AI and AI-assisted technologies in the writing process

During the preparation of this work, the authors used ChatGPT in order to improve language. After using this tool or service, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Deposited data

RNA data	GEO repository	GEO: GSE60052 ¹⁵
RNA data	GEO repository	GEO: GSE149507 ¹⁶
RNA data	George et al.¹⁸	https://pubmed.ncbi.nlm.nih.gov/26168399/
RNA data, scRNA-seq data	Chan et al.¹⁷	https://pubmed.ncbi.nlm.nih.gov/34653364/

Software and algorithms

R (version 4.4.3)	The R Project	https://www.r-project.org
Limma (R package)	Bioconductor	https://doi.org/10.18129/B9.bioc.limma
WGCNA (R package)	Langfelder et al.⁴²	https://www.rdocumentation.org/packages/WGCNA/versions/1.51
RandomForest (R package)	CRAN	https://doi.org/10.32614/CRAN.package.randomForest
pROC (R package)	CRAN	https://CRAN.R-project.org/package=pROC
ImageJ	Schindelin et al.⁴³	https://imagej.net/software/fiji/
FlowJo	BD Biosciences	https://www.flowjo.com/
Metascape	Metascape Team	https://metascape.org/
GraphPad Prism	GraphPad Software	https://www.graphpad.com/

Antibodies

B2M	Boster	A00456-2; BM4209; RRID: AB_3081128
β-catenin	Proteintech	Cat#66379-1-Ig; RRID: AB_2857358
Calreticulin	Boster	BM4228; RRID: Not Available
Calnexin	Boster	M03372-2; RRID: Not Available
Zombie NIR Fixable Viability Kit	Biolegend	Cat#423105; RRID: Not Available
CD45	Biolegend	Cat#157607; RRID: AB_2832553
CD3	Biolegend	Cat#100217; RRID: AB_1595597
IFNγ	Biolegend	Cat#505829; RRID: AB_10897937
CD8	Biolegend	Cat#100751; RRID: AB_2561389
H2D/H2K	Biolegend	Cat#114615; RRID: AB_2750195
HLA-ABC	Selleck	F2095; RRID: Not Available
GPX4	Proteintech	Cat#30388-1-AP; RRID: AB_3086304
FTH1	Proteintech	Cat#11682-1-AP; RRID: AB_3669131
GPX3	Proteintech	Cat#13947-1-AP; RRID: AB_3085426
RND2	Proteintech	Cat#13844-1-AP; RRID: AB_2181817
CAMK2N2	Sigma	SAB4300530; RRID: AB_10626632
TRIM36	Sigma	SAB2106623; RRID: AB_268447
ACTB	Proteintech	Cat#20536-1-AP; RRID: AB_10700003
Secondary antibodies	Servicebio	GB23303; RRID: AB_2811189
Cy3-conjugated goat anti-rabbit IgG	Servicebio	GB21303; RRID: AB_2861435
FITC-conjugated donkey anti-mouse IgG	Servicebio	GB22401; RRID: Not Available
SIINFEKL-Kb antibody	Biolegend	Cat#141603; RRID: AB_10897938

Experimental models: Cell lines

Mouse SCLC cell line	Professor Hongbin⁴⁴	https://pubmed.ncbi.nlm.nih.gov/34373217/
Human SCLC cell line H1688	Cell Bank of the Chinese Academy of Sciences	https://www.cellbank.org.cn/
Human SCLC cell line DMS114	Cell Bank of the Chinese Academy of Sciences	https://www.cellbank.org.cn/
HEK-293T	Cell Bank of the Chinese Academy of Sciences	https://www.cellbank.org.cn/

Experimental models: Organisms/strains

Mouse (C57BL/6J)	Shulaibao Biotechnology	RRID: IMSR_JAX:000664
Mouse (OT-I transgenic)	Shulaibao Biotechnology	RRID: IMSR_JAX:003831
Mouse (Rb1^fl/fl Trp53^fl/fl)	Professor Hongbin⁴⁴	https://pubmed.ncbi.nlm.nih.gov/34373217/

Reagent

RSL3 (Ferroptosis inducer)	MedChem Express	Cat# HY-100218A
Erastin (Ferroptosis inducer)	MedChem Express	Cat# HY-15763
Fer-1 (Ferroptosis inhibitor)	AbMole	Cat# M2698
PI (Propidium Iodide)	MedChem Express	Cat# HY-D0815
BODIPY 581/591 C11 (Lipid peroxidation probe)	MedChem Express	Cat# HY-D1301
RPMI-1640 Medium	Gibco	Cat# 11875093
Fetal Bovine Serum (FBS)	Gibco	Cat# 10099141
TUNEL Staining Kit	Beyotime	Cat# C1088
IFNγ ELISA Kit	Biolegend	Cat# 430804
TNFα ELISA Kit	Biolegend	Cat# 430904

Open in a new tab

Experimental model and study participant details

Animal models

Male C57BL/6 mice (6–8 weeks old) were used for subcutaneous tumor implantation and CD8⁺ T cell isolation, with OT-I transgenic mice (C57BL/6 background) employed for OVA-specific T cell experiments; all mice were purchased from Shulaibao Biotechnology Co. Ltd. and maintained under specific pathogen-free (SPF) conditions with a temperature of 22 ± 2°C, relative humidity of 50 ± 5%, and a 12 h light/dark cycle.

Ethical approval and regulatory compliance

All animal experiments were approved by the Ethics Committee of Huazhong University of Science and Technology (IACUC Number: 4847) and conducted in accordance with the Guide for the Care and Use of Laboratory Animals; publicly available human datasets were analyzed following the respective data access regulations and ethical guidelines of the GEO repository and original study publications.Influence of sex, gender on results.

Only male C57BL/6 mice were used in this study to avoid potential confounding factors from sex differences in immune responses and tumor progression; the exclusive use of male mice may limit the generalizability of the results to female individuals, which is acknowledged as a limitation of the study.

Additionally, when interpreting the association between sex and SCLC-related results, the experimental models of the publicly available datasets used should be considered: For GEO: GSE60052,¹⁵ GEO: GSE149507,¹⁶ and George’s cohort¹⁸ (experimental models: human SCLC clinical sample cohorts), although patient sex information was recorded, these datasets lacked sex-matched control designs or sex-stratified analytical frameworks (e.g., unbalanced sex distribution or no adjustment for confounding clinical factors like smoking history). Thus, it is not feasible to clarify the independent effect of sex on SCLC molecular features (e.g., gene expression profiles) based on these datasets. For Chan et al. (2021)¹⁷ dataset (experimental model: multi-lesion SCLC clinical samples), patient sex information was not documented, precluding the analysis of sex-related associations with SCLC heterogeneity or microenvironmental characteristics. These dataset-related limitations further highlight the need for future studies with sex-stratified experimental models (e.g., sex-matched clinical cohorts and both male/female mouse models) to clarify the role of sex in SCLC.

Cell lines

Mouse primary SCLC cell line was developed as follows: it was initially derived from de novo tumors of Rb1^fl/fl Trp53^fl/fl mice with a C57BL/6 background, which were kindly provided by Professor Hongbin Ji and were maintained under SPF conditions (22 ± 2°C, 50 ± 5% humidity, 12 h light/dark cycle). Tumors were induced in the mice via intratracheal administration of adenovirus expressing Cre recombinase at a dose of 1 × 10⁸ PFU per mouse.⁴⁵ Primary tumor tissues were subsequently dissected, minced, and digested to prepare single-cell suspensions. The primary tumor-derived cells were then subjected to immortalization through serial passage combined with selection in serum-free medium to enrich immortalized clones, thereby establishing the murine SCLC cell line RP1.

Human SCLC-A cell lines (H1688, DMS114) were purchased from the Cell Bank of the Chinese Academy of Sciences (https://www.cellbank.org.cn/). All cells were cultured in RPMI-1640 medium (Gibco) supplemented with 10% fetal bovine serum (FBS, Gibco) and maintained in a 37°C incubator with 5% CO₂.

The human SCLC cell lines H1688, DMS114 and mouse SCLC cell line RP1 were selected to construct cell lines with stable knockdown of CAMK2N2 (shCAMK2N2) and TRIM36 (shTRIM36) via lentiviral transfection. The short hairpin RNA (shRNA) sequences were cloned into the pLKO.1 lentiviral vector. Meanwhile, negative control (NC) and an untreated group (untreated) were set up. Stable cell lines were established following 2 weeks of selection with 1 μg/mL puromycin-containing medium. The target gene coding sequences of CAMK2N2 and TRIM36 were cloned into the pCDH/hygro lentiviral vector, and the plasmid was obtained through transformation, plating, single colony picking, and sequencing. Viruses were packaged using 293T cells, and the viral supernatant was used to infect CAMK2N2- and TRIM36-knockdown RP1 cells to construct the shCAMK2N2+rCAMK2N2 and shTRIM36+rTRIM36 cell lines. The expression levels of CAMK2N2 and TRIM36 were verified by qRT-PCR, and the primers are listed in Table S1.

Ethics approval and consent to participate

This research was approved by the Ethics Committee of Huazhong University of Science and Technology (IACUC Number: 4847).

Method details

Data sources

Gene expression datasets were retrieved from the Gene Expression Omnibus (GEO) (GSE60052 ¹⁵ and GSE149507 ¹⁶), George’s cohort¹⁸ and Chan et al.¹⁷ The detailed information of the three datasets is listed in Table S1.

These four public datasets all focus on human SCLC research. Their grouping methods and criteria can be uniformly summarized based on tissue origin and disease status: samples were classified into SCLC tumor tissues and normal lung tissues, where normal lung tissues included paired, unpaired, and adjacent normal samples. Each dataset was further subjected to stratified and grouped analysis integrating multiple factors, including patients’ baseline clinical characteristics (age, smoking history, TNM stage, treatment status), tumor molecular characteristics (SCLC-A, SCLC-N, SCLC-P subtypes), experimental detection techniques (RNA-seq, microarray, WGS, scRNA-seq, MIBI), and core molecular markers (SRSF1, PLCG2, ASCL1, NEUROD1, POU2F3). The core grouping criteria covered histopathological type, disease status, clinical features, molecular phenotype, and detection technology. Sample and data classification and integration were performed to address research objectives focused on SCLC genomic characteristics, molecular subtypes, tumor heterogeneity, tumor microenvironment, clinical prognosis, and therapeutic targets.

GEO: GSE60052 ¹⁵ is a human sample dataset released publicly in 2016, comprising 86 cases in total: 79 SCLC tumor samples and 7 normal lung tissue samples. RNA-seq was used to analyze the association between SRSF1 amplification or overexpression and SCLC prognosis or chemotherapy sensitivity. Notably, this dataset does not record baseline patient information such as age, gender, or smoking history; GEO: GSE149507 ¹⁶ is a paired human sample dataset released in 2020, including 36 cases: 18 SCLC tumor samples and 18 matched adjacent normal lung tissue samples. Microarray technology was applied to explore gene expression differences between SCLC tumor tissues and adjacent normal tissues. This dataset also contains baseline patient data, including gender, smoking status, and treatment history; George’s cohort¹⁸ consists of 152 human SCLC clinical specimens, including 110 SCLC tumor samples each paired with normal tissues. Whole-genome and transcriptome sequencing were used to characterize the genomic features and molecular subtypes of SCLC. Available baseline patient information includes gender, smoking history, and TNM stage; The study by Chan et al.¹⁷ enrolled 21 SCLC samples (including primary tumors, regional lymph node metastases, and distant metastases), together with 24 lung adenocarcinoma samples and 4 adjacent normal lung tissue samples. Technologies including scRNA-seq were used to reveal SCLC heterogeneity, subtype characteristics, and prognostic biomarkers. However, patient gender information was not specified in this study.

Differential analysis and functional enrichment analysis

Differential analysis of gene expression between tumor and normal groups was performed based on GEO: GSE60052 ¹⁵ RNA-seq data. The “limma” package in R was used to analyze the differences in gene expression between the tumor group and the normal group. The screening criteria for DEGs were set as |log2 Fold change (FC)| > 1 and p < 0.05. The “ggrepel” package was employed to generate volcano plots of DEGs. The R package “clusterProfiler” was used to perform GO and KEGG analyses on the DEGs, conduct functional annotation of the gene sets, and infer the main functions of the genes. The gene lists and the classification have been provided in the “Cluster” and “Expression profiles” sections of Supplementary Material 1.

Gene set enrichment analysis was performed using the online tool Metascape. Firstly, a list containing 2902 human Entrez gene IDs was submitted to Metascape. Subsequently, the tool conducted enrichment analysis to identify enriched ontology clusters, which were visualized through bar charts (showing the -log₁₀(p) values for each enriched term), network diagrams colored by cluster ID, and network diagrams colored by p-values. In addition, protein-protein interaction (PPI) network analysis was carried out to identify and visualize MCODE components within the PPI network. Finally, biological interpretations of the PPI network and MCODE components were obtained, providing annotations related to biological processes and pathways.

GSEA

GEO: GSE60052 ¹⁵ was applied for GSEA. GSEA was used to explore the enriched biological pathways and functions related to APM and ferroptosis in SCLC, so as to determine the expression differences of predefined gene sets under different biological states.

WGCNA

WGCNA was performed on the gene expression data of GEO: GSE60052.¹⁵ Firstly, outliers were excluded through sample clustering. Then, an appropriate soft threshold was selected to construct a scale-free network, and hierarchical clustering of genes was conducted using the dynamic pruning method to divide modules. Module membership and gene significance were calculated to screen for modules significantly associated with APM and ferroptosis. The hub genes within these modules were identified based on the intramodular connectivity. The genes within these modules were included in Supplementary Material 1.

Screening, analysis, and validation of key genes

The diagnostic efficacy of key genes was initially evaluated using GEO: GSE60052 ¹⁵ and further validated with GEO: GSE149507 ¹⁶ microarray data. Key genes were screened from DEGs and WGCNA module genes using the Support Vector Machine algorithm. The SVM-recursive feature elimination method was applied to optimize the feature set, and top-ranked genes were selected. Expression analysis was performed on the screened key genes. The R package “pROC” was used to draw ROC curves for evaluating their diagnostic performance, and AUC values were calculated to measure the sensitivity and specificity in distinguishing tumor samples from normal samples. Meanwhile, an external dataset (GEO: GSE149507 ¹⁶) was used for validation, and the expression patterns and diagnostic efficacy of key genes were compared to ensure the reliability of the results.

Survival analysis of key genes and their correlation with immune cells or APM genes

The data used for survival analysis is sourced from the George’s cohort.¹⁸ Patients were divided into high-expression and low-expression groups according to the expression levels of key genes. Then, the “survival” package in R language was used to conduct Kaplan-Meier survival analysis on the survival data of the two groups, draw survival curves, and compare the survival differences between the two groups through log rank test, and calculate the corresponding p value. The data used for genes’ correlation with immune cells is sourced from GEO: GSE60052.¹⁵ Correlation analyses were performed to investigate associations between immune cells and key genes. Spearman correlation coefficients were calculated to assess the relationships among immune cell subsets and between key gene expression levels and immune cell abundances, with the results visualized as heatmaps. Scatterplots were further generated to illustrate the correlation between key genes and effector immune cells, with linear regression lines, correlation coefficients, and p-values annotated to demonstrate the strength and statistical significance of the associations. The data used for genes’ correlation with APM Genes is sourced from GEO: GSE60052.¹⁵ The gene expression data and APM genes were obtained, and the average correlation between each target gene and APM genes as well as their own average expression levels were calculated. Then, a scatterplot was drawn using plotting tool “ggplot2”, with the horizontal axis representing the average correlation with APM genes, the vertical axis representing the average expression level of the gene, and each target gene was labeled. Pearson correlation analysis was used to test the correlation between key genes and APM genes, and visualization was carried out based on “ggplot2” and “cowplot”.

Single-cell landscape of key genes

The data was sourced from Chan et al.¹⁷ To characterize the cellular distribution and gene expression patterns in the SCLC microenvironment, single-cell RNA sequencing (scRNA-seq) data were analyzed as follows. For cell type clustering and UMAP visualization, after quality control of scRNA-seq data (filtering cells with mitochondrial gene ratio >20% or gene count <200), the Seurat R package was used for normalization, identification of highly variable genes, principal component analysis, and Uniform Manifold Approximation and Projection (UMAP) dimensionality reduction. Cells were clustered using the Leiden algorithm, and cell types were annotated based on canonical marker genes. The UMAP plot was generated to illustrate the spatial distribution of immune cells (T, B, Macrophages, Neutrophils, DC) and SCLC subtypes (SCLC-A, SCLC-N, SCLC-P). For key gene expression visualization on UMAP, the expression matrices of key genes (GPX3, RND2, TRIM36, CAMK2N2) were extracted from the scRNA-seq dataset. UMAP plots were generated to display the expression distribution of these genes across different cell clusters, with color gradients representing relative expression levels. We retained the SCLC patient cell populations and excluded cell types with extremely low cell counts, which were heterogeneous and could obscure the main cell type patterns, leading to confusing visualization results.

Subcutaneous tumor

Stably transfected RP1 cells (NC, shCamk2n2, shTrim36) were respectively inoculated into the right axilla of 6-8-week-old male C57BL/6 mice (1×10⁶ cells per mouse) to establish subcutaneous tumors. Tumor volume was measured every 4 days using the formula: $V = \frac{1}{2} \times L \times W^{2}$ ; where V = Tumor Volume, L = Length, and W=Width. Observations were continued for 24 days. On day 24, mice were euthanized, and tumor tissues were dissected. After weighing, part of the tissues was used for paraffin embedding, and the other part was prepared into single-cell suspensions. Then, the five groups of RP1 cells (NC + Vector, ShCamk2n2+Vector, ShCamk2n2+rCamk2n2, ShTrim36+Vector, ShTrim36+rTrim36) were respectively inoculated into the right axilla of 6-8-week-old male C57BL/6 mice (1×10⁶ cells per mouse) to establish subcutaneous tumors.

Immunofluorescence staining

After dewaxing the sections, they were blocked with goat serum, then incubated with primary antibodies against B2M (Boster: A00456-2, Servicebio: GB21303) and β-catenin (Proteintech: Cat#66379-1-Ig). Nuclei were stained with DAPI, and the expression of B2M and cell distribution were observed under a fluorescence microscope.

DMS114 cells were transfected with ShCAMK2N2 or NC, cultured for 48–72 h, then seeded on coverslips, and after adhering to the surface and growing to an appropriate density, they were fixed with 4% paraformaldehyde at room temperature for 15 min, washed 3 times with PBS (5 min each time), permeabilized with 0.2% Triton X-100 at room temperature for 10 min. After washing with PBS, the cells were blocked with 5% bovine serum albumin (BSA) for 30 min, followed by the addition of primary antibodies against Calreticulin (Boster: BM4228) and Calnexin (Boster: M03372-2) respectively, and incubated overnight at 4°C. The next day, the cells were rewarmed at room temperature for 1 h, washed with PBS, and then incubated with corresponding fluorescent secondary antibodies in the dark at room temperature for 1 h. Finally, the cells were stained with DAPI solution (blue, for nuclear labeling) at room temperature for 5 min. Images were observed and captured under a laser confocal microscope. The Coloc 2 plugin of ImageJ software was used to perform Pearson correlation coefficient analysis on the fluorescent images of Calreticulin and Calnexin to quantitatively evaluate the degree of their colocalization. All immunofluorescence experiments were independently repeated 3 times (n = 3), with at least 3 fields of view selected for statistical analysis in each replicate.

Western blot analysis

Total protein was isolated from either cultured cells or homogenized tumor tissues using ice-cold RIPA lysis buffer, which was supplemented with 1% protease inhibitor cocktail and 1% phosphatase inhibitor cocktail. After lysis on ice for 30 min, lysates were centrifuged at 12000rpm for 15 min at 4 °C, and supernatants were collected. The concentration of the extracted proteins was measured with a BCA Protein Assay Kit, strictly following the protocols provided by the manufacturer. Equal amounts of protein (20 μg per lane) were separated by 10% or 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) at 80–120 V for 60–90 min. Separated proteins were transferred onto 0.45 μm polyvinylidene difluoride (PVDF) membranes at 300 mA for 60–90 min using a wet transfer system. Membranes were blocked with 5% non-fat milk dissolved in Tris-buffered saline containing 0.1% Tween 20 (TBST) for 1 h at room temperature to reduce non-specific binding. Subsequently, the blocked membranes were incubated with primary antibodies specific to the target proteins at 4 °C overnight. After three rounds of washing with TBST to remove unbound primary antibodies, the membranes were incubated with HRP-conjugated secondary antibodies for 1 h at room temperature. Protein bands were visualized using an enhanced chemiluminescence detection system, and the intensity of each band was quantified using ImageJ software. The loading control (ACTB) and target proteins run on separate gels under identical electrophoresis and transfer conditions.

TUNEL staining

Sections were processed following the protocol of the Beyotime (C1088) kit. After inactivating endogenous enzymes with 3% hydrogen peroxide, biotin-labeled solution was added and incubated away from light. Streptavidin-HRP was used for color development. The proportion of TUNEL-positive cells was counted to evaluate cell apoptosis. The TUNEL staining experiment was independently repeated 3 times (n = 3), with 3 random fields of view counted per section in each replicate.

Flow cytometry

Tumor tissues were minced and digested with 0.2% collagenase IV (Sigma), 0.1% hyaluronidase (Sigma), and 0.01% DNase I (Roche) in RPMI-1640 medium at 37°C for 30 min with gentle shaking, followed by filtering through a 70 μm cell strainer. The suspensions were stained with fluorescent antibodies against Zombie NIR Fixable Viability Kit (Zombie NIR Kit) (Biolegend: Cat#423105), CD45 (Biolegend: Cat#157607), CD3 (Biolegend: Cat#100217), IFNγ (Biolegend: Cat#505829) and CD8 (Biolegend: Cat#100751). Flow cytometry was used to gate the CD45⁺ immune cell population, analyze the number of CD3⁺CD8⁺ T cells and IFNγ⁺CD8⁺ T cells to assess their infiltration level, and simultaneously detect the expression levels of H2D/H2K (Biolegend: Cat#114615) and B2M (Boster: A00456-2, Servicebio: GB21303) on the tumor cell surface.

Ferroptosis-related functional assays

Cell Viability Detection: DMS114 cells were treated with 1 μmol/L RSL3 (MedChem Express: HY-100218A) or 10 μmol/L Erastin (MedChem Express: HY-15763) for 6 h; stably transfected H1688 cells were treated with 5 μmol/L RSL3 or 15 μmol/L Erastin for 6 h. For each drug treatment group, a co-treatment group with 5 μmol/L ferroptosis inhibitor Fer-1 (AbMole: M2698) was set up, with the treatment duration consistently 6 h. Then, 5 μg/mL PI staining solution (MedChem Express: HY-D0815) was added, followed by photography using a fluorescence microscope. The CCK-8 assay was used to detect cell viability at different time points, with absorbance measured at 450 nm using a microplate reader. Cell viability detection (PI staining and CCK-8 assay) was independently repeated 3 times (n = 3).

Lipid Peroxidation Detection: After treatment with RSL3, cells were incubated in HBSS containing BODIPY 581/591 C11 (MedChem Express: HY-D1301). Flow cytometry was used to detect the level of lipid peroxides. Lipid peroxidation detection by flow cytometry was independently repeated 3 times (n = 3).

Colony Formation Assay: Cells were seeded in 24-well plates and treated with RSL3/Erastin alone or RSL3/Erastin combined with Fer-1 for 6 h. The drug-containing medium was aspirated, and cells were washed twice with PBS, followed by the addition of drug-free complete medium. The medium was changed every 3–4 days to ensure cell survival. Colony numbers were counted after 10 days. The colony formation assay was independently repeated 3 times (n = 3).

Detection of APM genes and ferroptosis-related genes

qRT-PCR: Total RNA was extracted using TRIzol reagent (Invitrogen), and RNA purity was verified by A260/A280 ratio (1.8–2.0). After reverse transcription, the mRNA expressions of H2D, H2K, B2M, TAP1, PSMB8, PSMB9, and NLRC5 were detected, with ACTB as the internal reference. qRT-PCR experiments were independently repeated 3 times (n = 3).

WB: Total proteins were extracted using RIPA lysis buffer (Beyotime) supplemented with 1% protease inhibitor cocktail (Roche). Protein concentration was measured via BCA assay (Thermo Fisher). The protein levels of HLA-ABC (Selleck: F2095), B2M (Boster: BM4209), GPX4 (Proteintech: Cat#30388-1-AP), FTH1 (Proteintech: Cat#11682-1-AP), GPX3 (Proteintech: Cat#13947-1-AP), RND2 (Proteintech: Cat#13844-1-AP), CAMK2N2 (Sigma, SAB4300530) and TRIM36 (Sigma, SAB2106623) were detected, with ACTB (Proteintech: Cat#20536-1-AP) as the internal reference. WB experiments were independently repeated 3 times (n = 3).

Flow cytometry: Stably transfected RP1 cells were stained with fluorescently labeled antibodies against H2D/H2K (Biolegend: Cat#114615) and B2M (Boster: A00456-2, Servicebio: GB21303) to detect the expression of APM genes on the cell membrane surface. Flow cytometry detection for APM genes was independently repeated 3 times (n = 3).

OVA antigen loading and CD8⁺ T cell coculture assay

Logarithmically growing RP1 cells were adjusted to a concentration of 1×10⁶ cells/mL with RPMI-1640 medium containing 10% fetal bovine serum, seeded into 6-well plates, and incubated at 37°C with 5% CO₂ for 12 h. Subsequently, OVA protein (chicken ovalbumin) was added to a final concentration of 10 μg/mL, and the cells were cultured for another 24h to complete antigen loading. Control group cells were treated with an equal volume of medium only. 6-8-week-old C57BL/6 background OT-I transgenic mice were sacrificed by cervical dislocation. Spleens and inguinal lymph nodes were aseptically isolated, ground, and passed through a 70 μm cell strainer to obtain single-cell suspensions. CD8⁺ T cells were purified using a CD8⁺ T cell magnetic bead sorting kit (purchased from Miltenyi Biotec) according to the manufacturer’s instructions and reserved for subsequent experiments. OVA-loaded SCLC cells (Untreated, NC, ShCAMK2N2, ShTRIM36) were seeded into 24-well plates at 5×10⁴ cells/well. After the cells adhered to the plate, purified OT-I CD8⁺ T cells were added at ratios of T cells: tumor cells = 2:1, 4:1, and 8:1, respectively, with a total volume of 1 mL per coculture system. The cocultures were incubated at 37°C with 5% CO₂ for 48 h. After coculture, tumor cells were collected, washed twice with PBS, and incubated with a fluorescently labeled SIINFEKL-Kb specific antibody (antibody against the complex of OVA-derived peptide and MHC-I molecule) at 4°C in the dark for 30 min. After washing with PBS, the expression level of SIINFEKL-Kb on the cell surface was detected by flow cytometry, and the antigen presentation efficiency was quantified by MFI. The coculture supernatants were collected, and the secretion levels of IFNγ and TNFα were detected by ELISA. The absorbance was measured at 450 nm strictly according to the kit instructions, and the cytokine concentrations were calculated based on the standard curve. The co-culture assay and subsequent ELISA, flow cytometry detection were independently repeated 3 times (n = 3).

Quantification and statistical analysis

All statistical analyses were performed using R software (version 4.4.3). For screening DEGs, the criteria were set as |log2 FC| > 1 and p < 0.05, where the p-values were obtained via linear model testing using the limma package. The statistical significance of GO and KEGG enrichment analyses was determined by adjusted p < 0.05. Survival analysis was conducted using the Kaplan-Meier method to generate survival curves, and differences in survival between groups were compared using the log rank test. Correlations between genes were analyzed using Pearson correlation analysis, while correlations between immune cells and between genes and immune cells were analyzed using Spearman correlation analysis, with p < 0.05 considered statistically significant. For evaluating diagnostic efficacy, ROC curves were plotted using the R package “pROC” to calculate AUC values, and the DeLong test was used to compare differences in AUC among different models. Data from cell experiments (e.g., CCK-8, colony formation assays) were expressed as mean ± SD from 3 independent experiments (n = 3). For animal experiments, n refers to the number of mice per group (n = 5 per group). Statistical significance: n.s: no statistical significance, ∗p < 0.05, ∗∗p < 0.01 and ∗∗∗p < 0.001. Cell and animal experiments data were analyzed using GraphPad Prism software. Multiple group comparisons were performed using One-way ANOVA. For pairwise comparisons, the Bonferroni test was used to correct for multiple comparisons.

Published: March 11, 2026

Footnotes

Supplemental information can be found online at https://doi.org/10.1016/j.isci.2026.115310.

Contributor Information

Zhongyuan Yin, Email: yzyunion@163.com.

Jun Nie, Email: jun_nie@aliyun.com.

Supplemental information

Document S1. Figures S1–S6

mmc1.pdf^{(32.9MB, pdf)}

Data S1. Key datasets and supplementary information for SCLC ferroptosis and antigen presentation analysis

mmc2.pdf^{(7.2MB, pdf)}

Table S1. Key information related to the study, including qRT-PCR primers, public datasets for bioinformatics analysis, antigen presentation and ferroptosis-related genes, differentially expressed genes in SCLC, WGCNA brown module genes, single-cell clustering results, and expression profiles of key genes

(A) qRT-PCR primers used in the study. (B) Public datasets for bioinformatics analysis and their detailed information. (C) Antigen presentation-related genes. (D) Ferroptosis-related genes. (E) Differentially expressed genes in SCLC. (F) WGCNA brown module genes. (G) Single-cell clustering results. (H) Expression profiles of key genes.

mmc3.xlsx^{(32.5MB, xlsx)}

References

1.Khan R., Coleman N. Challenges and opportunities in the immunotherapy era: balancing expectations with hope in small-cell lung cancer. Ther. Adv. Med. Oncol. 2024;16 doi: 10.1177/17588359241249627. [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Dai E., Han L., Liu J., Xie Y., Zeh H.J., Kang R., Bai L., Tang D. Ferroptotic damage promotes pancreatic tumorigenesis through a TMEM173/STING-dependent DNA sensor pathway. Nat. Commun. 2020;11:6339. doi: 10.1038/s41467-020-20154-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Dixon S.J., Lemberg K.M., Lamprecht M.R., Skouta R., Zaitsev E.M., Gleason C.E., Patel D.N., Bauer A.J., Cantley A.M., Yang W.S., et al. Ferroptosis: an iron-dependent form of nonapoptotic cell death. Cell. 2012;149:1060–1072. doi: 10.1016/j.cell.2012.03.042. [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Liu X., He J., Ying H., Chen C., Zheng C., Luo P., Zhu W., Wei T., Tang B., Zhang J. Targeting PFKFB4 Biomimetic Codelivery System Synergistically Enhances Ferroptosis to Suppress Small Cell Lung Cancer and Augments the Efficacy of Anti-PD-L1 Immunotherapy. Adv. Sci. 2025;12 doi: 10.1002/advs.202417374. [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Lei G., Zhuang L., Gan B. Targeting ferroptosis as a vulnerability in cancer. Nat. Rev. Cancer. 2022;22:381–396. doi: 10.1038/s41568-022-00459-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Chen H., Deng C., Gao J., Wang J., Fu F., Wang Y., Wang Q., Zhang M., Zhang S., Fan F., et al. Integrative spatial analysis reveals tumor heterogeneity and immune colony niche related to clinical outcomes in small cell lung cancer. Cancer Cell. 2025;43:519–536.e5. doi: 10.1016/j.ccell.2025.01.012. [DOI] [PubMed] [Google Scholar]
7.Hinshaw D.C., Shevde L.A. The Tumor Microenvironment Innately Modulates Cancer Progression. Cancer Res. 2019;79:4557–4566. doi: 10.1158/0008-5472.Can-18-3962. [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Zheng Y., Sun L., Guo J., Ma J. The crosstalk between ferroptosis and anti-tumor immunity in the tumor microenvironment: molecular mechanisms and therapeutic controversy. Cancer Commun. 2023;43:1071–1096. doi: 10.1002/cac2.12487. [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Efimova I., Catanzaro E., Van der Meeren L., Turubanova V.D., Hammad H., Mishchenko T.A., Vedunova M.V., Fimognari C., Bachert C., Coppieters F., et al. Vaccination with early ferroptotic cancer cells induces efficient antitumor immunity. J. Immunother. Cancer. 2020;8 doi: 10.1136/jitc-2020-001369. [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Meng J., Yang X., Huang J., Tuo Z., Hu Y., Liao Z., Tian Y., Deng S., Deng Y., Zhou Z., et al. Ferroptosis-Enhanced Immunotherapy with an Injectable Dextran-Chitosan Hydrogel for the Treatment of Malignant Ascites in Hepatocellular Carcinoma. Adv. Sci. 2023;10 doi: 10.1002/advs.202300517. [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Wiernicki B., Maschalidi S., Pinney J., Adjemian S., Vanden Berghe T., Ravichandran K.S., Vandenabeele P. Cancer cells dying from ferroptosis impede dendritic cell-mediated anti-tumor immunity. Nat. Commun. 2022;13:3676. doi: 10.1038/s41467-022-31218-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Cui Y., Qiu T., Wang J., Liu X., Luo L., Yin J., Zhi X., Wang W., Feng G., Wu C., et al. IFITM3 enhances immunosensitivity via MHC-I regulation and is associated with the efficacy of anti-PD-1/-L1 therapy in SCLC. Mol. Cancer. 2025;24:187. doi: 10.1186/s12943-025-02383-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Liang J.L., Jin X.K., Zhang S.M., Huang Q.X., Ji P., Deng X.C., Cheng S.X., Chen W.H., Zhang X.Z. Specific activation of cGAS-STING pathway by nanotherapeutics-mediated ferroptosis evoked endogenous signaling for boosting systemic tumor immunotherapy. Sci. Bull. 2023;68:622–636. doi: 10.1016/j.scib.2023.02.027. [DOI] [PubMed] [Google Scholar]
14.Cao W., Zhang X., Feng Y., Li R., Lu A., Li Z., Yu F., Sun L., Wang J., Wang Z., He H. Lipid Nanoparticular Codelivery System for Enhanced Antitumor Effects by Ferroptosis-Apoptosis Synergistic with Programmed Cell Death-Ligand 1 Downregulation. ACS Nano. 2024;18:17267–17281. doi: 10.1021/acsnano.4c04901. [DOI] [PubMed] [Google Scholar]
15.Jiang L., Huang J., Higgs B.W., Hu Z., Xiao Z., Yao X., Conley S., Zhong H., Liu Z., Brohawn P., et al. Genomic Landscape Survey Identifies SRSF1 as a Key Oncodriver in Small Cell Lung Cancer. PLoS Genet. 2016;12 doi: 10.1371/journal.pgen.1005895. [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Cai L., Liu H., Huang F., Fujimoto J., Girard L., Chen J., Li Y., Zhang Y.A., Deb D., Stastny V., et al. Cell-autonomous immune gene expression is repressed in pulmonary neuroendocrine cells and small cell lung cancer. Commun. Biol. 2021;4:314. doi: 10.1038/s42003-021-01842-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Chan J.M., Quintanal-Villalonga Á., Gao V.R., Xie Y., Allaj V., Chaudhary O., Masilionis I., Egger J., Chow A., Walle T., et al. Signatures of plasticity, metastasis, and immunosuppression in an atlas of human small cell lung cancer. Cancer Cell. 2021;39:1479–1496.e18. doi: 10.1016/j.ccell.2021.09.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
18.George J., Lim J.S., Jang S.J., Cun Y., Ozretić L., Kong G., Leenders F., Lu X., Fernández-Cuesta L., Bosco G., et al. Comprehensive genomic profiles of small cell lung cancer. Nature. 2015;524:47–53. doi: 10.1038/nature14664. [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Zhang X., Tang B., Luo J., Yang Y., Weng Q., Fang S., Zhao Z., Tu J., Chen M., Ji J. Cuproptosis, ferroptosis and PANoptosis in tumor immune microenvironment remodeling and immunotherapy: culprits or new hope. Mol. Cancer. 2024;23:255. doi: 10.1186/s12943-024-02130-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Lei G., Zhuang L., Gan B. The roles of ferroptosis in cancer: Tumor suppression, tumor microenvironment, and therapeutic interventions. Cancer Cell. 2024;42:513–534. doi: 10.1016/j.ccell.2024.03.011. [DOI] [PubMed] [Google Scholar]
21.Li Y., Yu J., Li R., Zhou H., Chang X. New insights into the role of mitochondrial metabolic dysregulation and immune infiltration in septic cardiomyopathy by integrated bioinformatics analysis and experimental validation. Cell. Mol. Biol. Lett. 2024;29:21. doi: 10.1186/s11658-024-00536-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Liu X., Yan C., Chang C., Meng F., Shen W., Wang S., Zhang Y. FOXA2 Suppression by TRIM36 Exerts Anti-Tumor Role in Colorectal Cancer Via Inducing NRF2/GPX4-Regulated Ferroptosis. Adv. Sci. 2023;10 doi: 10.1002/advs.202304521. [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Tong Q., Yi M., Kong P., Xu L., Huang W., Niu Y., Gan X., Zhan H., Tian R., Yan D. TRIM36 inhibits tumorigenesis through the Wnt/β-catenin pathway and promotes caspase-dependent apoptosis in hepatocellular carcinoma. Cancer Cell Int. 2022;22:278. doi: 10.1186/s12935-022-02692-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Zhao B., Qiao G., Li J., Wang Y., Li X., Zhang H., Zhang L. TRIM36 suppresses cell growth and promotes apoptosis in human esophageal squamous cell carcinoma cells by inhibiting Wnt/β-catenin signaling pathway. Hum. Cell. 2022;35:1487–1498. doi: 10.1007/s13577-022-00737-x. [DOI] [PubMed] [Google Scholar]
25.Zhao X., Zhou T., Wang Y., Bao M., Ni C., Ding L., Sun S., Dong H., Li J., Liang C. Trigred motif 36 regulates neuroendocrine differentiation of prostate cancer via HK2 ubiquitination and GPx4 deficiency. Cancer Sci. 2023;114:2445–2459. doi: 10.1111/cas.15763. [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Zhang J., Zou B., Geng Y., Yin H., Qin B., Gao W., Lin X., Sun N. TRIM36 serves as a prognostic indicator linked to immune infiltration in KIRC. Heliyon. 2025;11 doi: 10.1016/j.heliyon.2025.e42540. [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Chen M., Shi J., Liu T., Liu J., Liu Y., Li J., Luo Y., Luo J., Li X., Gong H., Fan X. Astragaloside IV ameliorates autism-like behaviors in BTBR mice by modulating Camk2n2-dependent OXPHOS and neurotransmission in the mPFC. J. Adv. Res. 2025;77:763–777. doi: 10.1016/j.jare.2025.01.030. [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Zhao Y.Y., Lian J.X., Lan Z., Zou K.L., Wang W.M., Yu G.T. Ferroptosis promotes anti-tumor immune response by inducing immunogenic exposure in HNSCC. Oral Dis. 2023;29:933–941. doi: 10.1111/odi.14077. [DOI] [PubMed] [Google Scholar]
29.Michalak M., Groenendyk J., Szabó É., Gold L.I., Opas M.J.T.B.j. Calreticulin, a multi-process calcium-buffering chaperone of the endoplasmic reticulum. Biochem. J. 2009;417:651–666. doi: 10.1042/BJ20081847. [DOI] [PubMed] [Google Scholar]
30.Li J., Marionneau C., Zhang R., Shah V., Hell J.W., Nerbonne J.M., Anderson M.E. Calmodulin kinase II inhibition shortens action potential duration by upregulation of K+ currents. Circ. Res. 2006;99:1092–1099. doi: 10.1161/01.RES.0000249369.71709.5c. [DOI] [PubMed] [Google Scholar]
31.Li T., Qiao T. Unraveling tumor microenvironment of small-cell lung cancer: Implications for immunotherapy. Semin. Cancer Biol. 2022;86:117–125. doi: 10.1016/j.semcancer.2022.09.005. [DOI] [PubMed] [Google Scholar]
32.Rudin C.M., Balli D., Lai W.V., Richards A.L., Nguyen E., Egger J.V., Choudhury N.J., Sen T., Chow A., Poirier J.T., et al. Clinical Benefit From Immunotherapy in Patients With SCLC Is Associated With Tumor Capacity for Antigen Presentation. J. Thorac. Oncol. 2023;18:1222–1232. doi: 10.1016/j.jtho.2023.05.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Zugazagoitia J., Osma H., Baena J., Ucero A.C., Paz-Ares L. Facts and Hopes on Cancer Immunotherapy for Small Cell Lung Cancer. Clin. Cancer Res. 2024;30:2872–2883. doi: 10.1158/1078-0432.Ccr-23-1159. [DOI] [PubMed] [Google Scholar]
34.Dora D., Rivard C., Yu H., Pickard S.L., Laszlo V., Harko T., Megyesfalvi Z., Gerdan C., Dinya E., Hoetzenecker K., et al. Protein Expression of immune checkpoints STING and MHCII in small cell lung cancer. Cancer Immunol. Immunother. 2023;72:561–578. doi: 10.1007/s00262-022-03270-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Nguyen E.M., Taniguchi H., Chan J.M., Zhan Y.A., Chen X., Qiu J., de Stanchina E., Allaj V., Shah N.S., Uddin F., et al. Targeting Lysine-Specific Demethylase 1 Rescues Major Histocompatibility Complex Class I Antigen Presentation and Overcomes Programmed Death-Ligand 1 Blockade Resistance in SCLC. J. Thorac. Oncol. 2022;17:1014–1031. doi: 10.1016/j.jtho.2022.05.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Zhan L., Zhang J., Zhang J., Liu X., Zhu S., Shi Y., He Y., Wang W., Fan Y., Tang Z., et al. LC3 and NLRC5 interaction inhibits NLRC5-mediated MHC class I antigen presentation pathway in endometrial cancer. Cancer Lett. 2022;529:37–52. doi: 10.1016/j.canlet.2021.12.031. [DOI] [PubMed] [Google Scholar]
37.Chen Y., Li H., Fan Y. Shaping the tumor immune microenvironment of SCLC: Mechanisms, and opportunities for immunotherapy. Cancer Treat Rev. 2023;120 doi: 10.1016/j.ctrv.2023.102606. [DOI] [PubMed] [Google Scholar]
38.Tong Q., Li D., Yin Y., Cheng L., Ouyang S. GBP5 Expression Predicted Prognosis of Immune Checkpoint Inhibitors in Small Cell Lung Cancer and Correlated with Tumor Immune Microenvironment. J. Inflamm. Res. 2023;16:4153–4164. doi: 10.2147/jir.S401430. [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Desai P., Takahashi N., Kumar R., Nichols S., Malin J., Hunt A., Schultz C., Cao Y., Tillo D., Nousome D., et al. Microenvironment shapes small-cell lung cancer neuroendocrine states and presents therapeutic opportunities. Cell Rep. Med. 2024;5 doi: 10.1016/j.xcrm.2024.101610. [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Jin Y., Wu Y., Reuben A., Zhu L., Gay C.M., Wu Q., Zhou X., Mo H., Zheng Q., Ren J., et al. Single-cell and spatial proteo-transcriptomic profiling reveals immune infiltration heterogeneity associated with neuroendocrine features in small cell lung cancer. Cell Discov. 2024;10:93. doi: 10.1038/s41421-024-00703-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Dora D., Rivard C., Yu H., Bunn P., Suda K., Ren S., Lueke Pickard S., Laszlo V., Harko T., Megyesfalvi Z., et al. Neuroendocrine subtypes of small cell lung cancer differ in terms of immune microenvironment and checkpoint molecule distribution. Mol. Oncol. 2020;14:1947–1965. doi: 10.1002/1878-0261.12741. [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Langfelder P., Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinf. 2008;9:559. doi: 10.1186/1471-2105-9-559. [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Schindelin J., Arganda-Carreras I., Frise E., Kaynig V., Longair M., Pietzsch T., Preibisch S., Rueden C., Saalfeld S., Schmid B., et al. Fiji: an open-source platform for biological-image analysis. Nat. Methods. 2012;9:676–682. doi: 10.1038/nmeth.2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Jin Y., Xiao T., Feng Y., Yang J., Guo C., Hu L., Ji H. A mesenchymal-like subpopulation in non-neuroendocrine SCLC contributes to metastasis. J. Genet. Genomics. 2021;48:571–581. doi: 10.1016/j.jgg.2021.05.007. [DOI] [PubMed] [Google Scholar]
45.Meuwissen R., Linn S.C., Linnoila R.I., Zevenhoven J., Mooi W.J., Berns A. Induction of small cell lung cancer by somatic inactivation of both Trp53 and Rb1 in a conditional mouse model. Cancer Cell. 2003;4:181–189. doi: 10.1016/s1535-6108(03)00220-4. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Document S1. Figures S1–S6

mmc1.pdf^{(32.9MB, pdf)}

Data S1. Key datasets and supplementary information for SCLC ferroptosis and antigen presentation analysis

mmc2.pdf^{(7.2MB, pdf)}

mmc3.xlsx^{(32.5MB, xlsx)}

Data Availability Statement

•
The datasets analyzed during the current study are available in the Gene Expression Omnibus (GEO) repository, with accession numbers GEO: GSE60052: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE60052) and GSE149507: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE149507). All datasets used in this study are fully publicly available with no access restrictions.
•
Custom code: Not applicable. No custom code was generated in this study.
•
This study did not generate new standardized datasets that require deposition in a Cell Press-approved repository.
•
Any additional information required to reanalyze the data reported in this article is available from the lead contact upon request.

[bib1] 1.Khan R., Coleman N. Challenges and opportunities in the immunotherapy era: balancing expectations with hope in small-cell lung cancer. Ther. Adv. Med. Oncol. 2024;16 doi: 10.1177/17588359241249627. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib2] 2.Dai E., Han L., Liu J., Xie Y., Zeh H.J., Kang R., Bai L., Tang D. Ferroptotic damage promotes pancreatic tumorigenesis through a TMEM173/STING-dependent DNA sensor pathway. Nat. Commun. 2020;11:6339. doi: 10.1038/s41467-020-20154-8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib3] 3.Dixon S.J., Lemberg K.M., Lamprecht M.R., Skouta R., Zaitsev E.M., Gleason C.E., Patel D.N., Bauer A.J., Cantley A.M., Yang W.S., et al. Ferroptosis: an iron-dependent form of nonapoptotic cell death. Cell. 2012;149:1060–1072. doi: 10.1016/j.cell.2012.03.042. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib4] 4.Liu X., He J., Ying H., Chen C., Zheng C., Luo P., Zhu W., Wei T., Tang B., Zhang J. Targeting PFKFB4 Biomimetic Codelivery System Synergistically Enhances Ferroptosis to Suppress Small Cell Lung Cancer and Augments the Efficacy of Anti-PD-L1 Immunotherapy. Adv. Sci. 2025;12 doi: 10.1002/advs.202417374. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib5] 5.Lei G., Zhuang L., Gan B. Targeting ferroptosis as a vulnerability in cancer. Nat. Rev. Cancer. 2022;22:381–396. doi: 10.1038/s41568-022-00459-0. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib6] 6.Chen H., Deng C., Gao J., Wang J., Fu F., Wang Y., Wang Q., Zhang M., Zhang S., Fan F., et al. Integrative spatial analysis reveals tumor heterogeneity and immune colony niche related to clinical outcomes in small cell lung cancer. Cancer Cell. 2025;43:519–536.e5. doi: 10.1016/j.ccell.2025.01.012. [DOI] [PubMed] [Google Scholar]

[bib7] 7.Hinshaw D.C., Shevde L.A. The Tumor Microenvironment Innately Modulates Cancer Progression. Cancer Res. 2019;79:4557–4566. doi: 10.1158/0008-5472.Can-18-3962. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib8] 8.Zheng Y., Sun L., Guo J., Ma J. The crosstalk between ferroptosis and anti-tumor immunity in the tumor microenvironment: molecular mechanisms and therapeutic controversy. Cancer Commun. 2023;43:1071–1096. doi: 10.1002/cac2.12487. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib9] 9.Efimova I., Catanzaro E., Van der Meeren L., Turubanova V.D., Hammad H., Mishchenko T.A., Vedunova M.V., Fimognari C., Bachert C., Coppieters F., et al. Vaccination with early ferroptotic cancer cells induces efficient antitumor immunity. J. Immunother. Cancer. 2020;8 doi: 10.1136/jitc-2020-001369. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib10] 10.Meng J., Yang X., Huang J., Tuo Z., Hu Y., Liao Z., Tian Y., Deng S., Deng Y., Zhou Z., et al. Ferroptosis-Enhanced Immunotherapy with an Injectable Dextran-Chitosan Hydrogel for the Treatment of Malignant Ascites in Hepatocellular Carcinoma. Adv. Sci. 2023;10 doi: 10.1002/advs.202300517. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib11] 11.Wiernicki B., Maschalidi S., Pinney J., Adjemian S., Vanden Berghe T., Ravichandran K.S., Vandenabeele P. Cancer cells dying from ferroptosis impede dendritic cell-mediated anti-tumor immunity. Nat. Commun. 2022;13:3676. doi: 10.1038/s41467-022-31218-2. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib12] 12.Cui Y., Qiu T., Wang J., Liu X., Luo L., Yin J., Zhi X., Wang W., Feng G., Wu C., et al. IFITM3 enhances immunosensitivity via MHC-I regulation and is associated with the efficacy of anti-PD-1/-L1 therapy in SCLC. Mol. Cancer. 2025;24:187. doi: 10.1186/s12943-025-02383-x. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib13] 13.Liang J.L., Jin X.K., Zhang S.M., Huang Q.X., Ji P., Deng X.C., Cheng S.X., Chen W.H., Zhang X.Z. Specific activation of cGAS-STING pathway by nanotherapeutics-mediated ferroptosis evoked endogenous signaling for boosting systemic tumor immunotherapy. Sci. Bull. 2023;68:622–636. doi: 10.1016/j.scib.2023.02.027. [DOI] [PubMed] [Google Scholar]

[bib14] 14.Cao W., Zhang X., Feng Y., Li R., Lu A., Li Z., Yu F., Sun L., Wang J., Wang Z., He H. Lipid Nanoparticular Codelivery System for Enhanced Antitumor Effects by Ferroptosis-Apoptosis Synergistic with Programmed Cell Death-Ligand 1 Downregulation. ACS Nano. 2024;18:17267–17281. doi: 10.1021/acsnano.4c04901. [DOI] [PubMed] [Google Scholar]

[bib15] 15.Jiang L., Huang J., Higgs B.W., Hu Z., Xiao Z., Yao X., Conley S., Zhong H., Liu Z., Brohawn P., et al. Genomic Landscape Survey Identifies SRSF1 as a Key Oncodriver in Small Cell Lung Cancer. PLoS Genet. 2016;12 doi: 10.1371/journal.pgen.1005895. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib16] 16.Cai L., Liu H., Huang F., Fujimoto J., Girard L., Chen J., Li Y., Zhang Y.A., Deb D., Stastny V., et al. Cell-autonomous immune gene expression is repressed in pulmonary neuroendocrine cells and small cell lung cancer. Commun. Biol. 2021;4:314. doi: 10.1038/s42003-021-01842-7. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib17] 17.Chan J.M., Quintanal-Villalonga Á., Gao V.R., Xie Y., Allaj V., Chaudhary O., Masilionis I., Egger J., Chow A., Walle T., et al. Signatures of plasticity, metastasis, and immunosuppression in an atlas of human small cell lung cancer. Cancer Cell. 2021;39:1479–1496.e18. doi: 10.1016/j.ccell.2021.09.008. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib18] 18.George J., Lim J.S., Jang S.J., Cun Y., Ozretić L., Kong G., Leenders F., Lu X., Fernández-Cuesta L., Bosco G., et al. Comprehensive genomic profiles of small cell lung cancer. Nature. 2015;524:47–53. doi: 10.1038/nature14664. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib19] 19.Zhang X., Tang B., Luo J., Yang Y., Weng Q., Fang S., Zhao Z., Tu J., Chen M., Ji J. Cuproptosis, ferroptosis and PANoptosis in tumor immune microenvironment remodeling and immunotherapy: culprits or new hope. Mol. Cancer. 2024;23:255. doi: 10.1186/s12943-024-02130-8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib20] 20.Lei G., Zhuang L., Gan B. The roles of ferroptosis in cancer: Tumor suppression, tumor microenvironment, and therapeutic interventions. Cancer Cell. 2024;42:513–534. doi: 10.1016/j.ccell.2024.03.011. [DOI] [PubMed] [Google Scholar]

[bib21] 21.Li Y., Yu J., Li R., Zhou H., Chang X. New insights into the role of mitochondrial metabolic dysregulation and immune infiltration in septic cardiomyopathy by integrated bioinformatics analysis and experimental validation. Cell. Mol. Biol. Lett. 2024;29:21. doi: 10.1186/s11658-024-00536-2. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib22] 22.Liu X., Yan C., Chang C., Meng F., Shen W., Wang S., Zhang Y. FOXA2 Suppression by TRIM36 Exerts Anti-Tumor Role in Colorectal Cancer Via Inducing NRF2/GPX4-Regulated Ferroptosis. Adv. Sci. 2023;10 doi: 10.1002/advs.202304521. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib23] 23.Tong Q., Yi M., Kong P., Xu L., Huang W., Niu Y., Gan X., Zhan H., Tian R., Yan D. TRIM36 inhibits tumorigenesis through the Wnt/β-catenin pathway and promotes caspase-dependent apoptosis in hepatocellular carcinoma. Cancer Cell Int. 2022;22:278. doi: 10.1186/s12935-022-02692-x. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib24] 24.Zhao B., Qiao G., Li J., Wang Y., Li X., Zhang H., Zhang L. TRIM36 suppresses cell growth and promotes apoptosis in human esophageal squamous cell carcinoma cells by inhibiting Wnt/β-catenin signaling pathway. Hum. Cell. 2022;35:1487–1498. doi: 10.1007/s13577-022-00737-x. [DOI] [PubMed] [Google Scholar]

[bib25] 25.Zhao X., Zhou T., Wang Y., Bao M., Ni C., Ding L., Sun S., Dong H., Li J., Liang C. Trigred motif 36 regulates neuroendocrine differentiation of prostate cancer via HK2 ubiquitination and GPx4 deficiency. Cancer Sci. 2023;114:2445–2459. doi: 10.1111/cas.15763. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib26] 26.Zhang J., Zou B., Geng Y., Yin H., Qin B., Gao W., Lin X., Sun N. TRIM36 serves as a prognostic indicator linked to immune infiltration in KIRC. Heliyon. 2025;11 doi: 10.1016/j.heliyon.2025.e42540. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib27] 27.Chen M., Shi J., Liu T., Liu J., Liu Y., Li J., Luo Y., Luo J., Li X., Gong H., Fan X. Astragaloside IV ameliorates autism-like behaviors in BTBR mice by modulating Camk2n2-dependent OXPHOS and neurotransmission in the mPFC. J. Adv. Res. 2025;77:763–777. doi: 10.1016/j.jare.2025.01.030. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib28] 28.Zhao Y.Y., Lian J.X., Lan Z., Zou K.L., Wang W.M., Yu G.T. Ferroptosis promotes anti-tumor immune response by inducing immunogenic exposure in HNSCC. Oral Dis. 2023;29:933–941. doi: 10.1111/odi.14077. [DOI] [PubMed] [Google Scholar]

[bib29] 29.Michalak M., Groenendyk J., Szabó É., Gold L.I., Opas M.J.T.B.j. Calreticulin, a multi-process calcium-buffering chaperone of the endoplasmic reticulum. Biochem. J. 2009;417:651–666. doi: 10.1042/BJ20081847. [DOI] [PubMed] [Google Scholar]

[bib30] 30.Li J., Marionneau C., Zhang R., Shah V., Hell J.W., Nerbonne J.M., Anderson M.E. Calmodulin kinase II inhibition shortens action potential duration by upregulation of K+ currents. Circ. Res. 2006;99:1092–1099. doi: 10.1161/01.RES.0000249369.71709.5c. [DOI] [PubMed] [Google Scholar]

[bib31] 31.Li T., Qiao T. Unraveling tumor microenvironment of small-cell lung cancer: Implications for immunotherapy. Semin. Cancer Biol. 2022;86:117–125. doi: 10.1016/j.semcancer.2022.09.005. [DOI] [PubMed] [Google Scholar]

[bib32] 32.Rudin C.M., Balli D., Lai W.V., Richards A.L., Nguyen E., Egger J.V., Choudhury N.J., Sen T., Chow A., Poirier J.T., et al. Clinical Benefit From Immunotherapy in Patients With SCLC Is Associated With Tumor Capacity for Antigen Presentation. J. Thorac. Oncol. 2023;18:1222–1232. doi: 10.1016/j.jtho.2023.05.008. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib33] 33.Zugazagoitia J., Osma H., Baena J., Ucero A.C., Paz-Ares L. Facts and Hopes on Cancer Immunotherapy for Small Cell Lung Cancer. Clin. Cancer Res. 2024;30:2872–2883. doi: 10.1158/1078-0432.Ccr-23-1159. [DOI] [PubMed] [Google Scholar]

[bib34] 34.Dora D., Rivard C., Yu H., Pickard S.L., Laszlo V., Harko T., Megyesfalvi Z., Gerdan C., Dinya E., Hoetzenecker K., et al. Protein Expression of immune checkpoints STING and MHCII in small cell lung cancer. Cancer Immunol. Immunother. 2023;72:561–578. doi: 10.1007/s00262-022-03270-w. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib35] 35.Nguyen E.M., Taniguchi H., Chan J.M., Zhan Y.A., Chen X., Qiu J., de Stanchina E., Allaj V., Shah N.S., Uddin F., et al. Targeting Lysine-Specific Demethylase 1 Rescues Major Histocompatibility Complex Class I Antigen Presentation and Overcomes Programmed Death-Ligand 1 Blockade Resistance in SCLC. J. Thorac. Oncol. 2022;17:1014–1031. doi: 10.1016/j.jtho.2022.05.014. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib36] 36.Zhan L., Zhang J., Zhang J., Liu X., Zhu S., Shi Y., He Y., Wang W., Fan Y., Tang Z., et al. LC3 and NLRC5 interaction inhibits NLRC5-mediated MHC class I antigen presentation pathway in endometrial cancer. Cancer Lett. 2022;529:37–52. doi: 10.1016/j.canlet.2021.12.031. [DOI] [PubMed] [Google Scholar]

[bib37] 37.Chen Y., Li H., Fan Y. Shaping the tumor immune microenvironment of SCLC: Mechanisms, and opportunities for immunotherapy. Cancer Treat Rev. 2023;120 doi: 10.1016/j.ctrv.2023.102606. [DOI] [PubMed] [Google Scholar]

[bib38] 38.Tong Q., Li D., Yin Y., Cheng L., Ouyang S. GBP5 Expression Predicted Prognosis of Immune Checkpoint Inhibitors in Small Cell Lung Cancer and Correlated with Tumor Immune Microenvironment. J. Inflamm. Res. 2023;16:4153–4164. doi: 10.2147/jir.S401430. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib39] 39.Desai P., Takahashi N., Kumar R., Nichols S., Malin J., Hunt A., Schultz C., Cao Y., Tillo D., Nousome D., et al. Microenvironment shapes small-cell lung cancer neuroendocrine states and presents therapeutic opportunities. Cell Rep. Med. 2024;5 doi: 10.1016/j.xcrm.2024.101610. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib40] 40.Jin Y., Wu Y., Reuben A., Zhu L., Gay C.M., Wu Q., Zhou X., Mo H., Zheng Q., Ren J., et al. Single-cell and spatial proteo-transcriptomic profiling reveals immune infiltration heterogeneity associated with neuroendocrine features in small cell lung cancer. Cell Discov. 2024;10:93. doi: 10.1038/s41421-024-00703-x. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib41] 41.Dora D., Rivard C., Yu H., Bunn P., Suda K., Ren S., Lueke Pickard S., Laszlo V., Harko T., Megyesfalvi Z., et al. Neuroendocrine subtypes of small cell lung cancer differ in terms of immune microenvironment and checkpoint molecule distribution. Mol. Oncol. 2020;14:1947–1965. doi: 10.1002/1878-0261.12741. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib42] 42.Langfelder P., Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinf. 2008;9:559. doi: 10.1186/1471-2105-9-559. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib43] 43.Schindelin J., Arganda-Carreras I., Frise E., Kaynig V., Longair M., Pietzsch T., Preibisch S., Rueden C., Saalfeld S., Schmid B., et al. Fiji: an open-source platform for biological-image analysis. Nat. Methods. 2012;9:676–682. doi: 10.1038/nmeth.2019. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib44] 44.Jin Y., Xiao T., Feng Y., Yang J., Guo C., Hu L., Ji H. A mesenchymal-like subpopulation in non-neuroendocrine SCLC contributes to metastasis. J. Genet. Genomics. 2021;48:571–581. doi: 10.1016/j.jgg.2021.05.007. [DOI] [PubMed] [Google Scholar]

[bib45] 45.Meuwissen R., Linn S.C., Linnoila R.I., Zevenhoven J., Mooi W.J., Berns A. Induction of small cell lung cancer by somatic inactivation of both Trp53 and Rb1 in a conditional mouse model. Cancer Cell. 2003;4:181–189. doi: 10.1016/s1535-6108(03)00220-4. [DOI] [PubMed] [Google Scholar]

PERMALINK

TRIM36 and CAMK2N2 regulate ferroptosis and antigen presentation in small cell lung cancer

Yifan Cai

Shuang Zhu

Honglin Wang

Fang Huang

Zhongyuan Yin

Jun Nie

Summary

Graphical abstract

Highlights

Introduction

Results

Study design and preliminary data analysis

Figure 1.

Significant differences in biological processes and signaling pathways between SCLC and normal tissues

Figure 2.

Screening of key genes based on DEGs and WGCNA

Correlation analysis between immune cells and key genes

Figure 3.

Expression differences and diagnostic efficacy of key genes

Figure 4.

Survival analysis of patients based on key genes and their correlation with APM genes

Figure 5.

TRIM36 and CAMK2N2 promote tumor growth in vivo and inhibit the expression of MHC I

Figure 6.

TRIM36 and CAMK2N2 promote ferroptosis resistance in SCLC cells

Figure 7.

Knockdown of CAMK2N2 or TRIM36 enhances ferroptosis susceptibility via downregulating GPX4/FTH1, with CAMK2N2 disrupting ER calcium homeostasis

TRIM36 and CAMK2N2 inhibit the expression of APM genes in SCLC cells

Figure 8.

TRIM36 and CAMK2N2 suppress antigen presentation in SCLC through ferroptosis-dependent and independent mechanisms, limiting T cell effector function

Figure 9.

Discussion

Limitations of the study

Resource availability

Lead contact

Materials availability

Data and code availability

Acknowledgments

Author contributions

Declaration of interests

Declaration of generative AI and AI-assisted technologies in the writing process

STAR★Methods

Key resources table

Experimental model and study participant details

Animal models

Ethical approval and regulatory compliance

Cell lines

Ethics approval and consent to participate

Method details

Data sources

Differential analysis and functional enrichment analysis

GSEA

WGCNA

Screening, analysis, and validation of key genes

Survival analysis of key genes and their correlation with immune cells or APM genes

Single-cell landscape of key genes

Subcutaneous tumor

Immunofluorescence staining

Western blot analysis

TUNEL staining

Flow cytometry

Ferroptosis-related functional assays

Detection of APM genes and ferroptosis-related genes

OVA antigen loading and CD8+ T cell coculture assay

Quantification and statistical analysis

Footnotes

Contributor Information

Supplemental information

References

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases

OVA antigen loading and CD8⁺ T cell coculture assay