IFITM3 enhances immunosensitivity via MHC-I regulation and is associated with the efficacy of anti-PD-1/-L1 therapy in SCLC

Abstract

Background: Most small cell lung cancer (SCLC) patients exhibit resistance to immune checkpoint inhibitors (ICIs) and demonstrate downregulation of major histocompatibility complex class I (MHC-I) molecules. This study aimed to elucidate the regulatory mechanisms underlying MHC-I expression and potential combination strategies.

Methods: Single-cell and bulk RNA sequencing data from SCLC patients were analyzed. Clinical data from SCLC patients treated with PD-1/PD-L1 inhibitors were used to investigate the associations between treatment efficacy and IFITM3 expression. In vitro and in vivo functional studies were conducted to evaluate the role and mechanisms of IFITM3 in modulating tumor sensitivity to PD-1 inhibitors.

Results: Integrative analysis of multiple real-world SCLC cohorts confirmed a significant positive association between IFITM3 expression and MHCI. IFITM3 overexpression upregulated MHC-I-related genes, enriched antigen presentation pathways, and increased CD8⁺ T-cell infiltration and cytotoxicity. Elevated IFITM3 expression was significantly associated with prolonged progression-free survival (PFS) in patients receiving chemoimmunotherapy but not in those treated with chemotherapy alone. Additionally, patients with high H-scores for IFITM3, as determined by immunohistochemistry, demonstrated better clinical outcomes with chemoimmunotherapy. Inducing IFITM3 expression directly or through treatment with ethyl gallate (EG), an IFITM3 inducer, effectively sensitized tumors to PD-1 blockade in SCLC mouse models. Mechanistic studies revealed that IFITM3 upregulates NLRC5, a key transcriptional activator of MHC-I, facilitating its nuclear translocation and thereby increasing MHC-I levels.

Conclusions: IFITM3 is associated with MHC-I expression and can predict the efficacy of anti-PD-1/-L1 therapy in SCLC patients. IFITM3 inducers potently improved the efficacy of anti-PD1 monotherapy in SCLC.

Supplementary Information: The online version contains supplementary material available at 10.1186/s12943-025-02383-x.

Keywords: IFITM3; Immunotherapy; MHC-I; Small cell lung cancer.

Fig. 1

IFITM3 is correlated with MHC-I expression in SCLC. (A) UMAP plot displaying 11,718 cancer cells from single-cell RNA sequencing of nine untreated SCLC tumors. (B) Dot plot representing MHC-I gene set scores calculated via the AUCell, UCell, and singscore algorithms. The dot size indicates the percentage of cells expressing the MHC-I gene set, and the color represents the average expression level. (C) UMAP plot categorizing cancer cells on the basis of MHC-I expression levels. (D) Pseudotime trajectory initiated from MHC-I low clusters toward MHC-I high clusters. Pie charts represent the proportion of cells within each category along the trajectory. (E) Heatmap showing 627 genes whose expression significantly changed along the pseudotime trajectory. Cells were ordered by pseudotime (left to right), with gene expression scaled by Z-score (bottom color bar). Clusters represent distinct temporal gene expression patterns: Cluster 1 genes were enriched in early pseudotime states (low-MHC-I cells), and Cluster 2 genes in late pseudotime states (high-MHC-I cells). Sample size: n = 11,718 cells from 9 patients. (F) Bar plot of enriched GO terms for the Cluster 2 gene set from (E). (G) Venn diagram showing overlapping MHC-I-related genes (correlation r > 0.5, P < 0.05) across five SCLC datasets. (H) Correlation matrix depicting IFITM3 expression and key MHC-I genes across the five datasets. Spearman correlation coefficients were calculated for each gene pair. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

Fig. 2

IFITM3 regulates MHC-I expression in SCLC cell lines. (A) Relative mRNA expression of MHC-I-related genes (HLA-A, HLA-B, HLA-C, TAP1, TAP2, and B2M) in NCI-H446 and NCI-H69 cells after overexpression or knockdown of IFITM3. (B-C) Western blot analysis of MHC-I protein levels in NCI-H446 and NCI-H69 cells after overexpression or knockdown of IFITM3. (D-E) Flow cytometry analysis of MHC-I surface expression in NCI-H446 and NCI-H69 cells after overexpression or knockdown of IFITM3, with or without IFN-γ treatment. The results are presented as the means ± SEMs, n = 3. Statistical comparisons were performed using unpaired two-tailed Student’s t-tests. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

Fig. 3

Clinical correlation of IFITM3 expression with immunotherapy outcomes in SCLC. (A) Scatter plot showing the correlation between IFITM3 expression and immune-related parameters via the IPS algorithm. The x-axis represents the spearman correlation coefficient. (B) Kaplan-Meier curves of PFS for SCLC patients from the IMpower133 cohort treated with anti-PD-L1 plus chemotherapy (left) or chemotherapy alone (right), stratified by the optimal IFITM3 expression cutoff. Statistical significance was assessed using the log-rank test. (C) Kaplan-Meier survival curve for PFS in the PH cohort stratified by the optimal IFITM3 expression cutoff. Statistical significance was assessed using the log-rank test. (D) Box plot comparing IFITM3 expression between patients with PD/NE and patients with CR in the Rochester cohort. Statistical significance was assessed using the two-sided Wilcoxon rank-sum test. (E) Schematic of clinical cohort. (F) IFITM3 H-scores across patients with PD, SD, or PR. Representative IHC images show IFITM3, MHC-I, and CD8 staining in PR vs. PD tumors (scale bar: 40 μm). Statistical significance was assessed using the two-sided Wilcoxon rank-sum test. (G) Scatter plot showing correlations between IFITM3 and MHC-I, as well as between IFITM3 and CD8 H-scores. Spearman correlation was used to assess statistical significance. (H) HRs with 95% confidence intervals for PFS among patients with high versus low IFITM3 expression, defined by increasing H-score percentile cutoffs. A range of thresholds (from 15–90%) was tested in the IHC-stained clinical cohort (n = 42). HRs and P values were calculated using Cox proportional hazards regression models. (I) Kaplan-Meier survival curve for PFS among the indicated groups, with a 75% H-score cutoff for IFITM3 and a 55% cutoff for MHC-I. Survival differences were assessed using the log-rank test. *P < 0.05, **P < 0.01, n.s., not significant

Fig. 4

IFITM3 enhances PD-1 inhibitor efficacy and tumor immune responses in SCLC mouse models. (A) A schematic view of the treatment plan. C57BL/6 mice were injected with RPP control or OE-IFITM3 cells, and a PD-1 mAb or vehicle was administered when the tumors reached approximately 100 mm³. (B) Representative images and weight plots of tumors, along with tumor volume plots measured every 3 days. (C) H&E staining and IFITM3 and MHC-I IHC analyses of tumor sections from the indicated treatment groups (scale bar: 100 μm). (D) Flow cytometry analysis of CD3⁺ CD45⁺ cells, CD8⁺ CD3⁺ cells, and CD69⁺, CD44⁺, and GZMB⁺ CD8⁺ cells. (E) A schematic view of the treatment plan. BALB/c nude mice were injected with RPP-control or OE-IFITM3 cells. (F) Representative images and weight plots of tumors harvested after the mice were euthanized, along with tumor volume plots measured every 2 days. The results are presented as the means ± SEMs, n = 4. Statistical comparisons between groups were performed using unpaired two-tailed Student’s t-tests. *P < 0.05, **P < 0.01, ***P < 0.001, n.s., not significant

Fig. 5

EG enhances the efficacy of PD-1 inhibition and antitumor immune responses in SCLC mouse models. (A) Volcano plots showing differential gene expression following EG treatment in NCI-H446 and NCI-H69 cells. IFITM3 is significantly upregulated upon EG treatment. Vertical dashed lines indicate log2FC = ± 1, and the horizontal dashed line represents the significance threshold at P = 0.05. (B) GSEA plots for the KEGG_ANTIGEN_PROCESSING_AND_PRESENTATION pathway in EG-treated NCI-H446 and NCI-H69 cells. NES and P values are shown. (C) Western blot analysis of IFITM3 and MHC-I protein levels in NCI-H446 and NCI-H69 cells treated with EG or transduced with sh-IFITM3, as indicated. Densitometric quantification is shown. (D) Western blot analysis of IFITM3 and MHC-I protein levels in RPP cells treated with EG (40 μm) for 48 h. (E) Schematic of the BALB/c nude mouse model for subcutaneous injection of RPP cells followed by oral EG treatment. Representative images and weight plots of tumors, along with tumor volume plots measured every 2 days. (F) C57BL/6 mice were implanted with RPP cells and received EG and PD-1 mAb treatment. Representative images and weight plots of tumors along with tumor volume plots measured every 3 days. (G) Flow cytometry analysis of CD3⁺ CD45⁺ cells, CD8⁺ CD3⁺ cells, and CD69⁺, CD44⁺, and GZMB⁺ CD8⁺ cells. The results are presented as the means ± SEMs, n = 4. Statistical comparisons between groups were performed using unpaired two-tailed Student’s t-tests. Statistical comparisons between groups were performed using unpaired two-tailed Student’s t-tests. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, n.s., not significant

Fig. 6

IFITM3 enhances MHC-I expression via NLRC5 upregulation and nuclear trafficking in SCLC cells. (A) Volcano plots showing DEGs between IFITM3-overexpressing NCI-H446 and NCI-H69 cells and control cells. Vertical dashed lines indicate log2FC = ± 1, and the horizontal dashed line represents the significance threshold at P = 0.05. (B) GSEA showing activated antigen processing and presentation pathways caused by IFITM3 overexpression in NCI-H446 and NCI-H69 cells. Enrichment significance was determined using the NES and nominal P values, as calculated by the GSEA algorithm. (C) Venn diagram of overlapping DEGs between NCI-H446 and NCI-H69 cells, accompanied by a bar plot of enriched GO terms. (D) Box plots displaying NLRC5 expression levels in NCI-H446 and NCI-H69 cells following IFITM3 overexpression. (E) Immunofluorescence images of NCI-H69 and NCI-H446 cells overexpressing MYC-NLRC5 (green) and/or HA-IFITM3 (red). Nuclei were counterstained with DAPI (blue). Scale bar: 20 μm. (F) Subcellular fractionation assays showing NLRC5 expression in cytoplasmic and nuclear fractions following IFITM3 overexpression. (G) Co-IP of IFITM3 with NLRC5 in NCI-H446 and NCI-H69 cells. The results are presented as the means ± SEMs. Statistical comparisons between groups were performed using unpaired two-tailed Student’s t-tests. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

Fig. 7

Mechanism of action of IFITM3 in regulating MHC-I expression to enhance immunotherapy sensitivity in SCLC