Boswellic Acids with Acetylcholinesterase Inhibitory
Properties from Frankincense
Masahiro Otaa,b and Peter J. Houghtona
aPharmaceutical Sciences Research Division, King’s College London, 150 Stamford St,
London, SE1 9NH, UK
bShiseido Life Science Research Center, 2-2-1 Hayabuchi, Tsuzuki-ku, Yokohama-shi,
224-8558, Japan
Received: August 1st, 2007; Accepted: August 15th, 2007
Dedicated to Professor Peter G Waterman, one of the pioneers of phytochemical research.
Frankincense, a gum resin secreted from barks of Boswellia species, is reputed in Arabian folk medicine to improve the memory. In this study, the acetylcholinesterase inhibitory effect of extracts of frankincense and their constituents were investigated. The ethyl acetate soluble fraction from methanol extracts showed the greatest inhibition of acetylcholinesterase. Bioassay-guided fractionation was carried out to isolate several boswellic acids, and their structures were determined spectroscopically. The boswellic acids showing the most inhibitory activity on acetylcholinesterase were 11α-hydroxy-β -boswellic acid (1) and 11-keto-β-boswellic acid (5), whilst others isolated i.e. 3α-acetyl-11-keto-β-boswellic acid (2), 3α-acetyl-11α-hydroxy-β-boswellic acid (3), 11α-methoxy-β-boswellic acid (4), β-boswellic acid (6) and α-boswellic acid (7) were inactive. Acetylcholinesterase inhibitory activity appears to be associated with the presence of either the free hydroxyl group or keto group at C-11 and of the free hydroxyl group at C-3 in the ursane skeleton.
Keywords: Frankincense, Boswellia, triterpenoids, acetylcholinesterase inhibition, memory improvement.
Alzheimer’s disease (AD) is the most common form of dementia [1,2], and is associated with low levels of the neurotransmitter acetylcholine (ACh) [3]. Inhibiting acetylcholinesterase (AChE), involved in the breakdown of ACh, increases its level in brain and provides the major symptomatic treatment of AD available at present. The archetypal AChE inhibitor is physostigmine from Physostigma venenosum, which was used traditionally in Africa as a ritual poison. However, it is only recently, with the introduction of the physostigmine derivative rivastigmine and the alkaloid galantamine as drugs in clinical use for AD, that attention has been drawn to natural products that boost ACh levels by inhibition of AChE, particularly as active constituents of plant species with a traditional use related to memory disorders [4,5]. The oleo gum resin frankincense, also called olibanum, is a product known for thousands of years
in the Middle East and parts of Africa. It is obtained from several species of Boswellia (Burseraceae), which grow in Oman, and has several uses in traditional Arabian medicine, including a reputation of improving memory [6]. It was, therefore, thought to be of interest to investigate frankincense for the possibility of AChE inhibitory compounds using the modified Ellman reaction to monitor extracts and fractions [7].
The results for the AChE inhibition of the total methanol (MeOH) extract and the fractions obtained by solvent partition show that the major activity resides in the ethyl acetate (EtOAc) fraction (Table 1). This fraction was, therefore, selected for isolation of its contained compounds in order to find those responsible for the effects. The activity of the fractions obtained from a silica gel CC separation showed that fractions 5-7 were the most
NPC Natural Product Communications
Vol. 3
No. 1
21 - 26
HOOC R R'O HOOC HO 7 R R' 1 α-OH H 2 = O Ac 3 α-OH Ac 4 α-OCH3 H 5 = O H 6 H H 7 21 30 29 28 22 8 3 6 20 19 17 18 27 25 24 23 1 2 4 5 9 10 13 16 14 11 12 15 26
Table 1: Percentage inhibition of AChE by total MeOH extract of frankincense and derived fractions
Extract/fraction Conc mg/mL % AChE inhibition
Total MeOH 0.93 64 extract 0.46 27 0.23 7 n Hexane fraction 0.93 28 0.46 16 EtOAc fraction 0.93 61 0.46 21 0.23 9 Water fraction 0.93 26 0.46 19 0.23 9 Physostigmine 0.08 50
Table 2: Percentage inhibition of AChE by 0.09 mg/mL fractions from silica gel CC of EtOAc fraction of MeOH extract of frankincense.
Fraction % AChE inhibition Fraction % AChE inhibition FKE 1 11.1 FKE 6 8.7 FKE 2 1.5 FKE 7 26.9 FKE 3 3.4 FKE 8 0.1 FKE 4 4.6 FKE 9 0 FKE 5 7.9 FKE 10 0
active (Table 2), and so these were selected for isolation of compounds. Together with MS data, the
13C NMR spectra were of most use in identifying the
compounds isolated. All appeared to be triterpenoids from the molecular formulae derived from HRMS, by their reaction on TLC after spraying with acidic
Table 3: 13C NMR chemical shifts of compounds 1-7 (CDCl
3, 125 MHz). C 1 2 3 4 5 6 7 1 35.8 34.6 36.4 35.0 33.9 33.9 33.6 2 26.4 23.5 23.7 26.5 26.2 26.2 26.2 3 70.6 73.1 73.2 70.6 70.5 70.7 70.8 4 47.5 46.5 46.7 47.6 47.2 47.4 47.4 5 49.0 50.4 50.6 49.1 48.8 49.1 49.1 6 19.6 18.7 19.5 19.5 18.8 19.7 19.7 7 33.9 32.8 33.8 33.9 32.9 33.1 32.8 8 43.3 43.8 43.2 43.0 45.1 40.1 39.8 9 54.6 60.3 54.4 50.6 60.4 46.9 46.8 10 38.7 37.5 38.6 38.7 37.5 37.5 37.6 11 68.8 199.3 68.6 77.1 199.5 23.4 23.6 12 129.0 130.5 128.8 124.4 130.5 124.5 121.8 13 142.8 165.0 142.9 143.6 165.1 139.6 145.2 14 42.4 45.1 42.3 42.3 43.8 42.3 41.9 15 26.5 27.2 26.6 26.6 27.2 26.5 26.0 16 27.9 27.5 27.9 28.0 27.5 28.1 26.9 17 33.7 34.0 33.7 33.8 34.0 33.8 32.5 18 58.3 59.0 58.2 58.7 59.0 59.2 47.3 19 39.4 39.3 39.4 39.4 39.3 39.7 46.8 20 39.3 39.3 39.4 39.5 39.3 39.6 31.1 21 31.1 30.9 31.1 31.2 30.9 31.3 34.7 22 41.3 40.9 41.3 41.4 40.9 41.5 37.1 23 24.5 23.9 24.1 24.6 24.3 24.2 24.1 24 181.5 181.9 182.1 183.2 181.9 182.9 182.6 25 14.2 13.2 14.2 14.4 13.2 13.3 13.1 26 18.0 18.4 18.0 18.3 18.3 16.9 16.7 27 23.0 20.5 23.0 22.5 20.5 23.3 25.9 28 28.7 28.9 28.7 28.7 28.9 28.8 28.4 29 17.5 17.4 17.5 17.5 17.4 17.5 33.3 30 21.3 21.2 21.3 21.3 21.1 21.4 23.7 3-O-Ac 170.3 21.4 170.4 21.4 11-OMe 54.1
anisaldehyde reagent and their 13C NMR spectroscopic data (Table 3). Compounds 6 and 7 (Table 4) were identified as the known compounds
β-boswellic acid and α-boswellic acid, respectively, by comparison of spectral properties with those reported in the literature [8]
.
Compound 1 was obtained as a white amorphous powder, which showed a pseudomolecular ion at m/z 495 [M+Na]+ in the positive ESI-MS, leading to
the molecular formula C30H48O4. The 13C NMR
spectral data showed signals for 30 carbons (Table 3). Multiplicity selection by DEPT NMR analysis and characteristic 13C chemical shift data led to
7 methyls, 8 methylenes, 7 methines, 5 quaternary carbons, one carboxylic carbonyl carbon at δc 179.7,
and two olefinic carbon atoms at δc 127.1 and 141.0.
The signals in the 1H NMR spectrum (Table 4)
showed 5 tert-methyl signals at δH 0.82 ( H-28), 1.09
(H-25), 1.13 (H-26), 1.21 (H-27), and 1.39 (H-23). The oxymethine proton signal at δH = 4.26 (H-11)
was also revealed. The correlation in the NOESY spectrum between H-11 and H-25, and between H-11 and H-26 indicated an α-β oriented hydroxyl group at C-11. The relative configuration of the hydroxyl group at C-3 was assigned to be α (axial) on the basis of the NOESY spectrum, which revealed the absence
Table 4: 1H-NMR chemical shifts of compounds 1-7 (CDCl
3, 500 MHz). H 1 2 3 4 5 6 7 1 1.49, 1.33 m 1.48, 1.33 m 1.48, 1.33 m 1.49, 1.33 m 1.49, 1.33 m 1.49, 1.33 m 1.45, 1.29 m 2 2.24, 1.61 m 2.24, 1.60 m 2.24, 1.61 m 2.24, 1.60 m 2.24, 1.60 m 2.24, 1.60 m 2.22, 1.59 m 3 4.08 t J =2.5 Hz 4.08 t J =2.5 Hz 4.08 t J = 2.5 Hz 4.08 t J = 2.5 Hz 4.08 t J = 2.5 Hz 4.08 t J = 2.5 Hz 4.08 t J = 2.5 Hz 5 1.50 m 1.51 m 1.50 m 1.51 m 1.49m 1.50 m 1.49 m 6 1.83, 1.70 m 1.83, 1.71 m 1.82, 1.70 m 1.83, 1.70 m 1.83, 1.70 m 1.82, 1.70 m 1.85, 1.70 m 7 1.58, 1.40 m 1.58, 1.41 m 1.58, 1.41 m 1.59, 1.41 m 1.58, 1.41 m 1.58, 1.40 m 1.52, 1.37 m 9 1.63 m 2.25 m 1.63 m 1.63 m 2.24 m 1.79 m 1.66 m 11 4.26 d J = 3.6 Hz 4.26 d J = 3.6 Hz 4.07 d J = 3.6 Hz 1.92, 1.18 m 1.88 m 12 5.14 t J = 3.5 Hz 5.77 s J = 3.5 Hz 5.15 t J = 3.5 Hz 5.14 t 5.77 s J 5.14= 3.5 Hzt J = 3.5 Hz 5.19 t 15 1.85, 1.02 m 1.86, 1.02 m 1.85, 1.02 m 1.86, 1.02 m 1.86, 1.03 m 1.86, 1.03 m 1.77, 1.00 m 16 2.01, 0.88 m 2.02, 0.89 m 2.02, 0.88 m 2.02, 0.88 m 2.02, 0.89 m 2.02, 0.88 m 2.00, 0.81 m 18 1.34 m 1.3 m 1.34 m 1.34 m 1.34 m 1.34 m 1.96 m 19 1.33 m 1.33 m 1.33 m 1.33 m 1.33 m 1.33 m 1.70, 1.02 m 20 0.94 m 0.94 m 0.94 m 0.94 m 0.94 m 0.94 m 21 1.41, 1.29 m 1.41, 1.29 m 1.41, 1.29 m 1.41, 1.29 m 1.41, 1.29 m 1.41, 1.29 m 1.33, 1.10 m 22 1.45, 1.27 m 1.45, 1.27 m 1.45, 1.27 m 1.45, 1.27 m 1.45, 1.27 m 1.45, 1.27 m 1.44, 1.22 m 23 1.39 s 1.35 s 1.35 s 1.39 s 1.39 s 1.37 s 1.35 s 25 1.09 s 1.14 s 1.14 s 1.09 s 1.09 s 1.09 s 0.89 s 26 1.13 s 1.20 s 1.20 s 1.13 s 1.13 s 1.13 s 1.13 s 27 1.21 s 1.24 s 1.24 s 1.21 s 1.21 s 1.21 s 1.21 s 28 0.82 s 0.83 s 0.83 s 0.82 s 0.82 s 0.82 s 0.81 s 29 0.96 s 0.96 d J= 5.5 Hz 0.96 d J = 5.5 Hz 0.96 d J = 5.5 Hz 0.96 d J = 5.5 Hz 0.96 d J = 5.5 Hz 0.91 s 30 0.95 s 0.96 d J = 6.0 Hz 0.96 d J = 6.0 Hz 0.96 d J = 6.0 Hz 0.96 d J = 6.0 Hz 0.96 d J = 6.0 Hz 0.91 s 3-O-H 2.01 bs 3-O-Ac 2.09 s 2.09 s 11-OH 2.11 bs 2.11 bs 11-OMe 3.24 s 24-COOH 2.55 bs 2.52 bs 2.52 bs 2.55 bs 2.55 bs 2.54 bs 2.53 bs
of interactions between H-3 and H-5, but correlation between H-3 and H-23. Based on the evidence given above and 2D-NMR experiments using COSY, HMQC, HMBC, and NOESY pulse sequence, the structure of 1 was determined as 11α-hydroxy-β -boswellic acid. It was successfully isolated, in spite of being a thermodynamically unstable compound easily transformed into 9,11-dehydro-β-boswellic acid [9]. The 23-carboxylic acid methyl ester of this compound has been previously isolated and its reported NMR signals are similar to those obtained for 1 [10].
Compound 2 was obtained as a white amorphous powder, which showed a pseudomolecular ion at m/z 535 [M+Na]+ in the positive ESI-MS, leading to the
molecular formula C32H48O5. The 13C NMR spectral
data showed signals for 32 carbons (Table 3). Multiplicity selection by DEPT NMR analysis and characterisitic 13C chemical shift data led to a similar
profile to 1, except for the presence of a ketone (δC 199.3) and the absence of the oxymethine
δH = 4.26. Extra signals were also observed indicating the presence of an acetyl group (δH 2.09,
δc 170.3, 21.4) and a signal at δH 5.53 instead of that seen at 4.26 in 1. implied that the 3-OH seen in 1 was acetylated. The signals in the 1H NMR spectrum
showed 5 tert-methyl signals at δH 0.83 (H-28), 1.14
(H-25), 1.20 (H-26), 1.24 (H-23), and 1.35 (H-27), and an acetyl signal at δH 2.09. The relative
configuration of the acetyl group at C-3 was assigned to be α (axial) for the same reasons as given for the OH group in 1. Based on the evidence given above and 2D-NMR experiments using COSY, HMQC, and HMBC, NOESY pulse sequence, 2 was determined to be 3α -O-acetyl-11-keto-β-boswellic acid [11]. Compound 3 showed a pseudomolecular ion at m/z 537 [M+Na]+ in the positive ESI-MS leading
to the molecular formula C32H50O5. The 13C NMR
spectral data were similar to those given by 2, apart from the absence of the carbonyl at 199.31 (Table 3). The oxymethine proton signal at δH 4.26
(H-11) corresponded with a CH signal at δ 68.64, and HMQC and HMBC spectra revealed that these signals were due to a hydroxy group at C-11, as also seen in 1. NOESY spectral evidence showed that this was in α orientation. Thus 3 was determined as 3α-O-acetyl-11α -hydroxy-β-boswellic acid [12]. Compound 4 showed a pseudomolecular ion at m/z 486.0 [M+H]+ in the positive ESI-MS leading to the
molecular formula C31H50O4. The 13C NMR spectral
methoxy signal at δc 54.13 and δH 3. 78. HMQC and
HMBC spectra revealed that these signals could only be due to a 11-methoxy group. The NOESY spectrum indicated an α-orientation for this group, so compound 4 was determined as 11α-methoxy-β -boswellic acid, previously known only as the 3-acetyl derivative [12].
Compound 5 showed a pseudomolecular ion at m/z 494.2 [M+Na]+ in the positive ESI-MS, leading to the
molecular formula C30H46O4 and gave similar NMR
spectra to 2, except for lack of signals seen corresponding to the 3-acetyl group. The molecular formula agreed with a compound with a 3-OH structure and the orientation of this group as β was confirmed by the NOESY spectrum, which showed the absence of interactions between H-3 and H-5, and correlation between H-3 and H-23.The structure of compound 5 was, therefore, determined as 11-keto-β-boswellic acid [11]
Table 5: Inhibitory activities on acetylcholinesterase of compounds from frankincense.
Compound % AChE inhibition at 0.75mM
11α-Hydroxyl-β-boswellic acid, 1 80 3α -O-acetyl-11-keto-β-boswellic acid, 2 9.1 3α -O-acetyl-11α -hydroxy- β -boswellic acid, 3 15
11α-Methoxy-β-boswellic acid, 4 -3.7
11-Keto-β-boswellic acid, 5 71
β-Boswellic acid, 6 21
α -Boswellic acid, 7 -2.5
Physostigmine (positive control) IC50 = 0.25μM
Of the boswellic acids tested, only 11-α-hydroxy-β -boswellic acid (1) and 11-keto-β-boswellic acid (5) inhibited AChE activity (Table 5). It is noteworthy that β-boswellic acid (6) and 11-methoxy-β-boswellic acid (4) showed no inhibitory activity, thus suggesting that an oxygen function at C-11 of the ursane skeleton is associated with inhibitory activity. In addition, 3α-O-acetyl-11-keto-β-boswellic acid (2), and 3α-O-acetyl-11α-hydroxy-β-boswellic acid (3) were inactive, in contrast to 11α-hydroxy-β -boswellic acid (1) and 11-keto-β-boswellic acid (5), and this indicates that a free OH group at C-3 is also necessary for the activity.
Not many triterpenoids appear amongst the large number of compounds now known to have some inhibitory effect on acetylcholinesterase [5] and the only other active compound with a similar skeleton so far described is ursolic acid [14].
The occurrence of the two compounds (1) and (5) could explain the activity of frankincense and,
thereby, provide an explanation for its reputation for improving a memory disorder.
Experimental
General: The structures of the isolated compounds
were determined from their NMR spectral data [1H and 13C on a Varian Unity Inova 500 spectrometer (500 MHz for 1H; 125 MHz for 13C)]. HMBC, HSBC and NOESY experiments were carried out using the same instrument. IR spectra were measured on a Perkin-Elmer Spectrum One® FT-IR spectrometer with diamond-attenuated total reflectance (DATR) technique. UV spectra were determined using a Hewlett Packard 8452A diode array spectrometer. EI mass spectra, were obtained from a JEOL JMS-AX505W spectrometer and HR-ESI-MS measured on a Bruker Apex III FT ion cyclotron resonance mass spectrometer.
Plant material: Frankincense (batch number 02553)
was purchased from the Herbal Apothecary, Leicester, U.K. A voucher specimen Bo 7 Y19 is deposited in the museum of the Pharmacy Department, King’s College London and showed an identical TLC profile with an authenticated specimen of Boswellia carterii resin Bo7 2 Y11 from the museum.
Extraction and isolation: Frankincense (520 g) was
extracted with MeOH at room temperature for 6 days. The extract was filtered and concentrated in vacuo at 40oC to dryness to give 241.6 g residue. The residue
(11 g) was re-dissolved in 10 mL of MeOH and partitioned between n-hexane: H2O 1:1. Removal of
the solvent from each phase gave the n-hexane-soluble fraction (7.91 g), and the H2O-soluble
fraction (0.08 g). A portion of the original MeOH extract (49 g) of the frankincense was partitioned between EtOAc: H2O 1:1 to yield the EtOAc-soluble
fraction (47.73 g). Subsequently, the EtOAc-soluble fraction was chromatographed over silica gel using n-hexane and increasing amounts of EtOAc up to 100% to give 10 fractions (FKE 1 – FKE 10), monitored by TLC. The three most active fractions (Table 2) were subjected to further preparative chromatography as follows. FKE 5 was subjected to HPLC [column: Capcell Pak C18 UG 120, MeOH :
H2O (5:1, v/v) → MeOH ] to give 2 (10.2 mg) and 3
(19.4 mg). FKE 6 was subjected to flash chromatography [Flash Master Parallel (Jones Chromatography)] with a step gradient of DCM: EtOAc: MeOH in the ratio 20:2:1, 20:2:11, v/v/v
(550 mL each) and MeOH up to 100 % (550 mL), successively collecting 50 mL fractions, the most active of which (data not shown) was further purified on HPLC [column: Capcell Pak C18 UG 120, MeOH :
H2O (4:1, v/v) → MeOH] to isolate compound 1 (4.5
mg). FKE 7 was subjected to flash chromatography and HPLC [column: Capcell Pak C18 UG 120, MeOH
: H2O : acetic acid (4:1:0.01, v/v/v) → MeOH: acetic
acid (5: 0.01, v/v)] to give compounds 1 (10.6 mg), 4 (15.8 mg), 5 (17.3 mg), 6 (25.3 mg) and 7 (20.2 mg). 11α-Hydroxy-β-boswellic acid (1) White crystals, C30H48O4.. Yield: 2.03 x10-5 %. IR (KBr) νmax: 3500, 1720 cm-1. UV (MeOH) λmax: 213 nm. 1
H NMR (CDCl
3): Table 4. 13C NMR (CDCl 3): Table 3. ESI-MS: m/z (rel.int.) = 472.36 [M+]; 472.3498, calc. 472.3552). 3α-O-Acetyl-11-keto-β-boswellic acid (2) White crystals, C32H48O5.. Yield: 1.96 x10-5 %. IR (KBr) νmax: 3500, 1720, 1740 cm-1. UV (MeOH) λmax: 251 nm. 1H NMR (CDCl
3): Table 4. 13C NMR (CDCl 3): Table 3.ESI-MS: m/z (rel.int.) = 512.36[M]+; 512.3509, calc. 512.3502). 3α-O-Acetyl-11α-hydroxy-β-boswellic acid (3) White crystals, C30H50O5. Yield: 3.73 x10-5 %. IR (KBr) νmax:3500, 1720, 1740 cm-1. UV (MeOH) λmax: 213 nm. 1
H NMR (CDCl
3): Table 4. 13C NMR (CDCl 3): Table 3.ESI-MS: m/z (rel.int.) = 514.38[M]+; 514.3663, calc. 514.3658). 11α-Methoxy-β-boswellic acid (4) White crystals, C31H50O4.. Yield: 3.04 x10-5 %. IR (KBr) νmax: 3500, 1720 cm-1. UV (MeOH) λmax: 206 nm. 1
H NMR (CDCl
3): Table 4. 13C NMR (CDCl 3): Table 3. ESI-MS: m/z (rel.int.) = 486.39 [M]+; 486.3714, calc. 486.3709). 11 Keto-β-boswellic acid (5) White crystals, C30H46O4. Yield: 3.32 x10-5 %. IR (KBr) νmax: 3500, 1720 cm-1. UV (MeOH) λmax: 213, 252 nm. 1H NMR (CDCl
3): Table 4. 13C NMR (CDCl 3): Table 3. ESI-MS: m/z (rel.int.) = 470. 34 [M]+; 470.3389, calc. 470.3396). β-Boswellic acid (6) White crystals, C30H48O3. Yield: 4.90 x10-5 % IR (KBr) νmax: 3500, 1720 cm-1. UV (MeOH) λmax: 206 nm. 1H NMR (CDCl
3): See Table 4. 13C NMR (CDCl 3): see Table 3. ESI-MS: m/z (rel.int.) = 456. 36 [M]+; 456.3601, calc. 456.3603). α-Boswellic acid (7) White crystals, C30H48O3. Yield: 3.88 x10-5 %. IR (KBr) νmax: 3500, 1700 cm-1. UV (MeOH) λmax: 206 nm. 1H NMR (CDCl
3): Table 4. 13C NMR (CDCl 3): Table 3. ESI-MS: m/z (rel.int.) = 456.36 [M]+; 456.3600, calc. 456.3603).Test for acetylcholinesterase inhibition: The method
followed that previously used in our laboratories [7], which is a modified Ellman method, based on the reaction of released thiocholine to give a colored product with a chromogenic reagent. AChE (40 μL of 0.86 U/mL phosphate buffer, pH 8) and either extracts or isolated compounds were mixed with 2.0 mL phosphate buffer (pH 8), and incubated at 40oC
for 20 min. The reaction was started by adding DTNB (20 μL of 0.05mM buffer, pH 7) and ATCI (20 μL of 0.06 mM phosphate buffer, pH 7) in phosphate buffer, pH 7, 20 μL at 37oC for 20 min.
The reaction was halted by placing the assay solution tubes in a bath at 4oC and adding physostigmine
(20 μL 0.018 mM in phosphate buffer, pH 7). The production of the yellow anion was detected at 412 nm. Blanks were used of reagents without extract and a positive control was set up that was the same as the blank, except that physostigmine (20 μL, 0.018M in buffer pH 7) was added. The inhibition rate (%) was calculated using the following formula:
Percent Inhibition = (Blank – Blank positive control) – (Experiment – Experiment control) / (Blank – Blank control) x 100
Acknowledgments – We thank Shiseido Co., Ltd. for financing this project.
References
[1] Evans DA, Funkenstein HH, Albert MS. (1989) Prevalence of Alzheimer’s disease in a community population of older persons.
Journal of the American Medical Association, 262, 2551-2556.
[2] Nordberg A. (1996) Pharmacological treatment of cognitive dysfunction in dementia disorder. Acta Neurologica Scandinavia
Supplement, 168, 87-92.
[3] Perry E, Tomlinson E, Blessed G, Thomlinson BE. (1986) Changes in brain cholinesterase in senile dementia of the Alzheimer type. Neuropathology and Applied Neurobiology, 4, 273-277.
[4] Mukherjee PK, Kumar V, Mal M, Houghton PJ. (2007) Acetylcholinesterase inhibitors from plants. Phytomedicine, 14,289-300. [5] Houghton PJ, Ren Y, Howes MR. (2006) Acetylcholinesterase inhibitors from plants and fungi. Natural Product Reports, 23,
181-199.
[6] Marshall S. (2003) Frankincense: festive pharmacognosy. Pharmaceutical Journal, 271, 862-864.
[7] Perry NSL, Houghton PJ, Theobald AE, Jenner P, Perry EK. (2000) In-vitro inhibition of human erythrocyte acetylcholinesterase
by Salvia lavandulaefolia essential oil and constituent terpenes. Journal of Pharmacy and Pharmacology, 52, 895-902.
[8] Culioli G, Mathe C, Archier P, Vieillescazes C. (2003) A lupane triterpene from frankincense. Phytochemistry, 62, 537-541. [9] Büchele B, Zugmaier W, Simmet T. (2003) Analysis of pentacyclic triterpenic acids from frankincense gum resin and related
phytopharmaceuticals by high-performance liquid chromatography. Identification of lupeolic acid, a novel pentacyclic triterpene.
Journal of Chromatography B, 791, 21-30.
[10] Corsano S, Iavarone C. (1964) Isolation of 3-acetyl-11-hydroxy-β-boswellic acid from incense. Gazzetta Chimica Italiana, 94, 328-335.
[11] Pardhy RS, Bhattacharya SC. (1978) Tetracyclic triterpene acids from Boswellia serrata Roxb. Indian Journal of Chemistry B, 16, 174.
[12] Schweizer S, von Brocke AFW, Boden SE, Bayer E, Ammon HPT, Safayhi H. (2000) Workup-dependent formation of 5-lipoxygenase inhibitory boswellic acid analogues. Journal of Natural Products, 63, 1058-1061.
[13] Chung YK, Heo HJ, Kim EK, Kim HK, Huh TL, Lim Y, Kim SK, Shin DH. (2001) Inhibitory effect of ursolic acid purified from
