Switching APOE4 to APOE2 in adult mice reduces Alzheimer’s pathology and improves memory

By Vijay Kumar MalesuReviewed by Susha Cheriyedath, M.Sc.Nov 12 2025

By flipping a single genetic switch from APOE4 to APOE2 in adult mice, researchers reversed key Alzheimer’s-like changes in the brain while revealing the metabolic trade-offs that future APOE-targeted therapies must safely navigate.

Study: APOE4 to APOE2 allelic switching in mice improves Alzheimer’s disease-related metabolic signatures, neuropathology and cognition. Image Credit: Lightspring / Shutterstock

In a recent study published in the journal Nature Neuroscience, a group of researchers investigated whether switching apolipoprotein E4 (APOE4) to apolipoprotein E2 (APOE2) in adult mice improves Alzheimer’s disease-relevant brain pathology and cognition, while also defining systemic lipid effects.

Genetic Risk Landscape of APOE Variants

One in three families knows someone living with memory loss, and late-onset Alzheimer’s disease affects tens of millions worldwide. The strongest common genetic risk factor is APOE: the E4 isoform raises risk up to thirteen-fold in homozygotes, while the E2 isoform lowers risk and slows decline.

Scientists have proposed gene-editing strategies to neutralize APOE4 or boost APOE2, however, in vivo tests of whole-body and cell-specific switching have been limited. What people want to know is simple: if the gene could be switched in adulthood, would the brain actually improve?

Further research is needed to evaluate efficacy, safety, and target cell types.

Engineering the APOE4-to-APOE2 Switch Model

Investigators engineered a knock-in "APOE switch" mouse line (APOE4s2) carrying a floxed human APOE4 coding exon followed by human APOE2. A tamoxifen-activable Cre recombinase Estrogen Receptor T (CreERT) enabled temporal control of allelic replacement.

Global switching (APOE4s2G) was achieved by crossing to a ubiquitously expressed CreERT; astrocyte-specific switching (APOE4s2A) used Aldehyde Dehydrogenase 1 Family Member L1 (Aldh1l1) promoter-driven Cre recombinase Estrogen Receptor T2 fusion (CreERT2) (Aldh1l1-CreERT2) to primarily restrict replacement to astrocytes in the central nervous system, though some peripheral expression occurred.

Validation of Allele Conversion and Systemic Lipid Profiling

Allele conversion was validated in the liver and brain using quantitative polymerase chain reaction (qPCR) and liquid chromatography–tandem mass spectrometry (LC-MS/MS) peptide profiling of APOE in both brain and plasma.

Systemic lipid effects were profiled under normal chow and Western diets with lipoprotein fractionation of very-low-density lipoprotein (VLDL), low-density lipoprotein (LDL), and high-density lipoprotein (HDL), and enzyme-linked immunosorbent assay (ELISA) quantification of ApoE (protein) and triglycerides.

Cerebral Lipidomics and Cell-Specific Transcriptomics

Cerebral lipidomics used untargeted LC-MS/MS and weighted gene coexpression network analysis (WGCNA) focusing on phosphatidylcholine (PC), phosphatidylethanolamine (PE), ceramide (CER), and lyso-phosphatidylcholine (LPC) species.

Whole-brain single-cell ribonucleic acid sequencing (scRNA-seq) defined differentially expressed genes and Gene Ontology pathways across astrocytes, microglia, oligodendrocytes, endothelial cells, and others. For disease relevance, APOE4s2A mice were crossed with the five familial Alzheimer’s disease mutation model (5xFAD).

Cognition was tested by fear conditioning and the Morris water maze. Histopathology quantified amyloid-beta (Aβ 40/42), glial fibrillary acidic protein (GFAP), ionized calcium-binding adapter molecule 1 (IBA1), major histocompatibility complex class II (MHC-II), plaque-associated ApoE, cerebral amyloid angiopathy (CAA), and zonula occludens-1 (ZO1).

Global Replacement Effects on Lipids and Brain Pathways

Global induction converted APOE4 to APOE2 efficiently at both the messenger RNA (mRNA) and protein levels in the liver and brain, as demonstrated by LC-MS/MS, which revealed that the vast majority of detected ApoE peptides were unique to ApoE2 post-switch.

Systemically, global switching elevated plasma ApoE and triglycerides and increased VLDL fractions, recapitulating features of type III hyperlipoproteinemia seen in a subset of APOE2 homozygous humans. In the brain, untargeted lipidomics revealed selective remodeling, with numerous phosphatidylcholine and ceramide species altered following the switch.

A WGCNA blue lipid module, enriched for PC and PE, correlated with the transition from ApoE4 to ApoE2, indicating coordinated rewiring of membrane and signaling lipids important for synaptic function and glial communication.

Adult Transcriptional Shifts After APOE Editing

ScRNA-seq revealed that a one-month switch from APOE4 to APOE2 in adulthood altered brain transcriptional programs across multiple cell types, most notably astrocytes, oligodendrocytes, microglia, and endothelial cells.

Differentially expressed genes overlapped with late-onset Alzheimer’s disease lists from human datasets and converged on metabolism, redox control, neurotransmission, and cytoskeleton or vascular integrity pathways. These findings demonstrate that signatures often attributed to lifelong APOE4 expression are surprisingly malleable in adulthood.

Astrocyte-Restricted APOE2 Replacement Effects

Because astrocytes are the principal source of central nervous system ApoE, the team tested an astrocyte-restricted replacement from APOE4 to APOE2. Cell sorting and qPCR confirmed efficient, cell-specific switching with a modest shift at the whole-brain level, consistent with astrocyte predominance.

Astrocyte-only switching recapitulated many of the global transcriptional changes, including substantial overlap with late-onset Alzheimer’s disease gene sets and pathway terms related to neurotransmission, redox status, and metabolite transport.

Notably, microglia and oligodendrocytes that still expressed APOE4 nevertheless exhibited numerous differentially expressed genes, indicating non-cell-autonomous effects of astrocyte-derived ApoE2.

APOE2 Switching Outcomes in the 5xFAD Model

In the 5xFAD model, astrocyte-only switching to ApoE2 resulted in improved fear conditioning, with female mice exhibiting stronger improvements in contextual and cued memory, but with limited effects in the Morris water maze. Amyloid pathology decreased: plaque burden fell (with higher baseline in females), and Aβ 40 and Aβ 42 by ELISA were reduced.

Cerebral amyloid angiopathy was not reduced, and tight junction protein ZO1 was unchanged. Gliosis dampened: GFAP and IBA1 declined, activated-response microglia decreased, plaque-associated ApoE diminished, and microglial disease-associated markers Triggering Receptor Expressed On Myeloid cells 2 (TREM2) and C-type Lectin Domain Containing 7A (CLEC7A) were downregulated.

Therapeutic Implications of Post-Developmental APOE Editing

Inducible post-developmental replacement of APOE4 with APOE2 in adult mice rapidly remodels metabolism, rewires lipid classes central to membranes and signaling, and reprograms cell-type-specific brain transcriptomes. Astrocyte-restricted switching alone lowers parenchymal amyloid, reduces plaque-proximal gliosis, shifts microglia away from activated phenotypes, lessens plaque-associated ApoE, and improves associative memory.

These data suggest that post-developmental, cell-targeted APOE editing could modify multiple Alzheimer’s disease pathways at once. Translation will require balancing central nervous system benefits with peripheral risks, such as type III hyperlipoproteinemia and other APOE2-associated disorders (e.g., melanoma, age-related macular degeneration). This will involve considering sex-specific responses and defining dose, timing, durability, and delivery strategies for safe and equitable clinical application.