Anay Mondal, Reetuparna Acharya, Smita Jain
Ferroptosis-targeted therapies in Alzheimer's disease (AD) have yielded consistent preclinical evidence; however, their translation into clinical outcomes has remained elusive. This mismatch reflects a fundamental flaw that conceptualizes this peroxidative pathway as a single, stage-invariant druggable entity. Existing frameworks conflate acyl-CoA synthetase long-chain family member 4 (ACSL4) activity, a slow, constitutive, amyloid-responsive process, with oxidative execution, an acute 5-lipoxygenase (5-LOX) process driven by neuroinflammatory triggers operating on incompatible time scales. This conflation makes it structurally impossible to match as a therapeutic strategy to any individual patient. This review proposes the Substrate-Engine Model to resolve this translational paradox: ACSL4 constitutes the substrate arm, establishing membrane vulnerability detectable years before cognitive symptom onset, while 5-LOX constitutes the engine arm, catalyzing peroxidative neuronal death under neuroinflammatory conditions in prodromal-to-dementia stages. These arms are mechanistically noninterchangeable. Systematic network analysis confirms that ACSL4 and 5-LOX occupy nonredundant positions in the ferroptosis-AD interactome and reside in distinct Louvain communities with divergent cellular and disease-stage expression profiles that preclude compensatory pathway rerouting and provide a rationale for dual-arm targeting. Mechanistic reanalysis of "REFLECT" and "TOMMORROW" trials identified factors responsible for substrate-arm therapeutic failure. These analyses yield a four-subtype precision stratification framework (Types A-D), anchored on two orthogonal biomarker axes. This framework repositions ferroptosis as a stage-specific therapeutic target, guiding the choice between prophylactic substrate depletion and acute engine suppression based on the patient's biomarker profile.