Longevity & AgingResearch PaperOpen Access

Mitochondrial Failure Starves Brain Cells of Antioxidants, Triggering Ferroptosis in Alzheimer's

New research links ATP depletion from mitochondrial dysfunction to glutathione loss and iron-driven cell death in Alzheimer's disease.

Saturday, May 30, 2026 0 views
Published in Adv Sci (Weinh)
Glowing mitochondria inside a fading neuron, surrounded by iron particles and depleted antioxidant molecules, dark neural background

Summary

Researchers analyzed brain proteomics from 625 Alzheimer's patients and found that mitochondrial protein loss is the dominant biochemical change in AD. Because glutathione (GSH) synthesis requires ATP, mitochondrial failure depletes GSH, removing the brain's key defense against ferroptosis — an iron-dependent form of cell death. Cellular experiments confirmed that ATP availability is rate-limiting for GSH production, and that mitochondrial inhibition increases or decreases ferroptosis depending on whether mitochondria are net GSH producers or consumers in that context. The study positions bioenergetic insufficiency as a direct physiological trigger for ferroptosis in AD, offering a mechanistic bridge between two long-recognized but previously unconnected disease hallmarks: energy deficit and oxidative stress.

Deep Dive Audio
0:00--:--

Detailed Summary

Alzheimer's disease (AD) is defined by amyloid plaques and tau tangles, but the downstream cellular events that actually kill neurons remain poorly understood. This study, led by researchers at the Florey Institute in Melbourne, proposes a compelling mechanistic chain: mitochondrial failure → ATP depletion → glutathione (GSH) deficit → ferroptosis vulnerability.

Using weighted gene co-expression network analysis (WGCNA) of proteomics data from 625 post-mortem inferior temporal cortex samples from the Memory and Aging Project (MAP), the team found that the protein module most strongly associated with clinical AD diagnosis and cognitive decline mapped almost entirely to oxidative phosphorylation and mitochondrial function. This mitochondrial module eigenvalue was significantly lower in AD patients versus controls, and low scores correlated with faster cognitive decline in the five years before death. The finding was replicated in an independent dorsolateral prefrontal cortex dataset.

Because ATP is unstable in post-mortem tissue, the team used GSH as a functional proxy for bioenergetic status, reasoning that GSH synthesis requires 2 ATP molecules per molecule produced. An unbiased correlation analysis of the same 625-subject dataset revealed that mitochondrial proteins were the strongest positive correlates of total GSH (t-GSH) levels in AD brain tissue — more so than cysteine availability or GPX4 expression. Genetic models (adenylate kinase overexpression) and pharmacological models (oligomycin, iodoacetate, the novel bacterial ATP nucleosidase CAP-17) confirmed in cellular systems that ATP is genuinely rate-limiting for GSH synthesis, independent of cysteine supply.

The role of mitochondria in ferroptosis proved context-dependent. When mitochondria were the primary ATP source — and thus GSH producers — their inhibition increased ferroptosis susceptibility. However, mitochondria also import GSH via the SLC25A39 transporter to protect their own membranes. In conditions where mitochondria acted as net GSH consumers (e.g., under erastin-induced cysteine deprivation), inhibiting mitochondrial GSH uptake actually reduced ferroptosis. This dual role explains some conflicting findings in the literature regarding mitochondria and ferroptosis.

The authors introduce CAP-17 as a novel research tool — a bacterial ATP nucleosidase that selectively depletes intracellular ATP in mammalian cells without directly targeting the GSH pathway, confirming that ATP scarcity alone is sufficient to sensitize neurons to ferroptotic death. Collectively, the findings frame low bioenergetic output as a physiological ferroptosis trigger in AD, and suggest that ferroptosis inhibitors (e.g., liproxstatins) or strategies to restore GSH could be clinically meaningful even without resolving amyloid or tau pathology.

Key Findings

  • Mitochondrial/oxidative phosphorylation proteins were the top proteomics module depleted in AD and correlated with faster cognitive decline.
  • GSH synthesis requires 2 ATP per molecule; ATP availability — not cysteine supply — is the rate-limiting factor for GSH in AD brain.
  • Mitochondrial proteins were the strongest positive correlates of total brain GSH levels across 625 AD-continuum subjects.
  • Mitochondrial inhibition increases ferroptosis when mitochondria are net GSH producers but decreases it when they are net GSH consumers via SLC25A39.
  • CAP-17 bacterial ATP nucleosidase is validated as a new tool to selectively deplete ATP and induce GSH-dependent ferroptosis vulnerability.

Methodology

Proteomic WGCNA analysis of 625 post-mortem inferior temporal cortex samples from the Rush Memory and Aging Project, stratified by clinical dementia status and CERAD pathology criteria. Findings were validated in an independent prefrontal cortex proteomics dataset and in cellular models using genetic (adenylate kinase), pharmacological (oligomycin, iodoacetate), and novel bacterial enzyme (CAP-17) ATP depletion approaches.

Study Limitations

All human data are cross-sectional post-mortem proteomics, precluding causal inference in living patients. Cellular models used pharmacological ATP depletion that may not fully replicate the chronic, partial bioenergetic decline seen in AD. The dual pro- and anti-ferroptotic roles of mitochondria depending on context add complexity that makes straightforward therapeutic targeting challenging.

Enjoyed this summary?

Get the latest longevity research delivered to your inbox every week.