Investigating the effects of mutations causative for early-onset familial Alzheimer’s disease using zebrafish as a model organism

Abstract

Development of an effective therapeutic for Alzheimer’s disease (AD) is currently a global health priority. To prevent, or at least delay, the onset of AD, we must understand the initial cellular stresses/changes that drive the disease. These initiating changes are likely subtle and occur decades before symptom onset. We cannot easily investigate these changes in humans, as pre-symptomatic brain material from individuals genetically pre-disposed to AD is inaccessible for detailed molecular analyses. For this, we must utilise animal models. The most frequently used animal models of AD are mice expressing one or more transgenes containing the sequences of human genes bearing mutations which cause AD. These transgenic models have been useful in elucidating some aspects of the pathogenic mechanisms of AD. However, they have not led to development of successful therapeutics. Mutations in a small number of genes cause early-onset familial forms of AD (EOfAD). These mutations can be introduced into the orthologous, endogenous genes of an animal (i.e. knock-in models). However, relatively few papers describe research with knock-in models. Transcriptome analysis is currently the most detailed form of molecular phenotyping and can give a largely unbiased view of the molecular state of the brains of young knock-in models of EOfAD. Surprisingly, this has not previously been performed using knock-in models of EOfAD-like mutations. To address this gap in our knowledge, the work presented in this thesis (along with previous work from the Alzheimer’s Disease Genetics Laboratory (ADGL)), describes the generation, and/or characterisation of a collection of zebrafish knock-in models of EOfAD-like mutations. The power of zebrafish as a model organism lies in this species’ ability to generate large families of synchronous siblings which can be raised together in the same tank. This has allowed the assessment of the effects of heterozygosity for EOfAD-like mutations (closely mimicking the genetic state of human EOfAD) or the effects of non-EOfAD-like mutations (such as frameshift mutations in presenilin genes) on the brain transcriptome with minimal external sources of “noise.” These analyses have revealed that the only cellular process predicted to be affected by EOfAD-like mutations in the heterozygous state, and not by non-AD-related mutations, is oxidative phosphorylation. Comparison of these transcriptomes with recent, publicly available brain transcriptomes from two knock-in mouse models of late onset AD risk alleles revealed similar affected processes, thereby supporting the findings from the zebrafish models. Preliminary non-transcriptomic characterisations of previously generated/novel zebrafish models were also performed. The effects of heterozygosity for EOfAD-like mutations on brain vasculature were assessed, as well as effects on spatial working memory. Only limited differences were observed in these studies. However, future work with greater statistical power and/or alternate study designs is recommended. Overall, the research described in this thesis demonstrates the value of unbiased, transcriptome analyses of young, knock-in animals models for understanding the early stages of AD pathogenesis.