Aims

Late onset Alzheimer’s disease (LOAD) is genetically complex but GWAS studies and histopathological data consistently implicate endocytic dysfunction which is an early disease phenotype (1,2,3). Studying the sporadic disease is hard using conventional models where relatively few genetic manipulations can be made. It remains essential for us to understand how multiple LOAD risk genes come together to drive the disease.

Methods

We have developed a novel endocytic pathway specific polygenic risk score (PRS) and applied it to a genetically well characterised cohort of LOAD individuals. We have stratified the patient cohort and selected individuals with high endocytic PRS and reprogrammed 20 PBMC samples to induced pluripotent stem cell lines to compare to healthy age matched controls.

Results

Our endocytic PRS can be used to predict LOAD with >70% accuracy alone or increased to >90% when combined with biomarker data. We have demonstrated that the PRS is not impacted by APOE status and suggest there are further endocytic risk genes to be discovered in association with AD. We have used these lines to examine underlying disease mechanisms by differentiating 20 lines from individuals with high endocytic PRS and 12 from healthy age matched controls with low PRS to microglia and have carried out transcriptomics, proteomics, live imaging assays to assess endocytic and phagocytic behaviour including amyloid beta uptake and assessed microglial interactions with neurons in coculture. This body of data is currently being analysed but will provide the first insights into how complex genetic changes associated with LOAD drive disease mechanisms which is key to developing better therapeutic targets.

Conclusions

Our work demonstrates it is possible to stratify patient cohorts on the basis on underlying disease mechanisms. This is important for studying the molecular basis of disease and for identifying patients most likely to respond to specific treatment strategies. We have demonstrated our endocytic pathway specific PRS has high accuracy at predicting LOAD, is not impacted by APOE status and suggests further endocytic risk genes are yet to be identified. We have developed a novel iPSC resource to study endocytic mechanisms of LOAD and have undertaken deep phenotypic characterisation of iPSC-derived microglia.