Abstract
Alzheimer’s disease (AD) is a neurodegenerative condition responsible for 60-70% of global dementia cases, that presents initially as memory loss before progressing to debilitating cognitive decline across many domains. AD creates an immense economic and societal burden, and despite many decades of research into preventative and curative treatments, current therapies are unable to provide little beyond minimal symptom relief at mid- to late-stage disease. AD pathology is characterised by the extracellular accumulation of amyloid-beta (Aβ) plaques, and intracellular accumulation of hyperphosphorylated tau protein tangles. While it was initially assumed that AD was caused by these aggregates themselves, it is now thought that dysfunction of other upstream processes, like the autophagy lysosomal pathway (ALP), may be significantly contributing to the accumulation of unwanted proteins. Thus, the ALP is now an area of focus for the development of novel AD therapeutics, particularly for preventative or early-stage treatments. Evolution of adeno-associated viral (AAV) vectors that are highly specialised at crossing the blood-brain barrier (BBB) can overcome a number of hurdles for treating neurological disease as they can be administered by a minimally invasive systemic injection, and can spread through the whole central nervous system (CNS) to treat a global brain condition like AD. The vector AAV-PHP.eB was evolved in C57BL/6 mice for this purpose, and produced substantially higher CNS transduction than the existing gold standard, AAV9.
The principal aim of this thesis was to utilise AAV-PHP.eB in a preclinical gene therapy treatment, which was intended to modulate autophagy in the APP/PS1 mouse model of AD. This was achieved through expression of a transgene encoding the master ALP regulator transcription factor EB (TFEB), under the control of a ubiquitous promoter. Because this project planned to use several novel elements, two pilot studies were initially conducted. First, since the particular APP/PS1 mouse model was newly introduced to the laboratory, the electrophysiology of synaptic function and neuropathology of this model was first assessed. Interestingly, synaptic function was not impaired in APP/PS1 compared to control mice, nor did it decline with age, despite an increasing Aβ plaque load with age in APP/PS1 tissue. In a further pilot study, the AAV-PHP.eB vector’s CNS transduction capacity was compared to AAV9’s after administration via multiple routes, and in C57BL/6 as well as B6C3 mice. Here it was shown that AAV-PHP.eB’s enhanced transduction was only afforded in C57BL/6 and not B6C3 mice, and it was no more effective than AAV9 via any administration route except intravenous injection.
Based on this pilot work, the principal study was then begun by administration of AAV-PHP.eB-TFEB via tail vein intravenous injection in APP/PS1 and C57BL/6 wildtype male mice. While this was intended to be carried out in young pre-symptomatic mice, delays meant animals instead ranged from 10-14 mo when treated. Surprisingly, the treatment had almost immediate toxic effects, causing the death of two animals within 12 days of treatment, and marked weight loss in all treated animals regardless of genotype or age. This toxicity occurred in mice treated with AAV-TFEB at both high and low doses but was not replicated in mice treated with the AAV-PHP.eB control vector. Splenic tissue from AAV-TFEB-treated mice had pathology indicative of a peripheral immune system reaction, and while neuropathology marker outcomes were varied, they suggest that intravenous AAV-TFEB treatment at the mid-late development stage of AD-like pathology may have caused further disruption to an already overloaded ALP system. TFEB overexpression was unable to be successfully detected after application of AAV-TFEB in vitro in neural cultures. However, when high dose AAV-TFEB was administered directly into the hippocampus by stereotaxic injection, the associated protein tag could be detected, mice also lost weight, and their hippocampal structure was severely disrupted. At a lower dose, this weight loss and hippocampal damage were not replicated, indicating a dose-related vulnerability of the hippocampus when injecting AAV directly into the brain parenchyma.
Collectively, these data provide some important lessons for preclinical gene therapy applications which may also translate to the clinic. Overcoming the hurdle of achieving strong, but also widespread therapeutic effect by a minimally invasive treatment method will revolutionise the field of gene therapy for AD. However, careful selection of transgenes, mitigation of off-target effects with cell-specific promoters, and immunosuppression will be crucial for the continued use of systemically administered viral vectors, particularly in humans where the risk of previous exposure to AAV is high.