Cerebrospinal fluid proteome profiling using machine learning shows a unique protein signature associated with APOE4 genotype

Artur Shvetcov,Shannon Thomson,Ann-Na Cho,Heather M Wilkins,Joanne H Reed,Russell H Swerdlow,David A Brown,Alzheimer's Disease Neuroimaging Initiative,Caitlin A Finney
DOI: https://doi.org/10.1101/2024.04.18.590160
2024-06-16
Abstract:APOE4 is the biggest genetic risk factor for Alzheimers disease (AD). Proteome-wide changes independent of AD brain pathology and whether these extend to APOE4 carriers irrespective of cognitive status remain unknown. To investigate APOE4-associated proteome changes in people with and without AD. Patient clinical, APOE genotype, and cerebrospinal fluid (CSF) proteome data for 735 participants was sourced from the Alzheimers Disease Neuroimaging Initiative (ADNI) database. Participants with no cognitive impairment, mild cognitive impairment (MCI), and AD were included. Diagnosis was defined using cognitive assessments. Supervised machine learning (classification and regression trees; CART) was used for proteome profiling to identify protein signatures. Enrichment analyses for brain regions and cells and peripheral immune cells were performed using NetworkAnalyst 3.0 and the Human Protein Atlas. CART revealed an APOE4 specific proteome signature consisting of 7 proteins that were independent predictors of APOE4 carriers (sensitivity, specificity, PPV, NPV, and AUC = 1.0) and 50 proteins that interacted together for prediction (sensitivity = 0.99, specificity = 0.74, PPV = 0.86, NPV = 0.98, AUC = 0.92). This signature was found across APOE4 carriers independently of diagnosis and was associated with an increased risk of progression to cognitive impairment over time. Proteins within the signature were enriched in brain regions including the caudate and cortex. Enriched brain cells included endothelial cells, oligodendrocytes, and astrocytes. Enriched peripheral immune cells included T cells, macrophages, and B cells. We identified an APOE4 specific CSF proteome signature of 57 proteins that was independent of cognitive status and was associated with an increased risk of progression to cognitive impairment. It was also associated with a strong immune and inflammatory phenotype across the brain and periphery. Our findings suggest that the APOE4 proteome signature is independent of AD-specific brain pathology and likely underlies APOE4 carriers vulnerability to cognitive decline and AD.
Neuroscience
What problem does this paper attempt to address?
The problem that this paper attempts to solve is: whether the changes in the cerebrospinal fluid (CSF) proteome are unique in carriers of the APOE4 genotype in the absence of Alzheimer's disease (AD) pathological features. Specifically, the research aims to explore: 1. **The relationship between the APOE4 genotype and proteome changes**: Whether APOE4 variant carriers have a specific proteome signature independent of Alzheimer's disease pathology and diagnosis. 2. **Proteome features independent of cognitive status**: Whether this proteome signature exists in APOE4 carriers with different cognitive statuses (including Alzheimer's disease, mild cognitive impairment, and no impairment). 3. **Progression risk assessment**: Whether this proteome signature in APOE4 carriers is associated with an increased risk of cognitive decline. By using machine - learning methods to analyze clinical, genotypic, and proteomic data from the Alzheimer's Disease Neuroimaging Initiative (ADNI) database, the research team hopes to reveal the unique proteome features of APOE4 carriers and explore whether these features are independent of Alzheimer's disease pathological markers and whether they are associated with an increased risk of cognitive decline. ### Main problem summary: - **Does an APOE4 - specific proteome signature independent of Alzheimer's disease pathology exist?** - **Does this feature exist in APOE4 carriers with different cognitive statuses?** - **Is this feature in APOE4 carriers associated with a higher risk of cognitive decline?** By answering these questions, the research team hopes to better understand how the APOE4 genotype affects proteome changes and its relationship with cognitive decline, thereby providing new directions for future research and treatment.