Manipulations of amyloid precursor protein cleavage disrupt the circadian clock in aging Drosophila

Highlights

• dBACE overexpression accelerates decline of rest-activity rhythms in aging Drosophila.

• PER oscillations in pacemaker neurons are dampened by elevated dBACE.

• Kuzbanian overexpression induces age-related rest-activity rhythm decline in flies.

• Neuronal dAICD expression induces age-related rest-activity rhythm decline.

• APPL expression in pacemaker neurons strengthens rest-activity rhythms in aging flies.

Abstract

Alzheimer's disease (AD) is a neurodegenerative disease characterized by severe cognitive deterioration. While causes of AD pathology are debated, a large body of evidence suggests that increased cleavage of Amyloid Precursor Protein (APP) producing the neurotoxic Amyloid-β (Aβ) peptide plays a fundamental role in AD pathogenesis. One of the detrimental behavioral symptoms commonly associated with AD is the fragmentation of sleep-activity cycles with increased nighttime activity and daytime naps in humans. Sleep-activity cycles, as well as physiological and cellular rhythms, which may be important for neuronal homeostasis, are generated by a molecular system known as the circadian clock. Links between AD and the circadian system are increasingly evident but not well understood. Here we examined whether genetic manipulations of APP-like (APPL) protein cleavage in Drosophila melanogaster affect rest-activity rhythms and core circadian clock function in this model organism. We show that the increased β-cleavage of endogenous APPL by the β-secretase (dBACE) severely disrupts circadian behavior and leads to reduced expression of clock protein PER in central clock neurons of aging flies. Our data suggest that behavioral rhythm disruption is not a product of APPL-derived Aβ production but rather may be caused by a mechanism common to both α and β-cleavage pathways. Specifically, we show that increased production of the endogenous Drosophila Amyloid Intracellular Domain (dAICD) caused disruption of circadian rest-activity rhythms, while flies overexpressing endogenous APPL maintained stronger circadian rhythms during aging. In summary, our study offers a novel entry point toward understanding the mechanism of circadian rhythm disruption in Alzheimer's disease.

Introduction

Alzheimer's disease (AD) is characterized by progressive neurodegeneration resulting in the loss of cognitive ability. The exact pathology of AD is not well understood and is widely debated. The amyloid cascade hypothesis suggests that abnormal production of neurotoxic Amyloid-beta (Aβ) in combination with tangles of phosphorylated Tau-microtubule associated protein lead to neuronal dysfunction, cell loss and thus cognitive decline (Hardy and Selkoe, 2002, Mandelkow and Mandelkow, 1998, Selkoe, 2000). The crucial enzyme in the production of Aβ is the rate-limiting Beta-site Amyloid Precursor Protein Cleaving Enzyme (BACE), which shows elevated expression in AD patients (Fukumoto et al., 2002, Vassar et al., 2009). Recently, knock-in of human BACE was shown to recapitulate many behavioral and physiological phenotypes of AD in a murine model (Plucinska et al., 2014). In addition to BACE, Amyloid Precursor Protein (APP) is cleaved by an α-secretase. Both cleavage pathways yield a secreted N-terminal fragment known as sAPP, sAPPβ for β-cleavage and sAPPα for α-cleavage. Subsequent ϒ-cleavage following β or α-cleavage results in either the Aβ or P3 fragment, respectively (De Strooper and Annaert, 2000, Selkoe, 2000, Turner et al., 2003). In addition to those two fragments, ϒ-cleavage also produces the APP intracellular domain (AICD), which has been shown to affect transcriptional regulation (Belyaev et al., 2010, Kimberly et al., 2001, Pardossi-Piquard and Checler, 2012, Turner et al., 2003). Both detrimental and positive effects of the AICD have been reported in cultured cells (Lu et al., 2000, Zhou et al., 2012); however, the physiological function of the AICD remains poorly understood.

One of the detrimental behavioral symptoms commonly associated with AD is the fragmentation of sleep-activity cycles with increased nighttime activity and daytime naps (Harper et al., 2005, Volicer et al., 2001, Wu and Swaab, 2007). Impaired rest-activity rhythms have also been reported in experimental AD model mice (Roh et al., 2012, Sterniczuk et al., 2010). Rest-activity cycles are generated by a molecular system known as the circadian clock. The circadian clock mechanism is based on negative feedback loops involving transcriptional activators and repressors, which are largely conserved from Drosophila to humans (Hardin and Panda, 2013). Circadian clocks generate daily rhythms in expression of many output genes leading to cellular, physiological, and behavioral rhythms. Loss of circadian rhythms is detrimental to health and adversely affects neuronal homeostasis in both murine (Hastings and Goedert, 2013, Kondratova and Kondratov, 2012, Reddy and O'Neill, 2010) and Drosophila models (Krishnan et al., 2012). Therefore, it is important to understand the connections between the circadian system and AD pathology.

The links between AD and circadian rhythms were inferred from transgenic model animals expressing pathogenic versions of human genes (Rezaval et al., 2008, Roh et al., 2012, Sterniczuk et al., 2010). Recent reports showed that pathogenic human amyloid peptides disrupt behavioral rest-activity rhythms in Drosophila but are not sufficient to disrupt molecular oscillations of PERIOD (PER) in central pacemaker neurons (Chen et al., 2014, Long et al., 2014). Here, we manipulated AD-related genes in the fruit fly, Drosophila melanogaster in order to understand endogenous pathways that may interfere with the circadian system. APP has a functional ortholog in Drosophila called Amyloid Precursor Protein-like (APPL) (Luo et al., 1990, Luo et al., 1992, Rosen et al., 1989). In addition, orthologs of BACE, the α-secretase, and the ϒ-secretase complexes have been identified in Drosophila. They are known as dBACE, Kuzbanian (KUZ) and Presenilin (PSN), respectively, and were shown to process APPL resulting in peptide fragments comparable to APP fragments (Bolkan et al., 2012, Carmine-Simmen et al., 2009, Greeve et al., 2004). This conservation made Drosophila a suitable model for our study. Because BACE is the rate-limiting enzyme in the production of Aβ, we first investigated if over-expression of dBACE is involved in the rest-activity disturbances characteristic of AD.

We found that increasing dBACE expression disrupted endogenous rest-activity rhythms. This effect was most severe in aged flies suggesting an age-dependent mechanism. Furthermore, dBACE expression resulted in the dampened oscillation of the core clock protein PER in central pacemaker neurons, which are master regulators of rest-activity rhythms. Surprisingly, over-expression of Kuzbanian (KUZ) also disrupted rest-activity rhythms suggesting a mechanism independent of endogenous Drosophila Aβ peptide (dAβ) production, which like its vertebrate ortholog has been shown to be neurotoxic (Carmine-Simmen et al., 2009). This suggested that rhythm deficits were due to a cleavage product of APPL conserved in both α and β cleavage pathways. Therefore, we expressed the APPL intracellular domain (dAICD) and this caused severe disruption of behavioral rest-activity rhythms. In contrast, expression of full-length APPL in central clock neurons protected against an age-dependent decline in rest-activity rhythms. Taken together, these findings suggest that APPL and specifically the dAICD can affect circadian behavior by interfering with molecular clock oscillations.