Brain-Wide Gene Therapy Restored Memory in Alzheimer's and Old Mice

This is animal research, not a human trial. It was reported in 2026 in Alzheimer's & Dementia: Translational Research & Clinical Interventions by Badesso and colleagues. The experiments used mice, and the findings say nothing yet about whether the approach is safe or works in people.

The team focused on an enzyme called PLA2G4E. In the brain it is found only in neurons. It helps make N-acyl phosphatidylethanolamines, fats that are precursors to natural compounds with protective and antioxidant properties. The authors had already found that raising this enzyme in one brain region helped older mice from a different Alzheimer's model.

In this study, they first used a standard viral vector to deliver the enzyme to the hippocampus, a memory region, in 7to 8-month-old APP NL-G-F mice. These mice carry changes that mimic Alzheimer's disease. Two months later, untreated mice of this type did worse than healthy mice on a fear memory test and a novel object recognition test. Mice given either the human or the mouse version of the enzyme did better. The paper describes the fear memory rescue as partial. On the object test, both versions significantly improved performance.

The hippocampus-only approach reached a limited area. So the team switched to a vector called AAV9P31, which is built to cross the blood-brain barrier. It was given by injection behind the eye, and it produced widespread expression across the brain, including cortical layers and the hippocampus. A dose-testing step showed that more vector gave more enzyme.

In aged normal mice, 18 to 20 months old, treated animals did significantly better on the fear memory test two months after injection. Brain scans showed higher glucose uptake across the whole brain, a sign of greater neuronal activity and metabolic function. The regional scans alone showed no significant differences. Levels of phosphorylated tau, a chemical change linked to tangles in Alzheimer's, were significantly lower in treated mice.

In Alzheimer's-model mice treated at 4 to 5 months old, results depended on timing. At two months after injection, memory improved only in males. The authors suggest females were still performing well at that age. At six months after injection, when the mice were 12 to 14 months old, the treatment fully reversed the memory deficits. Only female mice were analyzed at that point, because the males were used for fat analysis. The authors call the effect robust and long-lasting.

Amyloid did not change. Levels of the Aβ42 protein and staining for plaques were the same in treated and untreated mice. So the benefit appears to occur without clearing amyloid. In the earlier hippocampal experiment, tau changes were not statistically significant, so the tau finding is clearest in the brain-wide, aged-mouse work.

Brain fat analysis showed that disease-linked changes in lipids were partly normalized. The authors say this may help explain the recovery, but they note that the exact mechanisms are still unknown. They also saw shifts in proteins involved in glutamate signaling, a key chemical messaging system for learning and memory.

The limits are clear. These are mice, group sizes are not given in the portion of the paper available to us, and the authors list long-term safety, the cell types involved and the specific fat molecules as open questions. They also want to know whether a drug, rather than a gene therapy, could mimic the effect.

The finding is a different angle from lifestyle research. Our October 17 report on volunteering and slower memory decline in APOE ε4 carriers was an observational study in people, which cannot show cause. This mouse work adds a possible biological route, but the two studies are not connected.

What to watch next

Tests of long-term safety, studies that show which cells and fats matter, and whether a drug can raise PLA2G4E activity.

Two-Year Fish Oil Trial Finds No Memory or Brain-Shrinkage Benefit Detailed Data Confirm Personalized mRNA Melanoma Vaccine Cut Recurrence Risk 42% Blood Signatures of Fitness Linked to Lower Death Risk in UK Biobank Change in AI Face Age Between Photos Linked to Cancer Survival Untreated Sleep Apnea Linked to 69% Faster Cognitive Decline Over 10 Years Over 80 Nutrition Experts Urge White House to Release Ultra-Processed Food Definition Beetroot Juice and Postmenopausal Heart Oxygen Balance After Hard Exercise: A Small Trial Topical Urolithin A Lowered Eye Wrinkle Depth in Randomized Studies Short Breathing Exercises Blunted Heart Rate and Anxiety Under Stress Aerobic and Resistance Training Ended Up Equal for Type 2 Diabetes Plant and Dairy Protein Shakes Left Blood Chemistry Alike in Adults Over 50 Intensity Relative to Your Own Fitness Better Explained Fitness in Midlife Women Blood Flow Restriction Training Showed No Overall Advantage for Artery Health Cigarette Smoke Disrupts Eye Cell Mitochondria, Pointing to Macular Degeneration Risk Ultra-Processed Food Intake Leaves Distinct Blood Fat and Metabolite Signatures Gut-Made Kidney Toxins Linked to Tiny Brain Bleeds in Mice and Patients Flo App Survey of 17,494 Women Finds Gaps in Perimenopause Knowledge New Aging Framework Names Epigenetic Change the Dominant Driver of Aging Exercise Helps Blood Cancer Patients During Treatment, but Diet Evidence Lags Review: Protein Supplements Often Unnecessary Outside the Lab for Active Adults Ear Vagus Nerve Stimulation Linked to Less Inflammation After Surgery Fasting and Lean Mass Over 45: Short Trials, Small Differences Fragmented Sleep Tracked by Wrist Monitors Linked to More Brain Amyloid at Autopsy Blood Test for p-tau217 Predicted Decline Over 12 Years Steady Aerobic Exercise Raised Fitness on Average in Heart Disease Patients Rising Arterial Stiffness Linked to Progression of Heart-Kidney-Metabolic Disease Brain-Wide Gene Therapy Restored Memory in Alzheimer's and Old Mice Men With Heart and Metabolic Disease Show Faster Epigenetic Aging AI Spotted Heart Disease From Face and Tongue Photos in Small Test