BACE2 degradation is mediated by both the proteasome and lysosome pathways

Background Alzheimer’s disease is the most common neurodegenerative disease in the elderly. Amyloid-β protein (Aβ) is the major component of neuritic plaques which are the hallmark of AD pathology. β-site APP cleaving enzyme 1 (BACE1) is the major β-secretase contributing to Aβ generation. β-site APP-cleaving enzyme 2 (BACE2), the homolog of BACE1, might play a complex role in the pathogenesis of Alzheimer’s disease as it is not only a θ-secretase but also a conditional β-secretase. Dysregulation of BACE2 is observed in Alzheimer’s disease. However, the regulation of BACE2 is less studied compared with BACE1, including its degradation pathways. In this study, we investigated the turnover rates and degradation pathways of BACE2 in both neuronal cells and non-neuronal cells. Results Both lysosomal inhibition and proteasomal inhibition cause a time- and dose-dependent increase of transiently overexpressed BACE2 in HEK293 cells. The half-life of transiently overexpressed BACE2 protein is approximately 6 h. Moreover, the half-life of endogenous BACE2 protein is approximately 4 h in both HEK293 cells and mouse primary cortical neurons. Furthermore, both lysosomal inhibition and proteasomal inhibition markedly increases endogenous BACE2 in HEK293 cells and mouse primary cortical neurons. Conclusions This study demonstrates that BACE2 is degraded by both the proteasome and lysosome pathways in both neuronal and non-neuronal cells at endogenous level and in transient overexpression system. It indicates that BACE2 dysregulation might be mediated by the proteasomal and lysosomal impairment in Alzheimer’s disease. This study advances our understanding of the regulation of BACE2 and provides a potential mechanism of its dysregulation in Alzheimer’s disease.


Background
Alzheimer's disease (AD) is the most common neurodegenerative disease in the elderly. Amyloid-β protein (Aβ) is the major component of neuritic plaques which are the hallmark of AD pathology [1]. Deposition of Aβ is formed from amyloid-β precursor protein (APP) by sequential cleavage of βand γ-secretase [2].β-site APP cleaving enzyme 1 (BACE1) is the major β-secretase contributing to Aβ generation.β-site APP-cleaving enzyme 2 (BACE2), the homolog of BACE1, is a θ-secretase, which cleaves APP at Phe20 site to yield a CTF with 80 amino acids (CTFθ or C80) contributing to the generation of a truncated Aβ [3,4]. Moreover, BACE2 prevents neuronal apoptosis by cleaving a potassium channel at the surface of plasma membrane [5]. However, our recent study demonstrated that BACE2 can be converted into a β-secretase with comparable β-secretase activity to that of BACE1, implying that BACE2 could contribute to Aβ generation in AD [6].Consistently, increased BACE2 expression and activity is detected in neurons of AD brains [7]. Genetic data highly supports that BACE2 is associated with AD risk. For example, BACE2 haplotype associates with AD, while SNPs in BACE2 (e.g., rs2252576, rs2837990, rs7281733) predispose to early onset of AD in patients with Down syndrome [8,9]. Recently, the association between a number of SNPs in BACE2 and AD was detected in APOE ε4 non-carriers, which might be mediated by altered BACE2 expression-mediated Aβ generation and clearance [10]. It indicates that dysregulation of BACE2 might contribute to the pathogenesis AD.
It is important to elucidate the regulation of BACE2 expression as BACE2 homeostasis is critical to maintain the physiological function and counteract the pathogenesis of AD. In addition to the transcriptional regulation, protein degradation does play an important role in BACE2 homeostasis [11]. The ubiquitin-proteasome pathway (UPS) and the autophagy-lysosome pathway (ALP) are two major pathways for protein degradation in eukaryotic cells [12,13]. The impairment of the proteasome and lysomsome activity in AD has been reported in a number of studies, which might contribute to the dysregulation of BACE2 in AD [14]. However, the degradation of BACE2 remains elusive. To further elucidate the feature of BACE2 degradation, we investigated BACE2 degradation in both neuronal and non-neuronal cells. We found that both lysosomal inhibition and proteasomal inhibition cause the increase of transiently overexpressed BACE2 in HEK293 cells. Moreover, both lysosomal inhibition and proteasomal inhibition markedly increases endogenous BACE2 levels in HEK293 cells and mouse primary cortical neurons, indicating that BACE2 is degraded by both the proteasome pathway and lysosome pathway. This work advances our understanding of the regulation of BACE2 and provides a potential mechanism of its dysregulation in AD. It might provide a potential strategy for the treatment of AD by targeting the dysregulation of BACE2 in AD.
To further confirm the effect of proteasomal inhibition on BACE2 degradation, the proteasomal inhibitor ALLN was also applied to BACE2 transfected cells for 12 h. ALLN treatment significantly increased BACE2 levels to 5.107 ± 0.93, 4.78 ± 0.44 and 5.76 ± 0.71 fold at the dose of 5, 10 or 25 μM, respectively, compared to that in control cells, p < 0.05 ( Fig. 3e and f). For the time course assay, cells were treated with 10 μM ALLN for 0, 6, 12 and 24 h. BACE2 levels were significantly increased to 2.69 ± 0.20, 4.72 ± 0.32 and 6.46 ± 0.25 fold, respectively, compared to that in control cells, p < 0.05 ( Fig. 3g and h). It indicated that transiently overexpressed BACE2 is degraded by the proteasome pathway in HEK293 cells.

The half-life of endogenous BACE2 is approximately 4 h in HEK293 cells
To examine the half-life of endogenous BACE2, HEK293 cells were divided equally into four dishes. The cells were treated with 100 μg/mL CHX to block BACE2 synthesis [16][17][18]. The cells were harvested at 0, 4, 8 and 12 h time-point, respectively. Western blot analysis was performed to measure the level of remaining BACE2 protein relative to the BACE2 level at 0 h time-point. BACE2 protein levels were decreased to 41 ± 7.0%, 35 ± 4.6% and 24 ± 2.9% at 4, 8 and 12 h time-point, respectively, p < 0.05 ( Fig. 4a and b). Our data showed that the half-life of endogenous BACE2 is approximately 4 h in HEK293 cells.
Both lysosomal inhibition and proteasomal inhibition increases the expression of endogenous BACE2 in HEK293 cells HEK293 cells were divided equally into 6 cm dishes and treated with 10 uM MG-132 for 12 h and 25 mM NH 4 CL for 24 h, respectively. Endogenous BACE2 levels  were significantly increased to 3.21 ± 0.10 and 6.65 ± 0.20 fold by MG-132 and NH 4 CL treatments, respectively, compared to that in control cells, p < 0.05 ( Fig. 5a  and b). Our data demonstrated that endogenous BACE2 is degraded by both the lysosome pathway and the proteasome pathway in HEK293 cells.

The half-life of endogenous BACE2 is approximately 4 h in primary neurons
To examine the half-life of endogenous BACE2 in primary neurons, mouse primary cortical neurons were isolated and seeded equally into five dishes. 7 days later, the cells were treated with 100 μg/mL CHX to block BACE2 synthesis [16][17][18] Fig. 6a and b). Our data showed that the half-life of endogenous BACE2 is approximately 4 h in primary neurons.

Both lysosomal inhibition and proteasomal inhibition increases the expression of endogenous BACE2 in primary neurons
The mouse primary cortical neurons were isolated and equally seeded. 7 days later, the primary neurons were treated with 100 μg/ml CHL and 5 μM MG132, respectively. Endogenous BACE2 levels were significantly increased to 3.97 ± 0.21 and 8.09 ± 0.90 fold by MG-132 and CHL treatments, respectively, compared to that in control cells, p < 0.05 ( Fig. 7a and b). Our data demonstrated that endogenous BACE2 is degraded by both the lysosome pathway and the proteasome pathway in mouse primary cortical neurons.

Discussion
The impairment of proteasome and lysosome function is reported in AD, and the dysregulation of BACE2 is also observed in AD. It suggests that impaired protein degradation function might play an important role in BACE2 dysregulation in AD by inhibiting BACE2 degradation. However, the feature of BACE2 degradation remains elusive. Thus, it is critical to elucidate the feature of BACE2 degradation. Both BACE1 and BACE2 are transmembrane proteins, and their amino acid sequences are 75% homologous. It is possible that BACE2 degradation might behave similarly to BACE1 which is degraded by both the proteasome and lysosome pathways [3,19]. However, our previous study showed that BACE2 is only degraded by the autophagy-lysosome pathway in HEK293 cells stably overexpressing BACE2, but not by the proteasome pathway [18]. To further complement previous work performed in stable cells, we investigated BACE2 degradation at endogenous levels and under transient overexpression condition in both neuronal cells and non-neuronal cells. We found that both the lysosome and the proteasome pathways are implicated in the degradation of BACE2.The different conclusion drawn from the current study and the previous study might be caused by the following reasons. First, the clonal effect of stable cells might contribute to the difference as it might not represent the features of BACE2 in the population of cells. In addition, the high expression level and long-term overexpression might cause stronger stress on the cells in stable overexpression system. Moreover, long-term overexpression of BACE2 might have global effect on gene transcription including genes involved in protein degradation system, which inversely affects BACE2 degradation. For example, our previous study did show that long-term overexpression of transmembrane protein APP reduces proteasome activity by downregulating the expression of core subunits of the proteasome complex [20]. Furthermore, long-term protein overexpression might exhaust the capacity of the proteasome degradation system, resulting in proteasome impairment, thus, the proteasomal inhibitor has no effect on the proteasome-mediated BACE2 degradation in stable cells.
Our previous study showed that the half-life of stably overexpressed BACE2 is around 20 h [18]. In the current study, we showed that the half-life of transient overexpressed BACE2 is approximately 6 h, while the half-life of endogenous BACE2 is approximately 4 h. The dramatic difference between the half-life of stably overexpressed BACE2 and the half-life of transiently overexpressed BACE2 or endogenous BACE2 might be caused by the following reasons. First, the clonal effect of stable cells might contribute to the difference as it might not represent the features of BACE2 in the population of cells. Moreover, the relative long half-life of BACE2 in stable cells might be attribute to the impairment of proteasome pathway or the reduced proteasome activity as described above.

Conclusions
In conclusion, we found that BACE2 is degraded by both the proteasome and the lysosome pathways in both neuronal cells and non-neuronal cells at endogenous level or in transient overexpression system, indicating that BACE2 degradation by both the proteasome and lysosome pathways is a common feature of BACE2. Moreover, the impairment of the proteasome pathway and the lysosome pathway in AD might lead to BACE2 dysregulation in neurons, contributing to the pathogenesis of AD [10]. This study not only advances our  understanding of the regulation of BACE2 but also provides a potential mechanism of its dysregulation in AD. Moreover, it might provide a strategy for the treatment of AD by targeting the dysregulation of BACE2.

Cell culture and transfection
Human embryonic kidney HEK293 cells obtained from Dr. Weihong Song's lab were cultured in high-glucose DMEM containing 10% fetal bovine serum and 1% penicillin-streptomycin. C57BL/6 mice were obtained from Cyagen Bioscience. Mouse cortical neurons were prepared from E18 embryos as described previously [6]. Briefly, the cortices of embryos were dissected out and the meninges were completely removed. The cortices were digested with papain at 37°C for 20 min. Then, the papain solution was replaced with inactivation solution (MEM containing 0.6% D-[+]-glucose, 1 mM pyruvate, 10% horse serum, 2.5% bovine serum albumin [BSA], and 2.5% trypsin inhibitor) and the cells were dissociated by repeatedly pipetting. The isolated cells were equally seeded on poly-D-lysine-coated dishes with Neurobasal Media (Invitrogen) containing B27 [6]. All cells were maintained at 37°C with 5% CO 2 in an incubator as described previously [16,18,21]. pBACE2-mycHis refers to pZ-BACE2mycHis in this study, which is constructed previously [22]. Transient transfection was performed by using polyetherimide (PEI) method as described previously [23,24]. Briefly, HEK293 cells were seeded 24 h prior to transfection. The regular culture medium was replaced with high-glucose DMEM without serum 1 h prior to transfection. 6 h after transfection, the medium was replaced with regular culture medium.

Statistical analysis
The proteins expression was quantified by using ImageJ. One-way ANOVA followed by Tukey's test was used for data analysis with three or more independent experiments. P < 0.05 was considered as a significant difference.