Ferroptosis activates NK and CAR T cell-mediated anti-tumor immunity in Acute Myeloid Leukemia

Abstract Resistance to apoptosis remains a major clinical challenge in acute myeloid leukemia (AML), especially in relapsed/refractory (R/R) cases following standard therapies including venetoclax (VEN) -containing regimens. While immunotherapy has revolutionized the treatment of R/R lymphoid malignancies and certain solid tumors, its efficacy in AML remains limited due to the immunologically “cold” tumor microenvironment. These challenges underscore the urgent need to develop novel treatment strategies to overcome apoptosis resistance while eliciting a durable anti-AML immune response. Ferroptosis, an iron-dependent form of cell death negatively regulated by glutathione peroxidase 4 (GPX4),represents a promising alternative in therapy-resistant solid tumors. In AML, we recently demonstrated that inhibiting GPX4 induces a distinct, mitochondria-dependent form of ferroptosis termed “mitochondrial ferroptosis”, highlighting the new therapeutic potential (Leukemia;38:729-740, 2024).However, its effect on AML immunity remains unexplored. Emerging evidence suggests that ferroptosis can trigger immunogenic cell death (ICD) through the release of danger-associated molecular patterns (DAMPs), activating immune cells and promoting anti-tumor immunity, primarily in the context of solid tumors. As the immunogenicity of ferroptosis is highly context-dependent, it is critical to elucidate the immunogenic properties of ferroptosis specifically in AML to develop tailored immune-potentiating therapeutics. To investigate whether GPX4 inhibition induces ICD signaling in AML, we applied a doxycycline (dox)-inducible GPX4-knockdown (KD) system in MOLM-13 cells and assessed key DAMPs including ATP, HMGB1, and calreticulin (CRT), using CTG assay, ELISA, and flow cytometry, respectively. GPX4-KD significantly increased extracellular ATP (p 0.01) and HMGB1 (p 0.5) levels, as well as the surface expression of CRT (p 0.0001) on AML cells. These effects were reversed by a ferroptosis inhibitor, ferrostatin-1 (Fer-1), confirming ferroptosis dependence. Similar results were reproduced in OCI-AML3 cells. Pharmacological inhibition of GPX4 by RSL3 also demonstrated a time-dependent increase in ATP, HMGB1, and CRT levels in MOLM13 cells, all of which were also suppressed by Fer-1. Notably, treatment with venetoclax, which induces apoptosis, failed to trigger significant ICD signals except for HMGB1, suggesting that the DAMP profile induced by ferroptosis is distinct from and potentially more immunogenic than apoptosis-induced ICD. Consistent results were observed upon treatment with the selective GPX4 inhibitor ML210 in both MOLM13 and OCI-AML3 cells. Next, to test our hypothesis that ferroptosis enhances the anti-AML activity of immune effectors, we co-cultured NK cells with ferroptotic AML cells and assessed AML cell killing using the Incucyte live-cell imaging system. Co-culture of NK cells with dox-induced GPX4-KD or RSL3-treated AML cells significantly enhanced AML cell killing (p 0.0001, 3.4-fold increase in co-culture with NK cells alone, 5.9-fold increase with shGPX4 alone, and 11.6-fold increase in co-culture with NK cells and shGPX4, respectively, compared to untreated controls). This demonstrates that ferroptosis enhances NK cell-mediated anti-AML effects. Mechanistically, we observed that inhibition of GPX4 significantly reduces HLA class 1 surface expression in both AML cell lines (50% decrease compared to controls, p=0.0001) and primary samples (75% decrease compared to controls). This suggests that diminished HLA-dependent immune evasion may also contribute to NK cell-mediated AML cell killing in addition to DAMPs release. We next conducted similar co-culture experiments using CD123-targeted CAR T cells. CD123-CAR T cells exhibited significantly enhanced AML cell killing (p ≤ 0.01, 3.2-fold increase in co-culture with NK cell alone, 4.7-fold increase with shGPX4 alone, and 8.3-fold increase in co-culture with CAR T cells and shGPX4, respectively, compared to untreated controls). These results confirm that ferroptosis enhances CAR-T cell anti-AML activity. In summary, our findings demonstrate that ferroptosis enhances anti-AML immunity, likely through the release of ICD signals. These results provide a potential basis for developing ferroptosis-based immunotherapy in AML, warranting further investigation into the underlying molecular mechanisms and in vivo validation of the concept.