Clerodane diterpenes from Polyalthia longifolia var. pendula protect SK-N-MC human neuroblastoma cells from β-amyloid insult

Abstract

Three new diterpenes, polylongifoliaic A (1), polylongifoliaons A (4) and B (5), together with nine known diterpenes, were isolated from the unripe fruits of Polyalthia longifolia var. pendula. The structures of the new isolates were determined by extensive spectroscopic analysis. The effects of all the isolated compounds on the viability of human neuroblastoma SK-N-MC cells under β-amyloid (Aβ)-induced neurotoxicity were evaluated. Polylongifoliaic A (1), polylongifoliaon B (5) and 8–10 improved the viability of human neuroblastoma cells (SK-N-MC cells) under Aβ-induced neurotoxicity. Among the active compounds, polylongifoliaic A (1) and polylongifoliaon B (5) exhibited the most potent activity toward SK-N-MC cells with IC₅₀ values of 1.64 μM and 3.75 μM, respectively. In addition, the isolated diterpenes were found to possess potent promising acetylcholinesterase inhibitory activity, which was revealed by the TLC bioautographic assay.

---

Introduction

Polyalthia longifolia var. pendula known as “Indian Mast Tree” is an evergreen plant that is native to India.¹ It is cultivated in the tropical and subtropical regions of the world as ornamental or avenue trees. Since antiquity, it has been used by traditional healers in Asian folk medicine. Its name gives a clear indication of such therapeutic potential. Polyalthia is the Greek word for poly, meaning much or many, and althia from àltheo, meaning to cure.² The plant has been used as an anthelmintic and germicide, and in the treatment of pyrexia, skin diseases, diabetes and helminthiasis. Its methanolic extract showed potent activity in regulating high blood pressure.³ A series of clerodane diterpenoids were isolated from this plant with a wide range of biological activities such as cytotoxic,^4,5 anti-bacterial,^6,7 anti-fungal⁶ and anti-inflammatory activities.^8,9 Interestingly, recent studies have reported the potent neuroprotective activity of certain clerodane diterpenoids, demonstrating their potentiating effect on nerve growth factor (NGF) activity¹⁰ and their inhibitory effect on Aβ fibril accumulation.¹¹ However, the effect of this plant or its active secondary metabolites on acetylcholinesterase (AChE) has never been studied. AChE not only plays an important role in ailments related to cholinergic dysfunction, but can also accelerate the aggregation of Aβ into amyloid fibril, acting as an amyloid promoting factor.^12–14 In the present study, TLC bioautographic assay was utilized to evaluate the AChE inhibitory activity of different extracts of the unripe P. longifolia var. pendula fruits at a concentration of 10 μg mL⁻¹ (Fig. S14†). The preliminary bioassay results revealed the potent inhibitory effect of the methanolic extract on AChE activity. This active extract was further purified, resulting in the isolation of three new diterpenes, polylongifoliaic A (1), polylongifoliaons A (4) and B (5), together with nine known diterpenes (2, 3, 6–12). The structural elucidation of the new isolates was achieved by an extensive analysis of their NMR and mass spectroscopic data. The relative configurations of polylongifoliaons A (4) and B (5) were determined by CD experiments. The neuroprotective activity of the isolated compounds was evaluated by examining their effect on AChE activity and on the viability of human neuroblastoma SK-N-MC cells under Aβ-insult.

Results and discussion

The methanolic extract of unripe P. longifolia var. pendula fruits was fractionated using different organic solvents. Bioassay-guided fractionation utilizing the TLC bioautographic assay,¹⁵ for detecting the AChE inhibitory activity, indicated that the n-hexane layer is the most active layer and was subjected to further purification. From this layer, three new diterpenes, including one (4 → 2)-abeo-clerodane diterpene: polylongifoliaic A (1) and two 2-oxo-clerodane diterpenes: polylongifoliaons A (4) and B (5), along with nine known compounds, including: 3β,5β-dihydroxy-16α-methoxy-halima-13Z-en-15,16-olide (2),¹⁶ 3β,5β-dihydroxy-16β-methoxy-halima-13Z-en-15,16-olide (3),¹⁶ (4 → 2)-abeo-16-hydroxy-cleroda-2(4)E,13Z-dien-15,16-olide-2-al (6),¹⁷ 3α,16-dihydroxy-cleroda-4(18),13Z-dien-15,16-olide (7),¹⁸ 3β,16-dihydroxy-cleroda-4(18),13Z-dien-15,16-olide (8),¹⁹ 16(R&S)-hydroxy-3,13Z-kolavadien-15,16-olide-2-one (9),¹⁷ 2-oxokolavenic acid (10),^20,21 16-methoxy-cleroda-3,13Z-dien-15,16-olide (11)²² and 16-oxo-cleroda-3Z,13E-dien-15-oic acid methyl ester (12),²³ were isolated (Fig. 1). Among the known isolates, this is the first report of the isolation of 11 from a natural source.

Fig. 1 Isolated compounds from Polyalthia longifolia var. pendula.

Polylongifoliaic A (1) was isolated as a white wax. The molecular formula of 1 was established as C₂₀H₂₈O₅ by ¹³C NMR and HRFABMS data, representing seven indices of hydrogen deficiency (IHD). The UV absorption band at 216 nm and the IR absorption bands at 1749 and 1685 cm⁻¹ suggested the presence of α,β-unsaturated γ-lactone and α,β-unsaturated carboxylic acid, respectively. The ¹³C NMR experiment revealed the presence of 20 carbons including four methyls, five methylenes, three methines and eight quaternary carbons (Table S1†). The ¹H NMR spectrum indicated the presence of four methyl groups, including one olefinic methyl at δ_H 1.97 (3H, s, CH₃-18), two tertiary methyl groups with overlapping signals at δ_H 0.93 (6H, s, CH₃-19/CH₃-20) and a secondary methyl group at δ_H 0.86 (3H, d, J = 6.5 Hz, CH₃-17) (Table S1†). Two signals at δ_H 6.04 (1H, s, H-16) and 5.90 (1H, s, H-14), which are characteristic to the γ-hydroxyl α,β-unsaturated-γ-lactone moiety, were also detected, supporting the data from the UV and IR experiments (Table S1†). Comparing the spectral data of 1 with that of other known diterpenes, such as (4 → 2)-abeo-16-hydroxy-cleroda-2,13Z-dien-15,16-olide-2-al (6),¹⁷ revealed that 1 possesses a clerodane-type skeleton with a (4 → 2) rearranged ring A. This suggestion was further supported by HMBC correlations between δ_H 2.33 and 2.27 (m, 2H, H₂-1) to δ_C 128.9 (C-2), 165.2 (C-4), 51.5 (C-5) and 55.5 (C-10) (Fig. 2). In addition, the ¹³C NMR data revealed that the aldehyde group (δ_C 189.3) attached to C-3 in 6 is replaced by a carboxylic group in 1 (δ_C 173.7). The A/B ring junction in 1 was deduced to be trans because of the lack of a NOESY correlation between δ_H 0.93 (CH₃-19) and δ_H 1.67 (H-10). The key HMBC and COSY correlations are shown in Fig. 2. Therefore, the structure of 1 was assigned to be (4 → 2)-abeo-16-hydroxy-cleroda-2(4)E,13Z-dien-15,16-olide-2-oic acid.

Fig. 2 Selected HMBC and COSY correlations of 1–3.

Compounds 2 and 3 were isolated as a pair of diterpene enantiomers, and their stereochemistry was deduced by utilizing simulations of circular dichroism spectra by a time-dependent density functional theory.¹⁶ The isolation of these compounds (2 and 3) was previously reported but spectral data was not shown.¹⁶ Compound 2 was obtained as a white wax. Its molecular formula was calculated as C₂₁H₃₄O₅ from the analysis of its positive HRFABMS data. The UV and IR absorption bands of 2 indicated the presence of a hydroxyl (IR: 3483 cm⁻¹) and α,β-unsaturated lactone (UV: λ_max 207 nm; IR: 1758 cm⁻¹) functionalities. The ¹H and ¹³C NMR signals of 2 (Table S1†) indicated the presence of four methyl protons [δ_H 1.27 (3H, s, CH₃-18), 1.14 (3H, s, CH₃-19), 0.79 (3H, s, CH₃-20), 0.78 (3H, m, CH₃-17)], one methoxy [δ_H 3.57 (3H, s, CH₃-16)] and one oxymethine proton [δ_H 3.59 (1H, t, J = 2.5 Hz, H-3)]. The presence of the α,β-unsaturated lactone was supported by two characteristic proton signals [δ_H 5.87 (1H, s, H-14), 5.65 (1H, s, H-16)], as well as by typical carbon resonances [δ_C 104.4 (C-16), 170.8 (C-15), 117.6 (C-14), 168.5 (C-13)]. HMBC correlations [δ_H 3.59 (H-3)/δ_C 76.3 (C-5); δ_H 1.27 (H₃-18), 1.14 (H₃-19)/δ_C 76.3 (C-5), 76.3 (C-3), 41.3 (C-4); δ_H 1.75 (H-10β)/δ_C 76.3 (C-5)] confirmed the location of the quaternary hydroxylated carbon at C-5, which is next to the quaternary C-4 carbon with the geminal methyl groups (Fig. 2). The detailed analysis of the NMR spectral data of 2 revealed similarity to those of halimane-type diterpene, such as 3β,5β,16α-trihydroxy-halima-13Z-en-15,16-olide, except for an additional methoxy group in 2.⁵ The methoxy group was assigned to C-16 based on the key HMBC correlation between δ_H 3.57 (CH₃-16) and δ_C 104.4 (C-16). To determine the stereochemistry of compound 2, its NOESY and CD spectra were analyzed (Fig. S1† and 3). The NOESY correlation between δ_H 1.14 (CH₃-19) and 0.79 (CH₃-20) confirmed the A/B ring junction of this bicyclic diterpene as cis, suggesting a β-orientation for H-10 and 5-OH (Fig. S1†). In addition, a smaller coupling constant and the NOESY correlations (Fig. S1†) confirmed the β-orientation of 3-OH. The CD spectrum of 2 (Fig. 3) demonstrated a positive Cotton effect due to n–π* transition (235–250 nm) and a negative Cotton effect due π–π* (200–220 nm) transition, suggesting an S configuration at C-16.¹⁶ Thus, the structure of 2 was assigned to be 3β,5β-dihydroxy-16α-methoxy-halima-13Z-en-15,16-olide.

Fig. 3 The CD spectra of 2, 3, 4, and 5.

Compound 3 was isolated as a white wax and was assigned the molecular formula of C₂₁H₃₄O₅ based on its HRFABMS data. The UV, IR, 1D and 2D NMR spectroscopic data of 3 were identical to those of 2 (Table S1†), except for the CD spectrum (Fig. 3). The detected negative Cotton effect due to n–π* transition (235–250 nm) and the positive Cotton effect due to π–π* transition (200–220 nm) were opposite to those of 2, suggesting that 2 and 3 are epimers with 3 possessing the R configuration at C-16 (Fig. 3).¹⁶ Therefore, the structure of 3 was assigned to be 3β,5β-dihydroxy-16β-methoxy-halima-13Z-en-15,16-olide. The experimental CD spectra of 2 and 3 were in agreement with the ab initio calculations of CD spectra using Gaussian TDDFT (B3LYP) based on 6-31+G(d) and 6-311++G(d,p).¹⁶

Both polylongifoliaons A (4) and B (5) were isolated as white waxes. They had the same molecular formula (C₂₁H₃₀O₄) as revealed by their HRFABMS data, which indicated seven IHD. The IR spectra of both 4 and 5 showed characteristic absorption bands, suggesting the presence of a hydroxyl (3449 cm⁻¹), α,β-unsaturated γ-lactone (1757 cm⁻¹) and conjugated carbonyl moieties (1681 cm⁻¹). The ¹H and ¹³C NMR spectra of 4 showed signals attributable to a clerodane diterpene with an α,β-unsaturated γ-lactone, including four methyl groups [δ_H 1.95 (3H, d, J = 1.0 Hz, CH₃-18)/δ_C 19.3, δ_H 1.18 (3H, s, CH₃-19)/δ_C 18.6, δ_H 0.90 (3H, s, CH₃-20)/δ_C 18.1, δ_H 0.87 (3H, d, J = 6.2 Hz, CH₃-18)/δ_C 16.0] and olefinic and oxymethine protons [δ_H 5.98 (1H, s, H-14)/δ_C 118.6, δ_H 5.82 (1H, s, H-16)/δ_C 106.3] with three quaternary carbons [δ_C 173.0, 170.4] (Table S2†). A proton signal [δ_H 5.72 (1H, s, H-3), a tertiary carbon [δ_C 125.8 (C-3)] and two quaternary carbons [δ_C 202.8 (C-2), 176.4, (C-4)] suggested that the C-2 position in 4 should be oxidized, which was further supported by the HMBC correlations [δ_H 5.72 (H-3)/δ_C 41.3 (C-5), 19.2 (C-18); δ_H 2.48, 2.29 (H₂-1)/δ_C 202.8 (C-2), 47.0 (C-10)]. The detailed analysis of the NMR spectral data of 4 revealed its similarity to 16(R&S)-hydroxy-3,13Z-kolavadien-15,16-olide-2-one (9),¹⁷ except for an additional methoxy group [δ_H 3.55 (3H, s, CH₃-16) and δ_C 57.6]. The HMBC correlation between δ_H 5.82 (H-16) and δ_C 57.6 (CH₃-16) further confirmed the location of the methoxy group at C-16 in 4, instead of the hydroxyl group (16-OH) in 9.

The 1D and 2D NMR spectra of 4 and 5 were almost identical (Table S2†), but they showed different CD absorption (Fig. 3). Compound 4 displayed a positive Cotton effect at 291 (shoulder) and 240 nm due to n–π* excitation and a negative Cotton effect at 214 nm due to π–π* transition. On the other hand, 5 showed a positive Cotton effect at 291 (shoulder) and 240 (weak) nm for n–π* band and a strong negative Cotton effect at 203 nm for π–π* transition. By comparing these data with the CD data of known butenolides and applying the octant rule,²⁴ the orientation of the methyl groups at C-19 in 4 and 5 were suggested as α. However, the orientation of the methoxy group at C-16 in 4 was assigned as α, and as β in 5.¹⁶ Finally, the structures of 4 and 5 were elucidated to be 16α-methoxy-3,13Z-kolavadien-15,16-olide-2-one and 16β-methoxy-3,13Z-kolavadien-15,16-olide-2-one, respectively.

The nine known compounds were identified by comparing their UV, IR, ¹H NMR, ¹³C NMR and MS data with those reported in the literature.

The recent findings on the neuroprotective activity of certain clerodane diterpenoids have encouraged us to evaluate the activity of the isolated compounds on the viability of human neuroblastoma SK-N-MC cells under Aβ-induced neurotoxicity (Fig. 4). The results indicated that the isolates did not exhibit cytotoxicity toward SK-N-MC cells except for 6 and 12, which were derivatives with an aldehyde group. In Fig. 4, 1 and 5 showed significant protective activity (65–66%) against Aβ_1–42 insult at 10 μM. These compounds, 1 and 5, demonstrated their neuroprotective effect in a dose dependent manner with IC₅₀ values of 1.64 μM and 3.75 μM, respectively (Fig. 5). On the other hand, compounds 8–10 exhibited moderate protection (37.7–42%) (Fig. 4). The activity was comparable to epigallocatechin gallate (EGCG) (48.5%), which is a well-known neuroprotective agent. Several studies have demonstrated the neuroprotective activity of EGCG and revealed its mechanism.²⁵

Fig. 4 The effect of the selected diterpenes isolated from P. longifolia var. pedula on SK-N-MC cell viability. Results are presented as mean ± S.E.M. (n = 3).

Fig. 5 The effects of 1 and 5 in different dosages were tested on SK-N-MC cell viability.

To further understand the activity of the isolated diterpenoids, we sought to study the structure activity relationship (SAR) of these compounds. It was found that the substituent nature and its stereoconfiguration at C-16 in the 2-furanone ring significantly affects the neuroprotective activity of the isolated diterpenes. Compound 8 with an α hydroxyl group (electron withdrawing group) at C-16 exhibited more potent activity compared to 7 with a β hydroxyl group. On the other hand, 5 with a β methoxy group (electron donating group) was more active than 4 with an α methoxy group.

In addition, the inhibitory effect of the isolated compounds on AChE was examined. Compounds 1, 4, 5, 6, 9, 10 and 11 exhibited potent inhibitory effects on acetylcholinesterase (AChE) at a concentration of 20 μg mL⁻¹ (Fig. S15†). The effect of the active compounds on AChE was comparable to galantamine, a marketed second generation acetylcholinesterase inhibitor.²⁶ It is known that the accumulation of beta-amyloid peptides and the reduction in the activity of the cholinergic neurons are one of the well-known features of Alzheimer's disease (AD). However, no medication has been discovered to effectively delay or halt the progression of AD.²⁷ The potent neuroprotective activity of the isolated compounds warrants further investigation to determine the mechanism of action of Polyalthia longifolia diterpenes.

Experimental

Materials and methods

General experimental procedures. Melting points were measured on a Yanaco MP-500D melting point apparatus (Yanaco, Kyoto, Japan) and were used uncorrected. Optical rotations were measured on a JASCO P-1020 polarimeter (JASCO, Tokyo, Japan). UV spectra were recorded on a Hitachi U-2800 UV-Vis spectrophotometer (Hitachi, Tokyo, Japan). IR spectra were recorded on a Shimadzu IR Prestige-21 FT-IR spectrometer (Shimadzu, Nakagyo-ku, Japan). CD spectra were obtained on a JASCO J-715 spectropolarimeter (Bruker BioSpin GmbH, Karlsruhe, Germany). 1D and 2D NMR spectra were recorded with Bruker 400 AV, Bruker 500 AVII, and Varian Unity 600 NMR spectrometers (Bruker Daltonics, Bremen, Germany). HRFABMS data were measured with a Finnigan/Thermo Quest MAT 95XL spectrometer (Finnigan MAT LCQ, San Jose, CA, USA) and ESI-MS/MS data were obtained on a Bruker HCT ultra PTM Discovery system (Bruker Daltonics, Bremen, Germany). Sephadex LH-20 (Amersham Biosciences, Uppsala, Sweden) and Silica gel 60 (230–400 mesh or 70–230 mesh, Merck, Darmstadt, Germany) were used for column chromatography, and precoated Si gel plates (silica gel 60 F₂₅₄, Merck, Darmstadt, Germany) were used for analytical TLC. The spots were detected by spraying 50% H₂SO₄ aqueous solution followed by heating on a hot plate. HPLC was performed on a Hitachi L-2130 pump equipped with a Hitachi L-2420 UV-Vis detector (Hitachi, Tokyo, Japan). Discovery® HS C₁₈ (5 μm, 250 × 4.6 mm i.d., Supelco, Bellefonte, PA, USA) and semi-preparative Discovery® HS C₁₈ (5 μm, 250 × 10 mm i.d., Supelco, Bellefonte, PA, USA) columns were applied for analytical and preparative purposes, respectively.

Reagent and cell cultures. Amyloid β_1–42 was purchased from Sigma (München, Germany). The human neuroblastoma SK-N-MC cell line (American Type Culture Collection) was cultured in minimum essential medium (MEM), supplemented with penicillin (50 IU ml⁻¹), streptomycin (50 μg ml⁻¹), non-essential amino acids, sodium pyruvate (1 mM) and 10% heat-inactivated fetal bovine serum. Cell culture was performed according to the manufacturer's instructions.

The cell viability assay in SK-N-MC cells under Aβ-induced neurotoxicity. For the cell viability assay, cells were cultured in 24-well culture dishes, seeded at 2 × 10⁵ cells per well, and maintained in a logarithmic growth phase. The cells were treated with the indicated compounds (10 μM) with or without amyloid β_1–42 peptide (5 μM) in the culture medium. After incubation for 24 h, cell viability was measured by the modified 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) assay method using MEM without phenol red. The absorbance of the converted dye was measured under the wavelength of 550 nm with a microplate reader. All the tested compounds were re-purified by reversed-phase HPLC before the bioassay test (purity > 99%). All the results were expressed as mean ± S.E.M. from three different independent experiments (n = 3). Student's t-test was performed for statistical analyses; a value of P < 0.05 was considered statistically significant.

Plant material. The unripe fruits of P. longifolia var. pendula (500 g) were collected from Kaohsiung City, Taiwan, in September 2005. A voucher specimen (PLP-F) was deposited in the Department of Marine Biotechnology and Resource, National Sun Yat-sen University, Kaohsiung 804, Taiwan.

Extraction and isolation. The unripe fruits of P. longifolia var. pendula (500 g) were extracted with methanol (4 L × 4 times). After removing the solvent, the MeOH extract (27.9 g) was partitioned with n-hexane and water to yield n-hexane (8.0 g) and aqueous layers. The n-hexane layer was further separated into four fractions (Ha–Hd) by column chromatography (CC) on silica gel with n-hexane-CHCl₃ and CHCl₃–MeOH as an eluent. Fraction Hb (2.5 g) was applied on a silica gel column eluted with a solvent mixture of increasing polarity (n-hexane-CHCl₃ and CHCl₃–MeOH) to yield 30 fractions (Hb-1 ∼ Hb-30). Fraction Hb-17 (254.0 mg) was purified by reversed-phase HPLC (75 : 25 MeOH : H₂O + 0.05% TFA) to obtain 9 (3.4 mg, t_R 12.4 min), 6 (15.0 mg, t_R 16.0 min) and 1 (4.0 mg, t_R 19.5 min). Fraction Hb-18 (210.0 mg) was separated by reversed-phase HPLC (75 : 25 MeOH : H₂O + 0.05% TFA) to obtain 10 (4.8 mg, t_R 18.5 min). Compounds 4, 5 and 11 were isolated from the subfraction Hb-20 (139.0 mg) by reversed-phase HPLC eluted with 75 : 25 MeOH–H₂O + 0.05% TFA. Fraction Hb-21 (93.0 mg) was separated by reversed-phase HPLC (70 : 25 MeOH–H₂O + 0.05% TFA) to obtain 7 (1.0 mg, t_R 13.6 min) and 8 (2.4 mg, t_R 15.0 min). Fraction Hb-22 (59.0 mg) was separated by reversed-phase HPLC (73 : 27 MeOH : H₂O + 0.05% TFA) to obtain 3 (4.0 mg, t_R 29.0 min) and 2 (2.4 mg, t_R 34.0 min).

Characterization of compounds.
Polylongifoliaic A (1). White wax; [α]_D²⁵ + 4.0 (c 0.10, CHCl₃); UV (MeOH) λ_max (log ε) 216 (2.82) nm; IR (KBr) ν_max: 2961, 2927, 1749, 1685, 1642, 1405, 1384 cm⁻¹; ¹H NMR (CD₃OD, 400 MHz) and ¹³C NMR (CD₃OD, 100 MHz), see Table S1;† (+)-ESIMS m/z 371.2 [M + Na]⁺ (100), 349.3 [M + H]⁺ (12.6); (−)-ESIMS m/z 695.4 [2M − H]⁻ (17.1), 461.2 [M + TFA-H]⁻ (100), 347.3 [M − H]⁻ (32.7); HRFABMS m/z 349.4401 [M + H]⁺ (calcd for C₂₀H₃₁O₄⁺, 349.4407).
Polylongifoliaon A (4). White wax; [α]_D²⁵ − 8.6 (c 1.0, CHCl₃); UV (MeOH) λ_max (log ε) 236.0 (3.40), 216.0 (3.50) nm; IR (KBr) ν_max: 2959, 2933, 2876, 1763, 1653, 1448, 1383, 1118 cm⁻¹; CD (c 0.30 × 10⁻⁴ M, MeOH) Δε (nm): 6.29 (244.5), −4.18 (216.0), 3.80 (207.5); ¹H NMR (CD₃OD, 600 MHz) and ¹³C NMR (CD₃OD, 150 MHz), see Table S2;† (+)-ESIMS m/z 715.4 [2M + Na]⁺ (50.7), 369.2 [M + Na]⁺ (100), 347.2 [M + H]⁺ (0.2); (−)-ESIMS m/z 345.3 [M − H]⁻ (100); HRFABMS m/z 347.2216 [M + H]⁺ (calcd for C₂₁H₃₁O₄⁺, 347.2222).
Polylongifoliaon B (5). White wax; [α]_D²⁵ − 18.0 (c 1.0, CHCl₃); UV (MeOH) λ_max (log ε) 236 (3.21), 215 (3.30) nm; IR (KBr) ν_max: 2963, 2938, 1788, 1761, 1668, 1652, 1464, 1377, 1118 cm⁻¹; CD (c 0.30 × 10⁻⁴ M, MeOH) Δε (nm): 8.27 (238.0), −15.70 (203.0); ¹H NMR (CD₃OD, 500 MHz) and ¹³C NMR (CD₃OD, 125 MHz), see Table S2;† (+)-ESIMS m/z 715.4 [2M + Na]⁺ (49.3), 369.2 [M + Na]⁺ (100), 347.2 [M + H]⁺ (0.8); (−)-ESIMS m/z 345.3 [M − H]⁻ (73.4); m/z 346.2; HRFABMS m/z 347.2217 [M + H]⁺ (calcd for C₂₁H₃₁O₄⁺, 347.2222).

Conclusions

In summary, three new diterpenes polylongifoliaic A (1), polylongifoliaons A (4) and B (5), together with nine known diterpenes (2, 3, 6–12), were isolated from the methanolic extract of the unripe Polyalthia longifolia var. pendula fruits. Note that, this is the first report that extracts active secondary metabolites from this species, and exert acetylcholinesterase (AChE) activity and protective effects in human neuroblastoma cells (SK-N-MC cells) under Aβ-induced neurotoxicity. Among the separated metabolites, polylongifoliaic A (1) and polylongifoliaon B (5) exhibited the most potent activity toward SK-N-MC cells with IC₅₀ values of 1.64 μM and 3.75 μM, respectively. The in vivo neuroprotective activities of the active metabolites is under investigation in our laboratory and will be reported in due course.

Acknowledgements

This work was supported in part by grants from the National Science Council, Taiwan (NSC 102-2320-B-110-002) awarded to C.-C. L. We are grateful to the National Center for High-Performance Computing for computer time and facilities (Hsin-Chu), and the Proteomics Research Core Laboratory, Office of Research and Development at China Medical University for ESIMS measurements.

Notes and references

1. P. A. Kavitha, K. Pawan, M. T. P. Narasimah and S. M. Gopinath, Int. J. Latest Res. Sci. Technol., 2013, 2, 508–510.
2. K. V. Katkar, A. C. Suthar and V. S. Chauhan, Pharmacogn. Rev., 2010, 4, 62–68.
3. R. Saleem, M. Ahmed, S. I. Ahmed, M. Azeem, R. A. Khan, N. Rasool, H. Saleem, F. Noor and S. Faizi, Phytother. Res., 2005, 19, 881–884.
4. G. X. Zhao, J. H. Jung, D. L. Smith, K. V. Wood and J. L. McLaughlin, Planta Med., 1991, 57, 380–383.
5. C. Y. Chen, F. R. Chang, Y. C. Shih, T. J. Hsieh, Y. C. Chia, H. Y. Tseng, H. C. Chen, S. J. Chen, M. C. Hsu and Y. C. Wu, J. Nat. Prod., 2000, 63, 1475–1478.
6. M. Marthanda Murthy, M. Subramanyam, M. Hima Bindu and J. Annapurna, Fitoterapia, 2005, 76, 336–339.
7. K. V. Sashidhara, S. P. Singh and P. K. Shukla, Nat. Prod. Commun., 2009, 4, 327–330.
8. F. R. Chang, T. L. Hwang, Y. L. Yang, C. E. Li, C. C. Wu, H. H. Issa, W. B. Hsieh and Y. C. Wu, Planta Med., 2006, 72, 1344–1347.
9. T. H. Wu, Y. Y. Cheng, C. J. Chen, L. T. Ng, L. C. Chou, L. J. Huang, Y. H. Chen, S. C. Kuo, M. El-Shazly, Y. C. Wu, F. R. Chang and C. C. Liaw, Molecules, 2014, 19, 2049–2060.
10. W. Tang, H. Hioki, K. Harada, M. Kubo and Y. Fukuyama, J. Nat. Prod., 2008, 71, 1760–1763.
11. C. C. Chen, Y. J. Shiao, R. D. Lin, Y. Y. Shao, M. N. Lai, C. C. Lin, L. T. Ng and Y. H. Kuo, J. Nat. Prod., 2006, 69, 689–691.
12. N. C. Inestrosa, A. Alvarez, C. A. Perez, R. D. Moreno, M. Vicente, C. Linker, O. I. Casanueva, C. Soto and J. Garrido, Neuron, 1996, 16, 881–891.
13. N. C. Inestrosa, A. Alvarez and F. Calderon, Mol. Psychiatry, 1996, 1, 359–361.
14. H. Y. Zhang, H. Yan and X. C. Tang, Cell. Mol. Neurobiol., 2008, 28, 173–183.
15. A. Marston, J. Kissling and K. Hostettmann, Phytochem. Anal., 2002, 13, 51–54.
16. C. C. Liaw, J. L. Chang, S. F. Chen, J. H. Huang, J. F. Sie and Y. Y. Cheng, Chem. Phys. Lett., 2011, 517, 51–54.
17. A. Kijjoa, M. M. M. Pinto, P. M. M. Pinho, B. Tantisewie and W. Herz, Phytochemistry, 1990, 29, 653–655.
18. K. V. Sashidhara, S. P. Singh, J. Sarkar and S. Sinha, Nat. Prod. Res., 2010, 24, 1687–1694.
19. X. Ma, I. S. Lee, H. B. Chai, K. Zaw, N. R. Farnsworth, D. D. Soejarto, G. A. Cordell, J. M. Pezzuto and A. D. Kinghorn, Phytochemistry, 1994, 37, 1659–1662.
20. C. M. Hasan, T. M. Healey and P. G. Waterman, Phytochemistry, 1982, 21, 1365–1368.
21. R. Aquino, M. L. Ciavatta, N. De Tommasi and E. Gàcs-Baitz, Phytochemistry, 1992, 31, 1823–1825.
22. M. Chakrabarty and A. C. Nath, J. Nat. Prod., 1992, 55, 256–258.
23. A. P. Phadnis, S. A. Patwardhan, N. N. Dhaneshwar, S. S. Tavale and T. N. Guru Row, Phytochemistry, 1988, 27, 2899–2901.
24. W. Moffitt, R. B. Woodward, A. Moscowitz, W. Klyne and C. Djerassi, J. Am. Chem. Soc., 1961, 83, 4013–4018.
25. C. L. Lin, T. F. Chen, M. J. Chiu, T. D. Way and J. K. Lin, Neurobiol. Aging, 2009, 30, 81–92.
26. E. Scarpini, P. Scheltens and H. Feldman, Lancet Neurol., 2003, 2, 539–547.
27. C. Reitz, C. Brayne and R. Mayeux, Nat. Rev. Neurol., 2011, 7, 137–152.

---

Footnotes

† Electronic supplementary information (ESI) available: full details of TLC bioautographic assay, the 1D and 2D spectroscopic data of compounds 1–5 and 11. See DOI: 10.1039/c4ra01879f

‡ These authors contributed equally to this work.

---