Alzheimer's disease (AD) is the most common age-related neurodegenerative disorder, currently afflicting around 900,000 patients in the United Kingdom. However, therapeutic avenues and factors that can exacerbate AD pathologies remain unclear. Hence, there is an urgent need for a better understanding of AD at the molecular, organellar, cellular, and behavioural levels.

Here, I integrated insights from a fruit fly model of AD with data from AD patients. The accumulation of misfolded amyloid-β (Aβ) peptides as plaques is one of the main histopathological hallmarks of AD. In Chapter 3, I analysed the metabolomic changes in flies expressing human Aβ and uncovered alterations in nicotinamide adenine dinucleotide (NAD⁺) metabolism. NAD⁺, a component of vitamin B, is essential for both mitochondrial bioenergetics and DNA repair through NAD⁺-consuming poly (ADP-ribose) polymerases (PARPs). I found that increasing the bioavailability of NAD⁺ pharmacologically or by suppressing PARP is neuroprotective in fly models. Using UK Biobank data, I showed that polymorphisms in the human PARP1 gene or the intake of vitamin B, are associated with a decrease in the risk and severity of Alzheimer’s disease.

Building on these observations, I measured PAR levels in flies and found evidence of higher PARP activity. In Chapter 4, I hypothesised that higher PARP is linked to DNA misrepair, and validated this hypothesis in single-nuclei genome sequencing data from AD patients. I observed an imbalance in the nucleotide pool and showed that pharmacological or genetic enhancement of nucleotide metabolism is neuroprotective in AD.

In Chapter 5, I conducted a proteomics analysis and revealed significant alterations in one-carbon (1C) metabolism. The 1C pathway facilitates the transfer of methyl groups for biosynthetic pathways and mitochondrial metabolism. I found that enhancing one-carbon metabolism pharmacologically or genetically decreases AD-related impairments in a fly model of AD, and corroborated these results using human data.

During my PhD, I also observed and quantified large changes in sleep patterns in fly models, which recapitulate clinical phenotypes in AD patients. Upon comparing proteomics, metabolomics and single-cell RNA data in Chapter 6, I identified the redox-sensitive protein Hyperkinetic and its substrate NADP⁺ as potentially having a causative role in sleep changes in AD. I showed that overexpressing this gene and manipulating redox in the brain modulates sleep duration and AD pathology.

Since my work highlighted the importance of mitochondrial function in AD, I tested whether drugs that cause mitochondrial dysfunction could have any toxic effects. In Chapter 7, I showed that the antipsychotic aripiprazole, which is taken by AD patients to treat psychotic symptoms, is neurotoxic through inhibiting mitochondrial complex I.

Taken together, my work integrates multiomic data from animal models and human AD patients to gain insights into disease mechanisms, drug safety and therapeutic mechanisms. These results offer translational opportunities and call for further clinical validations to develop effective interventions for AD.