Re: 신경성장인자(nerve growth factor, BDNF) 자극 한약재 탐구

[문형철]

beyond reason

Critical review

Nonpeptide neurotrophic agents useful in the treatment of neurodegenerative diseases such as Alzheimer's disease

Author links open overlay panelMasaakiAkagiaNobuakiMatsuiaHarukaAkaeaNanaHirashimaaNobuyukiFukuishiaYoshiyasuFukuyamaaReikoAkagib

Show more

https://doi.org/10.1016/j.jphs.2014.12.015Get rights and content

Under a Creative Commons license

open access

Abstract

Developed regions, including Japan, have become “aged societies,” and the number of adults with senile dementias, such as Alzheimer's disease (AD), Parkinson's disease, and Huntington's disease, has also increased in such regions. Neurotrophins (NTs) may play a role in the treatment of AD because endogenous neurotrophic factors (NFs) prevent neuronal death. However, peptidyl compounds have been unable to cross the blood–brain barrier in clinical studies. Thus, small molecules, which can mimic the functions of NFs, might be promising alternatives for the treatment of neurodegenerative diseases. Natural products, such as or nutraceuticals or those used in traditional medicine, can potentially be used to develop new therapeutic agents against neurodegenerative diseases. In this review, we introduced the neurotrophic activities of polyphenols honokiol and magnolol, which are the main constituents of Magnolia obovata Thunb, and methanol extracts from Zingiber purpureum (BANGLE), which may have potential therapeutic applications in various neurodegenerative disorders.

• Previous article in issue
• Next article in issue

Keywords

Nonpeptide neurotrophic agents

Natural products

Senescence-accelerated mice

Rat cortical neurons

1. Introduction

The percentage of older adults is increasing with the medical advances that have prolonged the lifespan of individuals. The extensively accepted standard indicates that numerous countries, including Japan and Germany, have become “aged societies” (Table 1). In Japan, the population ratio of people aged more than 65 years is estimated to reach 29.1% by 2020 and further increase to 38.8% by 2050 (according to data from World Population Prospects: The 2010 Revision, Japan).

Table 1. The increase of people over 65-years old in various countries.

Ratio of aged 65+ (%)Year2020203020402050

World	9.4	11.7	14.3	16.2
More developed regions	19.0	22.4	24.5	25.7
Less developed regions	7.5	9.8	12.5	14.7
Japan	29.1	31.6	36.1	38.8
China	12.0	16.5	23.3	25.6
India	6.3	8.3	10.5	13.5
United States of America	16.2	19.9	20.9	21.2
United Kingdom	18.7	21.1	23.0	23.6
Germany	23.0	28.0	31.0	30.9
France	20.3	23.1	24.9	24.9

Note: data from “World Population Prospects: The 2010 Revision” by United Nations.

More developed regions: Europe, Northern America, Australia/New Zealand and Japan.

The data for China don not include Hong Kong and Macao, Special Administrative Regions (SAR) of China.

Japan: “2010 Population Census” by the Ministry of Internal Affairs and Communications. Future population Population & Household Projection (medium-mortality assumption) by the National Institute of Population and Social Security Research.

The number of adults with senile dementias, such as Alzheimer's disease (AD), Parkinson's disease (PD), and Huntington's disease (HD), has increased owing to the advent of this rapidly aging society. In Japan, the elderly population with dementia will increase to 4,700,000 by 2025 unless current medical treatments are not significantly improved (Table 2). Dementia causes serious health problems to the elderly and burdens the overall society. Although AD is not a normal consequence of aging, the greatest risk factor for AD is aging. AD is the most preval‎ent dementia (accounting for 50%–80% of dementias, according to the Alzheimer's Association, www.alz.org) and some of its causes have recently been discovered. However, we still lack effective preventative or curative therapies.

Table 2. Patients with dementia (x 10,000) and the increasing trend.

Year2010201520202025

Patients with dementia	280	345	410	470
% of people above 65 years old	9.5	10.2	11.3	12.8

Note: data from National Institute of Population and Social Security Research, “Japan's Future Estimated Population: 2006 median Estimate”.

The senile dementias (AD, PD, and HD), brain trauma, and amyotrophic lateral sclerosis are neurodegenerative diseases that are often associated with synaptic loss and neuronal death. In most cases, neuronal injuries occur after self-degeneration, toxic insults, or lack of nutrients (1). AD is characterized by neuronal loss, and the overall treatment strategy for AD is to prevent neuronal death or to produce new neuronal cells in order to replenish the cells in the degenerated regions. Neurotrophins (NTs) are expected to play a role in the treatment of AD because neuronal death is prevented by endogenous neurotrophic factors (NFs), such as nerve growth factor (NGF) and brain-derived neurotrophic factor (BDNF) (2). Therapeutic applications of NFs by intracranial injections, transplantation of cells secreting NFs, or gene therapy have shown promising results in animal models of neuronal degeneration with learning and memory impairments as well as in clinical trials with AD patients. Thus, NFs are likely to be useful agents for ameliorating neurodegeneration (3). However, NFs were unable to cross the blood–brain barrier in clinical studies and were easily metabolized. Moreover, they caused immune responses because of their peptidyl properties. Therefore, small molecules, which can mimic the functions of NFs, might be promising alternatives for the treatment of neurodegenerative diseases (4). In vitro, NTs can also promote process outgrowth and survival. Thus, rat cortical neurons were cultured to investigate the small neurotrophic molecules in natural products, which resulted in the discovery of several active compounds (5). There are some reports on the elucidation of the mechanisms of natural products, but the detailed mechanism is yet unclear. The present review focuses the neurotrophic activity of polyphenols, honokiol and magnolol, the main constituents of Magnolia obovata Thunb, and methanol extracts from Zingiber purpureum (BANGLE), which may have potential therapeutic applications in various neurodegenerative disorders.

2. AD: pathology and therapeutics2.1. Pathology of AD

AD is a progressive neurodegenerative disorder characterized by accelerated neuronal cell death resulting to dementia. The hallmark pathological features are extracellular amyloid plaques and intraneuronal fibrillar structures. The latter includes neurofibrillary tangles (NFTs), neuropil threads, and dystrophic neuritis that invade amyloid plaques. The extracellular deposition of proteolytic fragments of amyloid precursor protein (APP) is termed amyloid β (Aβ), whereas the fibrillar pathologies are composed of the microtubule-associated protein, tau, assembled into polymeric filaments, forming amyloid plaques (6). APP is a large type-1 transmembrane multidomain protein that performs multiple cellular activities. The most frequently expressed isoforms in the brain are APP695, APP751, and APP770. Secretases (α, β, and γ) catabolize APP, forming non-amyloidal and/or amyloid-derived products (7). β-secretase, also known as beta-site amyloid precursor protein cleaving enzyme 1 (BACE1), cleaves APP at Met 671 to produce soluble amyloid precursor protein β (sAPPβ) and a C-terminal 99 fragment. This is then cleaved by γ-secretase at Val 711 and Ala 713, resulting in the formation of APP intracellular domain (AICD), Aβ40, and Aβ42 peptides (8). Aβ fibrils are formed by the aggregation of these amyloid peptides. On the other hand, α-secretase cleaves APP at Lys 687 to form the soluble amyloid precursor protein α (sAPPα) and a C-terminal fragment of 83 amino acid residues. This fragment is cleaved by γ-secretase to produce non-amyloidogenic Aβ17–42 [also known as protein 3 (p3)] and AICD. γ-Secretase is a heterotetrameric membrane-embedded aspartyl protease consisting of four subunits: nicastrin, presenilin, anterior pharynx, and presenilin enhancer 2 (9). When α- or β-secretase cannot cleave APP, it is cleaved by γ-secretase, forming the soluble amyloid precursor protein γ (sAPPγ) and AICD (10). Some drugs that target APP processing are complex and challenging to design because of the multiple functional enzymes and substrates involved. Polyphenols are potential activators of α-secretase and inhibitors of β- and γ-secretase, resulting in a reduction in amyloid fibril deposition in the brain (11).

In AD, the tau that is present in NFLs is aberrantly phosphorylated and often proteolytically truncated at its C terminus (12). Cleavage of tau at Glu 391 by a yet-to-be-identified protease is one such proteolytic event. Tau truncated at this site is observed in the NFTs in brains of AD patients (12). Such post-translational modifications are believed to impair the ability of tau to bind/stabilize microtubules.

The AD brain exhibits evidence of oxygen radical-mediated damage. However, the ability to directly implicate this mechanism in AD has been a difficult task because of the evanescent nature of oxygen radicals and the use of non-specific analytical approaches. Oxidative stress increases in the AD brain. The central nervous system (CNS) is particularly vulnerable to oxidative damage because it has a high energy requirement, high oxygen consumption rate, and relative deficit of antioxidant defense systems compared with that of other organs (13).

Brain insulin/insulin-like growth factor (IGF) resistance and deficiency are accompanied by impairments in glucose utilization and disruption of energy metabolism, with attendant increases in the mitochondrial dysfunction, oxidative stress, production of reactive oxygen species (ROS), and DNA damage. All of these situations drive pro-apoptosis, pro-inflammatory, and pro-AβPP-Aβ cascades. Insulin stimulates the expression‎ of glucose transporter protein 4 (GLUT4) and protein trafficking from the cytosol to the plasma membrane to modulate glucose uptake and utilization. The stimulation of GLUT4 with insulin is critical for regulating neuronal metabolism and energy production, which are needed for memory and cognition (14). Deficiencies in energy metabolism that are caused by the inhibition of insulin/IGF signaling promote oxidative stress, mitochondrial dysfunction, and pro-inflammatory cytokine activation. Oxidative stress results in the generation and accumulation of reactive oxygen and nitrogen species, which attack subcellular organelles and cause chemical modifications of lipids, proteins, DNA, and RNA. For example, adducts formed with macromolecules can compromise the structural and functional integrity of neurons due to the loss of membrane functions, disruption of the cytoskeleton, dystrophy of synaptic terminals, deficits in neurotransmitter functions and plasticity, and perturbation of signaling pathways for energy metabolism, homeostasis, and cell survival. The expression‎ and phosphorylation of tau are regulated by insulin and IGF. In AD, brain insulin/IGF resistance reduces signaling through phosphoinositol-3-kinase (PI3K) and Akt and increases activation of glycogen synthase kinase 3β (GSK-3β) (15). Overactivation of GSK-3β is partly responsible for the hyperphosphorylation of tau, thus promotes tau misfolding and fibril aggregation. The consequences include failure to generate sufficient quantities of normal soluble tau in relation to accumulation of hyperphosphorylated insoluble fibrillar tau and attendant exacerbation of cytoskeletal collapse, neurite retraction, and synaptic disconnection. Finally, intracellular AβPP-Aβ directly interferes with PI3K activation of Akt, impairing neuronal survival and increasing GSK-3β activation and hyperphosphorylation of tau. Consequently, hyperphosphorylated tau is prone to misfolding, aggregation, and ubiquitination, prompting the formation of dementia-associated paired-helical filament-containing neuronal cytoskeletal lesions.

2.2. Therapeutics for AD

Alzheimer's disease is a process in which toxic Aβ fragments are generated (7) due to uncontrolled cleavage of APP by unknown inducing factors. Alzheimer's disease is also characterized by amyloid fibril and phosphorylated tau aggregates and tangles. At present, there are two major problems with AD drug discovery.

(1)

There is no available animal model that reflects all the pathological events observed in the human AD brain.

(2)

There is a lack of reliable biomarkers to detect and understand the progression of AD.

Scientists are trying to develop drugs that can simultaneously perform multiple tasks such as reducing inflammation, inhibiting β-secretase, activating α-secretase, preventing Aβ and tau aggregation, and driving the disintegration of preformed fibrils. However, no perfect drug or treatment of AD has been discovered so far, and several drugs have failed in recent clinical trials.

2.2.1. Experimental models to screen neurotrophic compounds

The brain contains many types of neurons having different degrees of selectivity for NTs. Neurotrophic activity is experimentally model-dependent; for example, septal cholinergic neurons in cultures were extremely sensitive to NGF, whereas septal dopaminergic neurons were not (16). Small molecular weight neurotrophic compounds also have different degrees of selectivity for neuronal cells. Therefore, a compound may be significantly sensitive to NTs in one experimental model and extremely insensitive in another model. Currently, several in vivo and in vitro screening models have been designed to examine the different degrees of selectivity in neurotrophic studies (Table 3).

Table 3. Experimental models to screen neurotrophic compounds.

Cell lines	PC12 cells (17), SH-SY5Y cells (18), P19 cells (19), Neuro2A cells (20) et al.
Primary cultured cells	cortical neurons (21), hippocampal neurons (21),
	dorsal root ganglion neurons (22), spinal ventral horn neurons (23),
	medial septal neurons (24), retinal ganglion cells (25), astrocytes (26) et al.
Tissue slices	hippocampal slices (27) et al.
Whole animals	senescence-accelerated mice 1 (SAMP1) (28), SAMP8 (29),
	olfactory bulbectomized mice (30) et al.

A small molecule can mimic NTs by directly activating or upregulating the NT receptors. Furthermore, a small molecule can upregulate the expression‎ of NTs, enhance the intracellular NT signaling pathway, and/or inhibit the degradation of the trophic signal. The physiological structure and function of the neurons are maintained by the NTs when the neurons completely differentiate into their neuronal phenotypes. Therefore, a lack of NTs in the body or withdrawal of NTs from the culture medium results in deleterious signals, which cause neuronal apoptosis.

2.2.2. Small molecular weight compounds with neurotrophic activity

Compounds such as 1-{3-[2-(1-benzothiophen-5-yl)ethoxy]propy-l}azetidin-3-ol maleate (T-817-MA), xaliproden, and DMXB-A have neurotrophic activities and are in the clinical testing stage. The first of these has been developed as a therapeutic agent to treat neurodegenerative disorders such as AD by modulating endogenous antioxidative mechanisms. Phase II clinical trials are being conducted in the United States to assess the efficacy of T-817-MA in treating AD (31).

Xaliproden {SR57746A; 1-[2-(naphth-2-yl)ethyl]-4-(3-trifluoromethylphenyl)]-1,2,5,6-tetrahydropyridine hydrochloride} is a selective and potent 5-HT1A receptor agonist with low nanomolar affinity for the 5-HT1A receptors. Studies on xaliproden have focused on the pro-neurotrophic properties of this compound in cellular models, including rat pheochromocytoma PC12 cells, primary neuronal cells, and animal models of neurodegeneration (32).

DMXB-A {GTS-21; 3-[(2,4-dimethoxy)benzylidene]-anabaseine dihydrochloride}is a derivative of the naturally occurring alkaloid, anabaseine (33). It is a partial agonist of human α7 nicotinic receptors and, at higher concentrations, a weak antagonist of α4β2 receptors and serotonin 5-HT3 receptors (33). Approximately 40% of the oral dose is absorbed within 1 h of administration (34). Metabolites with hydroxyl substituents at positions 2 and 4 are more efficacious agonists when they are bound to the receptor, but the extent of their production in the human brain in unknown (34). DMXB-A improves memory in several animal models, and it normalizes the inhibition of auditory responses in rodents with significantly less tachyphylaxis than that with nicotine. Mecamylamine, an α4β2 antagonist, has a deleterious effect on conditioned learning in young rabbits. The eyeblink classical conditioning acquisition is improved in young rabbits administered with mecamylamine and DMXB-A (35). In aged rats, DMXB-A improves one-way active avoidance and the performance in Lashley III maze testing and 17-arm radial maze test (35). Passive avoidance deficits are normalized in rats and ischemia-induced hippocampal cell death and impaired passive avoidance performance in gerbils are prevented by treatment with DMXB-A (36). DMXB-A also improves performance on the Morris water maze (35). Positive neurocognitive effects, particularly on attention, were observed in healthy volunteers administered with DMXB-A (37). All subjects were eval‎uated for performance on a computerized test battery to measure the effect of treatment on cognitive functions, including changes in attention [simple reaction time (RT), choice RT, and digit vigilance], numeric and spatial working memory, secondary episodic recognition memory (word and picture recognition and immediate and delayed word recall), and visual tracking. DMXB-A significantly improved performance on simple RT, choice RT, correct detection during digit vigilance, both word and picture recognition memory, and both immediate and delayed word recall. DMXB-A improved subject performance speed on numeric and spatial working memory tasks. Improvement was generally observed with a 25-mg dose, with further improvement at a 75-mg dose, and an equivalent effect at a 150-mg dose (37).

2.2.3. Neurotrophic natural products

Natural products that are used in traditional medicine or as nutraceuticals could potentially be used to develop new therapeutic agents against neurodegeneration. Illicium, the sole genus of the family Illiciaceae, is distributed in the southern parts of China (38). Majority of these plants are considered toxic; however, compounds from these plants have demonstrated neuroactivity and are characterized according to their neurotrophic profiles. To characterize the plants according to their neurotrophic profiles, rat pheochromocytoma PC12 cells (17), primary cultured rat cortical/hippocampal neurons (21), and human neuroblastoma SH-SY5Y cells (18) were used (Table 3). The rat PC12 cells are easy to culture and have been extensively used as a neuronal model because they allow the screening of several compounds. Primary cultured neurons bear majority of the in vivo neuronal properties, whereas the SH-SY5Y cells contain human biochemical characteristics. These three cellular models (PC12 cells, primary cultured neurons, and SH-SY5Y cells) were utilized, and the biological characteristics of the neurotrophic compounds discovered in the screening system were classified. As a result, several neurotrophic compounds (i.e., alkaloids, terpenes, and phenols) were revealed. However, it is unknown whether the neurotrophic activities that occurred in vitro would also transpire in vivo. Thus, it would be important to determine the amount of neurotrophic compounds that would be required for systemic administration. Table 4 shows the compounds that have been investigated in our laboratory.

Table 4. Representative small-molecular weight natural compounds with neurotrophic activity identified in our group.

SourceCompoundReferences

Apiospora montagnei	TMC-95A	(39)
Aristolochia arcuata MASTERS	talaumidin, nectandrin A, isonectandrin B, nectandrin B, galgravin, aristolignin	(40)
Garcinia subelliptica	garsubellin A	(41)
Illicium anisatum	illicinin A, 4-allyl-2-methoxy-6-(2-methylbut-3-en-2-yl)phenol	(42)
Illicium fargesii	isodunnianol	(43)
Illicium jiadifengpi	jiadifenoxolane A, jiadifenolide, (2R)-hydroxy-norneomajucin	(44)
Illicium merrillianum	merrilactone A	(45)
Illicium tashiroi	isodunnianin, tricycloillicinone, bicycloillicinone asarone acetal, tricycloillicinone	(46)
Leonurus heterophyllus	leonurusoleanolide A, B, C, D	(47)
Magnolia ovobata	honokiol, magnolol	(28), (29), (48)
Mastigophora diclados	mastigophorenes A, B, D	(49)
Phellinus ribis	ribisin A, B, C, D	(50)
Phytolacca americana	americanoic acid A methyl ester	(51)
Piper retrofractum	piperodione	(5)
Plagiochila fruticosa	plagiochilal B	(27)
Ptychopetalum olaacoides	ptychonal hemiaccetal, ptychonal	(52)
Viburnum awabuki	neovibsanin, neovibsanins A, B	(53)
Viburnum sieboldii	neovibsanins A, B	(54)
Zingiber purpureum (BANGLE)	trans-3-(3′,4'-dimethoxyphenyl)-4-[(E)-3″,4″-dimethoxystyryl]cyclohex-1-ene, cis-3-(3′,4'-dimethoxyphenyl)-4-[(E)-3″,4″-dimethoxystyryl]cyclohex-1-ene	(30)

2.2.3.1. Polyphenolic compounds

Recent research is focusing on natural products, such as the water extract of the leaves of Caesalpinia crista and Centella asiatica, Ginkgo biloba extract, and extracts prepared from the medicinal herb, Paeonia suffruticosa. These products are alternatives for the treatment of AD (55). The water extract of the leaves of C. crista prevents Aβ aggregation from monomers and disintegrates preformed Aβ fibrils. C. asiatica prevents synuclein aggregation (56), G. biloba extract inhibits the formation of oligomers (55), and extracts prepared from the medicinal herb P. suffruticosa inhibit Aβ fibril formation and also destabilize the preformed amyloid fibrils (57). The anti-amyloidogenic properties observed are attributed to the polyphenolic compounds present in these extracts. The chemical structure of polyphenols includes two or more phenol rings with hydroxyl groups in ortho or para positions. The phenol rings are essential for redox reactions (58). There is a direct positive correlation between the antioxidant capacity and number of hydroxyl groups present in the polyphenols' structure. Thus, an increase in the amount of hydroxyl groups in the polyphenol chemical structure is associated with an increase in redox potential and antioxidant activity (59).

Polyphenols are powerful anti-amyloidogenic compounds owing to their physicochemical features such as the presence of aromatic rings, molecular planarity, capacity to form hydrogen bonds, presence of an internal double bond, and molecular weights below 500 g/mol. The molecular weight allows for potential inhibition of APP pathways that, in turn, reduces amyloid load (60).

Several members of the disintegrin and metalloproteinase (ADAM) family (ADAM9, 10, and 17) have been proposed to be physiologically active α-secretases. It has been demonstrated that ADAM10 has the highest α-secretase activity in vivo. Moreover, it has been suggested that the upregulation of ADAM10 could be a potential therapeutic target for the treatment of AD. ADAM10 has a potential neuroprotective role because it promotes the non-amyloidogenic pathway (61). This enzyme is activated by the removal of the prodomain, which is probably promoted by the action of proprotein convertases. Phlorotannins, curcumin, and epigallocatechin-3-gallate (EGCG) have been shown to increase the overexpression‎ of sAPPα through activation of α-secretase, favoring neuroprotection (62).

The amino acids, Asp 32 and Asp 228, and two water molecules are essential for the catalytic activity of β-secretase, and all these components are located in the catalytic binding site. One of these water molecules participates in enzymatic proteolysis, whereas the second water molecule is important for stabilizing a tetrahedral intermediate that is essential for protein cleavage (63). Computational modeling studies have clarified that Asp 32, not Asp 228, is protonated (63). Proteolytic β-secretase activity is initiated through a nucleophilic attack on the carbonyl group of the peptide bond by a water molecule (63). The inhibition of β-secretase represents a potential therapeutic target in AD treatment because it decreases the Aβ load. Isophthalamides have shown greater inhibitory effects on β-secretase (63), and myricetin is a potential β-secretase inhibitor (64). The proposed mechanism is that the displacement of a water molecule inhibits the polyphenol, and the water molecule then participates in a hydrogen-bonding network with Asp 32 and Asp 228, which is essential for β-secretase proteolytic activity (63).

Presenilin I (PS1) protein is a member of the aspartic protease family, which is implicated in the regulation of intramembrane proteolysis (9), (65). Presenilin I has been identified as the catalytic subunit of the γ-secretase complex. This protein is composed of nine transmembrane domains, and domains 6 and 7 form the catalytic site. Mutations in PS1 have been linked to familial AD (FAD). Therefore, PS1 is a potential target in the design of drugs against AD (66). The catalytic activity of PS1 requires two opposing aspartyl groups (Asp 257 and Asp 385) in the active site. One of these aspartates is deprotonated and acts as a base, which activates a water molecule present in the catalytic site. The other aspartate donates a proton to the carbonyl group of the substrate (66). Therefore, it is suggested that polyphenols that can occupy the active site of γ-secretase and displace the water molecule required by the enzyme for catalysis, may disable the enzyme and reduce Aβ formation (66).

Aggregation of Aβ42 is the central point of AD pathology. Consequently, the inhibition of Aβ42 aggregation is a key focus in drug discovery. There are multiple steps involved in the biology of Aβ42 aggregation, i.e., monomers form oligomers, protofibrils, and mature fibrils. Drug discovery efforts are focused on preventing the formation of either oligomers or fibrils. However, studies that explore the binding of drugs to Aβ42 and prevent fibril formation are restricted because more than 50% of the amino acids in this peptide are hydrophobic residues. It has been suggested that the hydrophobic core is essential for fibrillogenesis (67). The fibrillization of Aβ, which begins with the formation of Aβ oligomers that are constituted by 24 monomers, is a multistep process (68). Amyloid fibril formation depends on concentration of Aβ42 peptide, low pH, the time of incubation, and the length of the carboxyl chain (69). The Aβ structure is stabilized by hydrophobic forces, aromatic stacking, and electrostatic interactions (70). The main physicochemical properties of molecules with the potential to inhibit amyloidal fibril formation might be due to the presence of aromatic rings in their chemical structure and the ability to form noncovalent interactions with amino acid residues of the Aβ peptide sequence (60). Moreover, the planarity of the inhibitor is essential for increasing surface contact with the Aβ peptides (71). Most polyphenols have more than two aromatic rings, which are essential for π–π stacking interactions with the hydrophobic amino acid residues of Aβ (Tyr and Phe), and at least three hydroxyl groups that form hydrogen bonds with the hydrophilic amino acid residues of Aβ (His6, Ser8, Tyr10, His14, and Lys16) (72). The resonance structure of polyphenols provides enough planarity to penetrate the Aβ fibril hydrophobic groove, thus disturbing the fibril structure (73).

Polyphenolic compounds, such as resveratrol, curcumin, and myricetin, have demonstrated anti-Aβ aggregation properties (74), (75). Furthermore, flavanols possess more anti-amyloidogenic activity than isoflavonoids because they contain more hydroxyl groups that are able to form hydrogen bonds with Aβ peptides. Curcuminoids are more hydrophobic than flavonoids. This physicochemical property might enhance their affinity for binding with the hydrophobic core of the Aβ fibril, resulting in an increased anti-amyloidal activity (76). Tannins are complex polyphenols that have the highest number of hydroxyl groups of all the polyphenolic compounds; therefore, they have the strongest anti-aggregation activity (77). However, their large molecular weight reduces their suitability as a therapeutic drug (78).

Our understanding of the bioavailability and health benefits of polyphenols remains unclear. However, it has been shown that some polyphenols can cross the blood–brain barrier and contribute to neuroprotection (79).

2.2.3.2. Neurotrophic activity of honokiol and magnolol2.2.3.2.1. Neurite outgrowth promotes activities in cortical neurons

Honokiol and magnolol are bioactive compounds found in the stem bark of the M. obovata Thunb (Wakoboku) and Magnolia officinalis Rhed (Houpu) (80). Honokiol and magnolol are herbal drugs (called “Kouboku”) used in China and Japan for the treatment of digestive and nervous complaints.

Neurons were placed in cultures that contained honokiol or magnolol. The neurons that were in honokiol-containing cultures had extremely long neurites and more prominent, darker-stained neuronal soma compared with those of the neurons in the control culture. In addition, one of the processes was much longer and displayed more neurite branches. If bFGF were added to this cell culture, the neurite extension could be significantly enhanced. The neurons that were in magnolol-containing cultures also had long neurites. The length of neurites increased in the cultures that were treated with honokiol or magnolol in a dose-dependent manner and the neurites reached their maximum length at 10 μM. Magnolol had a weaker effect on neurite extension compared with that of honokiol. This difference in the trophic effects caused by honokiol and magnolol presumably results from the different positions of an aryl–aryl bond that is formed between two units of allylphenol (Fig. 1) (81). These results are consistent with previous studies that have shown that honokiol is more effective than magnolol at producing various CNS-related effects (82).

• Download : Download full-size image

Fig. 1. Structures of honokiol and magnolol.

2.2.3.2.2. Prevention of learning and memory impairment in senescence-accelerated-prone mice (SAMP)

Senescence-accelerated mice (SAM) are selected by brother–sister mating of AKR/J mice and are characterized by a rapid accumulation of senile features and a shortened lifespan (1–1.5 years compared with that of 2–2.5 years in control animals) (83). SAM are composed of nine inbred strains of senescence-accelerated-prone mice (SAMP), and three strains of senescence-accelerated-resistant mice (SAMR), the latter of which show normal aging. Neither SAMP1 nor SAMR1 mice are behaviorally or morphologically different until they are 4 months old. At that age, the SAMP1 mice begin to rapidly accumulate senile changes. These mice have decreases in orientation ability and learning capacity and the appearance of age-specific morphological changes such as lordokyphosis, hair loss, frequent corneal cataracts, and mucosal inflammation (28). Among the various SAMP substrains, SAMP8 mice exhibit early-onset impairment of learning and memory (84). Recent studies have reported that aged SAMP8 mice had cholinergic neuron deficits in the septal region of the forebrain (85). These findings suggest that the SAMP8 mouse might be a useful animal model for investigating the cholinergic hypothesis of human dementia in advanced aging and AD.

Magnolol (5 and 10 mg/kg) was orally administered once a day for 14 days to 2- or 4-month-old mice. The mice were examined at 4 and 6 months of age. In the SAMP1 mice, the density of the neurofibrils in the stratum radiatum of the CA1 region in the hippocampus decreased with age. This result was not observed in the SAMR1 mice. When 2-month-old mice were treated with magnolol, the decrease in neurofibrils in the CA1 region was significantly prevented. However, magnolol administration at 4 months of age did not result in a protective effect. These findings suggest that magnolol must be administered before neuronal loss begins in order to have a protective effect on the hippocampus (28).

Magnolol or honokiol was orally administered once a day for 14 days to 2-month-old SAMP8 mice. Passive avoidance tests and object recognition tests were used to eval‎uate learning and memory at 4 and 6 months of age. Learning and memory in the SAMP8 mice were markedly impaired at 4 and 6 months of age. However, this age-related learning and memory impairment was prevented by the pretreatment with magnolol (10 mg/kg) or honokiol (1 mg/kg). Immunohistochemical analysis of choline acetyltransferase (ChAT) was used to eval‎uate the forebrain cholinergic neuron densities in the medial septum and vertical limb of the diagonal band. Significant cholinergic deficits were observed in 6-month-old SAMP8 mice. However, this age-related cholinergic deficit was prevented by the pretreatment with magnolol (10 mg/kg) or honokiol (1 mg/kg) (29).

2.2.3.2.3. Neurotrophic mechanisms of honokiol and magnolol

Fig. 2 illustrates the signal pathways of the NTs in neuronal cells (86). Neurotransmitters, such as NGF, BDF, and NT-3, bind to the extracellular domain of the tyrosine kinase receptors, TrkA, TrkB, and TrkC, respectively, and activate tyrosine kinase in the intracellular domain. The autophosphorylation of multiple tyrosine residues within the intracellular segments produces several binding sites for intracellular signaling proteins. When the target signal proteins bind to tyrosine kinase, they are phosphorylated to adopt active forms and then activate downstream products (87). The NT-activated signal molecules {Ca2+, MAPK [Extracellular signal-regulated kinases (ERK)], and Akt} in Fig. 2 are indispensable for transferring intracellular signals to the nuclei.

• Download : Download full-size image

Fig. 2. Neurotrophic pathways of neurotrophins (NTs).

The effects of honokiol and magnolol have been explained by their effects on GABAA receptors (88). In primary cultured rat cortical neurons and human SH-SY5Y cells, honokiol and magnolol increased cytoplasmic free Ca2+ with a characteristic lag phase. The increase in cytoplasmic free Ca2+ was independent of extracellular Ca2+, but dependent on phospholipase C (PLC) and inositol 1,4,5-trisphosphate (IP3) receptor activation. These results suggest that honokiol and magnolol increase cytoplasmic free Ca2+ through a PLC-mediated pathway. Furthermore, the release of Ca2+ from intracellular stores mainly contributes to an increase in cytoplasmic free Ca2+. Thus, honokiol and magnolol may be involved in the activation mechanism associated with intracellular Ca2+ mobilization (18). ERK1/2 and Akt might be related to the honokiol-induced neurite outgrowth and intracellular Ca2+ mobilization. This is suggested because the outgrowth in the cultured rat cortical neurons was markedly reduced by PD98059, a mitogen-activated protein kinase kinase (MAPKK, MAPK/ERK kinase MEK, directly upstream from ERK1/2) inhibitor but not by LY294002, a PI3K (upstream from Akt) inhibitor. Honokiol also highly enhanced the phosphorylation of ERK1/2 in a concentration-dependent manner. The phosphorylation of ERK1/2 was enhanced by honokiol and inhibited by KN93, PD98059, Ca2+/calmodulin-dependent kinase II (CAMK II) inhibitor, and MAPKK inhibitor. Products of the phosphoinositide specific PLC-derived IP3 and 1,2-diacylglycerol (DAG) were measured after honokiol treatment (48). These results suggest that the signal transduction from PLC, IP3, Ca2+, and CAM II to ERK1/2 is involved in the honokiol-induced neurite outgrowth.

The Akt pathway is a prominent regulator of NT-mediated survival responses in neurons. Activated Akt inhibits apoptosis through multiple mechanisms and negatively regulates the phosphorylation and inactivation of the fork head transcription factors. Akt activity was decreased in the forebrain of SAMP8 mice at 2 months of age compared with that of SAMR1 mice. However, treatment with either magnolol or honokiol increased Akt phosphorylation in the forebrain of SAMP8 mice (29).

Regulation of oxidative stress may also explain the protective effects of magnolol and honokiol in preventing age-related impairment in SAMP8 mice. Magnolol and honokiol are known to have antioxidant effects (89), (90). The onset of increased lipid hydroperoxide in the brain coincides with the onset of cognitive impairment in SAMP8 mice that are 2 months old. Thus, oxidative stress may be important in the development of impaired learning and memory (91).

2.2.3.3. BANGLE: eval‎uation of hippocampal neurogenesis in olfactory bulbectomized (OBX) mice, an experimental depression and dementia animal model

In the early stages of AD, the olfactory system is significantly impaired (92); it is particularly rich in acetylcholine and other neurotransmitters and has decreased ChAT activity (93). These results indicate that degeneration of cholinergic neurons may occur in the olfactory system. Patients with AD have increased numbers of senile plaques containing Aβ and NFTs in the olfactory bulbs (OBs) and the anterior olfactory nuclei (93). Previous studies have shown that when the OBs are removed in animals (i.e., OBX mice), the result is impaired learning and memory (94) with increased exploratory behaviors (95). These results suggest that OBX mice are an animal model for the investigation of experimental depression and dementia. In mice, the intracerebroventricular application of fibroblast growth factor-2 (FGF-2) ameliorates the reduction in hippocampal neurogenesis and reverses depressive behaviors induced by OB removal (96). The olfactory bulbectomization procedure was performed according to a previously described method by Hozumi et al. (97). Two butenoid dimmers, trans-3-(3′,4′-dimethoxyphenyl)-4-[(E)-3″,4″-dimethoxystyryl]cyclohex-1-ene (compound 1) and cis-3-(3′,4′-dimethoxyphenyl)-4-[(E)-3″,4″-dimethoxystyryl] cyclohex-1-ene (compound 2) (Table 4) were isolated as neurotrophic molecules from an Indonesian medicinal plant, Z. purpureum (Bangle). These two compounds were administered once a day on days 15–28 after. Chronic treatment with compounds 1 and 2 significantly increased the number of BrdU/NeuN double-labeled cells (30). These results indicate that compounds 1 and 2 have neurotrophic effects and have the potential to modify depression and dementia.

2.2.3.4. Structure-activity relationships of natural products

Understanding the structure–activity relationship of NFs is important. However, using natural products to search for NF-like molecules with similar structures is difficult. Thus, these molecules should be synthesized. Although this type of research has not yet progressed, we have introduced it below.

The morphological effects of honokiol and its analogs on the cultured rat cortical neurons were eval‎uated by the anti-MAP2 staining method. The 4′-hydroxy group and the 5′-allyl group were deemed responsible for the honokiol-mediated neurite outgrowth-promoting activity. However, the 3′-allyl group was not considered essential for neurite growth (Fig. 1) (98).

Illicinin A (A) and its derivatives (Fig. 3) were synthesized for the structure–activity relationship studies. Sequential Stille reactions introduced a prenyl and allyl group to the benzene ring of A and its derivatives. Thus, an allylphenyl moiety in the A molecule is essential for its neurotrophic properties (42).

• Download : Download full-size image

Fig. 3. Structures of illicinin A (A), and Aa–d.

To further understand the structure–activity relationship, the neurotrophic properties of Aa–Ad were eval‎uated by measuring the longest neurites among the rat cortical neurons cultured in the presence of each analog. Among the screened compounds, Ac decreased neurite outgrowth, whereas Aa, Ab, and Ad were comparable to A in terms of neurite outgrowth. These results indicate that an allyl function in A is essential for preserving neurotrophic activity. Among the active analogs, the activities of Aa and Ab, which possess a 1,2-free diol group, slightly increased. This implies that the presence of two vicinal oxy groups is important for neurotrophic activity. However, the methylenedioxy forms of Aa and Ab did not enhance their neurotrophic activities. In addition, the presence and position of the prenyl group in A did not affect its neurotrophic activity (42).

3. Conclusions

Several small molecules, which can mimic the functions of NFs, have been discovered in natural products. However, the neurotrophic activity has only been eval‎uated in vivo for a few compounds. Magnolol, honokiol, and BANGLE extracts can supply the amount of neurotrophic compounds that is required for adequate systemic administration. Promising compounds with neurotrophic activity may eventually be used for systemic administration. The rapid discovery and production of a synthetic method for producing an effective compound in large quantities would be highly desirable for the treatment of AD.

Conflicts of interest

The authors declare no conflict of interest.

References(1)

R.M. Friedlander