Researchers at Army Medical University (Third Military Medical University) in Chongqing, China report that liver-directed delivery of the APOE3 Christchurch variant reduced amyloid burden, tau pathology, and neuroinflammation in a transgenic mouse model of APOE4 Alzheimer's disease, according to a study published May 6, 2026 in Neuron. The findings position peripheral APOE manipulation as a mechanistically distinct approach to APOE4-associated disease.
Rationale for targeting the liver in Alzheimer's disease research
APOE4 is the strongest known genetic risk factor for late-onset Alzheimer's disease, yet no approved therapy directly addresses its downstream consequences. The liver is the primary source of circulating apolipoprotein E, and peripheral APOE participates in the clearance of amyloid-beta from the bloodstream. This liver-brain axis has attracted growing interest as a route for therapeutic intervention that avoids the complexity of direct central nervous system gene delivery.
The study tested whether replacing hepatic APOE4 activity with apolipoprotein E3 treatment — specifically the rare APOE3 Christchurch variant — could reduce brain pathology by enhancing peripheral amyloid-beta reduction. APOE3 Christchurch carries a single arginine-to-serine substitution at position 136 (R136S), a change that has been associated with dramatic protection against Alzheimer's disease in a previously reported case involving a Colombian family carrying a high-penetrance presenilin mutation.
Study design and experimental approach
The study is entirely preclinical. Researchers used APP/PS1 transgenic mice expressing human APOE4, a model that develops amyloid pathology with age. Adeno-associated virus serotype 8 (AAV8) carrying a liver-specific thyroxine-binding globulin promoter was used to drive hepatic expression of either human APOE3, human APOE3 Christchurch (R136S), or GFP as a control. Mice were injected at four months of age and assessed at ten months, providing a six-month treatment window that preceded and then spanned the period of active plaque accumulation.
Cognitive outcomes were measured using the Morris Water Maze, novel object recognition, and Y-maze tests. Neuropathological endpoints included amyloid plaque immunostaining, soluble and insoluble amyloid-beta levels by ELISA, phosphorylated tau, and markers of astrogliosis and microglial activation. Bulk RNA sequencing was used to characterize neuroinflammatory gene expression. Peripheral clearance was assessed through ex vivo phagocytosis assays using primary hepatocytes, Kupffer cells, splenic macrophages, and the RAW264.7 monocyte cell line. In vivo behavioral and neuropathological assessments were conducted in male mice only.
Key results: cognition, amyloid, and neuroinflammation
Mice receiving APOE3 Christchurch via liver-directed gene therapy showed improved performance across all three cognitive assessments compared with GFP controls. In the Morris Water Maze, APOE3 Christchurch-treated animals displayed reduced escape latency and improved spatial memory, approaching levels observed in wild-type mice. Novel object recognition and Y-maze spontaneous alternation scores were also higher in the APOE3 Christchurch group. Standard APOE3 treatment produced intermediate or non-significant improvements on most measures, consistently underperforming relative to the Christchurch variant.
Amyloid pathology was reduced in cortical and hippocampal tissue of APOE3 Christchurch-treated animals, with lower soluble and insoluble amyloid-beta 40 and 42 levels detected by ELISA. Neuroinflammatory markers were attenuated: GFAP immunoreactivity, a marker of astrogliosis, and IBA1 staining, reflecting microglial activation, were both reduced. Bulk RNA sequencing revealed downregulation of pro-inflammatory transcriptional programs in APOE3 Christchurch brains, with gene expression profiles more closely resembling wild-type animals than either control or standard APOE3 groups. Phosphorylated tau at the p-Tau231 epitope was also reduced, and neuronal density, synaptic markers including PSD95, and dendritic integrity markers including MAP2 were preserved.
Peripheral clearance as the operative mechanism
A central finding of the study concerns the mechanism through which hepatically expressed APOE3 Christchurch produces brain-level effects. The researchers report that APP processing enzymes, amyloid-beta transport receptors, and brain-resident degrading enzymes were unchanged across groups, ruling out altered central production or degradation as explanatory factors.