Structure-Activity Relationships for Bergenin Analogues As Β-Secretase

Structure-activity Relationships for Bergenin Analogues as β-Secretase (BACE1) Inhibitors Yusei Kashima and Mitsuo Miyazawa＊ Department of Applied Chemistry, Faculty of Science and Engineering, Kinki University (3-4-1, Kowakae, Higashiosaka-shi, Osaka 577-8502, JAPAN)

Abstract: Here we evaluated the inhibitory effects of bergenin analogues (2-10), prepared from naturally occurring bergenin, (1) on b-secretase (BACE1) activity. All the bergenin analogues that were analyzed inhibited BACE1 in a dose-dependent manner. 11-O-protocatechuoylbergenin (5) was the most potent

inhibitor, with an IC50 value of 0.6 ± 0.07 mM. The other bergenin analogues, in particular, 11-O-3′,4′- dimethoxybenzoyl)-bergenin (6), 11-O-vanilloylbergenin (7), and 11-O-isovanilloylbergenin (8), inhibited

BACE1 activity with IC50 values of <10.0 mM. BACE1 inhibitory activity was influenced by the substituents of the benzoic acid moiety. To the best of our knowledge, this is the first report on the structure-activity relationships (SAR) in the BACE1 inhibitory activities of bergenin analogues. These bergenin analogues may be useful in studying the mechanisms of Alzheimer’s disease.

Key words: bergenin, bergenin analogues, b-secretase, antioxidant activity, structure-activity relationships

1 INTRODUCTION in adult mice was without any significant effect on brain Alzheimer’s diseas（e AD）is a neurodegenerative disorder, neuregulin processing9）, indicating that BACE1 inhibitors with symptoms such as memory loss and disruption in could be established as therapeutic targets for AD. judging, reasoning, and emotional stability. AD is pathologi- Oxidative stress also a cause of AD has been proposed to cally characterized by the accumulation of senile plaques, contribute to Aβ generation and the formation of NFT10）. neurofibrillary tangle（s NFT）, synaptic loss, and neuronal The β-amyloid peptide（Aβ）generates free radicals in a death. Much of AD research has been focused on the metal-catalyzed reaction, inducing neuronal cell death by amyloid cascade hypothesis, which states that β-amyloid reactive oxygen specie（s ROS）, which peroxidize mem- peptide（Aβ）, a proteolytic derivative of the large trans- brane lipids and oxidize proteins, producing drastic cellular membrane protein amyloid precursor protein（APP）, plays damages. Upregulation of lipid peroxidation leads to amy- an early role in the pathogenesis of AD1）. Aβ peptides, the loidogenesis through increased expression and activity of major constituent of senile plaques found in the brain of BACE111, 12）. Expression of BACE1 increases in conditions patients with AD, are generated from the cleavage of APP of oxidative stress caused by the lipid peroxidation product by β-secretas（e BACE1: β-site APP cleaving enzyme-1）and 4-hydroxynonena（l HNE）and hydrogen peroxide, indicat- γ-secretase. β-Secretase cleavage results in the production ing a correlation between BACE1 activity and oxidative of soluble APPβ and a membrane-associated C-terminal stress marker in AD brain13, 14）. In addition, microglia acti- fragment, C99 2, 3）. After the β-secretase cleavage, vated by oxidative damage release proinflammatory and γ-secretase cleaves C99 within its transmembrane domain free radicals, leading to various inflammatory reactions that to generate Aβ peptides. Thus, these secretases have been cause damage to neurons. Enhanced inflammation induces recognized as the key enzymes that commit APP catabo- the generation of ROS in ambient neurons, which contrib- lism to the amyloidogenic pathway4－6）. However, inhibition ute to Aβ generation15）. It is generally accepted that oxida- of γ-secretase may elicit unwanted side effects as it is in- tive damage in cellular structures precedes the phenome- volved in Notch processing7, 8）. Therefore, β-secretase is non of other pathological hallmarks of AD, and therefore, considered a better target than γ-secretase for the develop- oxidative stress as well as Aβ is an early causative event in ment of anti-AD agents with less severe side effects. More- the pathogenesis and progression of AD. On the basis of over, a recent report showed that inhibition of β-secretase these data, it can be concluded that drugs specifically scav-

＊Correspondence to: Mitsuo Miyazawa, Department of Applied Chemistry, Faculty of Science and Engineering, Kinki University, 3-4-1, Kowakae, Higashiosaka-shi, Osaka 577-8502, JAPAN E-mail: [email protected] Accepted February 5, 2013 (received for review December 1, 2012) Journal of Oleo Science ISSN 1345-8957 print / ISSN 1347-3352 online http://www.jstage.jst.go.jp/browse/jos/ http://mc.manusriptcentral.com/jjocs

391 Y. Kashima and M. Miyazawa

enging oxygen radicals could be useful for either the pre- tentially active. Furthermore, its esterified analogues occur vention and treatment of AD16）. widely in several plants, especially those used in traditional To have therapeutic potential, inhibitors should be able medicines36－39）, and are commonly found in plant extracts40－43）. to penetrate the blood-brain barrier and thus have molecu- In the present study, bergenin analogue（s 2-10）were pre- lar weights below 700 Da16）. Therefore, large peptide-based pared from naturally occurring bergenin（1）（Fig. 1）and in- inhibitors are not proper drug candidates, but the second- vestigated for their inhibitory effects against BACE1 activi- ary metabolites of plants, which have relatively low molecu- ty. To the best of our knowledge, this is the first report on lar weights and high lipophilicity, may be potential drugs the correlation between BACE1 inhibition and the struc- against AD. Polyphenols are a group of phytochemicals tures of bergenin and its analogues. that exhibit a wide range of physiological and therapeutic properties. Phenolic compounds could be a major determi- nant of the potential of food as a source of natural antioxi- dants17）. Phenolic compounds might be good candidates for 2 EXPERIMENTAL PROCEDURES BACE1 inhibitors; however, natural product inhibitors have 2.1 Materials rarely been reported18－23）. Here, we focused on the analysis Thin layer chromatograph（y TLC）was performed on pre- of these antioxidant substances. coated plate（s silica gel 60 F254, 0.25 mm, Merck, Darmstadt, Bergenin, a unique phenolic compound, has been isolat- Germany）. Column chromatography was carried out using ed from the Bergenia species, roots of Caesalpinia 70-230 mesh silica ge（l Kieselgel 60, Merck, Germany）. digyna, bark of Corylopsis spicata, and bark of Mallotus Melting point（s m.p）were measured on an MP-5000D philippinensis24－27）. It exhibits various biological activi- melting point apparatus and were uncorrected. Optical ro- ties, such as antioxidant27, 28）, anti-inflammatory29）, anti-ar- tations were measured on a LTDDIP-1000 polarimeter thritis30）, hypolipidemic31）, anti-HIV32）, antiarrhythmic33）, （Japan Spectroscopic Co.）. 1H and 13C NMR data were all hepatoprotective34）, and antinociceptive effects35）. Bergenin obtained with a JEOL ECA-400（400 MHz）spectrometer in contains 5 hydroxyl groups, which are considered to be po- DMSO-d6 or CDCl3 with tetramethylsilan（e TMS）as internal

Fig. 1 Structure of bergenin and its analogues (1-10).

392 J. Oleo Sci. 62, (6) 391-401 (2013) BERGENIN ANALOGUES AS b-SECRETASE INHIBITOR

standard（chemical shift in δ, ppm）. J values were reported in hertz. Infrared（IR）spectra were obtained with a Jasco FT/IR-470 plus Fourier transform infrared spectrometer. Electron ionization mass spectrometry（EI-MS）spectra were obtained on a JEOL JMS-700 Tandem MS station （Japan Electron Optics Laboratory Co., Ltd.）. The absorbance or ﬂuorescence was measured with an MTP-800Lab microplate reader. A BACE1（recombinant human BACE1） assay kit was purchased from PanVera Co（. USA）. 1,1-Di- phenyl-2-picrylhydrazy（l DPPH）, dibutylhydroxytoluene （BHT）, benzoic acid, p-hydroxybenzoic acid, p-anisic acid, protocatechuic acid, 3,4-dimethoxy benzoic acid, vanillic acid, isovanillic acid, 3,5-dimethoxy benzoic acid, syringic acid, diisopropyl azodicarboxylate（DIAD）, tetraki（s triphe-

nyl phosphine）palladium［Pd（PPh3）4］, triphenylphosphine

（Ph3P）, sodium iodid（e NaI）, morpholine, and allyl bromide were purchased from Tokyo Kasei Kogyo. All solvents were purchased from Kanto Chemical.

2.2 Isolation of bergenin The bark（5.0 kg）of Bergenia ligulata was extracted once with MeOH at room temperature. The solution was evaporated until it dried in vacuo to obtain a MeOH Fig. 2 Isolation scheme of bergenin (1) from Bergenia extrac（t 480 g）. The residue was re-extracted successively ligurata. with n-hexane, CH2Cl2, EtOAc, n-BuOH, and water. Each fraction was concentrated to dryness in vacuo to give n-

hexane extrac（t 12.5 g）, CH2Cl2 extrac（t 40.5 g）, EtOAc 10.8, 1.8 Hz, 2×＝CHa）, 5.40（2H, dd, J＝17.2 Hz, 1.8, 2× extrac（t 88.2 g）, n-BuOH extrac（t 98.1 g）, and water ＝CHb）, 5.97-6.12（2H, m, 2×－CH＝）, 7.34（1H, s, H-7）. fraction（240.7 g）. The n-BuOH extract was fractionated to fractions 1-3 by silica gel column chromatography with 2.4 Synthesis of bergenin derivatives（2-4, 6-10）

CH2Cl2-MeOH（9:1, 8:2, 7:3, v/v）as eluents. Fraction 2 was First, various benzoic acids and Ph3P（2 equiv.）were dis-

recrystallized from CH2Cl2-MeOH（12:1, v/v）, and bergenin solved in a solution of 1a in anhydrous tetrahydrofuran （1）（3.7 g）was isolated（Fig. 2）. The structure of 1 was （THF）（2 ml/mmol）. DIAD（1.7 equiv.）was added dropwise identified by the comparison of its physical and spectral at 0℃. The mixture was stirred under nitrogen for 2 h and data with those described in the literature44）. then concentrated in vacuo. The residue was chromatographed on silica gel to afford benzoic acid ester of 1a 2.3 Synthesis of 8,10-diallyloxy-bergenin（1a）（yield 79-92％）as a white powder.

The phenolic hydroxy groups in bergenin（1）（700 mg, Next, the allyl protected esters and Pd（PPh3）（4 1 mol％, 2.13 mmol）were selectively allylated. Allyl bromide and freshly prepared）were dissolved in degassed anhydrous

NaI were added to a solution of bergenin and K2CO3 in an- THF（2 ml/mmol）, and morpholin（e 10 equiv. per allyl group hydrous dimethylformamide（DMF）（5 mL）. After the to be cleaved）was added dropwise. The mixture was mixture was stirred under nitrogen at 55℃ for 2 h, it was stirred at room temperature and concentrated in vacuo. concentrated in vacuo. The residue was treated with The residue was dissolved in EtOAc. The organic layer was wate（r 15 mL）and extracted with ethyl acetat（e 3×10 mL）. washed several times with small amounts of 1 N HCl, dried, The organic extract was washed successively with brin（e 3 and concentrated. The crude material was puriﬁed by silica

×15 mL）. The extract was dried over Na2SO4 and concen- gel column chromatography to obtain the bergenin deriva- trated in vacuo. The residue was chromatographed on tive（s 2-4, 6-10）. silica gel to afford 1a（554 mg, 64％）as a white powder. 11-O-benzoylbergenin（2） 1 25 8,10-diallyloxy-bergenin（1a）White powder; yield: 84％; H Amorphous powder; yield 45％;［ α］D ＋66.8（˚ c＝0.15, －1 NMR（CDCl3, 400 MHz）δ 3.54-3.59（2H, m, H-2, H-3）, 3.68- MeOH）; mp: 248.8-249.1℃; IR（KBr）νmax cm : 3474（OH）, 3.73（1H, m, H-11a）, 3.83-4.01（1H, m, H-4）, 3.91（3H, s, 1722（ester）, 1610（arom. C＝C）; EI-MS, m/z 432［M］＋; 1H

OMe）, 3.98（1H, d, J＝9.2 Hz, H-11b）, 4.41-4.55（4H, m, 2× NMR（DMSO-d6, 400 MHz）δ 3.41（1H, dd, J＝9.2, 8.8 Hz,

O-CH2）, 4.72（1H, d, J＝10.4 Hz, H-10b）, 5.26（2H, dd, J＝ H-3）, 3.69（1H, m, H-4）, 3.74（3H, s, 9-OMe）, 3.93（1H, ddd,

393 J. Oleo Sci. 62, (6) 391-401 (2013) Y. Kashima and M. Miyazawa

J＝9.2, 7.2, 2.0 Hz, H-2）, 4.03（1H, t, J＝10.0 Hz, H-4a）, Hz, H-5’）, 7.48（1H, d, J＝1.6 Hz, H-2’）, 7.63（1H, dd, J＝ 13 4.34（1H, dd, J＝12.4, 7.2 Hz, H-11a）, 4.78（1H, dd, J＝12.4, 8.4, 1.6 Hz, H-6’）; C NMR（DMSO-d6, 100 MHz）δ 55.5（3’ 2.0 Hz, H-11b）, 5.03（1H, d, J＝10.8 Hz, H-10b）, 7.00（1H, s, -OMe）, 55.7（4’- OMe）, 59.7（9-OMe）, 64.0（C-11）, 70.5 H-7）, 7.54（2H, m, H-3’）, 7.67（1H, m, H-4’）, 8.01（2H, m, （C-3）, 72.1（C-10b）, 73.5（C-4）, 78.5（C-2）, 79.5（C-4a）, 13 H-2’, 6’）; C NMR（DMSO-d6, 100 MHz）δ 59.7（9-OMe）, 109.6（C-7）, 111.2（C-2’）, 111.7（C-5’）, 115.7（C-10a）, 118.1 64.0（C-11）, 70.2（C-3）, 72.2（C-10b）, 73.5（C-4）, 78.5（C-2）, （C-6a）, 121.6（C-6’）, 123.4（C-1’）, 140.6（C-9）, 148.0 79.6（C-4a）, 109.5（C-7）, 115.8（C-10a）, 118.1（C-6a）, 128.8 （C-10）, 148.4（C-3’）, 151.0（C-8）, 153.1（C-4’）, 163.3（C-6）, （C-3’, 5’）, 129.2（C-2’, 6’）, 129.5（C-1’）, 133.5（C-4’）, 140.6 165.3（C-7’）. （C-9）, 148.0（C-10）, 150.9（C-8）, 163.4（C-6）, 165.6（C-7’）. 11-O-vanilloylbergenin（7） 25 11-O-p-hydroxybenzoylbergenin（3） Amorphous powder; yield: 40％;［ α］D ＋56.4（˚ c＝0.1, 25 －1 Amorphous powder; yield 48％;［ α］D ＋18.3（˚ c＝0.15, MeOH）; mp: 268.5-270.5℃; IR（KBr）νmax cm : 3365（OH）, －1 MeOH）; mp: 214.6-215.5℃; IR（KBr）νmax cm : 3380（OH）, 1717（ester）, 1599（arom. C＝C）, 1222（C-O）; EI-MS: m/z ＋ 1 1712（ester）, 1608（arom. C＝C）, 1238（C-O）; EI-MS: m/z 478［M］; H NMR（DMSO-d6, 400 MHz）δ 3.39（1H, m, H-3）, ＋ 1 448［M］; H NMR（DMSO-d6, 400 MHz）δ 3.40（1H, m, H-3）, 3.69（1H, dd, J＝9.6, 8.8 Hz, H-4）, 3.73（3H, s, 9-OCH3）,

3.69（1H, t, J＝9.2 Hz, H-4）, 3.73（3H, s, 9-OCH3）, 3.87（1H, 3.85（3H, s, 3’-OCH3）, 3.89（1H, ddd, J＝8.0, 7.6, 2.0 Hz, ddd, J＝9.2, 6.4, 1.6 Hz, H-2）, 4.03（1H, t, J＝9.8 Hz, H-4a）, H-2）, 4.04（1H, dd, J＝10.4, 9.6 Hz, H-4a）, 4.19（1H, dd, J 4.26（1H, dd, J＝12.0, 6.4 Hz, H-11a）, 4.71（1H, dd, J＝12.0, ＝11.6, 8.0 Hz, H-11a）, 4.81（1H, dd, J＝11.6, 2.0 Hz, 1.6 Hz, H-11b）, 5.02（1H, d, J＝10.0 Hz, H-10b）, 6.86（2H, d, H-11b）, 5.05（1H, d, J＝10.8 Hz, H-10b）, 6.88（1H, d, J＝8.4 J＝8.8 Hz, H-3’, 5’）, 7.00（1H, s, H-7）, 7.85（2H, d, J＝8.8 Hz, H-5’）, 7.00（1H, s, H-7）, 7.48（1H, d, J＝2.0 Hz, H-2’）, 13 13 Hz, H-2’, 6’）; C NMR（DMSO-d6, 100 MHz）δ 59.7（9-OMe）, 7.51（1H, dd, J＝8.4, 2.0 Hz, H-6’）; C NMR（DMSO-d6, 100 63.5（C-11）, 70.2（C-3）, 72.2（C-10b）, 73.5（C-4）, 78.6（C-2）, MHz）δ 55.6（3’-OMe）, 59.7（9-OMe）, 63.8（C-11）, 70.5 79.6（C-4a）, 109.5（C-7）, 115.4（C-3’, 5’）, 115.8（C-10a）, （C-3）, 72.1（C-10b）, 73.5（C-4）, 78.5（C-2）, 79.6（C-4a）, 118.1（C-6a）, 120.0（C-1’）, 131.6（C-2’, 6’）, 140.6（C-9）, 109.6（C-7）, 112.6（C-2’）, 115.2（C-5’）, 115.8（C-10a）, 118.1 148.0（C-10）, 150.9（C-8）, 162.2（C-4’）, 163.4（C-6）, 165.5 （C-6a）, 120.2（C-1’）, 123.7（C-6’）, 140.6（C-9）, 147.4（C-3’）, （C-7’）. 148.0（C-10）, 151.0（C-8）, 151.8（C-4’）, 163.3（C-6）, 165.4 （C-7’）. 11-O-p-methoxybenzoylbergenin（4） 25 Amorphous powder; yield: 44％;［ α］D ＋44.4（˚ c＝0.15, 11-O-isovanilloylbergenin（8）－1 25 MeOH）; mp: 258.9-260.4℃; IR（KBr）νmax cm : 3373（OH）, Amorphous powder; yield: 36％;［ α］D ＋42.4（˚ c＝0.1, －1 2959（C-H）, 1720（ester）, 1609（arom. C＝C）, 1237（C-O）; MeOH）; mp: 226.6-227.1℃; IR（KBr）νmax cm : 3376（OH）, ＋ 1 EI-MS: m/z 462［M］; H NMR（DMSO-d6, 400 MHz）δ 3.40 1714（ester）, 1608（arom. C＝C）, 1237（C-O）; EI-MS: m/z ＋ 1 （1H, t, J＝8.8 Hz, H-3）, 3.69（1H, t, J＝8.8 Hz, H-4）, 3.74 478［M］; H NMR（DMSO-d6, 400MHz）δ 3.39（1H, m, H-3）,

（3H, s, 9-OCH3）, 3.83（3H, s, 4’-OCH3）, 3.89（1H, m, H-2）, 3.69（1H, dd, J＝9.2, 8.8 Hz, H-4）, 3.73（3H, s, 9-OCH3）,

4.04（1H, t, J＝10.0 Hz, H-4a）, 4.29（1H, dd, J＝12.0, 6.8 3.82（3H, s, 4’-OCH3）, 3.89（1H, ddd, J＝7.6, 6.4, 2.0 Hz, Hz, H-11a）, 4.74（1H, dd, J＝12.0, 2.1 Hz, H-11b）, 5.03（1H, H-2）, 4.02（1H, dd, J＝10.4, 9.2 Hz, H-4a）, 4.28（1H, dd, J d, J＝10.0 Hz, H-10b）, 7.00（1H, s, H-7）, 7.05（2H, d, J＝8.8 ＝12.0, 6.4 Hz, H-11a）, 4.70（1H, dd, J＝12.0, 2.0 Hz, Hz, H-3’, 5’）, 7.95（2H, d, J＝8.8 Hz, H-2’, 6’）; 13C NMR H-11b）, 5.03（1H, d, J＝10.4 Hz, H-10b）, 7.00（1H, s, H-7）,

（DMSO-d6, 100 MHz）δ 59.7（9-OMe）, 63.7（C-11）, 70.2 7.02（1H, d, J＝8.4 Hz, H-5’）, 7.40（1H, d, J＝2.0 Hz, H-2’）, 13 （C-3）, 72.2（C-10b）, 73.5（C-4）, 78.5（C-2）, 79.6（C-4a）, 7.47（1H, dd, J＝8.4, 2.0 Hz, H-6’）; C NMR（DMSO-d6, 100 109.5（C-7）, 114.1（C-3’, 5’）, 115.8（C-10a）, 118.1（C-6a）, MHz）δ 55.7（4’-OMe）, 59.7（9-OMe）, 63.8（C-11）, 70.1 121.7（C-1’）, 131.4（C-2’, 6’）, 140.6（C-9）, 148.0（C-10）, （C-3）, 72.1（C-10b）, 73.5（C-4）, 78.6（C-2）, 79.6（C-4a）, 150.9（C-8）, 163.3（C-4’）, 163.4（C-6）, 165.4（C-7’）. 109.5（C-7）, 111.5（C-5’）, 114.4（C-2’）, 115.8（C-10a）, 118.1 （C-6a）, 121.6（C-1’）, 122.0（C-6’）, 140.6（C-9）, 146.3（C-3’）, 11-O（- 3’, 4’-dimethoxybenzoyl）-bergenin（6） 148.0（C-10）, 151.0（C-8）, 152.1（C-4’）, 163.3（C-6）, 165.5 25 Amorphous powder; yield: 48％;［ α］D ＋32.6（˚ c＝0.025, （C-7’）. －1 MeOH）; mp: 260.6-262.7℃; IR（KBr）νmax cm : 3395（OH）, 1716（ester）, 1598（arom. C＝C）, 1225（C-O）; EI-MS: m/z 11-O（- 3’, 5’-dimethoxybenzoyl）-bergenin（9）＋ 1 25 492［M］; H NMR（DMSO-d6, 400 MHz）δ 3.38（1H, m, H-3）, Amorphous powder; yield: 46％;［ α］D ＋157.9（˚ c＝0.025, －1 3.70（1H, t, J＝9.6 Hz, H-4）, 3.73（3H, s, 9-OCH3）, 3.83（3H, MeOH）; mp: 185.9-194.2℃; IR（KBr）νmax cm : 3355（OH）,

s, 3’-OCH3）, 3.84（3H, s, 4’-OCH3）, 3.91（1H, m, H-2）, 4.04 1717（ester）, 1602（arom. C＝C）, 1237（C-O）; EI-MS: m/z ＋ 1 （1H, dd, J＝10.0, 9.6 Hz, H-4a）, 4.21（1H, dd, J＝12.0, 7.2 492［M］; H NMR（DMSO-d6, 400 MHz）δ 3.38（1H, m, H-3）,

Hz, H-11a）, 4.83（1H, brd, J＝12.0 Hz, H-11b）, 5.05（1H, d, 3.69（1H, m, H-4）, 3.73（3H, s, 9-OCH3）, 3.81（6H, s, 3’, 5’-

J＝10.0 Hz, H-10b）, 7.00（1H, s, H-7）, 7.09（1H, d, J＝8.4 OCH3）, 3.91（1H, ddd, J＝9.2, 7.6, 1.6 Hz, H-2）, 4.04（1H,

394 J. Oleo Sci. 62, (6) 391-401 (2013) BERGENIN ANALOGUES AS b-SECRETASE INHIBITOR

dd, J＝10.4, 9.6 Hz, H-4a）, 4.23（1H, dd, J＝12.0, 7.6 Hz, formed essentially according to the modified method of H-11a）, 4.90（1H, dd, J＝12.0, 1.6 Hz, H-11b）, 5.05（1H, d, Sumino et al（. 2002）46）. Two hundred microliters of 400 μM J＝10.4 Hz, H-10b）, 6.79（1H, t, J＝2.4 Hz, H-4’）, 7.00（1H, DPPH methanolic solution was mixed with 100 mM Tris- s, H-7）, 7.19（2H, d, J＝2.4 Hz, H-2’, 6’）; 13C NMR（DMSO- HCl buffe（r pH 7.4, 50 μL）, distilled wate（r 50 μL）, and test

d6, 100 MHz）δ 55.5（3’, 5’- OMe）, 59.7（9-OMe）, 64.1 sample（50 μL）at different concentrations. The mixture （C-11）, 70.5（C-3）, 72.1（C-10b）, 73.5（C-4）, 78.3（C-2）, was mixed well and allowed to stand for 20 min in the dark. 79.5（C-4a）, 105.5（C-4’）, 106.9（C-2’, 6’）, 109.6（C-7）, 115.7 The absorbance at 590 nm was measured using a micro- （C-10a）, 118.1（C-6a）, 131.5（C-1’）, 140.6（C-9）, 148.0 plate reader. BHT was used as a positive control. The per- （C-10）, 151.0（C-8）, 160.5（C-3’, 5’）, 163.3（C-6）, 165.2 centage of radical scavenging activity was calculated using （C-7’）. the following equation:

11-O-syringylbergenin（10） Radical scavenging activit（y ％）＝（absorbance of control 25 Amorphous powder; yield: 40％;［ α］D ＋24.5（˚ c＝0.1, －absorbance of sample/absorbance of control）×100 －1 MeOH）; mp: 240.7-241.3℃; IR（KBr）νmax cm : 3371（OH）,

1718（ester）, 1610（arom. C＝C）, 1235（C-O）; EI-MS: m/z EC50 was calculated by the application of the Reed and ＋ 1 47） 508［M］; H NMR（DMSO-d6, 400 MHz）δ 3.38（1H, m, H-3）, Muench method , as follows:

3.69（1H, m, H-4）, 3.72（3H, s, 9-OCH3）, 3.85（6H, s, 3’, 5’- EC50＝antilo｛g A＋［（B/C）×D］｝ OCH3）, 3.92（1H, ddd, J＝10.0, 8.1, 2.4 Hz, H-2）, 4.06（1H, t, J＝10.0 Hz, H-4a）, 4.12（1H, dd, J＝11.6, 8.8 Hz, H-11a）, where A is log concentration below 50％ scavenging, B is 4.89（1H, dd, J＝11.6, 2.4 Hz, H-11b）, 5.07（1H, d, J＝10.8 50－scavenging below 50％, C is scavenging above 50％－ Hz, H-10b）, 7.00（1H, s, H-7）, 7.26（2H, s, H-2’, 6’）; 13C NMR scavenging below 50％, and D is log concentration above

（DMSO-d6, 100 MHz）δ 56.1（3’, 5’-OMe）, 59.7（9-OMe）, 50％－log concentration below 50％. Data were obtained 64.0（C-11）, 70.8（C-3）, 72.1（C-10b）, 73.5（C-4）, 78.5（C-2）, from Figs. 3 and 4. All the tests were performed in tripli- 79.5（C-4a）, 107.0（C-2’, 6’）, 109.6（C-7）, 115.7（C-10a）, cate. 118.1（C-6a）, 119.0（C-1’）, 140.6（C-9）, 141.1（C-4’）, 147.6 （C-3’）, 147.7（C-5’）, 148.0（C-10）, 151.1（C-8）, 163.2（C-6）, 2.7 β-Secretase inhibitory activity 165.3（C-7’）. BACE1（β-secretase）assay was performed according to the supplied manual with modifications18）. Briefly, a 2.5 Synthesis of 11-O-protocatechuoylbergenin（5） mixture of 10 μL of assay buffe（r 50mM sodium acetate, pH The phenolic hydroxy groups in protocatechualdehyde 4.5）, 10 μL of BACE1（1.0 U/ml）, 10 μL of the substrate were selectively allylated. Then, 3,4-diallyloxy-protocate- （750 nM Rh-EVNL-DAEFK-Quencher in 50 mM ammonium chuic acid was generated according to the method from a previous study45）. The same procedure as for 2-4, 6-9 was used by starting from 1a.

11-O-protocatechuoylbergenin（5） 25 Amorphous powder; yield: 26％;［ α］D ＋25.2（˚ c＝0.25, －1 MeOH）; mp: 167.2-173.5℃; IR（KBr）νmax cm : 3364（OH）, 1709（ester）, 1610（arom. C＝C）, 1227（C-O）; EI-MS, m/z ＋ 1 464［M］; H NMR（DMSO-d6, 400 MHz）δ 3.39（1H, t, J＝9.2

Hz, H-3）, 3.69（1H, t, J＝9.2 Hz, H-4）, 3.74（3H, s, 9-OCH3）, 3.86（1H, m, H-2）, 4.01（1H, t, J＝10.0 Hz, H-4a）, 4.26（1H, dd, J＝11.6, 6.0 Hz, H-11a）, 4.67（1H, brd, J＝11.6 Hz, H-11b）, 5.05（1H, d, J＝10.4 Hz, H-10b）, 6.81（1H, d, J＝8.4 Hz, H-5’）, 7.00（1H, s, H-7）, 7.35（1H, d, J＝8.4 Hz, H-6’）, 13 7.38（1H, d, J＝1.6 Hz, H-2’）; C NMR（DMSO-d6, 100 MHz） δ 59.8（9-OMe）, 63.4（C-11）, 70.1（C-3）, 72.2（C-10b）, 73.5 （C-4）, 78.6（C-2）, 79.6（C-4a）, 109.5（C-7）, 115.3（C-5’）, 115.8（C-10a）, 116.4（C-2’）, 118.（1 C-6a）, 120.（2 C-1’）, 122.0 （C-6’）, 140.6（C-9）, 145.1（C-3’）, 148.0（C-10）, 150.7（C-4’）, 150.9（C-8）, 163.4（C-6）, 165.6（C-7’）.

2.6 DPPH radical scavenging activity Fig. 3 DPPH radical scavenging activities of bergenin DPPH（1,1-Diphenyl-2-picrylhydrazyl）assay was per- and its analogues (1-4, 6-10).

395 J. Oleo Sci. 62, (6) 391-401 (2013) Y. Kashima and M. Miyazawa

effect of samples, the sample solution was added to the reaction mixture C, and any reduction in ﬂuorescence by the

sample was then investigated. IC50 was calculated by the application of the Reed and Muench method47）, as follows:

IC50＝antilo｛g A＋［（B/C）×D］｝ where A is the log concentration below 50％ inhibition, B is 50－inhibition below 50％, C is inhibition above 50％－inhibition below 50％, and D is log concentration above 50％－log concentration below 50％. Data were obtained from Fig. 5. All data are the means of 3 experiments.

3 RESULTS AND DISCUSSION 3.1 DPPH radical scavenging activity of bergenin and its analogues（1-10） The antioxidant activities of compounds 1-10 were eval- Fig. 4 DPPH radical scavenging activities of 11-O- uated by their EC5（0 50％ Effective Concentration）in DPPH protocatechuoylbergenin (5) and BHT. radical scavenging activity as shown in Figs. 3 and 4. Com- pounds 2, 3, and 5-10 were found to be stronger than 1, which is a known antioxidant. Among them, 11-O-protocatechuoylbergenin（5）, which contains a catechol moiety,

showed the highest antioxidant activity（EC50: 5.0 μM）. Therefore, it is a better DPPH radical scavenger than BHT

（EC50: 21.7 μM）. By contrast, 4 presented the lowest activity in the assay. Compound 2 with no substituents of benzoic acid moiety showed greater activity than 1. When a functional group was placed at the 4’ position, compounds 3, 5, 7, and 10, which bear an OH group, showed higher activities than 4, 6, and 8, which have an OMe group at the 4’ position. Furthermore, when an OH group was located at the 4’ position, and the meta-hydrogen was substituted by an OH or OMe group（5, 7, and 10）, the antioxidant activities were enhanced. In particular, the radical scavenging activity of 5 markedly increased. Compounds 4, 6, and 8, which have an OMe group at the 4’ position, showed low activities; among them, only 11-O-p-methoxy- Fig. 5 β-Secretase inhibitory effects of bergenin ana- benzoylbergenin（4）showed no effective in radical scaveng- logues (2-10). ing activity, indicating that an additional functional group at the 3’ position increases the antioxidant activity. bicarbonate）, and 10 μL of sample dissolved in 30％ DMSO However, the antioxidant activity of 9 was lower than that was incubated for 60 min at room temperature in the dark. observed for compound 10. Thus, these results indicate The mixture was irradiated at 550 nm, and the emission in- that an OH group at the 4’ position and a functional group tensity at 590 nm was recorded. The inhibition ratio was at the 3’ or 5’ position are important factors for DPPH obtained using the following equation: radical scavenging activity. The results are summarized in Table 1. Inhibition（％）＝［1－（S－S0）（/ C－C0）］×100 where C was the fluorescence of the contro（l enzyme, 3.2 Inhibition of BACE1 activity by bergenin and its ana- buffer and substrate）after 60 min of incubation, C0 was the logues（1-10）. fluorescence of the control at zero time, S was the fluores- The BACE1 inhibitory activities of bergenin and its ana- cence of the tested sample（s enzyme, sample solution, and logue（s 1-10）were investigated. All bergenin analogues in- substrate）after incubation, and S0 was the fluorescence of hibited BACE1 in a dose-dependent manne（r Fig. 5）. From the tested samples at zero time. To allow for the quenching the regression curves of concentration versus the percent-

396 J. Oleo Sci. 62, (6) 391-401 (2013) BERGENIN ANALOGUES AS b-SECRETASE INHIBITOR

Table 1 Effects of bergenin and its analogues against DPPH radical.

a a Compound EC50 (μM) Compound EC50 (μM) bergenin (1) 865.5±2.20 11-O-(3',4'-dimethoxybenzoyl)-bergenin (6) 740.5±1.47 11-O-benzoylbergenin (2) 246.8±2.05 11-O-vanilloylbergenin (7) 547.5±1.89 11-O-p-hydroxybenzoylbergenin (3) 715.3±1.35 11-O-isovanilloylbergenin (8) 773.0±3.15 11-O-p-methoxybenzoylbergenin (4) >1000 (42.1%) b 11-O-(3',5'-dimethoxybenzoyl)-bergenin (9) 840.8±1.28 11-O-protocatechuoylbergenin (5) 5.0±0.96 11-O-syringylbergenin (10) 430.9±2.57 BHTc 21.8±1.55 a All compounds were examied in a set of experiments repeated three times. The results are the mean±SD of three experiments. b Effect (%) at 1000 μM. c BHT was used as the positive control.

Table 2 BACE1 inhibitory activities of bergenin and its analogues.

a a Compound IC50 value (μM) Compound IC50 value (μM) Bergenin analogues 1 > 400 (22.1%)b 6 6.9±0.9 2 66.1±2.9 7 8.8±0.6 3 23.8±0.6 8 5.0±0.5 4 26.8±1.3 9 20.8±1.9 5 0.6±0.07 10 21.0±1.5 Benzoic acids protocatechuic acid > 400 (8.9%)b isovanillic acid > 400 (-1.1%)b 3,4-dimethoxybenzoic acid > 400 (8.0%)b vanillic acid > 400 (1.9%)b Referencec 0.2±0.01 a All compounds were examined in a set of experiments repeated three times. The results are the mean±SD of three experiments. b Inhibition (%) at 400 μM. c Lys-Thr-Glu-Glu-lle-Ser-Glu-Val-Asn-(statine)-Val-Ala-Glu-Phe-OH was used as the positive control.

age of control activity, the IC50 value（s 50％ inhibitory con- parison of 3 and 4, which have IC50 values of 23.8 and 26.8 centrations）of bergenin（1）and each bergenin analogues μM, respectively, indicated that the introduction of an OH （2-10）were estimated and are presented in Table 2. Struc- group was slightly beneficial for the activity. This effect tural change was markedly involved in these activities. As was also observed between 6 and 8. The inhibitory potency shown in Table 2, all bergenin analogue（s 2-10）exhibited a improved when functional groups were substituted for hy-

signiﬁcant degree of β-secretase inhibition（IC50 values: 0.6- drogen atoms at C-3’ and C-4’. Compounds 6, 7, and 8, 66.1 μM）in in vitro BACE1 activity assays. In particular, which have functional groups only at C-3’ and C-4’, showed 11-O-protocatechuoylbergenin（5）showed excellent inhibi- stronger inhibition of BACE1 activity than did the other

tion of BACE1 with extremely high potency, and its IC50 bergenin analogue（s 2-4, 9, and 10）, with IC50 values of 6.9, value was calculated as 0.6 μM. 11-O-Protocatechuoylber- 8.8, and 5.0 μM, respectively. When an OMe group was in- genin（5）, having 2 OH groups at C-3’ and C-4’, showed the troduced at the C-4’ position on bergenin analogue moiety, strongest inhibitory activity among the bergenin analogues and the meta-hydrogen was substituted by an OH or OMe （2-10）. The other active compounds, 2-4 and 6-10, also group（6 and 8）, the inhibitory activities of BACE1 were

exhibited more potent activities than bergenin（1）. The much more potent than that of 4（IC50 values: 5.06 and 6.9 structure-activity analysis showed that 3 and 4, which have vs. 26 μM）. On the other hand, in the case of 3, which has an OH or OMe group at C-4’, led to an approximately 3 an OH group at the C-4’ position, an additional OMe group times increase in BACE1 inhibition compared with 2, which at the C-3’ position（7）also led to an increase in inhibitory

has no substituents of benzoic moiety. Furthermore, com- activity（IC50 value: 8.8 μM）. Furthermore, compound 5,

397 J. Oleo Sci. 62, (6) 391-401 (2013) Y. Kashima and M. Miyazawa

with 2 OH groups at the C-3’ and C-4’ positions of the benzoic acid analogues to elucidate whether the inhibitory benzoic moiety, showed a marked increase in inhibitory activities of 5-8 are caused by the benzoic acid moiety or activit（y IC50 value: 0.6 μM）. Comparison of the BACE1 in- the bergenin moiety. Protocatechuic acid, 3,4-dimethoxy- hibitory activities of compounds 5 and 7 showed that the benzoic acid, vanillic acid, and isovanillic acid（inhibition: potency increased in the order of 5＞7. This observation 1.1-8.9％ at 400 μM）and bergenin（inhibition: 22.1％ at 400 revealed that the presence of an OH group at the C-3’ posi- μM）exhibited low inhibitory activities; these compounds tion led to significantly higher activity than the presence of exhibited relatively poor inhibitory activities compared an OMe group. These findings, at suggest that at least a with the bergenin analogues. Therefore, we assumed that functional group at the C-3’ position of the benzoic moiety the ester linkage of bergenin and benzoic acid as well as is essential for the stronger inhibitory activity, and an addi- bergenin moiety and benzoic acid moiety influenced the in- tional OH or OMe group at the C-4’ position enhances the hibitory activity. The inhibitory effect of each compound inhibitory activity. However, an additional OMe group at on BACE1 activity is shown in Table 2. The inhibition ki- the C-5’ position seems to have no significant influence as netics of compounds 5-8 against BACE1 were also analyzed evidenced by the comparison of the structure and activity by Dixon plo（t Fig. 6）. The Dixon plot is a graphical of 7 and 10. These results show that functional substitu- method［plot of 1/enzyme velocity（1/V）vs. inhibitor con- tions on the benzoic moiety selectively enhance or de- centration（I）with varying concentrations of the substrate］ crease inhibition of BACE1 activity. used to determine the type of enzyme inhibition and the 48） Furthermore, we evaluated the inhibitory activities of inhibition constan（t Ki）for an enzyme-inhibitor complex .

Fig. 6 Dixon plots for inhibition of bergenin analogues 5, 6, 7, and 8 (panels A, B, C, and D, respectively) on BACE1 activity. The BACE1 substrate concentrations used were 1.5 mM (◆), 0.75 mM (▲), and 0.375 mM (■).

398 J. Oleo Sci. 62, (6) 391-401 (2013) BERGENIN ANALOGUES AS b-SECRETASE INHIBITOR

The data gave a family of straight lines passing through the ACKNOWLEDGMENT same point at the second quadran（t Fig. 6）. The Ki values This work was supported by Grant-in Aid from the Japan were determined on the basis of the interpretation of the Society for the Promotion of Science（No. 24658055）.

Dixon plot, where the value of the x-axis implies－Ki. The data in panel A suggested that the most active compound, 11-O-protocatechuoylbergenin（5）, was a competitive-type inhibitor of BACE1, since the curves intersected to the left References of the y-axis and above the x-axis. From the analysis, it can 1） Walsh, D. M.; Selkoe, D. J. Aβ oligomers-a decade of be concluded that compound 5 competes with the sub- discovery. J. Neurochem., 101, 1172-1184（2007）. strate for its active site on the enzyme. The Ki value of 5 （2010）. estimated from the Dixon plot was 0.42 μM. As shown in 2） Kurosumi, M.; Nishio, Y.; Osawa, S.; Kobayashi, H.; panels B to D, compounds 6-8 are also competitive-type in- Iwatsubo, T.; Tomita, T.; Miyachi, H. Novel Notch-spar- hibitors with Ki values of 6.8, 8.7, and 6.4 μM, respectively. ing g-secretase inhibitors derived from a peroxisome

Usually, the lower the Ki value, the tighter the binding with proliferator-activated receptor agonist library. Bioorg. enzyme and the more effective the inhibitor, indicating that Med. Chem. Lett., 20, 5282-5285（2010）. these compounds, especially compound 5, may be signifi- 3） Vassar, R.; Bennett, B. D.; Babu-Khan, S.; Kahn, S.; cant BACE1 inhibitors. Mendiaz, E. A.; Denis, P.; Teplow, D. B.; Ross, S.; Ama- In this study, the bergenin analogues were shown to have rante, P.; Loeloff, R.; Luo, Y.; Fisher, S.; Fuller, J.; DPPH radical scavenging and BACE1 inhibitory activities. Edenson, S.; Lile, J.; Jarosinski, M. A.; Biere, A. L.; AD is a complex neurodegenerative disorder caused by Curran, E.; Burgess, T.; Louis, J. C.; Collins, F.; Tre- various factors including active oxygen species. Aβ oligo- anor, J.; Rogers, G.; Citron, M. β-Secretase clavage of mers in the brain increase intracellular ROS, which causes Alzheimer’s amyloid precursor protein by the trans- neuronal cell death in AD. Therefore, ROS play key roles in membrane aspartic protease BACE. Science, 286, the initiation and progression of many neurodegenerative 735-741（1999）. diseases of the central nervous system such as Alzheimer’s 4） Hardy, J.; Selkoe, D. J. The amyloid hypothesis of Al- and Parkinson’s disease49）. In addition, a previous study zheimer’s disease: progress and problems on the road showed that antioxidants such as flavonoid glycosides and to therapeutics. Science, 297, 353-356（2002）. a triterpene ester exert neuroprotective activity50）. These 5） Crouch, P. J.; Harding, S. M. E.; White, A. R.; Camak- findings indicate that the neuroprotective effects against aris, J.; Bush, A. I.; Masters, C. L. Mechanisms of Aβ AD might be mediated by reduced intracellular ROS gener- mediated neurodegeneration in Alzheimer’s disease. ation. Moreover, Takahashi et al. showed that esterification Int. J. Biochem. Cell Biol,. 40, 181-198（2008）. of hydroxy groups at C11 of bergenin with various fatty 6） Tian, X. Y.; Zhao, Y.; Yu, S. S.; Fang, W. S. BACE1（Be- acids enhanced antioxidant and neuroprotective activi- ta-secretase）inhibition phenolic acids and a novel ses- ties50）. Almost all bergenin analogues, in particular 11-O- quiterpenoid from Homalomena occulta. Chem. Bio- protocatechuoylbergenin（5）, possess stronger antioxida- diversity, 7, 984-992（2010）. tive properties than that of bergenin, a known antioxidant. 7） Artavanis-Tsakonas, S.; Rand, M. D.; Lake, R. J. Notch Therefore, the esterified bergenin analogues, which have signaling: cell fate control and signal integration in de- antioxidative effects as well as BACE1 inhibitory activities, velopment. Science, 284, 770-776（1999）. may be useful in the study of the mechanisms of AD. 8） Citron, M. β-Secretase inhibition for the treatment of In conclusion, this study suggests a correlation between Alzheimer’s disease-promise and challenge. Trends BACE1 inhibition and the structures of bergenin analogues Pharmacol. Sci., 25, 92-97（2004）. occurring widely in a number of plants. The inhibitors ana- 9） Sankaranarayanan, S.; Price, E. A.; Wu, G.; lyzed in this study might not be directly considered as a Crouthamel, M. C.; Shi, X. P.; Tugusheva, K.; Tyler, K. drug candidates because their IC50 values are low, although X.; Kahana, J.; Ellis, J.; Jin, L.; Steele, T.; Stachel, S.; they are higher than that of a peptide-derived inhibitor. To Coburn, C.; Simon, A. J. In vivo β-secretase 1 inhibi- the best of our knowledge, this is the first study to report tion leads to brain Aβ lowering and increased on the BACE1 inhibitory activity of bergenin analogues. We α-secretase processing of amyloid precursor protein believe that the studied compounds, which inhibited not without effect on neuregulin-1. J. Pharmacol. Exp. only BACE1 but also active oxygen species involved in the Ther., 324, 957-969（2008）. brain immune system, may be effective therapeutic re- 10） Christen, Y. Oxidative stress and Alzheimer disease. agents for further drug development in AD. Am. J. Clin. Nutr., 71, 621S-629S（2000）. 11） Miranda, S.; Opazo, C.; Larrondo, L. F.; Muñoz, F. J.; Ruiz, F.; Leighton, F.; Inestrosa, N. C. The role of oxidative stress in the toxicity induced by amyloid

399 J. Oleo Sci. 62, (6) 391-401 (2013) Y. Kashima and M. Miyazawa

β-peptide in Alzheimer’s disease. Prog. Neurobiol., β-secretase inhibitor. Bioorg. Med. Chem., 20, 784- 62, 633-648（2000）. 788（2012）. 12） Praticò, D.; Uryu, K.; Leight, S.; Trojanoswki, J. Q.; 24） Singh, D. P.; Srivastava, S. K.; Govindarajan, R.; Rawat, Lee, V. M. Y. Increased lipid peroxidation precedes A. K. S. High-performance liquid chromatographic de- amyloid plaque formation in an animal model of Al- termination of bergenin in different Bergenia species. zheimer amyloidosis. J. Neurosci., 21, 4183-4187 Acta Chromatogr., 19, 246-252（, 2007）. （2001）. 25） Arfan, M.; Amin H.; Karamać, M.; Kosińska, A.: Wicz- 13） Tamagno, E.; Guglielmotto, M.; Aragno, M.; Borghi, R.; kowsk, W.; Amarowicz, R. Antioxidant activity of phe- Autelli, R.; Giliberto, L.; Muraca, G.; Danni, O.; Zhu, X.; nolic fractions of Mallotus philippinensis bark ex- Smith, M. A.; Perry, G.; Jo, D. G.; Mattson, M. P.; Taba- tract. Czech J. Food Sci., 27, 109-117（2009）. ton, M. Oxidative stress activates a positive feedback 26） Jung, J. C.; Lim, E.; Kim S. H.; Kim, N. S.; Jung, M.; between the γ- and β-secretase cleavages of the Oh, S. Practical synthesis and biological evaluation of β-amyloid precursor protein. J. Neurochem., 104, bergenin analogs. Chem. Biol. Drug Des., 78, 725-729 683-695（2008）. （2011）. 14） Jo, D. G.; Arumugam, T. V.; Woo, H. N.; Park, J. S.; 27） Kumar, R.; Patel, D. K.; Prasad, S. K.; Laloo, D.; Krish- Tang, S. C.; Mughal, M.; Hyun, D. H.; Park, J. H.; Choi, namurthy, S.; Hemalatha, S. Type 2 antidiabetic activi- Y. H.; Gwon, A. R.; Camandola, S.; Cheng, A.; Cai, H.; ty of bergenin from the roots of Caesalpinia digyna Song, W.; Markesbery, W. R.; Mattson, M. P. Evidence Rottler. Fitoterapia, 83, 395-401（2012）. that γ-secretase mediates oxidative stress-induced 28） Mduka, H. C. C.; Okoye, Z. S. C. The effect of Sacoglot- β-secretase expression in Alzheimer’s disease. Neuro- tis gabonensis stem bark extract, a Nigerian alcoholic biol. Aging, 31, 917-925（2010）. beverage additive, on the natural antioxidant defences 15） Cai, Z.; Zhao, B.; Ratka, A. Oxidative stress and during 2,4-dinitrophenyl hydrazine-induced mem- β-amyloid protein in Alzheimer’s disease. Neuromol. brane peroxidation in vivo. Vasc. Pharmacol., 39, Med., 13, 223-250（2011）. 21-31（2002）. 16） Jun, M.; Lee, S. H.; Choi, S. H.; Bae, K.; Seong, Y. H.; 29） Swarnalakshmi, T.; Sethuraman, M. G.; Sulochana, N.; Lee, K. B.; Song, K. S. Plant phenolics as β-secretase Arivudainambi, R. A note on the anti-inflammatory ac- （BACE1）inhibitors. Food Sci. Biotechnol., 4, 612-624 tivity of bergenin. Curr. Sci., 53, 917（1984）. （2006）. 30） Nazir, N.; Koul, S.; Qurishi, M. A.; Taneja, S. C.; Ah- 17） Parr, A. J.; Bolwell, G. P. Phnols in the plant and in mad, S. F.; Bani, S.; Qazi, G. N. Immunomodulatory ef- man. The potential for possible nutritional enhance- fect of bergenin and norbergenin against adjuvant-in- ment of the diet by modifying the phenols content or duced arthritis-A flow cytometric study. J. profile. J. Sci. Food Agric., 80, 985-1012（2000）. Ethnopharmacol., 112, 401-405（2007）. 18） Jeon, S. Y.; Bae, K.; Seong, Y. H.; Song, K. S. Green tea 31） Jahromi, M. A. F.; Chansouria, J. P. N.; Ray, A. B. Hy- catechins as a BACE1（β-secretase）inhibitor. Bioorg. polipidemic acitivity in rats of bergenin, the major Med. Chem. Lett., 13, 3905-3908（2003）. constituent of Flueggea microcarpa. Phytother. Res., 19） Choi, S. H.; Hur, J. M.; Yang, E. J.; Park, H. J.; Lee, K. B.; 6, 180-183（1992）. Moon, E.; Song, K. S. β-Secretase（BACE1）inhibitors 32） Piacente, D.; Pizza C.; DeTommasi, N.; Mahmood, N. from Perilla frutescens var. acuta. Arch. Pharm. Constituents of Ardisia japonica and their in vitro Res., 31, 183-187（2008）. anti-HIV activity. J. Nat. Prod., 59, 565-569（1996）. 20） Hwang, E. M.; Ryu, Y. B.; Kim, H. Y.; Kim, D. G.; Hong, S. 33） Pu, H. L.; Huang, X.; Zhao, J. H.; Hing, A. Bergenin is G.; Lee, J. H.; Curtis-Long, M. J.; Jeong, S. H.; Park, J. the antiarrhythmic principle of Fluggea virosa. Planta Y.; Park, K. H. BACE1 inhibitory effects of lavandulyl Med., 68, 372-374（2002）. flavanones from Sophora flavescens. Bioorg. Med. 34） Lim, H. K.; Kim, H. S.; Choi, H. S.; Oh, S.; Choi, J. Hep- Chem., 16, 6669-6674（2008）. atoprotective effects of bergenin, a major constituent 21） Sasaki, H.; Miki, K.; Kinoshita, K.; Koyama, K.; Ju- of Mallotus japonicas, on carbon tetrachloride-intoxi- liawaty, L. D.; Achmad, S. A.; Hakim, E. H.; Kaneda, M.; cated rats. J. Ethnopharmacol., 72, 469-474（2000）. Takahashi, K. β-Secretase（BACE-1）inhibitory effect 35） DeOliveira, C. M.; Nonato, F. R.; DeLima, F. O.; Couto, of bioflavonoids. Bioorg. Med. Chem. Lett., 20, 4558- R. D.; David, J. P.; David, J. M.; Soares, M. B. P.; Villar- 4560（2010）. real, C. F. Antinociceptive properties of bergenin. J. 22） Takahashi, T.; Miyazawa, M. Serotonin derivatives as Nat. Prod., 74, 2062-2068（2011）. inhibitors of β-secretase（BACE 1）. Pharmazie 66, 36） Yoshida, T.; Seno, K.; Takama, Y.; Okuda, T. Bergenin 301-305（2011）. derivatives from Mamallotus japonicas. Phytochem- 23） Marumoto, S.; Miyazawa, M. Structure-activity rela- istry, 21, 1180-1182（1982）. tionships for naturally occurring coumarins as 37） Saijo, R.; Nonaka, G.; Nishioka, I. Gallic acid esters of

400 J. Oleo Sci. 62, (6) 391-401 (2013) BERGENIN ANALOGUES AS b-SECRETASE INHIBITOR

bergenin and norbergenin from Mallotus japonicas. carbon-coupled dimeric bergenin derivative biotrans- Phytochemistry, 29, 267-270（1990）. formed by Pleurotus ostreatus. Bioorg. Med. Chem. 38） Jia, Z.; Mitsunaga, K.; Koike, K.;Ohmoto, T. New ber- Lett., 15, 4073-4075（2005）. genin derivatives from Ardisia crenata. Nat. Med., 45） Pearl, I. A.; Vanillic acid. Org. Synth., 4, 972（1963）. 49, 187-189（1995）. 46） Sumino, M; Sekine, T.; Ruangrungsi, N.; Igarashi, K.; 39） Fuji, M.; Miyaichi, Y.; Tomimori, T. Studies on Nepalese Ikegami, F. Ardisiphenols and other antioxidant prin- crude drugs. XXII on the phenolic constiruents of the ciples from the fruits of Ardisia colorata. Chem. rhizome of Bergenia ciliate（Haw.）STERNB. Nat. Pharm. Bull., 50, 1484-1487（2002）. Med., 50, 404-407（1996）. 47） Al-Tel, T. H.; Semreen, M. H.; Al-Qawasmeh, R. A.; 40） Lee, Y. Y.; Jang, D. S.; Jin, J. L.; Yun-Choi, H. S. Anti- Schmidt, M. F.; El-Awadi, R.; Ardah, M.; Zaarour, R.; platelet aggregating and anti-oxidative activities of Rao, S. N.; El-Agnaf, O. Design, synthesis, and qualita- 11-O（- 4’-O-methylgalloyl）-bergenin, a new compound tive structure-activity evaluations of novel β-secretase isolated from Crassula cv.‘ Himaturi’ Planta. Med., inhibitors as potential Alzheimer’s drug leads. J. Med. 71, 776-777（2005）. Chem., 54, 8373-8385（2011）. 41） Li, C.; Chen, X.; Fang, D.; Li G. A new bergenin deriva- 48） Dixon, M. The determination of enzyme inhibitor con- tive from Corylopsis willmottiae. Chem. Nat. stants. Biochem. J., 55, 170-171（1953）. Compd., 47, 194-196（2011）. 49） Uttara, B.; Singh, A. V.; Zamoboni, P.; Mahajan, R. T. 42） Zamarrud; Ali I.; Hussain, H.; Ahmad, V. U.; Qaiser, M.; Oxidative stress and neurodegenerative disease: A re- Amyn, A.; Mohammad, F. V. Two new antioxidant ber- view of upstream and downstream antioxidant thera- genin derivatives from the stem of Rivea hypocrateri- peutic options. Curr. Neuropharmacol., 7, 65-74 formis. Fitoterapia, 82, 722-725（2011）. （2009）. 43） Siddiq, F.; Fatima, I.; Malik A.; Afza, N.; Iqbal, L.; La- 50） Takahashi, H.; Kosaka, M.; Watanabe, Y.; Nakade, K.; teef, M.; Hameed, S.; Khan, S. W. Biologically active Fukuyama, Y. Synthesis and neuroprotective activity bergenin derivatives from Bergenia stracheyi. Chem. of bergenin derivatives with antioxidant activity. Biodiversity, 9, 91-98（2012）. Bioorg. Med. Chem., 11, 1781-1788（2003）. 44） Wang, D.; Zhu, H. T.; Zhang,Y. J.; Yang, C. R. A carbon-

401 J. Oleo Sci. 62, (6) 391-401 (2013)