NADPH controls a cell state vulnerability towards ferroptosis in acute lymphoblastic leukaemia

Abstract Background and significance: Ferroptosis is a distinct form of non-apoptotic cell death characterised by iron-dependent lipid peroxidation of membrane phospholipids. The master regulator of ferroptosis is the selenoprotein glutathione peroxidase 4 (GPX4) which quenches oxidised membrane phospholipids. Lipid biosynthesis can also modulate ferroptosis sensitivity by altering monounsaturated fatty acid (MUFA) and squalene levels (via stearoyl-CoA desaturase (SCD) and squalene monooxygenase (SQLE), respectively), thereby altering initiation, propagation, and termination of phospholipid peroxidation. Solid tumours consistently demonstrate enhanced ferroptosis susceptibility in metastasising cancer cells and therapy resistant persister cells.However, the underlying mechanisms explaining shifts in susceptibility are conflicting. Furthermore, these cell state dependent vulnerabilities have never been reported in haematological malignancy. Here, we report a pan-ALL vulnerability to ferroptosis confined to low-density cell states such as post-chemotherapy measurable residual disease (MRD). Mechanistically, we show this is driven by nicotinamide adenine dinucleotide phosphate (NADPH)-sensors regulating GPX4, SCD and SQLE activity to switch between ferroptosis resistant and prone cell states. This can be successfully targeted in vivo, offering novel therapeutic approaches to efficiently eradicate MRD in ALL. Results: We observed that ALL cells cultured at low-density were exquisitely sensitive to ferroptosis-inducing agents but became resistant at high-density. In vitro analysis revealed significant density dependent alterations in SCD and GPX4 activity at high-density, accompanied by a reduction in SQLE activity by protein truncation. Metabolomic analysis of low- and high-density cultures revealed a pronounced shift toward glycolysis, akin to the Warburg effect, at high cell density. This shift significantly enhanced NADPH production through the pentose phosphate pathway. Indeed, genetically or pharmacologically impeding NADPH production in ALL cells reversed the density-dependent regulation of SCD, GPX4 and SQLE and re-sensitised cells to ferroptosis. We next identified and interrogated the effects of CRISPR/Cas9 gene deletion of two protein sensors of NADPH, the E3 ubiquitin ligase MARCHF6 and the m6A RNA demethylase FTO. Loss of MARCHF6 activity led to squalene depletion and reduced GPX4 expression both in vitro and in vivo. In parallel, inactivation of FTO reduced SCD expression, suppressing MUFA synthesis and further sensitising cells to ferroptosis both in vitro and in vivo, together revealing a distinct anti-ferroptosis role for the Warburg effect in ALL through NADPH sensing. Strikingly, single-cell RNA sequencing in both B- and T-ALL revealed a consistent downregulation of genes involved in glycolysis and NADPH regeneration in MRD samples compared to diagnostic samples. Importantly, MRD samples displayed lower FTO expression and MARCHF6 activity, indicative of a therapy-induced ferroptosis vulnerability. This phenotype was faithfully recapitulated in vitro through treatment of ALL cell lines and patient derived xenografts (PDXs) with Vincristine or Dexamethasone. Finally, using a PDX murine model, we demonstrate that GPX4 inhibition administered after induction chemotherapy (Vincristine Sulphate, Dexamethasone and Daunorubicin) prevented disease relapse, whereas front-line GPX4 inhibition in the setting of bulky disease had no significant effect — confirming a context-specific, treatment-induced susceptibility to ferroptosis. Conclusions: Through comprehensive in vitro and in vivo studies, we have uncovered a distinct mechanism by which ALL cells acquire resistance to ferroptosis. Our findings show that glycolytic metabolism confers protection via NADPH generation—a process that can be transiently disrupted by standard-of-care chemotherapy. As a result, MRD ALL cells become highly sensitised to ferroptosis, presenting a novel therapeutic window for targeting GPX4 through pharmacologic, dietary, or other means of intervention. Furthermore, the mechanisms described here prompt a re-evaluation of ferroptosis regulation in solid tumours and may offer a concise pro-survival rationale for the Warburg effect.