Nrf2, a transcriptional activator of cell protection genes, is an attractive therapeutic target for the prevention of neurodegenerative diseases, including Alzheimer’s disease (AD). Current Nrf2 activators, however, may exert toxicity and pathway over-activation can induce detrimental effects. An understanding of the mechanisms mediating Nrf2 inhibition in neurodegenerative conditions may therefore direct the design of drugs targeted for the prevention of these diseases with minimal side-effects. Our study provides the first in vivo evidence that specific inhibition of Keap1, a negative regulator of Nrf2, can prevent neuronal toxicity in response to the AD-initiating Aβ42 peptide, in correlation with Nrf2 activation. Comparatively, lithium, an inhibitor of the Nrf2 suppressor GSK-3, prevented Aβ42 toxicity by mechanisms independent of Nrf2. A new direct inhibitor of the Keap1-Nrf2 binding domain also prevented synaptotoxicity mediated by naturally-derived Aβ oligomers in mouse cortical neurons. Overall, our findings highlight Keap1 specifically as an efficient target for the re-activation of Nrf2 in AD, and support the further investigation of direct Keap1 inhibitors for the prevention of neurodegeneration in vivo.

The greatest risk factor for Alzheimer's disease (AD) is age, and changes in the ageing nervous system are likely contributors to AD pathology. Amyloid beta (Aβ) accumulation, which occurs as a result of the amyloidogenic processing of amyloid precursor protein (APP), is thought to initiate the pathogenesis of AD, eventually leading to neuronal cell death. Previously, we developed an adult-onset Drosophila model of AD. Mutant Aβ42 accumulation led to increased mortality and neuronal dysfunction in the adult flies. Furthermore, we showed that lithium reduced Aβ42 protein, but not mRNA, and was able to rescue Aβ42-induced toxicity. In the current study, we investigated the mechanism/s by which lithium modulates Aβ42 protein levels and Aβ42 induced toxicity in the fly model. We found that lithium caused a reduction in protein synthesis in Drosophila and hence the level of Aβ42. At both the low and high doses tested, lithium rescued the locomotory defects induced by Aβ42, but it rescued lifespan only at lower doses, suggesting that long-term, high-dose lithium treatment may have induced toxicity. Lithium also down-regulated translation in the fission yeast Schizosaccharomyces pombe associated with increased chronological lifespan. Our data highlight a role for lithium and reduced protein synthesis as potential therapeutic targets for AD pathogenesis.

Lithium has long been used for the treatment of psychiatric disorders, due to its robust beneficial effect as a mood stabilizing drug. Lithium’s effectiveness for improving neurological function is therefore well-described, stimulating the investigation of its potential use in several neurodegenerative conditions including Alzheimer’s (AD), Parkinson’s (PD) and Huntington’s (HD) diseases. A narrow therapeutic window for these effects, however, has led to concerted efforts to understand the molecular mechanisms of lithium action in the brain, in order to develop more selective treatments that harness its neuroprotective potential whilst limiting contraindications. Animal models have proven pivotal in these studies, with lithium displaying advantageous effects on behavior across species, including worms (C. elegans), zebrafish (Danio rerio), fruit flies (Drosophila melanogaster) and rodents. Due to their susceptibility to genetic manipulation, functional genomic analyses in these model organisms have provided evidence for the main molecular determinants of lithium action, including inhibition of inositol monophosphatase (IMPA) and glycogen synthase kinase-3 (GSK-3). Accumulating pre-clinical evidence has indeed provided a basis for research into the therapeutic use of lithium for the treatment of dementia, an area of medical priority due to its increasing global impact and lack of disease-modifying drugs. Although lithium has been extensively described to prevent AD-associated amyloid and tau pathologies, this review article will focus on generic mechanisms by which lithium preserves neuronal function and improves memory in animal models of dementia. Of these, evidence from worms, flies and mice points to GSK-3 as the most robust mediator of lithium’s neuro-protective effect, but it’s interaction with downstream pathways, including Wnt/β-catenin, CREB/brain-derived neurotrophic factor (BDNF), nuclear factor (erythroid-derived 2)-like 2 (Nrf2) and toll-like receptor 4 (TLR4)/nuclear factor-κB (NFκB), have identified multiple targets for development of drugs which harness lithium’s neurogenic, cytoprotective, synaptic maintenance, anti-oxidant, anti-inflammatory and protein homeostasis properties, in addition to more potent and selective GSK-3 inhibitors. Lithium, therefore, has advantages as a multi-functional therapy to combat the complex molecular pathology of dementia. Animal studies will be vital, however, for comparative analyses to determine which of these defense mechanisms are most required to slow-down cognitive decline in dementia, and whether combination therapies can synergize systems to exploit lithium’s neuro-protective power while avoiding deleterious toxicity.

Several species of β-amyloid peptides (Aβ) exist as a result of differential cleavage from amyloid precursor protein (APP) to yield various C-terminal Aβ peptides. Several N-terminal modified Aβ peptides have also been identified in Alzheimer’s disease (AD) brains, the most common of which is pyroglutamate-modified Aβ (AβpE3-42). AβpE3-42 peptide has an increased propensity to aggregate, appears to accumulate in the brain before the appearance of clinical symptoms of AD, and precedes Aβ1-42 deposition. Moreover, in vitro studies have shown that AβpE3-42 can act as a seed for full length Aβ1-42. In this study, we characterized the Drosophila model of AβpE3-42 toxicity by expressing the peptide in specific sets of neurons using the GAL4-UAS system, and measuring different phenotypic outcomes. We found that AβpE3-42 peptide had an increased propensity to aggregate. Expression of AβpE3-42 in the neurons of adult flies led to behavioural dysfunction and shortened lifespan. Expression of AβpE3-42 constitutively in the eyes led to disorganised ommatidia, and activation of the c-Jun N-terminal kinase (JNK) signaling pathway. The eye disruption was almost completely rescued by co-expressing a candidate Aβ degrading enzyme, neprilysin2. Furthermore, we found that neprilysin2 was capable of degrading AβpE3-42. Also, we tested the seeding hypothesis for AβpE3-42 in vivo, and measured its effect on Aβ1-42 levels. We found that Aβ1-42 levels were significantly increased when Aβ1-42 and AβpE3-42 peptides were co-expressed. Furthermore, we found that AβpE3-42 enhanced Aβ1-42 toxicity in vivo. Our findings implicate AβpE3-42 as an important source of toxicity in AD, and suggest that its specific degradation could be therapeutic.Electronic supplementary materialThe online version of this article (doi:10.1186/s40478-016-0380-x) contains supplementary material, which is available to authorized users.