Liver-derived protein helps protect brain cells and improves memory in Alzheimer's mice

By Vijay Kumar MalesuReviewed by Susha Cheriyedath, M.Sc.Nov 12 2026

Scientists uncover how proteins from the bloodstream reach specific brain cells, raising new questions about what happens when this communication falters during the early stages of Alzheimer's disease.

Study: The blood-brain barrier imports diverse neuroprotective proteins through pathways impaired early in Alzheimer’s disease models. Image Credit: AtlasStudio / Shutterstock

In a recent study published in the journal Cell, researchers characterized blood-to-brain protein transport, identified its alterations in Alzheimer’s disease models, and investigated the neuroprotective effects of liver-derived apolipoprotein A1 (APOA1).

Scientists from Gladstone Institutes, the University of California, San Francisco, the Buck Institute for Research on Aging, and Emory University in the United States conducted the research in collaboration with the Hong Kong Center for Neurodegenerative Diseases and the Hong Kong University of Science and Technology in China.

Background

The brain is not completely isolated from the rest of the body. The blood-brain barrier (BBB) protects the central nervous system (CNS) by restricting the entry of circulating substances, while regulating the passage of proteins and other necessary substances.

Estimates suggest that more than 98% of small-molecule drugs and most biologics cannot cross the BBB. This poses a major challenge for developing therapies for neurological disorders.

Recent studies suggest that proteins produced outside the nervous system influence brain health and neurodegeneration. Yet, the proteins transported across the BBB and their functions remain incompletely understood. Further research is needed to characterize these transport pathways and how they are disrupted in neurodegenerative diseases.

About the study

The researchers developed a bioorthogonal proteomics approach to identify circulating proteins entering the brain. Plasma proteins from albumin-deficient mice were labeled with trans-cyclooctene (TCO) and injected intravenously into wild-type mice three times over six hours.

Brain tissues were collected, and labeled proteins were captured using tetrazine-PEG4-biotin followed by streptavidin-based enrichment. Liquid chromatography-tandem mass spectrometry was used to identify proteins in brain vessels and brain parenchyma. Proteins detected across these compartments were classified according to their transport patterns.

The researchers also analyzed protein uptake using immunofluorescence, flow cytometry, cultured brain cells, and single-cell ribonucleic acid (RNA) sequencing. Functional experiments evaluated APOA1-mediated cholesterol transport in astrocytes and its effects on neuronal synaptic function.

To study disease-associated pathological changes, the authors compared wild-type (healthy) mice, lipopolysaccharide-treated mice, and 18-week-old male 5xFAD (Alzheimer’s disease) mice. They then increased hepatic APOA1 using an adeno-associated virus serotype 8 (AAV8) vector.

Changes in brain pathology, inflammation, synaptic integrity, and amyloid deposits were assessed using molecular, biochemical, and imaging techniques, while separate behavioral tests evaluated locomotor activity and memory.

Study results

The proteomic analysis identified 191 plasma proteins that interacted with the brain vasculature; under the experimental conditions, 92 crossed the BBB and entered the brain parenchyma. The remaining 99 proteins were detected only in brain vessels, while 670 proteins showed no detectable BBB interaction.

The transported proteins originated predominantly from the liver, adipose tissue, and intestine. They were enriched for carrier proteins, modulators of protein-binding activity, and enzymes involved in lipid and carbohydrate metabolism. Their transport was associated with specific molecular characteristics, including higher lysine and alpha-helical content.

APOA1 preferentially accumulated in astrocytes, whereas alpha-1 antitrypsin (SERPINA1) predominantly entered microglia. Hemopexin (HPX) and alpha-2-HS-glycoprotein (AHSG) were also detected in brain vessels and parenchymal tissues.

Further experiments demonstrated that scavenger receptor class B type 1 (SR-B1) mediated APOA1 uptake by brain endothelial cells. After crossing the BBB, APOA1 interacted with astrocytic ATP-binding cassette transporter A1 (ABCA1). This receptor-mediated hand-off helped direct APOA1 toward astrocytes.

APOA1 treatment altered astrocytic gene expression, upregulating pathways associated with cholesterol efflux and high-density lipoprotein particle assembly while downregulating those involved in lipid synthesis and storage. It reduced intracellular lipid levels in astrocytes and increased cholesterol release into astrocyte-conditioned medium.

Blocking ABCA1 prevented APOA1 uptake and its lipid-regulating effects in cultured astrocytes. Medium collected from APOA1-treated astrocytes increased neuronal cholesterol levels, postsynaptic density protein 95 (PSD-95) expression, and dendritic spine density, indicating improved synaptic characteristics. Direct APOA1 treatment did not reproduce these neuronal benefits, indicating that astrocyte-mediated cholesterol transport contributed to the synaptic effects. APOA1 also promoted greater cholesterol efflux and preserved synaptic markers more effectively than lipid-free APOE4 under amyloid-beta stress in cell-culture experiments. SERPINA1, meanwhile, promoted a disease-associated microglial state and increased phagocytosis of experimental particles and amyloid-beta peptides.

Lipopolysaccharide-induced inflammation resulted in widespread, nonspecific protein entry. Early-stage 5xFAD mice showed selective transport changes without widespread disruption of the physical barrier. APOA1 uptake decreased, especially in astrocytes, and SERPINA1 uptake increased in microglia. Reduced SR-B1 in brain endothelial cells and reduced ABCA1 in astrocytes were associated with impaired APOA1 delivery and uptake.

To investigate whether increasing APOA1 could improve brain function, the researchers increased its production in the liver of 5xFAD mice. Following treatment, APOA1 levels increased in the liver, blood, and brain. Astrocytic lipid accumulation and glial fibrillary acidic protein (GFAP) levels decreased, as did brain interleukin-1 beta (IL-1β) concentrations. Brain tumor necrosis factor-alpha (TNF-α) levels remained unchanged. Soluble vascular amyloid-beta 42 (Aβ42) levels and hippocampal dentate gyrus plaque burden also decreased, but soluble parenchymal amyloid-beta and plaque burden in other examined regions remained unchanged. Synaptic density increased in the hippocampus and cortex.

Behavioral assessments further demonstrated reduced hyperactivity and improved preference for a novel arm, indicating better spatial memory. Novel-object preference improved, whereas novel-place recognition did not change significantly.

Conclusions

The findings establish the BBB as an active transport interface that delivers circulating proteins to specific brain cell populations. Selective disruption of these pathways occurs early in the 5xFAD Alzheimer's disease mouse model, reducing the delivery and cellular uptake of neuroprotective APOA1.

Increasing liver-derived APOA1 production raised its availability in the brain, improved lipid homeostasis in astrocytes, reduced selected inflammatory and amyloid-related abnormalities, increased synaptic density, and improved some behavioral outcomes in the mice.

These findings identify the liver-plasma-brain APOA1 pathway as a potential therapeutic target. However, further investigations are needed to determine whether these mechanisms and treatment effects translate to humans and other neurological disorders. 

The proteomics screen may have missed some proteins, and the role of astrocytic ABCA1 in living mice and the effects of APOA1 treatment at advanced disease stages remain uncertain.