Amyloid precursor proteins are protective in Drosophila models of progressive neurodegeneration

Abstract

The processing of Amyloid Precursor Proteins (APPs) results in several fragments, including soluble N-terminal ectodomains (sAPPs) and C-terminal intracellular domains (AICD). sAPPs have been ascribed neurotrophic or neuroprotective functions in cell culture, although β-cleaved sAPPs can have deleterious effects and trigger neuronal cell death. Here we describe a neuroproprotective function of APP and fly APPL (Amyloid Precursor Protein-like) in vivo in several Drosophila mutants with progressive neurodegeneration. We show that expression of the N-terminal ectodomain is sufficient to suppress the progressive degeneration in these mutants and that the secretion of the ectodomain is required for this function. In addition, a protective effect is achieved by expressing kuzbanian (which has α-secretase activity) whereas expression of fly and human BACE aggravates the phenotypes, suggesting that the protective function is specifically mediated by the α-cleaved ectodomain. Furthermore, genetic and molecular studies suggest that the N-terminal fragments interact with full-length APPL activating a downstream signaling pathway via the AICD. Because we show protective effects in mutants that affect different genes (AMP-activated protein kinase, MAP1b, rasGAP), we propose that the protective effect is not due to a genetic interaction between APPL and these genes but a more general aspect of APP proteins. The result that APP proteins and specifically their soluble α-cleaved ectodomains can protect against progressive neurodegeneration in vivo provides support for the hypothesis that a disruption of the physiological function of APP could play a role in the pathogenesis of Alzheimer's Disease.

Fig. 1

Neuroprotective effects of APPL. (A) Paraffin head section of a 5 d old loe control fly, carrying the UAS-APPL construct, but not Appl-GAL4. The vacuole formation characteristic for loe mutants of this age is easily detectable. (B) In contrast, a loe fly expressing full-length APPL via Appl-GAL4 shows much less vacuolization. (C) Mean area of vacuoles in μm² in the different genotypes tested. SEMs and the number of brain hemispheres analyzed are indicated, *<0.05, **<0.01, ***≪0.001. All flies were 5 d old females. (D) Schematic of the different APPL constructs used. Scale bar in A, B=50 μm. re = retina, la = lamina, me = medulla, lo = lobula, lb = lobula plate.

Fig. 2

Expression of kuzbanian or human APP₆₉₅ ameliorates the loe phenotype. (A) Effects of altering the processing of endogenous APPL in loe flies. The mean area of vacuoles in μm² in loe flies overexpressing dBACE is increased whereas KUZ overexpression suppresses vacuole formation. (B) Expressing full-length APP₆₉₅ also reduces the mean area of vacuoles in loe flies. Co-expression of human BACE1 with APP₆₉₅ significantly reduced the suppressing effect compared to expression of APP₆₉₅ alone. SEMs are indicated, * < 0.05, **< 0.01, ***≪0.001. (C) A Western Blot using anti-HA reveals that all soluble fragments are expressed when induced via Appl-GAL4 and secreted into the media (M) after the brains were kept in culture. L = lysate from cultured brains.

Fig. 3

The protective function is mediated by full-length APPL. (A) Loss of APPL enhances the degeneration in Appl^d;loe double mutants. Expression of full-length APPL in Appl^d;loe double mutants can rescue the effects of this loss and decreases the vacuolization. In contrast, expression of sAPPL has no effect in the in Appl^d;loe double mutant. Flies were 4 d old males. SEMs are indicated. (B) Western Blot using α-APPL-AICD on immunopreciptates from Kc cells expressing HA-tagged soluble fragments and untagged full-length APPL. Similar amounts of full-length APPL in the input (upper panel) indicate comparable transfection efficiencies in these samples. Whereas a significant amount of full-length APPL can be found in the eluate (lower panel) from sAPPL expressing cells (lane 1), showing that it can bind to full-length APPL. In contrast, only marginal amounts of full-length APPL are detectable in the co-immunoprecipitations using the human sAPPα or sAPPβ fragment. A control that does not contain an HA-tagged construct also reveals marginal amounts of full-length APPL.

Fig. 4

APPL is protective in several neurodegenerative mutants. (A) A four week old futsch^olk1 fly shows the characteristic vacuolization in the mechanosensory area of the antennal lobe. (B) Mean area of vacuoles in μm² in 4 week old futsch^olk1 flies with or without sAPPL expression. (C) vap¹ mutants show vacuoles scattered throughout the brain after two weeks of aging. (D) Also in this mutant the mean area of vacuoles is significantly reduced by expression of the secreted APPL ectodomain. (E) A section from a 14 d old sws¹ fly reveals vacuoles in all brain areas. (F) In contrast to vap¹ and futsch^olk1, this phenotype is not significantly altered after induction of sAPPL or full-length APPL. All flies were males. *<0.05, **<0.01and the bars indicate SEMs. Scale bar in A, C, E=50 μm. ant. nerve = antennal nerve, dn = deutocerebral neuropil.

Fig. 5

LOE interferes with APPL processing whereas vap or futsch^olk do not. (A) A Western Blot using anti-APPL-AICD reveals a decrease in the α-cleaved CTF (arrowhead) in loe mutants compared to y w (the genetic background of loe) whereas the production of this fragment is increased after additional LOE expression. Full-length APPL is indicated by the arrow. (B) In contrast, neither vap¹ nor futsch^olk1 seem to interfere with the production of the α-CTF or the levels of full-length APPL when compared to their appropriate background Oregon R (OR). A quantification of the α-cleaved APPL CTF (±SEM) obtained from at least three independent Western blots, is shown underneath the blot. Loading controls using anti-actin are shown below.

Fig. 6

Flies lacking APPL show degeneration and reduced survival. (A) A four week old female wild type fly does not show obvious signs of neurodegeneration whereas a few vacuoles have formed in the CNS of an age-matched Appl^d female (B, arrows). (C, D) Measuring the survival rate of Appl^d flies revealed a dramatically reduced life span to 59% of the life span of wild type in females and 39% in male flies. At least 10 independent experiments were performed for each genotype and gender.