Genetic background influences the 5XFAD Alzheimer's disease mouse model brain proteome.

• Authors: Cheyenne D Hurst, Emory University; Amy R Dunn, Jackson Laboratory, Bar Harbor, ME; Eric B Dammer, Emory University; Duc M Duong, Emory University; Nicholas T Seyfried, Emory University; Catherine C Kaczorowski, University of Michigan; Erik C. B Johnson, Emory University
• Journal Title: bioRxiv
• Publisher: CSH | 2023-06-13
• Type of Work: Article | Preprint: Prior to Peer Review
• Publisher DOI: http://doi.org/10.1101/2023.06.12.544646
• Subjects: Chemistry, Biochemistry; Biology, Neuroscience
• Keywords: Alzheimer’s disease; Mouse models; Proteomics; 5XFAD; B6xD2; Translational

Abstract

There is a pressing need to improve the translational validity of Alzheimer's disease (AD) mouse models. Introducing genetic background diversity in AD mouse models has been proposed as a way to increase validity and enable discovery of previously uncharacterized genetic contributions to AD susceptibility or resilience. However, the extent to which genetic background influences the mouse brain proteome and its perturbation in AD mouse models is unknown. Here we crossed the 5XFAD AD mouse model on a C57BL/6J (B6) inbred background with the DBA/2J (D2) inbred background and analyzed the effects of genetic background variation on the brain proteome in F1 progeny. Both genetic background and 5XFAD transgene insertion strongly affected protein variance in hippocampus and cortex (n=3,368 proteins). Protein co-expression network analysis identified 16 modules of highly co-expressed proteins common across hippocampus and cortex in 5XFAD and non-transgenic mice. Among the modules strongly influenced by genetic background were those related to small molecule metabolism and ion transport. Modules strongly influenced by the 5XFAD transgene were related to lysosome/stress response and neuronal synapse/signaling. The modules with the strongest relationship to human disease-neuronal synapse/signaling and lysosome/stress response-were not significantly influenced by genetic background. However, other modules in 5XFAD that were related to human disease, such as GABA synaptic signaling and mitochondrial membrane modules, were influenced by genetic background. Most disease-related modules were more strongly correlated to AD genotype in hippocampus compared to cortex. Our findings suggest that genetic diversity introduced by crossing B6 and D2 inbred backgrounds influences proteomic changes related to disease in the 5XFAD model, and that proteomic analysis of other genetic backgrounds in transgenic and knock-in AD mouse models is warranted to capture the full range of molecular heterogeneity in genetically diverse models of AD.