The adipocyte differentiation protein APMAP is an endogenous suppressor of Aβ production in the brain

Abstract

The deposition of amyloid-beta (Aβ) aggregates in the brain is a major pathological hallmark of Alzheimer's disease (AD). Aβ is generated from the cleavage of C-terminal fragments of the amyloid precursor protein (APP-CTFs) by γ-secretase, an intramembrane-cleaving protease with multiple substrates, including the Notch receptors. Endogenous modulation of γ-secretase is pointed to be implicated in the sporadic, age-dependent form of AD. Moreover, specifically modulating Aβ production has become a priority for the safe treatment of AD because the inhibition of γ-secretase results in adverse effects that are related to impaired Notch cleavage. Here, we report the identification of the adipocyte differentiation protein APMAP as a novel endogenous suppressor of Aβ generation. We found that APMAP interacts physically with γ-secretase and its substrate APP. In cells, the partial depletion of APMAP drastically increased the levels of APP-CTFs, as well as uniquely affecting their stability, with the consequence being increased secretion of Aβ. In wild-type and APP/ presenilin 1 transgenic mice, partial adeno-associated virus-mediated APMAP knockdown in the hippocampus increased Aβ production by ∼20 and ∼55%, respectively. Together, our data demonstrate that APMAP is a negative regulator of Aβ production through its interaction with APP and γ-secretase. All observed APMAP phenotypes can be explained by an impaired degradation of APP-CTFs, likely caused by an altered substrate transport capacity to the lysosomal/autophagic system.

Figure 1.

Identification of endogenous modulators of γ-secretase and Aβ production. (A) Analysis by LC-MS/MS of the protein components of the highly purified γ-secretase complex. (B and C) siRNA knockdown of TSPAN6, RTN4 and APMAP affects APP-CTFs in HeLa cells (B) and HEK cells overexpressing APP bearing the Swedish mutation that causes early-onset familial Alzheimer's disease (HEK-APPSwe; C). Biological duplicates are shown for each siRNA condition. (D) Correlation between APP-CTFs levels and Aβ40 production in HEK-APPSwe cells with reduced TSPAN6, RTN4 and APMAP expression. APP-CTFs levels were estimated by densitometric analysis, while Aβ40 levels were quantified by ELISA. Student's t-test was applied for statistical analysis; significance is shown as mean ± SD, *P < 0.05; **P < 0.01; Aβ40: n = 6/group; APP-CTFs: n = 4/Scramble, TSPAN6and APMAP groups; n = 3/RTN4 group. β-Actin served as a protein loading control. mAPP-FL and iAPP-FL: mature and immature APP full-length.

Figure 2.

APMAP interacts physically and co-localizes with γ-secretase, APP-FL and APP-CTFs. (A) Velocity co-sedimentation and co-immunoprecipitation of APMAP with the γ-secretase complex, APP-FL and APP-CTFs. Total membrane protein extracts from HEK-APPSwe cells transiently overexpressing hAPMAP1 or hAPMAP1-Flag were sedimented on an 18–28% glycerol gradient containing 0.1% CHAPSO. Each fraction was collected and analyzed by western blot for APMAP1, APP-FL, APP-CTFs and mature and immature γ-secretase (top panels). Next, proteins interacting with hAPMAP1-Flag (Flag) were affinity-precipitated in the fractions labeled in red with M2 anti-Flag affinity resin (lower panel). Untagged APMAP (hAPMAP1, also labeled ‘-’ in the figure) served as a control for the specific co-precipitation. (B) Immunohistochemical co-localization of APMAP (green) with the γ-secretase subunit Nicastrin (red, upper panel) or APP (red, lower panel) in 14 days in vitro mouse primary cortical neurons. Scale bar: 10 µm. Both confocal images (left panels) and Z-stack projections (right panels) are shown with a microscope objective magnification of 40×. For comparison, un-merged images for APMAP, NCT, APP-CTFs and DAPI are shown in Supplementary Material, Fig. S10. mNCT and iNCT, mature and immature Nicastrin.

Figure 3.

APMAP controls the levels and the stability of APP-CTFs. (A) siRNA knockdown of APMAP in HeLa cells quantitatively and qualitatively impacts APP-CTFs. Reduced APMAP expression correlates with increased APP-CTFs and the formation of three forms of APP-CTFα (CTFα1, α2 and α3), as revealed on a Tris-Tricine urea gel (left panel). This phenotype is further amplified in the presence of 1 µm GSI DAPT (right panel). Biological triplicates are shown. (B) MALDI-TOF mass spectrometric analysis of APP-CTFα1, α2 and α3 immunoprecipitated from DAPT-treated cells with siRNA-reduced APMAP expression. (C) Peptide sequences of N-terminal truncated APP-CTFα1 (C80), -CTFα2 (C74) and -CTFα3 (C71).

Figure 4.

APMAP modulates Aβ production in vivo. Five-week-old wild-type (A and B) or APP/PS1 transgenic (C and D) male mice were injected in the dorsal hippocampus with AAV9 expressing APMAP shRNA or a scrambled control shRNA, together with a GFP reporter. Four weeks post-injection, AAV9 transduction is highly efficient and is mainly restricted to the hippocampus (A and C). Wild-type (B) or APP/PS1 (D) males displayed significant ∼50% and ∼35% decreases of APMAP expression (mean ± SD; ***P < 0.001; n = 6/group), associated with significant ∼20 and ∼55% increases in total Aβ levels in the hippocampus (*P < 0.05; **P < 0.01; n = 6/group). Student's t-test was applied for the statistical analysis. β-Actin served as a protein loading control. (A and C) Scale bar: 500 µm.

Figure 5.

APMAP regulates cellular APP-CTFs levels through the lysosomal–autophagic pathway. HEK-APPSwe (A) and HeLa (B) cells treated either with a scrambled control siRNA or with APMAP siRNA were incubated for 12 h with the lysosomal inhibitor Chloroquine (25 and 50 µM, respectively). Total membrane protein extracts were prepared and analyzed by western blot for APMAP1, APP-FL and APP-CTFs (left panels). Biological duplicates are shown and β-actin served as a protein loading control. Next, APP-CTFs bands were quantified by densitometry (right panels) and Student's t-test was applied for statistical analysis with n = 4/group. **P < 0.01; ***P < 0.001.

Figure 6.

Molecular hypothesis for the regulation of APP-CTFs/Aβ levels by APMAP. In this hypothesis, APMAP is implicated in the transport of APP-CTFs to the lysosomal–autophagic system, where it undergoes substrate degradation. Consistent with the observed cellular phenotypes (Figs 1 and 3), depletion of APMAP would impair this transport capacity and consequently trigger increased APP-CTFs levels, which in turn cause increased Aβ production. Impaired degradation would further make APP-CTFs substrates accessible to yet unidentified proteases, explaining the N-terminal truncated species observed in cells with depleted APMAP (Fig. 3).