The E2F1-HMGCR axis promotes ferroptosis resistance in immune refractory tumor cells

Abstract

During cancer immunoediting, cancer cells deregulate cell death executioner mechanisms to escape immunotherapy-induced antitumor immunity. Ferroptosis, a type of regulated necrosis triggered by lipid peroxidation, plays a pivotal role in the anti-tumor activity of T cell-based immunotherapies; however, mechanisms for the modulation of ferroptosis in immune-refractory tumor cells are unclear. In this study, using preclinical models of immune refractory tumors obtained following the course of immunoediting by PD-1 blockade and adoptive T cell therapy (ACT), we find that T cell-based immunotherapy drives the development of ferroptosis resistance of tumor cells. In this process, E2F1 is upregulated by immunotherapy and it in turn binds to the promoter of the HMGCR gene to upregulate HMGCR, thereby contributing to the resistance to ferroptosis. Notably, HMGCR inhibition renders immune-refractory tumor cells susceptible to ACT and PD-1 blockade. Thus, our results reveal a mechanism by which cancer cells modulate ferroptosis to acquire resistance to immunotherapy and implicate the E2F1-HMGCR axis as a central molecular target for controlling ferroptosis resistance of immune-refractory cancer.

Fig. 1. HMGCR upregulation following ICB-mediated immune selection promotes ferroptosis resistance.

B16 P0, P3 cells and TC-1 P0, P3 cells were treated with indicated concentrations of RSL3 (a) or Erastin (b) for 24 h. Cell viability was measured by Trypan blue exclusion assay. The concentrations causing 50% inhibition of cell viability (IC₅₀ values) were determined. c, d B16 P0, P3 cells and TC-1 P0, P3 cells were treated with RSL3 (1 or 2 μM) in the absence or presence of zVAD (10 μM) or Lip-1 (1 μM) for an additional 20 h. The percentage of 7-AAD⁺ cells c and relative lipid ROS d were measured by flow cytometry. The data are representative of those from 3 independent experiments with triplicate. e The real time PCR array analysis for ferroptosis negative regulator genes in B16 P0, P3 and TC-1 P0, P3. f Venn diagram showing the overlap of ferroptosis negative regulator genes between upregulated in B16 P3 versus P0 (red) and those upregulated in TC-1 P3 versus P0 (green). g HMGCR protein levels in B16 P0, B16 P3, TC-1 P0, and TC-1 P3 was determined by Western blot. β-actin was included as an internal loading control. h The HMG-CoA reductase activity in B16 P0, B16 P3, TC-1 P0, and TC-1 P3 (n = 3; in duplicate). Flow cytometry analysis of the 7AAD⁺ cells (i) and relative lipid ROS (j) in indicated cells treated with RSL3 with or without Lip-1 (1 μM) for 20 h. All data are representative of those from 3 independent experiments with triplicate. The error bars represent mean ± SD. All p value are presented exactly in figure. The p values by two-way ANOVA (c, d, i, j), and unpaired, two-tailed Student’s t test (e, g, h) are indicated. Source data are provided as a Source data file.

Fig. 2. HMGCR depletion reverses immune-refractory features by inducing the anti-PD-1-mediated ferroptotic cell death.

a Schematic of the therapy regimen in mice implanted with B16 P3 cells. b Tumor growth during 18 days after mice inoculated with B16 P3 then treated with the indicated reagents. The percentage of 7-AAD⁺ cells (c) and relative lipid ROS (d) measured by flow cytometry. B16 P3 tumor-bearing mice were administered siGFP or siHmgcr#1 with anti-PD-1 or anti-PD-1 plus Lip-1 as indicated. e Flow cytometric profiles of tumor-infiltrating CD3⁺ CD8⁺ T cells. f Percentage of IFNγ⁺ to tumor-infiltrating CD3⁺ CD8⁺ T cells. For the in vivo experiments, 7 mice from each group were used, and randomly selected 5 samples were analyzed. The error bars represent mean ± SD. All p value are presented exactly in figure. The p values by two-way ANOVA (b) and one-way ANOVA (c–f) are indicated. Source data are provided as a Source data file.

Fig. 3. CTL-mediated immune selection triggers ferroptosis-resistance by upregulating HMGCR.

a A375 P0 and P3 cells were treated with indicated concentrations of RSL3 for 24 h. Cell viability was measured by Trypan blue exclusion assay and the concentrations of RSL3 causing 50% inhibition of cell viability (IC₅₀ values) were determined. A375 P0 and P3 cells were treated with RSL3 (200 nM) in the absence or presence of Lip-1 (1 μM) for an additional 20 h, and then the percentage of 7-AAD⁺ cells (b) and relative lipid ROS (c) were measured by flow cytometry. d The HMGCR mRNA levels of immune-resistance A375 cells were determined by qRT-PCR. e The levels of HMGCR protein in A375 P0 and P3 cells transfected with siRNA targeting GFP or HMGCR were measured by Western blot. Flow cytometry analysis of the 7AAD⁺ cells (f) and relative lipid ROS (g) in indicated cells treated with RSL3 (200 nM) with or without Lip-1 (1 μM) for 24 h. h, i Tumor cells were incubated with the supernatant from tumor-specific CTLs with or without anti-IFNγ. The lipid ROS was measured by flow cytometry. All in vitro experiments were performed in triplicate. The error bars represent mean ± SD. All p value are presented in figure. The p values by two-way ANOVA (b, c, f, g, h, i), and unpaired, two-tailed Student’s t test (d) are indicated. Source data are provided as a Source data file.

Fig. 4. E2F1 directly regulates HMGCR, thereby promoting ferroptosis resistance of immune refractory tumor cells.

a The expression of HMGCR up-regulator in non-responders relative to responders to anti-PD-1 therapy. The horizontal dashed line indicates p-value cutoffs at the 0.05 level. Red dot indicates the E2F1 gene. Vertical dashed lines indicate a fold change cut off (1.5FC). The p-values were determined by unpaired, two-tailed Student’s t test. b Schematic diagram of E2F1 binding site on HMGCR promoter and sequence homology across different species. c The E2F1 and HMGCR proteins in A375 P0 and P3 cells transfected with siRNA targeting GFP or E2F1 were detected using Western blot. d The HMGCR mRNA levels in A375 P3 transfected with siRNA targeting GFP or E2F1. e Diagram of the human HMGCR promoter region containing the E2F1 binding elements. The arrows indicate the ChIP amplicons corresponding to the two adjacent E2F1-binding sites. f Luciferase activities in A375 P0 and P3 cells first transfected with siRNA targeting GFP or E2F1 then transfected with the pGL3-basic or pGL3-HMGCR WT plasmid as indicated. g Luciferase activities in A375 P3 cells transfected with the pGL3-basic, pGL3-HMGCR WT, pGL3-HMGCR E1Mut or E2Mut plasmids. h ChIP assay was carried out using A375 P0 and P3 cells. Cross-linked chromatin was immunoprecipitated with an anti-E2F1 antibody or an IgG control. The value of ChIP data represents relative ratio to the input, followed by qPCR analysis. Flow cytometry analysis of the 7AAD⁺ cells (i) and relative lipid ROS (j) in cells treated for 24 h with RSL3 (200 nM) with or without Lip-1 (1 μM) for an additional 20 h. HMGCR and HA-E2F1 protein levels (k), analyzed by western blots, and the percentage of 7-AAD⁺ cells (l) and relative lipid ROS (m), measured by flow cytometry, in A375-no insert and A375-HA-E2F1 cells transfected with siGFP or siHMGCR#1 as indicated. All in vitro experiments were performed in triplicate. The error bars represent mean ± SD. All p value are presented exactly in figure. The p values by unpaired, two-tailed Student’s t test (d, g) and two-way ANOVA (f, h, i, j, l, m) are indicated. Source data are provided as a Source data file.

Fig. 5. E2F1-HMGCR axis is associated with poor response to anti-PD-1 therapy.

Comparisons of expressions levels of HMGCR (a) or E2F1 (b) in responder (R, n = 9) and non-responder (NR, n = 33). c Comparisons of HMGCR expression in patients with low levels (n = 11) and high levels (n = 31) of E2F1. d Combined level of E2F1^high/HMGCR^high was significantly associated with resistance to anti-PD-1 therapy in patients. e Kaplan–Meier analysis of overall survival (calculated as months to death or the last follow-up) of patients whose melanoma have low levels (n = 7) or high levels (n = 27) of both E2F1 and HMGCR. In the dot plots, the center lines indicate the medians; and the ends of the whiskers indicate the maximum and minimum values, respectively. The top and bottom edges of boxes indicate the first and third quartiles, respectively. c–e Optimal cut-off values were determined by performing receiver operating characteristic (ROC) analysis using MedCalc statistical software. The error bars represent mean ± SD. All p value are presented exactly in figure. a–c The p values by unpaired, two-tailed Student’s t test are indicated. d, e The p-values were analyzed by a Student’s two-side t test. The p-values were determined by Mann–Whitney U test (d) and Gehan–Breslow–Wilcoxon test (e). Source data are provided as a Source Data file.

Fig. 6. Simvastatin overcomes ferroptosis resistance in Immune-refractory tumors via Inhibition of the HMGCR–CoQ10 Axis.

Cell death percentage of B16 P0 and P3 cells (a) or A375 P0 and P3 cells (b) treated with indicated concentrations of RSL3 with or without simvastatin (Sim, 1 μM) for 24 h. The percentage of cell death was determined by trypan blue exclusion assay. The percentage of 7AAD⁺ cells or relative lipid ROS in B16 P3 cells (c) and in A375 P3 cells (d) treated for 48 h with RSL3 and Sim with or without Lip-1 (1 μM) for an additional 20 h. e The mevalonate pathway and its metabolic branches in mammalian cells, highlighting key enzymatic steps catalyzed by HMGCR, COQ2, FDFT1, and SQLE, along with their respective inhibitors. f The levels of CoQ10 and cholesterol in A375 P0 and P3 cells. The percentage of 7AAD⁺ tumor cells and lipid ROS levels were measured in A375 P3 cells treated with RSL3 (200 nM) alone or in combination with ZA (20 μM) (g) or 4-NB (1 mM) (h), with or without Lip-1 (1 μM). i The levels of CoQ10 in A375 cells treated with or without simvastatin. j The percentage of 7AAD⁺ tumor cells and relative lipid ROS levels were assessed in A375 P3 cells treated with RSL3 (200 nM) alone or in combination with Sim (1 μM), with or without CoQ10 (200 μM) supplementation. All in vitro experiments were performed in triplicate. The error bars represent mean ± SD. All p value are presented exactly in figure. The p values by two-way ANOVA (a–d, g, h, j), and unpaired, two-tailed Student’s t test (f, i) are indicated. Source data are provided as a Source data file.

Fig. 7. HMGCR inhibition renders the tumor susceptible to ferroptosis response to T cell mediated immune therapy.

a Schematic of the therapy regimen in NOD/SCID mice implanted with A375 P3 cells. Tumor growth (b) and the percentage of 7AAD⁺ tumor cells (c) and lipid peroxidation (d), determined by flow cytometry, in mice inoculated with A375 P3 cells and treated with Sim or Lip-1 with or without NY-ESO1-specific T cell adoptive transfer (ACT). e Schematic of the therapy regimen in C57BL6 mice implanted with B16 P3 cells. Tumor growth (f), the percentage of 7AAD⁺ tumor cells (g), lipid peroxidation (h), the level of HMGB1 (i), and CRT (j) determined by flow cytometry, flow cytometry profiles of tumor-infiltrating cDC cells (k), CD8⁺ T cells (l), and the percentage of IFNγ⁺ in CD8⁺ T cells (m) in mice inoculated with B16 P3 cells and treated with the indicated reagents. For the in vivo experiments, 7 mice from each group were used, and randomly selected 5 samples were analyzed (b–d, f–m). All p value are presented exactly in figure. The p values by two-way ANOVA (b, f), and one-way ANOVA (c, d) and (g–m) are indicated. The error bars represent mean ± SD. Source data are provided as a Source data file.

Fig. 8. Scheme for the role of E2F1-HMGCR axis in ferroptosis resistance in immune refractory tumor cells.

E2F1-HMGCR axis drives refractoriness against immunotherapy by triggering ferroptosis resistance.