Premium
In vivo testing of the role of Alzheimer’s disease coding variants
Author(s) -
Sasner Michael,
Oblak Adrian L.,
Garceau Dylan,
Kotredes Kevin P.,
Preuss Christoph,
Pandey Ravi S.,
Uyar Asli,
Lamb Bruce T.,
Carter Gregory W.,
Howell Gareth
Publication year - 2020
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1002/alz.043896
Subject(s) - transcriptome , biology , trem2 , computational biology , phenotype , disease , genetics , model organism , bioinformatics , medicine , gene , cell , gene expression , myeloid cells
Background Numerous AD risk loci and variants have been identified by large‐scale genetic studies, but few have been functionally verified and models to study their mechanism of action are lacking. The Model Organism Development and Evaluation for Late‐onset AD (MODEL‐AD) Center was created to develop, characterize, and distribute more precise preclinical models for late‐onset AD (LOAD) by engineering these disease‐associated variants into mouse models, and characterizing the resulting phenotypes using clinically relevant assays. Method Diverse criteria including replication in multiple studies, predicted pathogenicity of variant, sequence conservation between human and mouse, and expression in relevant cell types were used to prioritize risk variants, which were then engineered into a mouse model that expresses the AD risk variants APOE4 and Trem2*R47H. Primary screening was done by transcriptomic analysis using the nanoString Mouse AD Panel at 4, 8, and 12 months. Result Bioinformatics studies enabled us to prioritize coding variants in the Abca7, Clasp2, Mthfr, Mtmr4, Plcg2, Shc2, Slc6a17, Snx1, and Sorl1 loci. Knockouts of Abca7 , Ceacam1 , Il1rap , and Plcg2 have also been created to model human loss of function variants. Transcriptomic analysis demonstrated that the Abca7*A1527G and Mthfr*C677T models exhibited some age‐dependent similarities to transcriptomic changes seen in post‐mortem samples from the AMP‐AD cohort. These will move on to deep phenotyping out to 24 months of age. Analysis of other models are still in progress and will be presented. Humanized APP and Tau ( MAPT ) models are also in progress. Conclusion We have prioritized Abca7*A1527G and Mthfr*C677T models to be comprehensively characterized up to 24 months of age in the MODEL‐AD program. We will discuss our approach to prioritize genetic variants, develop novel models relevant to LOAD, and use clinically relevant measures to determine which models are most useful for translational studies. All models are made available for both academic and for‐profit use from The Jackson Laboratory, and all validation data will be shared via the AMP‐AD knowledge portal. For more information see www.model‐ad.org.