Publication Details
Abstract
Congenital Ferroptosis suppressor protein 1 (FSP1) reduces coenzyme Q10 (CoQ10) to a lipophilic radical-trapping antioxidant and, in parallel with GPX4, protects lung adenocarcinoma (LUAD) from iron-dependent lipid peroxidation. FSP1 is an NAD(P)H-dependent oxidoreductase, yet the metabolic sources that regenerate the reducing equivalents required to sustain the FSP1–CoQ10 axis under nutrient stress remain incompletely defined, and bulk metabolite measurements cannot distinguish whether cytosolic or mitochondrial one-carbon flux preferentially supports this activity. Here we present a compartment-resolved framework in KRAS/LKB1-mutant LUAD organoids in which methionine restriction (MR) is used as a controlled metabolic stressor. Using [U-¹³C₃]-serine, [2,3,3-²H₃]-serine and [U-¹³C₂]-glycine tracing coupled to subcellular fractionation, together with SHMT1 (cytosolic) versus SHMT2 (mitochondrial) perturbation and FSP1 epistasis, we quantify how serine–glycine one-carbon metabolism sustains NAD(P)H, the CoQ10H2/CoQ10 ratio and ferroptosis resistance. In the accompanying experimental dataset, MR reprogrammed one-carbon metabolism toward the mitochondrial branch (mitochondrial:cytosolic flux ratio increasing under MR), and mitochondrial one-carbon flux correlated tightly with CoQ10 redox state (r = 0.96). SHMT2 loss, but not SHMT1 loss, lowered NADPH and CoQ10H2/CoQ10, raised lipid peroxidation and triggered ferroptosis; the effect required FSP1 and was rescued by FSP1 re-expression, establishing FSP1–CoQ10 as the downstream effector. Combining MR, FSP1 inhibition and a ferroptosis inducer produced greater-than-additive killing (excess over Bliss independence = 0.18). These findings position compartment-specific one-carbon flux as a targetable metabolic determinant of FSP1–CoQ10-dependent ferroptosis resistance in aggressive LUAD.