APOE isoforms split the brain’s antioxidant response long before overt disease
Apolipoprotein E is usually discussed in the context of Alzheimer’s disease, but this preprint asks a more basic question: what does each isoform do to the aging brain before pathology is obvious? The answer is not a single linear effect. Across the ancestral-to-derived sequence ApoE4→ApoE3→ApoE2, antioxidant regulation followed distinct trajectories, and the differences were visible in both mouse and human systems.
Young humanized ApoE4 mice already showed a biochemical disadvantage. Antioxidant protein abundance was lower, and oxidized phospholipid reduction was blunted before detectable oxidative damage appeared. With age, however, the same animals accumulated more lipid peroxidation products despite evidence of antioxidant remodeling. The pattern points to a system that can sense stress yet fails to convert that signal into durable biochemical protection.
Stress signaling rose, but protective protein output lagged in ApoE4 cells
Cellular experiments tightened that picture. ApoE4 cells mounted the strongest antioxidant transcriptional response to sulforaphane, a finding that fits with the idea of a highly inducible stress-sensing program. The catch was in translation and protein accumulation. Even with a strong transcriptional push, these cells accumulated less protein from targets tied to proteostasis, glutathione synthesis, and ferroptosis suppression.
One sentence captures the central tension in the data: transcription rose faster than functional defense.
The mismatch matters because antioxidant biology is not just about turning genes on. Cells must transcribe the right targets, translate them efficiently, and maintain the proteins long enough to detoxify reactive species, preserve membranes, and prevent iron-dependent lipid injury. The preprint links that failure of coordination to altered eIF2α signaling and translation kinetics, suggesting that APOE4 skews not only the stress response itself but the ability to convert mRNA programs into usable protein reserves.
Age shifted antioxidant bias away from ApoE4 and toward ApoE2
The trajectory did not stay fixed across the lifespan. Antioxidant transcription shifted from an ApoE4 bias in youth toward ApoE2 with age. In other words, the isoform that appeared most reactive to oxidative stress early in life was not the one that dominated later. Adult switching from APOE4 to APOE2 redirected expression in astrocytes and microglia, indicating that isoform context can remodel cell-state behavior even after development.
This is a useful clue for neurobiology because astrocytes and microglia do much of the brain’s oxidative housekeeping. Astrocytes support glutathione production and metabolic buffering, while microglia respond to damage signals and inflammatory cues. If APOE4 alters how those cells handle stress signaling and protein production, the consequence could be a brain that spends energy responding to insult while underperforming on actual repair.
Human brain tissue showed regional antioxidant depletion despite preserved cognition
The human data moved the question out of the mouse model and into aging cortex. Cognitively normal human brains showed regional reductions in nuclear Nrf2 and antioxidant proteins, with transcriptomic analyses supporting context-dependent organization. Nrf2 is a central regulator of cellular antioxidant defense, and its nuclear localization is a practical readout of whether the program is poised to respond to stress.
Here, the regional nature of the effect is important. The brain is not one metabolic unit. Different regions age at different rates, vary in lipid composition and neuronal demand, and do not necessarily deploy antioxidant defense in parallel. A decline in nuclear Nrf2 in selected regions suggests that antioxidant readiness can erode before clinical symptoms emerge, leaving vulnerable circuits exposed to cumulative oxidative injury for years.
The preprint does not show disease onset, dementia conversion, or a therapeutic intervention. It does show a biologically plausible pathway by which APOE4 may confer risk: weaker baseline antioxidant protein abundance, altered stress-responsive translation, and less efficient conversion of transcriptional activation into protein output. Those features can create a chronic gap between damage sensing and damage control.
Lipid oxidation and proteostasis sit at the center of the mechanism
The emphasis on oxidized phospholipids and lipid peroxidation products is mechanistically revealing. Membrane lipids are among the first macromolecules to suffer from oxidative stress, and their damage can propagate inflammation, protein misfolding, and mitochondrial dysfunction. When a genotype accumulates lipid peroxidation products despite an active antioxidant transcriptome, the problem is not simply “too little response.” It is a failure in the quality and timing of the response.
The same logic applies to glutathione synthesis and ferroptosis suppression. Glutathione is one of the brain’s major redox buffers, and ferroptosis is an iron-dependent form of cell death driven by lipid oxidation. If ApoE4 cells generate less protein from pathways that prevent ferroptosis, then the tissue may become more vulnerable to oxidative membrane injury even when stress signals are intact. The result is a system that appears responsive on paper but underpowered in practice.
What the data can and cannot support
These findings come from preclinical and postmortem human tissue analyses, so they do not establish a clinical endpoint, a dementia risk estimate, or a treatment effect. They do, however, sharpen the biological argument for why APOE genotype shapes neurodegenerative vulnerability long before memory loss is evident. The work also suggests that age is not a passive background variable; it changes which isoform dominates antioxidant transcription and how effectively those signals are translated into protein.
For translational work, the unanswered problem is not whether oxidative stress matters. It does. The harder question is how to restore coupling between transcription, translation, and protein abundance in the right cells and at the right stage of aging. This preprint argues that APOE isoform biology belongs squarely inside that question, especially in astrocytes, microglia, and regionally vulnerable human brain networks.