Studying Alzheimer’s disease using the next generation

Alzheimer’s disease (AD) is a prevalent and hugely impactful disease affecting millions. Despite the large public health impact, only two pharmacological treatments have been approved. Both of these treatments target amyloid protein, a key tenant in the amyloid hypothesis which describes the potential aetiology of AD. The amyloid hypothesis is a well-established concept in AD research and is well connected to the genetic risk factors of Mendelian AD. However, non-Mendelian forms of AD are much more common and are likely to be caused by 100-10000s of genetic risk factors in combination with many environmental factors. Prior to the start of this thesis, only tens of genomic regions had been linked with AD, leaving the vast majority of genetic risk factors unidentified. This leaves an opportunity for further exploration of the genetic risk factors of AD and how they combine in biological functions. The work in this thesis aimed to identify genetic risk factors for AD and to highlight biological processes relevant to AD. In Chapter 2, we performed the largest common variant genome wide association study (GWAS) meta-analysis to date and found 38 genomic regions associated with AD, including 7 novel loci. We found that genes enriched for association with AD were overexpressed in human microglia, providing genetic support for the role of microglia in AD. Conditional gene-set analysis highlighted astrocyte activation, immune cell recruitment, amyloid catabolism and neurofibrillary tangles as biological processes relevant to AD. Both of these findings support the idea that the immune system is relevant to AD. In Chapter 3, we performed the first exome wide association study (ExWAS) of proxy AD/dementia to show that using a proxy phenotype based on parental AD/dementia status can capture rare variants associated with AD. This study identified three well known genes enriched in rare variants associated with AD and outlined the potential of including proxy AD/dementia datasets in future rare variant studies. In Chapter 4, we performed common variant GWAS and rare variant ExWAS to compare methods for generating a proxy AD/ dementia phenotype. We showed that there were very small differences in GWAS and ExWAS results from the different phenotypes and suggested that the choice of proxy phenotype methods was unlikely to influence loci discovery in downstream meta-analyses. In Chapter 5, we performed a multivariate meta-analysis and local genetic correlation analyses to look for genetic risk factors common to four neurodegenerative disorders. We were unable to find shared genetic risk factors for AD, amyotrophic lateral sclerosis, Lewy body dementia, and Parkinson’s disease; however, we did find that the human leukocyte antigens region and TMEM175 were relevant to three of the four diseases. This suggests that the immune system and lysosome maintenance may be relevant for many neurodegenerative disorders. Overall, the research in this thesis identified novel genomic regions relevant to AD, highlighted the potential of using proxy phenotypes to aid in identifying genomic regions relevant to AD, and provided support for the role of the immune system and protein degradation in AD and other neurodegenerative disorders.