Dieckol Ameliorates Aβ Production via PI3K/Akt/GSK-3β Regulated APP Processing in SweAPP N2a Cell

The proteolytic processing of amyloid precursor protein (APP) by β-secretase (BACE1) and γ-secretase releases amyloid-β peptide (Aβ), which deposits in amyloid plaques and contributes to the initial causative events of Alzheimer’s disease (AD). In the present study, the regulatory mechanism of APP processing of three phlorotannins was elucidated in Swedish mutant APP overexpressed N2a (SweAPP N2a) cells. Among the tested compounds, dieckol exhibited the highest inhibitory effect on both intra- and extracellular Aβ accumulation. In addition, dieckol regulated the APP processing enzymes, such as α-secretase (ADAM10), β-secretase, and γ-secretase, presenilin-1 (PS1), and their proteolytic products, sAPPα and sAPPβ, implying that the compound acts on both the amyloidogenic and non-amyloidogenic pathways. In addition, dieckol increased the phosphorylation of protein kinase B (Akt) at Ser473 and GSK-3β at Ser9, suggesting dieckol induced the activation of Akt, which phosphorylated GSK-3β. The specific phosphatidylinositol 3-kinase (PI3K) inhibitor LY294002 triggered GSK-3β activation and Aβ expression. In addition, co-treatment with LY294002 noticeably blocked the effect of dieckol on Aβ production, demonstrating that dieckol promoted the PI3K/Akt signaling pathway, which in turn inactivated GSK-3β, resulting in the reduction in Aβ levels.


Introduction
Alzheimer's disease (AD) is a chronic neurodegenerative disorder accompanied with progressive memory loss and cognitive impairment. Amyloid-β (Aβ) peptides deposition is considered to be a primary hallmark in AD pathology. Insoluble Aβ is highly toxic and leads to synaptic dysfunction, glial cell activation and neuronal loss [1,2]. In addition, Aβ affects mitochondrial redox imbalance, causing oxidative stress and calcium overload, to induce cytokine release, and to disrupt acetylcholine and glutamate neurotransmission [3][4][5].
Aβ is a cleavage product of the amyloid-β precursor protein (APP) through the amyloidogenic pathway by β-site APP cleaving enzyme 1 (BACE1) and γ-secretase. BACE1 produces an extracellular APPβ (sAPPβ) fragment and a C-terminal fragment (C99). The C99 is subsequently cleaved by γ-secretase, leading to the production of Aβ peptides of various lengths (38-43 residues). Conversely, through a non-amyloidogenic pathway, APP is cleaved by α-secretase, mainly constituting a disintegrin and metalloproteinase domaincontaining protein 10 (ADAM10), causing it to release soluble APPα (sAPPα), which thus Mar. Drugs 2021, 19,152 3 of 13 levels after the treatment of all the tested compounds ( Figure 2c). Therefore, these concentrations were used in the following studies.
(a) (b) (c) Mar. Drugs 2021, 19, 152 3 of 13 levels after the treatment of all the tested compounds ( Figure 2c). Therefore, these concentrations were used in the following studies.
Both extracellular and intracellular Aβ play critical roles in the onset of AD. It is commonly considered that extracellular Aβ is produced by the BACE1-mediated processing of APP and released in the extracellular matrix, which is a pivotal cause of AD [27]. However, recently, it has been shown that APP proteolysis can also occur in the endoplasmic reticulum, golgi apparatus and endosomes to produce intracellular Aβ. Moreover, Aβ can be taken up from extracellular sources through subsequent internalization [28][29][30]. In addition, according to previous studies, Aβ1-42 was considered to be the most neurotoxic Aβ species because it was more prone to aggregate and to initiate the formation of pathological oligomers, fibrils and plaques compared to the other forms [31][32][33]. Our present findings exhibited that hexamers of phloroglucinol, such as dieckol and 8,8′-bieckol, were more effective Aβ inhibitors than eckol, a trimeric compound of phloroglucinol.  Both extracellular and intracellular Aβ play critical roles in the onset of AD. It is commonly considered that extracellular Aβ is produced by the BACE1-mediated processing of APP and released in the extracellular matrix, which is a pivotal cause of AD [27]. However, recently, it has been shown that APP proteolysis can also occur in the endoplasmic reticulum, golgi apparatus and endosomes to produce intracellular Aβ. Moreover, Aβ can be taken up from extracellular sources through subsequent internalization [28][29][30]. In addition, according to previous studies, Aβ  was considered to be the most neurotoxic Aβ species because it was more prone to aggregate and to initiate the formation of pathological oligomers, fibrils and plaques compared to the other forms [31][32][33]. Our present findings exhibited that hexamers of phloroglucinol, such as dieckol and 8,8 -bieckol, were more effective Aβ inhibitors than eckol, a trimeric compound of phloroglucinol.

Effects of Phlorotannins on APP Proteolytic Enzymes Expression and Activity
To confirm whether phlorotannins regulate the APP processing enzymes or not, protein expression of α-secretase (ADAM10), β-secretase (BACE1), γ-secretase and presenilin-1 (PS1) was evaluated. As shown in Figure 4a,b, the expression of ADAM10 was significantly increased by the pretreatment of all tested compounds. It is noteworthy that dieckol remarkably attenuated BACE1 expression even at the lowest concentration ( Figure 4c). Additionally, 8,8 -bieckol significantly decreased BACE1 protein level (p < 0.05), while eckol showed no effect on BACE1. The results were consistent with our previous study showing the strong BACE1 inhibitory effect of dieckol with an IC 50 value of 2.3 ± 0.1 in in vitro enzymatic system [25]. In the PS1 expression, shown in Figure 4d, statistically significant changes were detected as caused by dieckol at 10 and 50 µM (p < 0.05) and 8,8 -bieckol at 50 µM (p < 0.05).
For further investigation into the effects of the tested compounds on the activity of ADAM10 and BACE1, their proteolytic products, including sAPPα and sAPPβ, were measured. Consistent with the results of ADAM10 expression shown in Figure 4a,b, the protein levels of sAPPα were increased after dieckol and 8,8′-bieckol pretreatments (Figure 5). In particular, the expression level of sAPPα was dramatically elevated more than twofold after 50 μM dieckol pretreatment. The expression levels of sAPPβ were significantly reduced with the pre-treatment of all tested compounds.
It was demonstrated that ADAM10 and BACE1 compete for the same substrate, APP, and they have opposite effects on Aβ formation [34]. A previous study by May et al. (2011) revealed that the treatment of a non-peptidic BACE1 inhibitor, LY2811376, reduced sAPPβ and increased sAPPα in the human CNS, implying that the inhibition of BACE1 resulted in a compensatory augmentation in the cleavage of APP at the α-site [35]. Previously, a phlorotannin-rich extract from Ecklonia cava augmented the C-terminal fragment α (CTF-α) and sAPPα in SweAPP-HEK293 cells, which activity was similar to BACE inhibitor Ⅳ, suggesting that the extract acted as relative BACE inhibitor or TACE activator [36]. In a follow-up study by the same research group, the extract regulated the direct expression and substrate activity of α-and γ-secretase, leading to Aβ reduction [37]. These Aβ regulatory effects of the extract on APP processing were probably, in part, associated with our tested compound.   For further investigation into the effects of the tested compounds on the activity of ADAM10 and BACE1, their proteolytic products, including sAPPα and sAPPβ, were measured. Consistent with the results of ADAM10 expression shown in Figure 4a,b, the protein levels of sAPPα were increased after dieckol and 8,8 -bieckol pretreatments ( Figure 5). In particular, the expression level of sAPPα was dramatically elevated more

Regulation of Aβ Production by PI3K/Akt/GSK-3β Signaling Pathway
GSK-3β, highly expressed in the brain, has been identified as an in vivo substrate of the Akt pathway. The phosphorylation of GSK-3β via Akt is essential for the inhibition of GSK-3β during neuronal survival [38]. In addition, the regulation of GSK-3β by Akt is likely to affect other signaling events where GSK-3β is important, such as the hyperphosphorylation of tau [39]. It was demonstrated that ADAM10 and BACE1 compete for the same substrate, APP, and they have opposite effects on Aβ formation [34]. A previous study by May et al. (2011) revealed that the treatment of a non-peptidic BACE1 inhibitor, LY2811376, reduced sAPPβ and increased sAPPα in the human CNS, implying that the inhibition of BACE1 resulted in a compensatory augmentation in the cleavage of APP at the α-site [35]. Previously, a phlorotannin-rich extract from Ecklonia cava augmented the C-terminal fragment α (CTF-α) and sAPPα in SweAPP-HEK293 cells, which activity was similar to BACE inhibitor IV, suggesting that the extract acted as relative BACE inhibitor or TACE activator [36]. In a follow-up study by the same research group, the extract regulated the direct expression and substrate activity of αand γ-secretase, leading to Aβ reduction [37]. These Aβ regulatory effects of the extract on APP processing were probably, in part, associated with our tested compound.

Regulation of Aβ Production by PI3K/Akt/GSK-3β Signaling Pathway
GSK-3β, highly expressed in the brain, has been identified as an in vivo substrate of the Akt pathway. The phosphorylation of GSK-3β via Akt is essential for the inhibition of GSK-3β during neuronal survival [38]. In addition, the regulation of GSK-3β by Akt is likely to affect other signaling events where GSK-3β is important, such as the hyperphosphorylation of tau [39].
The GSK-3β signaling pathway is a critical regulator of APP processing, including BACE1 expression [17]. To determine whether the PI3K/Akt/GSK-3β signaling pathway could mediate the inhibitory effects of the tested phlorotannins on Aβ generation, phosphorylated levels of Akt at Ser473 and its downstream target GSK-3β at Ser9 were measured. Figure 6a, dieckol obviously increased the phosphorylation of Akt more than two-fold compared to the control group (p < 0.001). In addition, dieckol noticeably augmented the phosphorylation levels of GSK-3β at the Ser9 residue, subsequently inhibiting GSK-3β activity, suggesting the activated Akt downregulated GSK-3β activity (p < 0.001, Figure 6a). On the contrary, eckol or 8,8 -bieckol did not alter the phosphorylated levels of Akt and GSK-3β. mediate the inhibitory effects of the tested phlorotannins on Aβ generation, phosphorylated levels of Akt at Ser473 and its downstream target GSK-3β at Ser9 were measured. As shown in Figure 6a, dieckol obviously increased the phosphorylation of Akt more than two-fold compared to the control group (p < 0.001). In addition, dieckol noticeably augmented the phosphorylation levels of GSK-3β at the Ser9 residue, subsequently inhibiting GSK-3β activity, suggesting the activated Akt downregulated GSK-3β activity (p < 0.001, Figure 6a). On the contrary, eckol or 8,8′-bieckol did not alter the phosphorylated levels of Akt and GSK-3β.

As shown in
For further examination, the cells were exposed to PI3K inhibitor LY294002. LY294002 significantly decreased the level of phospho-GSK-3β compared to the control without LY294002. Interestingly, the combination of dieckol and LY294002 led to a markedly decreased phosphorylation of Akt and GSK-3β compared with the treatment with dieckol alone, suggesting that dieckol directly downregulated GSK-3β via the enhancement of PI3K/Akt.  For further examination, the cells were exposed to PI3K inhibitor LY294002. LY294002 significantly decreased the level of phospho-GSK-3β compared to the control without LY294002. Interestingly, the combination of dieckol and LY294002 led to a markedly decreased phosphorylation of Akt and GSK-3β compared with the treatment with dieckol alone, suggesting that dieckol directly downregulated GSK-3β via the enhancement of PI3K/Akt. Dieckol and 8,8 -bieckol significantly decreased the Aβ 1-42 levels (67.33 ± 6.86% and 73.17 ± 0.54%, respectively) compared with the SweAPP N2a control group (p < 0.01, Figure 6b). However, the inhibitory effect of dieckol on Aβ was obviously blocked, when co-treated with LY294002, demonstrating that dieckol decreased Aβ via the PI3K/Aktmediated inactivation of GSK-3β. These findings agree with a recent study exhibiting the reduction in Aβ 1-42 levels upon GSK-3β inhibition in an APP/tau double transgenic mouse model [16]. To study the effects of GSK-3β reduction on Aβ formation, Ly et al. found that the reduced GSK-3β activity is involved in reducing the BACE1-mediated cleavage of APP and Aβ production by decreasing BACE1 gene transcription and expression in cell cultures and APP23/PS45 double transgenic mice [17]. In addition, the GSK-3β inhibition not only attenuated intracellular neurofibrillary tangles (NFTs), but also recovered memory deficits. These findings indicate that GSK-3β appears to be a common molecular connection in both amyloidogenesis and tau abnormalities.
Several recent studies on active compounds within marine organism and plants have paid attention to the modulation of the Akt/GSK-3β signaling pathway. Eckmaxol, one of the unique phlorotannins, significantly reversed the decreased expression of phosphorylated GSK-3β in Aβ-oligomer-evoked neurotoxicity. Eckmaxol revealed a favorable interaction in the ATP binding site of GSK-3β [44]. Fucoxanthin, a marine carotenoid, attenuated Aβ-oligomer-induced neurotoxicity via the PI3K/Akt/GSK-3β and ERK pathway [45]. Polypeptides from Achyranthes bidentate suppressed neuronal apoptosis via regulating the PI3K/Akt/GSK-3β pathway [46]. Curcumin attenuated cognitive function by inhibiting the hyperphosphorylated Tau protein through Akt/GSK-3β signaling [47]. Puerarin, a major isoflavone from Pueraria lobata, protected neuronal damage via the Akt/GSK-3β signaling pathway in cerebral ischemic animal models [48]. In addition, berberine, an alkaloid derived from Berberis species, was known to restore the Akt/GSK-3β pathway in primary cultured neuron and diabetic encephalopathy rats [49].
Glutamate is the major excitatory neurotransmitter in the CNS. Glutamatergic synaptic transmission has been implicated in learning and memory, and synaptic plasticity. However, high levels of extracellular glutamate lead to neuronal death under certain pathological conditions. In addition, Aβ promotes glutamatergic excitotoxicity and potently disrupts synapses and plasticity, providing an explanation for the cognitive deficits in AD [52]. Cui et al. reported that dieckol exerted neuroprotective effects in glutamate-induced mitochondrial-dependent apoptosis by activating the nuclear factor-like 2 (Nrf2)/heme oxygenase-1 (HO-1) pathway in both the primary cortical neurons and the HT22 neurons [53].
In terms of the structure-activity relationship, the numbers and positions of hydroxyl groups on the phenol ring are usually responsible for the neuroprotective properties of phenolic compounds. Dieckol (IC 50 , 2.2 µM) revealed a five-fold higher BACE1 inhibitory activity compared with that of eckol (IC 50 , 12.2 µM) [25]. Furthermore, the molecular size of phlorotannins is a vital factor for a strong interaction with the targeted enzymes, demonstrating that hexamers of phloroglucinol act as a better inhibitor [54]. The Aβ protective property of dieckol with a diphenyl ether linkage was greater than that of 8,8bieckol with a biaryl linkage, although these two compounds are dimers of eckol [26]. A recent study suggested that dieckol had a four-fold stronger inhibitory property against Aβ 25-35 self-aggregation than eckol, which was achieved by interrupting the formation of the β-sheet-rich Aβ structures [24].
The present study revealed that dieckol effectively regulated the PI3K/Akt pathway, which in turn down-regulated GSK-3β activity, resulting in the suppression of Aβ production by modulation of the expression and activity of the APP processing enzymes. Although further in vivo study is definitely required, the present study itself is still meaningful in the terms of discovering potential anti-AD agents from marine organisms. As a next step towards expanding our research, neuronal primary cultures as well as animal studies will be performed.

Aβ ELISA Analysis
The SweAPP N2a cells were cultured in 6-well plates at a density of 3 × 10 6 cells per well. Conditioned media from the phlorotannins-pretreated and -untreated cells were collected. The levels of Aβ 1-40 and Aβ 1-42 were quantified using an ELISA kit according to the manufacturer's protocol (Invitrogen, Carlsbad, CA, USA). The absorbance was measured using a microplate reader at 450 nm. The concentrations of Aβ 1-40 and Aβ  were determined by comparison with their standard curves.

Western Blotting Assay
SweAPP N2a cells were plated in 6-well plates at the density of 3 × 10 6 cells/well and treated with eckol, dieckol, and 8,8 -bieckol (1, 10, and 50 µM) for various durations. The cells were washed with phosphate buffered saline (PBS) and then lysed in extraction buffer (Cell Signaling Technology Inc., Beverly, MA, USA) containing protease inhibitor cocktail for 1 h (Tech & Innovation, Chuncheon, Korea). The protein concentrations were quantified by the BCA method. Equal protein solutions (25 µg) were mixed with a loading buffer and heated at 95 • C for 5 min. The proteins were separated by 10% SDS-PAGE, then transferred to polyvinylidene fluoride (GE, Health Care Life Sciences, Piscataway, NJ, USA) and blocked in 5% skim milk for 2 h, at RT. The membranes were incubated with primary antibodies against APP (1:1000, Immuno-Biological Laboratories (IBL), Fujioka, Japan), The blots were visualized using an enhanced chemiluminescence (ECL) western blotting kit (Amersham Biosciences, Piscataway, NJ, USA) and Atto EZ-capture (Tokyo, Japan). β-Actin was probed as an internal control to confirm that an equal amount of protein was loaded in each lane. In order to avoid quantification and measurement bias, all images were obtained under the same setting conditions. Densitometric analysis and quantification of the protein bands were performed using the ImageJ software (NIH, Bethesda, MD, USA).

Statistical Analysis
All data were expressed as mean ± SD and were representative of the results obtained from three independent experiments. Statistical analyses were performed using SAS software (version 9.3, SAS Institute, Cary, NC, USA). One-way analysis of variance (ANOVA) with post hoc Tukey test was used to assess for the multiple comparisons. Graphs were constructed with the GraphPad Prism 9.0.2 software (GraphPad Software Inc., San Diego, CA, USA).

Conclusions
The present study is the first to demonstrate that dieckol regulated the APP proteolytic processing and Aβ production via the regulation of the PI3K/Akt/GSK-3β signaling pathway. Furthermore, the addition of LY294002 counteracts all the effects of dieckol, demonstrating that Akt/GSK-3β is the primary pathway mediating Aβ production in SweAPP N2a cells. The present findings supported a better understanding of the vital role of dieckol in preventing AD, and its potential to be used as a promising source of anti-AD agents.