Honokiol Attenuates Oligomeric Amyloid β_1-42-Induced Alzheimer’s Disease in Mice Through Attenuating Mitochondrial Apoptosis and Inhibiting the Nuclear Factor Kappa-B Signaling Pathway

Dementia is a kind of central nervous system (CNS) disease characterized by progressive memory loss and cognitive decline. Almost 50 million people worldwide are living with dementia, and approximately 60-70% was resulted from Alzheimer’s disease (AD) (https://blog.alz.co.uk/2016/12/18/alzheimer-community-between-disappointment-and-hope-regarding-drug-development-at-ctad-conference-2016/ accessed in March 2017). A major pathological hallmark of AD is extracellular senile plaques in brain, which is formed by accumulation of amyloid β-protein (Aβ)_1-42 [1]. Aβ_1-42 is derived from the large amyloid precursor protein (APP) precursor in a process called APP processing. In general, β-site APP-cleaving enzyme 1 (BACE1) and α-secretase mediates amyloidogenic and non-amyloidogenic pathway, respectively. In the amyloidogenic pathway, APP is cleaved by BACE1 and produces Aβ. The Aβ is released into the extracellular space and self-aggregates to form oligomers that toxic to neuronal cells. In the non-amyloidogenic pathway, APP was cleaved by α-secretase within the Aβ sequence to produce a soluble secretory amyloid precursor protein (sAPPα) [2, 3]. Thus, factors that trigged overwhelming amyloidogenic pathway and inhibited non-amyloidogenic pathway will result in overproduction of Aβ. Several forms of Aβ_1-42, such as soluble Aβ oligomers (AβO), Aβ fibril and amyloid plaques, are well studied in the pathological mechanisms of AD. Nowadays, soluble AβO, rather than Aβ fibrils or amyloid plaques, has been believed to be neurotoxic [4, 5]. Studies have demonstrated that AβO targeted to synaptic terminals [6] and its accumulation was associated with the spatial learning deficits in AD transgenic mice without amyloid plaque formation [7, 8]. Therefore, direct targeting AβO or attenuation of AβO-induced neurotoxicity is a common approach of AD treatment.

Honokiol (3′,5-di-2-propenyl-1, 1′-biphenyl-2, 4′-diol) is isolated from the barks of Magnolia officinalis, and has mutiple pharmacological activities. Previously, studies has shown that honokiol exhibited activities against anxiety [9, 10], depression [11, 12] and cerebral ischemia/reperfusion-induced memory deficiency [13]. In addition, honokiol has been found to improve learning and memory through preserving cholinergic neurons in the brain of senescence-accelerated prone mouse strain 8 (SAMP8) mice [14] and inhibiting the activity of acetylcholinesterase scopolamine-treated mice [15]. Importantly, Wang and colleagues have demonstrated that honokiol has the ability to cross the blood-brain barrier (BBB) [16], which enabled it to be a candidate for AD treatment. However, no study has reported the effects of honokiol in Aβ-accumulation-based cognition deficiency. Only an in vitro studies has reported that honokiol could attenuate aggregated form of Aβ_1-42 induced oxidative stress, mitochondrial dysfunction and apoptosis in PC-12 cells [17]. The present study aims to evaluate the effects of honokiol on AβO-induced cognitive dysfunction and investigate the possible mechanisms.

Male wild-type C57BL/6 mice (19-22 g, 6 weeks old) were purchased from the Experimental Animal Center of Dalian Medical University (Dalian, China). The mice were housed in standard animal house under a 12-h light/dark cycle at 22-23 °C with 60 ± 10 % humidity and were provided with food and water ad libitum. All of the methods and experimental procedures were approved by the Institutional Ethics Committee of Dalian Medical University.

Oligomers were generated as described elsewhere [18]. Aβ_1-42 (GL Biochem Ltd., Shanghai, China) was dissolved in cold HFIP at a concentration of 1 mg/mL for 2 – 4 h at room temperature. HFIP was removed by evaporation under N2, and the thin transparent film of peptides at the internal surface of the tube was stored at 80 °C. For oligomer preparation, HFIP-treated Aβ_1-42 peptide was dissolved in dimethylsulfoxide (DMSO) at 5 mM and diluted to 500 µM in artificial cerebrospinal fluid (aCSF) or cell culture medium and aged by incubation at 4 °C for 24 h.

The animals were randomly divided into 5 groups (n=6): control (con), Aβ_1-42 oligomer treatment (Aβ), honokiol (0.7, 7, and 70 µg/kg, GL Biochem Ltd., Shanghai, China) treatment (H-L, H-M, H-H, respectively) groups. The mice were anesthetized with 10 % chloral hydrateand fixed on a stereotaxic instrument. Then oligomeric Aβ_1-42 (2 µg/µL, 2 µL) was injected into both lateral hippocampi in the model and honokiol groups via a 5-µL micro-injector daily for 7 days. The mice in the control group received aCSF injection. The injection site was posterior from the bregma (AP) 2 mm, mediolateral from the midline (MR) 1.6 mm, and dorsoventral from the skull (DV) 1.5 mm. After 7 days injection, mice in the honokiol group were administered intraperitoneally with indicated dosage of honokiol daily for 14 days, while mice in the sham and model groups were administered with equal volume of 1% DMSO. The mice were sacrificed after the MWM test and the hippocampal tissues were collected for the following examinations.

The ability of spatial learning and memory of the mice was assessed using a MWM test. The test was performed on the 14th day after honokiol administration. The test was performed in a circular water tank (120 cm in diameter) containing water (23±1 °C) at a depth of 24 cm and divided into four quadrants. A hidden escape platform (9 cm in diameter) was placed in the center of one quadrant 1 cm under the surface of the water. Each mouse received four trials per day for five consecutive days. The mice were initially placed in the water facing the wall of the tank in the three quadrants which did not have the platform. The escape paths and latency were recorded by a camera. If the mouse failed to find the platform within 60 s, its escape latency was recorded as 60 s. The probe test was carried out at day 5 by removing the platform and the mice were placed at the starting point at the quadrant opposite to the platform quadrant and allowed to swim freely for 60 s. The time that the mice spent in the target quadrant and the number of times the mouse crossed over the platform location was recorded.

Primary cultures of hippocampal neurons were generated from fetal brains (embryonic day 18; E18) obtained from female Sprague-Dawley rats (the Experimental Animal Center of Dalian Medical University, Dalian, China). The brains were removed and hippocampi were dissected in dulbecco’s modified eagle medium (DMEM) (Invitrogen, Carlsbad, CA) containing 0.125% trypsin solution (Beyotime, Haimen, China) for 10 min at 37°C. Fetal bovine serum (10%, Gibco, Carlsbad, CA) was then added to terminate the digestion. The dispersed tissues were centrifuged at 300 g for 5 min and were resuspended in DMEM (Invitrogen) containing 2 % B27 supplement (Gibco) and 0.5mM L-glutamine (Gibco). For the hoechst staining, neurons were plated in glass cover slips. For the enzyme linked immunosorbent assay (ELISA), real-time PCR, and Western blotting analysis, neurons were plated onto six-well plates coated with poly-D-lysine. The neurons were maintained in a humidified atmosphere of 95% air and 5% CO₂ at 37°C. The culture medium was half-replaced every 3 days. Neurons cultured for 8 days were treated with indicated concentrations of honokiol (10, 50 or 100 µM) for 24 h and followed by incubated with 10 µM AβO for 12 h.

The hippocampal tissues were homogenized in cooled PBS and subjected to repeated freezing and thawing in liquid nitrogen. The homogenate was centrifuged at 10, 000 g for 10 min at 4 °C. The supernatant was used for ELISA. The culture medium were collected and centrifuged at 150 g for 20 min. The supernatant was used for ELISA. The levels of pro-inflammatory cytokines tumor necrosis factor-α (TNF-α), interleukin (IL)-1β and IL-6 in the hippocampal tissues and culture medium were determined using commercial ELISA kits according to the manufacture’s protocols (Wuhan Boster Biological Technology, Ltd., Wuhan, China).

The hippocampal tissues were fixed in 4 % paraformaldehyde (PFA) for 24 h and embedded in paraffin. The paraffin blocks were cut into 5-µm thick sections. The sections were deparaffinized in xylene, hydrated using a series of ethanol washes and incubated with permeabilization solution (0.1% Triton X-100 in 0.1% sodium citrate) for 8 min. Then, the hippocampal sections were stained using an In Situ Cell Death Detection Kit (Wanleibio, Shenyang, China) and then counterstained with haematoxylin. The sections were observed under an optic microscope (DP73; Olympus).

The neurons grown on poly-L-lysine-coated glass cover slips were staining using a Hoechst 33258 Staining Kit (Beyotime). Briefly, the neurons were fixed in the fixing solution in the kit for 20 min at room temperature. After a wash stage with PBS, the neurons were then staining with the Hoechst-33258 for 5 min at room temperature. Finally, the cells were mounted with anti-fade fluorescence mounting medium and observed under a fluorescence microscope (IX53, Olympus, Tokyo, Japan).

Intracellular ROS was assessed using commercial ROS assay kit (Nanjing Jiancheng Bioengineering Institute, Nanjing, China) by DCF-DA method and MMP was assessed using commercial MMP assay kit (Beyotime) by JC-1 method according to the manufactures’ protocols.

The hippocampal tissues and neurons were homogenized in cooled RIPA buffer and centrifuged at 12, 000 g for 10 min at 4 °C. Protein concentration was determined using a BCA protein assay kit (Beyotime). Tissue homogenates (40 µg protein) from each mouse were boiled in loading buffer and electrophoresed on 10% sodium dodecyl sulfate-polyacrylamide gels and transferred to polyvinylidene difluoride membranes (Millipore, Billerica, MA, USA). The membranes were blocked with 5 % nonfat milk and incubated with primary antibodies overnight at 4 °C. After incubation at 37°C for 45 min with HRP (horseradish peroxidase)-conjugated secondary antibodies (1: 5000; Beyotime), the membranes were developed using the enhanced chemiluminescence reagent (Beyotime) and visualized with autoradiography film. The intensity of each band was quantified with Gel-Pro-Analyzer software (Media Cybernetics, Bethesda, MD, USA). The primary antibodies used in the present study were listed in table 1.

Nuclear proteins were extracted from the hippocampal tissues and neurons using a Nuclear and Cyto-plasmic Protein Extraction Kit (Beyotime), followed by determination of protein concentrations using the BCA protein assay kit (Beyotime). The levels of NF-κB activation in each group hippocampal tissues and neurons were determined by EMSA using the nonradioactive NF-κB EMSA kit (Viagene biotech Inc., Changzhou, China) following the manufacturer’s instructions.

Data were expressed as mean ± standard deviation (SD). Statistical analysis of differences between groups was carried out by one-way ANOVA followed by the LSD test. P values < 0.05 were considered significant.

The spatial learning abilities of Aβ oligomer-treated mice were evaluated with MWM. The results showed that over the 5 training days, the time to find the hidden platform progressively declined in all groups. However, Aβ oligomer-treated mice showed significantly higher escape latencies than the control group at day 4 and 5 (P<0.01), implying learning deficits in the Aβ oligomer-induced mice. From the fourth day, the high-dose of honokiol-treated mice exhibited a significantly better performance than mice in the Aβ group. At day 5, the middle-dose of honokiol-treated mice also showed significantly lower escape latencies than the Aβ group (Fig. 1A). In the spatial probe test, as shown in Fig. 1B, compared with the control group, the time that the mice in the Aβ group spent in the target zone was significantly shorter (P<0.01). The treatment with honokilol dose-dependently increased the percentage of time in the target zone in Aβ oligomer-induced mice. The representative pathways of mice in the spatial probe test are shown in Fig. 1C. These results indicate that honokilol treatment significantly attenuates spatial learning impairments in Aβ oligomer-induced mice.

In the Nissl staining, obvious neuronal loss and nucleus disappearance were observe in the CA1 region of hippocampus Aβ oligomer-treated mice (Fig. 2A). The number of neurons was significantly decreased in the CA1 region in the Aβ group compared with the control group (P<0.01). However, the decreased neurons were dose-dependently increased by the treatment of honokiol. In the NeuN-TUNEL double staining, the number of apoptotic neurons was dramatically increased after Aβ_1-42 oligomer injection, which was significantly decreased by the treatment of honokiol (Fig. 2B).

We then investigated the expressions of the proteins involved in mitochondrial-apoptosis. As shown in Fig. 3A, a significant decrease in Bcl-2 expression and increase in Bax expression were found in the hippocampus of Aβ oligomer-treated mice compared with the control mice, and the expressions of cleaved-caspase 9 and 3 were also increased notably compared with the control mice. Moreover, cytochrome C (Cyt C) leakage was found, as evidenced by the upregulated cytoplasm Cyt C and the downregulated mitochondrial Cyt C. Treatment with honokiol dose-dependently enhanced the expression of Bcl-2 and reduced that of Bax, and the expressions of cleaved-caspase 9 and 3 were also reversed compared with the Aβ oligomer-treated mice. The abnormal expressions of cytoplasm mitochondrial Cyt C were also restored by the treatment of honokiol.

To confirm the effects of honokiol on mitochondrial functions, MMP and ROS generation were also evaluated in the hippocampus of Aβ oligomer-induced mice. Compared with the control group, the MMP was significantly reduced and the ROS production was significantly increased in the Aβ group (Fig. 3B and C). Treatment with high-dose of honokiol effectively enhanced the MMP and suppressed the ROS production in the hippocampus, indicating the protective effects in mitochondrial of honokiol.

The NF-κB signaling pathway, which is involved in the pathological mechanisms of Aβ-associated AD [19], was also investigated in the present study. Compared with the control mice, the expression of IκBα was downregulated and the expression of p-p65 was upregulated in the hippocampus of Aβ oligomer-treated mice (Fig. 4A and B). In addition, nuclear p65 expression was also upregulated, and the activity of NF-κB-DNA binding was increased (Fig. 4C and D). Honokiol dose-dependently reversed these changes. Moreover, the protein expressions of two target genes of NF-κB, amyloid precursor protein (APP) and β-site APP cleaving enzyme (BACE1) [20-22], were also upregulated in the hippocampus after Aβ oligomer-stimulation and were restored after the treatment of holokiol (Fig. 4E and F).

To investigate the cellular target of honokiol, hippocampal neurons were primary cultured and the associated mechanisms were studied. Fig. 5A shows the effects of honokiol on the neuronal apoptosis. Aβ oligomer induced obvious apoptosis in primary neurons, as evidenced by the significantly increased number of condensed nuclei and high-stained neurons. The treatment of honokiol reduced the number of apoptotic neurons in a dose-dependent manner (Fig. 5A). Consistently, Aβ oligomer also induced the expressions of mitochondrial-apoptosis-associated proteins made a pro-apoptotic shift. The expression level of the anti-apoptotic protein Bcl-2 was downregulated and the pro-apoptotic proteins Bax, cleaved-caspase 9 and 3 were upregulated (Fig. 5B). In addition, the expression level of Cyt C was increased in the cytoplasm and decreased in the mitochondrial, indicating the Cyt C leakage in the neurons. Honokiol effectively restored these changes. These findings are similar to that in the hippocampal tissues and confirm that honokiol attenuates Aβ oligomer-induced neuronal apoptosis.

MMP and ROS were also examined to reflect the mitochondrial function in primary hippocampal neurons (Fig. 5C and D). In consistent with the observation in vivo, Aβ oligomer leads to reduced MMP and enhanced ROS and pretreatment with honokiol dose-dependently reversed these changes.

The effect of honokiol on the NF-κB signaling pathway was also investigated in primary hippocampal neurons. In line with the in vivo study, the expression of IκBα (Fig. 6A) was inhibited and the p65 phosphorylation was enhanced by Aβ oligomer (Fig. 6B), thereby promoting the p65 nuclear translocation, as evidenced by the increased expression of nuclear p65 (Fig. 6C). Meanwhile, NF-κB-DNA binding activity was increased after Aβ oligomer exposure (Fig. 6D). In addition, the protein expressions of APP (Fig. 6E) and BACE1 (Fig. 6F) were also upregulated in the Aβ group. The pretreatment with honokiol effectively inhibited the activation of the NF-κB signaling pathway and the protein expressions of APP and BACE1.

In the present study, we found that honokiol improved AβO-induced cognitive dysfunction in mice. In addition, honokiol attenuated mitochondrial dysfunction and mitochondrial-pathway apoptosis in hippocampus of AβO-treated mice and AβO-stimulated hippocampal neurons. Furthermore, honokiol inhibited AβO-activated NF-κB signaling pathway and therefore downregulated the expressions of APP and BACE1. These mechanisms may contribute to the neuroprotective effects of honokiol.

Hippocampal neurogenesis is essential for learning and memory. The cognitive deterioration in AD is closely associated with hippocampal injury induced by Aβ [23]. In the present study, we observed that mice with AβO intrahippocampal injection had the delayed escape latency in MWM. In addition, Nissl staining and TUNEL-NeuN double staining showed the neuronal damage and apoptosis in CA1 region where AβO had been injected. Apoptosis plays a critical role in neurologic disorders including AD. Many factors will induce neuronal apoptosis during AD. Although the underlying mechanism is elusive, accumulating evidence shows that AβO induces neuronal apoptosis and contributes for the pathophysiology of AD [24-26]. Honokiol effectively protected neurons against AβO in vivo and in vitro, as evidenced by the markedly improved neuronal survival in hippocampus and the reduced apoptotic neurons in hippocampus and primary neurons exposed to AβO. This neuroprotective effects may contribute to the learning and memory improvement in mice.

In general, cellular apoptosis is triggered via two major pathways, the mitochondrial (intrinsic) pathway and the death receptor-mediated (extrinsic) pathway. Here, we focused on the mitochondrial pathway, which is mediated by Cyt C, the Bcl-2 family and caspases. In this apoptotic pathway, Cyt C is released from mitochondria, assembles with Apaf-1and pro-caspase 9, and activates caspase-3 to trigger apoptosis [27]. Bcl-2 family is involved in the regulation of the Cyt C mediated mitochondrial apoptotic pathway [28, 29]. BH3-only proteins (apoptosis initiator), Bcl-2 (anti-apoptotic protein) and Bax (pro-apoptotic proteins) and Bak (Bcl-2 antagonist) form the tripartite apoptotic switch to control the release of apoptogenic factors, particularly Cyt C [30]. In our study, honokiol dose-dependently inhibited AβO-induced downregulation of Bcl-2, upregulation of Bax, Cyt C leakage and elevated levels of cleaved caspase-9 and 3 in hippocampus and cultured neurons, which indicates that honokiol attenuates neuronal apoptosis by modulating mitochondrial apoptotic pathway-associated proteins in AβO-exposed hippocampus.

As the finding of honokiol inhibited mitochondrial Cyt C leakage in the AβO-exposed hippocampal neurons, we evaluated the effects of honokiol on the MMP and ROS generation, two typical parameters that reflect mitochondrial functions, in AβO-exposed hippocampus and neurons. Mitochondria are the main source of cellular ATP production via oxidative phosphorylation (OXPHOS). Besides, mitochondria also produce ROS as signaling molecules and participate in diverse cellular processes [31]. However, stress leads to excessive ROS production in mitochondria, which in turn damages the ROS-produced cells and exacerbates the stress. During AD, Aβ has deleterious effects on mitochondrial function and contributes to energy deficiency and overproduction of ROS in brain [32]. In agreement with the previous studies, we found that AβO results in reduced MMP and elevated ROS generation in the AβO-exposed hippocampus and primary neurons. Predictably, honokiol improved mitochondrial functions by increasing MMP and reducing ROS production. Based on our findings and the accepted knowledge [32-34], AβO interfered with the OXPHOS and respiratory chain, destroyed the proton gradient, decreased the MMP and increased the permeability of mitochondrial membrane, which led to Cyt C leakage and excessive ROS generation. Meanwhile, the Bcl-2 was inhibited and the Bax was activated, which finally induced apoptosis together with the Cyt C and ROS. Although our findings did not reveal a definite molecular target of honokiol, we can conclude that honokiol can protect neurons from AβO-induced mitochondrial dysfunction, which may be involved in its anti-apoptotic effects.

NF-κB is a key transcriptional regulator of various physiopathologic processes including learning and memory [19]. In normal state, NF-κB is inhibited in cytoplasm by IκBα. When IκBα is phosphorylated by IκB kinase (IKK), NF-κB is released and translocates to the nucleus, where it binds to the target DNAs and regulates the gene expressions [35]. Several factors can activate NF-κB in AD brain [19]. Here we found that AβO induced markedly nuclear translocation of NF-κB and this should be, at least partly, resulted from the increased ROS [36]. In addition, APP and BACE1, two target genes of NF-κB [21, 37], were found to be upregulated in the hippocampus and primary neurons treated by AβO. Based on the previous reports, the effects of NF-κB activation on APP and BACE1 are controversial. Yang et al. found that TNF-α-induced NF-κB activation did not affect APP in primary hippocampal neurons [38], while others reported that NF-κB activation upregulates promoter activity of APP in HEK293 cells [22]. Both Chami and Chen discovered that NF-κB mediated elevated BACE1 expression in HEK293 cells [22, 37], while Bourne et al. reported that factor NF-κB served as a suppressor in neuronal cell lines but as an activator of BACE1 transcription in astrocytes [39]. These discrepancies may due to the difference of the cells and stimulation factors used in the studies. However, it is noteworthy that the effect of AβO-induced NF-κB activation on APP and BACE1 expressions has not been reported. In our study, NF-κB activation and upregulation of APP and BACE1 were found in AβO-treated hippocampus and primary neurons, honokiol treatment restored these changes. These findings may lead to a hypothesis that the AβO-induced NF-κB activation upregulates APP and BACE1 expressions, thereby resulting in elevated Aβ generation, which forms a vicious cycle and exacerbates the neuronal damage. Honokiol blocked the cycle by inhibiting the NF-κB signaling. Unfortunately, the present study did not directly reveal the regulatory effect of NF-κB on APP and BACE1 under AβO stimulation, which is not the aim of this study and should be further studied in the future. Our study indicates that honokiol may protect the hippocampal neurons against AβO by inhibiting NF-κB activation.

Honokiol attenuates AβO-induced learning and memory dysfunction in mice. The neuroprotective effects of honokiol may be associated with the neuronal apoptotic attenuation, mitochondrial functions preservation and NF-κB inactivation. Our findings provide evidence for the utility of honokiol in treating AβO-based cognitive dysfunction.

This study was supported by a grant from the Natural Science Foundation of Liaoning Province (No. 2015020302).

The authors declare no conflict of interest.