The first suggestions that the cleaved intracellular domain

The first suggestions that the cleaved intracellular domain of APP, AICD, might transit to the nucleus and hence selectively regulate gene transcription arose from analogies with the Notch receptor signalling system where similar intra-membrane proteolysis occurs [84], [85] and it was later confirmed that AICD in a complex with the nuclear adaptor protein Fe65 and the histone acetyltransferase Tip60 regulates reporter gene transcription [86]. A host of subsequent studies identified specific neuronal genes regulated in this way, including APP itself, BACE-1, the amyloid-degrading enzyme, neprilysin (NEP) and transthyretin [10], [82], [87].
Role of APP in regulation of AChE Our first insight into possible regulatory interactions of APP with AChE RN486 receptor came from earlier studies of the mammalian Fe65 orthologue (the product of the feh-1 gene), which regulates pharyngeal contraction in Caenorhabditis elegans and which, in feh-1 mutant worms, is associated with decreased AChE activity [88]. The feh-1 gene is also expressed in the nervous system in the organism where the gene product was shown to interact with APL-1 protein, the orthologue of mammalian APP [89]. The decreased contraction in worms was accompanied by reduced expression of ace-1 and ace-2 transcripts, which code for the two major AChE activities in the nematode. Given that the major anti-AD therapies currently involve use of AChE inhibitors, such a mechanistic link between APP and AChE could provide an important bridge in the pathways to neuronal loss in AD. These, and associated observations, suggested that the mammalian AChE gene might be the target of transcriptional up-regulation by a Fe65-APP like complex, possibly mediated via AICD/Fe65 as shown previously by us for neprilysin [87]. We therefore set out to explore this possibility. As we had predicted, modulation of the expression levels of the neuronal isoform of APP (APP695) did indeed modulate mRNA, protein, and activity levels of AChE [90]. However, whereas knock-down of feh-1 expression in C. elegans resulted in a reduction in corresponding AChE levels [88], in human neuronal cell lines APP knock-down using siRNA up-regulated AChE expression [90]. Conversely, over-expression of APP695 resulted in substantial reductions in AChE mRNA, protein and activity levels. Furthermore, the AChE membrane anchor protein, PRiMA, appeared to be co-ordinately regulated by the same mechanism, whereas BChE and the high-affinity ACh transporter remained unaffected [90]. This novel mechanism turned out to be α-, β- and γ-secretase- (and hence sAPPα, sAPPβ, and AICD-) independent, as shown by the use of selective secretase inhibitors. In particular, release of AICD requires the action of γ-secretase and therefore its inhibition would prevent any AICD-dependent gene regulation. Furthermore, this effect appeared to be mediated via the E1 domain of APP, particularly its copper-binding domain, as indicated by site-directed mutagenesis experiments [90]. At the time, such a mode of regulation of gene expression by full length, not cleaved, APP was unknown but a subsequent independent study provided another example whereby holo-APP could modulate gene expression. In this study APP was shown to negatively regulate expression of EGR-1, an immediate early gene involved in memory formation [91]. This effect was demonstrated both in vivo and in vitro and shown to be γ-secretase and AICD-independent using a similar strategy. The mechanism involved an epigenetic regulation with enriched acetylation of histone H4 found on the EGR-1 promoter in APP−/- neurons and mice [91]. Since two EGR-1 sites in the AChE promoter regulate AChE expression [92], its regulation by APP may well be mediated through EGR-1, although this hypothesis remains to be tested. APP overexpression in SH-SY5Y cells has also been shown to decrease expression of histone deacetylases HDAC1 and HDAC3 [74], which might lead to changes in expression of a variety of genes, including EGR-1 and AChE.