Ferroptosis, as the most enriched programmed cell death process in glioma, induces immunosuppression and immunotherapy resistance

Abstract

Background: Immunosuppressive microenvironment is a major cause of immunotherapeutic resistance in glioma. In addition to secreting compounds, tumor cells under programmed cell death (PCD) processes release abundant mediators to modify the neighboring microenvironment. However, the complex relationship among PCD status, immunosuppressive microenvironment, and immunotherapy is still poorly understood.

Methods: Four independent glioma cohorts comprising 1,750 patients were enrolled for analysis. The relationships among PCD status, microenvironment cellular components, and biological phenotypes were fully explored. Tissues from our hospital and experiments in vitro and in vivo were used to confirm the role of ferroptosis in glioma.

Results: Analyses to determine enriched PCD processes showed that ferroptosis was the main type of PCD in glioma. Enriched ferroptosis correlated with progressive malignancy, poor outcomes, and aggravated immunosuppression in glioblastoma (GBM) patients. Enhanced ferroptosis was shown to induce activation and infiltration of immune cells but attenuated antitumor cytotoxic killing. Tumor-associated macrophages (TAMs) were found to participate in ferroptosis-mediated immunosuppression. Preclinically, ferroptosis inhibition combined with Programmed Cell Death 1 (PD-1) and Programmed Cell Death Ligand-1 (PD-L1) blockade generated a synergistic therapeutic outcome in GBM murine models.

Conclusions: This work provides a molecular, clinical, and biological landscape of ferroptosis, suggesting a role of ferroptosis in glioma malignancy and a novel synergic immunotherapeutic strategy that combines immune checkpoint blockade treatment with ferroptosis inhibition.

Keywords: ICB; ferroptosis; immune microenvironment; immunotherapy; programmed cell death.

Graphical abstract

Fig. 1

Ferroptosis was identified as the predominant PCD type in glioma. (A) The landscape of PCD score among serious clinical factors in CGGA cohort (n = 325). (B–D) The association between PCD score with other clinical factors in CGGA cohort (Student’s t test). (E–F) Kaplan-Meier (log-rank test) and Cox regression survival analysis of glioma patients based on PCD score in CGGA cohort. (G) Illustration of five different PCD scores in four databases. (H) The ratio of glioma patients with different predominant PCD types in four databases. (Data are presented as means ± standard deviations; *indicates P < .05, **indicates P < .01, ***indicates P < .001, ****indicates P < .0001.)

Fig. 2

The description of ferroptosis in glioma. (A) The landscape of ferroptosis score with clinical and molecular features in CGGA cohort (n = 325). (B) The distribution of ferroptosis score for different clinical factors in CGGA cohort (Student’s t test). (C-E) Kaplan-Meier survival analysis of ferroptosis score in different group (log-rank test). (Data are presented as means ± standard deviations; *indicates P < .05, **indicates P < .01, ****indicates P < .0001, ns indicates no statistical significance.)

Fig. 3

Ferroptosis score was associated with immunosuppressive microenvironment in GBM in CGGA cohort. (A) GO functional analysis of ferroptosis score by using Metascape tools. (B) Differences in various steps of Cancer-Immunity Cycle (Student’s t test). (C) GSEA results showed enrichment of dampened antitumor immunity in high ferroptosis score group. (D) The differential expression of immune checkpoint molecules (Student’s t test). (E) High score group had higher immune score and lower purity (Student’s t test). (F) High ferroptosis score associated with several immunosuppressive cell infiltrations (Student’s t test). (G) The correlation between immunosuppressive cell enrichment and dampened antitumor immunity (Term1: GO_NEGATIVE_REGULATION_OF_IMMUNE_SYSTEM_PROCESS; Term2: GO_NEGATIVE_REGULATION_OF_ALPHA_BETA_T_CELL_ACTIVATION; Term3: GO_NEGATIVE_REGULATION_OF_T_CELL_PROLIFERATION; Term4: GO_NEGATIVE_REGULATION_OF_T_CELL_DIFFERENTIATION) in GBM with high ferroptosis score (Pearson correlation). (Date are presented as means ± standard deviations; *indicates P < .05, **indicates P < 0.01, ****indicates P < .0001).

Fig. 4

High ferroptosis levels in GBM caused more macrophage infiltration and M2-like polarization. (A) IHC staining of TUNEL, GPX4, IBA1, CD86 and CD163 in clinical GBM tissues (n = 28, scale bars: 50 μm). (B–C) The migration ability of THP1-derived macrophages under different CM treatments (Student’s t test, n = 3, scale bars: 50 μm). (D–F) PCR and flow cytometry results of detecting the polarization of macrophages under different CM treatments (Student’s t test, n = 3). (G) Survival of SB-bearing C57BL6 mice under PBS or ferrostatin-1 treatment (log-rank test, n = 6). (H) IHC staining of IBA1 and CD86 in tumors from SB-bearing C57BL6 mice (n = 3, scale bars: 50 μm). (I) Flow cytometry identified macrophage polarization in tumors from SB-bearing C57BL6 mice (Student’s t test, n = 5). (Data are presented as means ± standard deviations; *indicates P < .05, **indicates P < .01, ****indicates P < .0001.)

Fig. 5

Ferroptosis inhibition reversed the immunosuppressive phenotype in tumor from SB-bearing C57BL6 mice. (A) Schematic illustration of CyTOF analysis of tumors from SB-bearing mice under different treatments (n = 4). (B) t-SNE plot of tumor-infiltrating immune cells. (C–D) The expression of CD80 and CD206 in macrophages under different treatments (Data are presented as means ± 95% confidence interval; Student’s t test.). (E) The expression of several markers in T cells under different treatments (Student’s t test). (****indicates P < .0001).

Fig. 6

Ferroptosis inhibition sensitized GBM to anti-PD1/L1 immunotherapy. (A) Schematic illustration of combination ferrostatin-1 with anti-PDL1 in SB-bearing C57BL6 mice. (B) Survival of SB-bearing C57BL6 mice under different treatments (log-rank test, n = 6). (C) IHC of IBA1, CD86, CD3, CD4, and CD8 in tumors from SB-bearing C57BL6 mice under different treatments (Student’s t test, n = 3). (D–F) Flow cytometry representing the ratio of CD3⁺, CD4⁺, and CD8⁺ T cells in tumors from SB-bearing C57BL6 mice under different treatments (Student’s t test, n = 6). (G) Schematic illustration of combination ferrostatin-1 with anti-PD1 in SB-bearing C57BL6 mice. (H) Survival of SB-bearing C57BL6 mice under different treatments (log-rank test, n = 6). (I) IHC of IBA1, CD86, CD206, CD3, CD4, and CD8 in tumors from SB-bearing C57BL6 mice under different treatments (Student’s t test, n = 3). (J) Survival of SB-bearing SCID mice under different treatments (log-rank test, n = 6). (K) The rate of response patients was lower in high ferroptosis score group and patients with response owned higher ferroptosis score in IMvigor210 database (response: PR/CR; non-response: PD/SD). (Date are presented as means ± standard deviations; *indicates P < .05, **indicates P < .01, ***indicates P < .001, ns indicates no statistical significance).