Computational identification of precision therapeutics for Alzheimer’s disease specific to endocrine and APOE genotype status

Abstract

AbstractBackgroundAlzheimer’s disease (AD) affects 6.5 million people in the US, two‐thirds of which are women. Despite extensive research, there are currently no highly effective treatments for AD. The greatest risk factors for AD are age, chromosomal female sex and APOEε4 genotype. A key driver of chronological and endocrinological aging is the activation of innate and adaptive immune systems occurring in midlife female brain. To effectively prevent and treat AD, it is crucial to develop computational methods that take into account these factors and screen for personalized therapeutics.MethodRNAseq data of 15‐month acyclic vs 9‐month regular APOEε3/ε3 and APOEε4/ε4 female mouse models were analyzed to identify APOE‐specific Differentially Expressed Genes (DEGs, pvalue<.01). STRING Protein‐Protein Interaction (PPI) database was used to extract protein interactors for the two APOEε3/ε3 and APOEε4/ε4 DEG lists and construct two APOE‐specific AD PPI networks. Target Proteins (TPs) candidates in each network were identified using the DrugBank database by selecting only proteins targeted by at least one FDA‐approved drug. A score of drug synergy was developed to rank FDA‐approved drugs based on the number of target proteins in‐ vs out‐ of the APOE‐AD specific networks and their distance to DEG nodes.ResultFrom RNAseq analyses, we identified a total of 265 DEGs for APOEε3/ε3 and 452 DEGs for APOEε4/ε4 genotype. The APOE‐AD‐specific PPI networks resulted in 1,711 nodes (proteins) and 21,445 edges (PPIs) for the APOEε3/ε3 genotype and 2,841 nodes and 35,493 edges for APOEε4/ε4 genotype. Using the DrugBank database, 342 and 568 potential therapeutic targets were identified in the APOEε3/ε3 and APOEε4/ε4 networks, respectively. Finally, drug synergy score enabled ranking 99 therapeutics for APOEε3/ε3 network and 102 drugs for APOEε4/ε4 network, of which 19 resulted in both lists.ConclusionThis study presents a network‐based approach for identifying personalized therapeutics for AD based on endocrine transition changes and APOE genotype. The results suggest that there is greater transcriptomic dysregulation in female APOEε4/ε4 brains compared to APOEε3/ε3 brains. The proposed computational approach has the potential to advance precision medicine for the prevention and treatment of AD, and future research will focus on identifying the most effective combinations of therapeutics.