Phenolic Constituents of Chinese Quince Twigs (Chaenomeles sinensis Koehne) and Their Anti-Neuroinflammatory, Neurotrophic, and Cytotoxic Activities

Abstract

Chaenomeles sinensis has been used as a food and traditional medicines. However, most of research on discovering bioactive constituents from this plant have been focused on its yellow fruit, Chinese quince, due to its wide usage. Here, we isolated and characterized three new phenolic compounds (1, 9, and 11) and 21 known compounds (2−8, 10, and 12−24) from the twigs of C. sinensis. Their chemical structures were established by spectroscopic and spectrometric data analysis including 1D and 2D NMR, high-resolution mass spectrometry (HRMS), electronic circular dichroism (ECD), and LC-MS analysis. Some of the isolated compounds (1−24) showed anti-neuroinflammatory effects on nitric oxide (NO) production in lipopolysaccharide (LPS)-activated BV-2 cells, neurotrophic activity in C6 cells through the secretion of nerve growth factor (NGF) and/or cytotoxicity against four human cancer cell lines (A549, SK-OV-3, SK-MEL-2, MKN-1).

1. Introduction

Phenolic compounds are one of the largest natural products groups, especially in the plant kingdom, and are characterized by the presence of at least a benzene ring with hydroxy group(s) attached to it [1]. They can be classified into several subgroups according to the number of carbon atoms and their arrangement as follows: phenolic acids (C₆–C₁), acetophenones and phenylacetic acids (C₆–C₂), phenylpropanoids (C₆–C₃), naphthoquinones (C₆–C₄), xanthones (C₆–C₁–C₆), stilbenoids (C₆–C₂–C₆), flavonoids (C₆–C₃–C₆), and lignans (C₆–C₃–C₃–C₆) [1,2]. Many of these well-known natural products in the phenolic compounds class have been reported to possess diverse pharmacological activities including anti-inflammatory (e.g., salicylic acid), neurotrophic (e.g., 6-shogaol) [2,3], antioxidant (e.g., resveratrol, quercetin) [2], and anticancer (e.g., podophyllotoxin) [2] effects.

Neurodegenerative diseases such as dementia and Parkinson’s disease are exploding due to rapid aging populations [4]. Despite various efforts to overcome these neurodegenerative diseases, there are not sufficient reports on drugs that modify the disease [5]. Therefore, it is very important to draw meaningful leading compounds from natural resources in order to regulate neurodegenerative diseases. As neurodegenerative diseases are highly complex and incurable things, there are some aspects that make it difficult to overcome diseases with a single target such as amyloid beta [5]. Thus, we should think about the study of exploring phytochemicals that overcome neurodegenerative diseases by regulating multiple targets. In this study, the effects of phytochemicals with anti-neuroinflammation and neurotrophic factor potentiation will be shown. They are important multi-targets that cause the induction and progression of neurodegenerative diseases.

Chaenomeles sinensis Koehne (or Pseudocydonia sinensis, Chinese Quince, Rosaceae) is a semi-evergreen tree in the family Rosaceae and is widely distributed in East Asia, including Korea, mainland China, and Japan. The yellow fruit of this plant has been consumed as a tea, in a candied form, or in liquor [6], and also used as a Korean traditional medicine over 400 years to treat myalgia, beriberi, vomiting, and diarrhea [7]. Due to the wide range of usages of C. sinensis fruit, the phytochemical and biological research on this plant has been mainly focused on the fruit [8,9,10,11,12,13] rather than the other parts such as twigs [14,15] or leaves. Based on these previous studies, it was found that the major compound classes of C. sinensis were triterpenoids [9,14,15,16] and flavonoids [8,13,15]. However, minor components such as non-flavonoid phenolic compounds in the C. sinensis twigs with anti-neurodegerative effects remain largely unknown. In this study, 24 phenolic compounds (1–24) including three new compounds (1, 9, and 11) were isolated and characterized from the twigs of C. sinensis (Figure 1). Their structures were elucidated by conventional NMR techniques coupled with simulation of ¹H NMR peak, analysis of HRMS and ECD data, and hydrolysis followed by LC-MS analysis. The isolates (1−24) were tested for their anti-neuroinflammatory, neurotrophic, and cytotoxic activities, and herein, we report the isolation and structure elucidation of phenolic phytochemicals from the twigs of C. sinensis as well as their biological activities.

2. Materials and Methods

2.1. General Experimental Procedures

JASCO P-1020 polarimeter equipped with the sodium D line (590 nm) (JASCO, Easton, MD, USA) was performed to measure specific rotations. Bruker AVANCE III 700 NMR spectrometer at 700 MHz (¹H) and 175 MHz (¹³C) was carried out to measure the NMR spectra (¹H, ¹³C, COSY, HSQC, and HMBC) at 700 MHz (¹H) and 175 MHz (¹³C) with chemical shifts given in ppm (δ) (Bruker, Karlsruhe, Germany). High-resolution fast atom bombardment mass spectroscopy (HRFABMS) were measured on either a Waters SYNAPT G2 (Milford, MA, USA) or a JEOL JMS700 mass spectrometer (Tokyo, Japan). LC-MS data were measured using an Agilent 1290 Infinity II HPLC instrument (Foster City, CA, USA) coupled to a G6545B quadrupole time-of-flight (Q-TOF) mass spectrometer (Agilent Technologies) with a Kinetex C₁₈ 5 µm column (250 mm length × 4.6 mm i.d.; Phenomenex, Torrance, CA, USA). The semipreparative high-performance liquid chromatography (HPLC) furnished with a Gilson 306 pump (Middleton, WI, USA) and a Shodex refractive index detector (New York, NY, USA) was used. Column chromatography was performed employing either silica gel 60 (70−230 and 230−400 mesh; Merck, Darmstadt, Germany) or RP-C₁₈ silica gel (Merck, 230−400 mesh). Sephadex LH-20 (Pharmacia, Uppsala, Sweden) was utilized for molecular sieve column chromatography. Merck precoated silica gel F₂₅₄ plates and RP-C₁₈ F_254s plates (Merck, Darmstadt, Germany) were used for thin-layer chromatography (TLC) analysis. Spots were detected on TLC under UV light or by heating after spraying with anisaldehyde−sulfuric acid.

2.2. Plant Material

Twigs of two-year-old C. sinensis (7.0 kg) were purchased at Yangjae Flower Market in Seoul, Korea, in January 2012. A voucher specimen of the plants (SKKU-NPL 1206) was authenticated by Prof. Kang Ro Lee and deposited in the herbarium of the School of Pharmacy, Sungkyunkwan University, Suwon, Korea.

2.3. Extraction and Isolation

Twigs of C. sinensis (7.0 kg) were extracted with 80% aqueous MeOH (each 10 L × 1 day, 3 times) under reflux and filtered. The filtrate was evaporated in vacuo to yield a MeOH extract (320 g). The 80% MeOH extract was suspended in distilled water and successively partitioned with n-hexane (3 g), CHCl₃ (15 g), EtOAc (6 g), and n-BuOH (30 g). Repeated column chromatographical fractionation and isolation of CHCl₃, EtOAc, and n-BuOH soluble layers afforded the compounds 1–24. (See Supplementary Materials for the detailed isolation process).

(7S,8R)-4,7,9,9′-tetrahydroxy-3,5,3′,5′-tetramethoxy-7′-oxo-8-O-4′-neolignan (1). Colorless gum; [α]${}_{\mathrm{D}}^{25}$−12 (c 0.1, MeOH); ECD (MeOH) λ_max (Δε) 244 (−2.3) nm; ¹H and ¹³C NMR data, see

Table 1. ¹H (700 MHz) and ¹³C (175 MHz) NMR data of compounds 1–3, and 9.

Pos.	1 1		2 1		3 1		9 2
	δC	δH (multi, J in Hz)	δC	δH (multi, J in Hz)	δC	δH (multi, J in Hz)	δC	δH (multi, J in Hz)
1	130.3				130.6		134.3
2	102.7	6.58 (s)		6.97 (d, 1.6)	102.6	6.59 (s)	105.1	6.75 (s)
3	147.2				147.2		149.5
4	134.2				134.0		136.4
5	147.2			6.86 (d, 8.1)	147.2		149.5
6	102.7	6.58 (s)		6.75 (dd, 8.1, 1.6)	102.6	6.59 (s)	105.1	6.75 (s)
7	73.1	4.99 (brs)	72.6	5.00 (d, 3.5)	72.7	4.99 (d, 3.2)	85.2	5.03 (d, 8.4)
8	87.5	4.23 (ddd, 6.6, 3.5, 3.0)	87.2	4.11 (dt, 6.7, 2.9)	87.2	4.10 (dt, 6.8, 3.1)	51.8	2.51 (m)
9a	60.7	3.93 (overlap)	60.8	3.87 (overlap)	60.8	3.87 (overlap)	67.8	3.82 (m)
9b		3.54 (m)		3.48 (overlap)		3.46 (m)		3.57 (dd, 9.9, 4.9)
1′	133.1		133.1		133.1		134.7
2′/6′	105.6	7.29 (s)	105.5	6.48 (s)	105.5	6.49 (s)	104.8	6.76 (s)
3′/5′	153.6		153.3		153.3		149.5
4′	139.9		138.8		138.9		136.3
7′	199.1		34.4	2.70 (m)	34.4	2.71 (m)	84.5	5.06 (d, 8.4)
8′	40.4	3.23 (t, 5.4)	32.8	1.91 (m)	32.8	1.92 (m)	54.6	2.38 (m)
9′a	58.3	4.06 (t, 5.3)	62.3	3.71 (t, 6.3)	62.3	3.72 (t, 6.3)	61.6	3.78 (dd, 11.2, 5.2)
9′b								3.69 (dd, 11.2, 4.4)
1′′							102.3	4.70 (d, 1.3)
2′′							72.3	3.81 (m)
3′′							72.7	3.65 (dd, 4.5, 3.2)
4′′							74.1	3.40 (m)
5′′							70.3	3.61 (dq, 9.5, 6.5)
6′′							18.2	1.27 (d, 6.5)
3-OCH3	56.5	3.88 (s)	56.1	3.90 (s)	56.5	3.88 (s)	57.0	3.90 (s)
5-OCH3	56.5	3.88 (s)			56.5	3.88 (s)	57.0	3.90 (s)
3′,5′-OCH3	56.6	3.96 (s)	56.3	3.87 (s)	56.3	3.88 (s)	57.0	3.90 (s)
4-OH		5.47 (s)		5.55 (s)		5.45 (s)

¹ Measured in chloroform-d; ² Measured in methanol-d₄.

; HRFABMS m/z 451.1600 [M − H]⁻ (calcd for C₂₂H₂₇O₁₀, 451.1599) (Figures S1–S7).

(7R,8S)-4,7,9,9′-tetrahydroxy-3,3′,5′-trimethoxy-8-O-4′-neolignan (2). Colorless gum; ECD (MeOH) λ_max (Δε) 241 (+2.1) nm; ¹H and ¹³C NMR data, see Table 1.

1:1 mixture of (7S,8R)- and (7R,8S)-4,7,9,9′-tetrahydroxy-3,5,3′,5′-tetramethoxy-8-O-4′-neolignan (3). Colorless gum; ¹H and ¹³C NMR data, see Table 1.

7R,7′R,8S,8′S-icariol A2 9′-O-α-l-rhamnopyranoside (9). Colorless gum; [α]${}_{\mathrm{D}}^{25}$+30 (c 0.1, MeOH); ECD (MeOH) λ_max (Δε) 246 (+5.3), 208 (−10.1) nm; ¹H and ¹³C NMR data, see Table 1; HRFABMS m/z 581.2231 [M − H]⁻ (calcd for C₂₈H₃₇O₁₃, 581.2229) (Figures S8–S15).

β-hydroxypropiosyringone 4′-O-β-d-glucopyranoside (11). Colorless gum; [α]${}_{\mathrm{D}}^{25}$+24 (c 0.25, MeOH); ¹H and ¹³C NMR data, see

Table 2. ¹H (700 MHz) and ¹³C (175 MHz) NMR data of compound 11 in methanol-d_4.

Pos.	11
	δC	δH (multi, J in Hz)
1	134.6
2/6	107.6	7.23 (s)
3/5	154.4
4	140.7
7	200.0
8	42.1	3.11 (t, 6.1)
9	58.8	3.86 (t, 6.1)
1′	104.6	5.00 (d, 7.5)
2′	75.9	3.40 (dd, 8.8, 7.5)
3′	78.0	3.33 (t, 8.8)
4′	71.5	3.30 (t, 8.8)
5′	78.6	3.12 (ddd, 8.8, 5.3, 2.3)
6′a	62.7	3.67 (dd, 11.9, 2.3)
6′b		3.55 (dd, 11.9, 5.3)
3,5-OCH3	57.3	3.81 (s)

; HRFABMS m/z 411.1263 [M + Na]⁺ (calcd for C₁₇H₂₄O₁₀Na, 411.1262) (Figures S16–S22).

2.4. ¹H NMR Simulation for H-7 of Compound 1

MestReNova Version 14.1.2-25024 was used to simulate the H-7 peak. The “Spin Simulation” under the “Prediction” tab was clicked to open the dialog box. The “Spin Groups” was set to be “2” (default value), the “Shifts (ppm)” values of “A” and “B” were given “4.987” and “4.232”, respectively, with the “Line Width (Hz)” of both “3.0”. After the coupling constant between “A” and “B” was set as “3.5” or “6.7”, the icon with “New Simulation” at the top left was clicked to generate the simulated ¹H NMR peaks of H-7 and H-8 of 1.

2.5. Acid Hydrolysis of Compounds 6 and 11 and Sugar Analysis

The sugar analysis experiment was performed as described in previous communication by Tanaka et al. with some modifications [17]. Compounds 9 and 11 (each of 0.5 mg) were individually hydrolyzed with 1 N HCl (1.0 mL) for 2 h at 100 °C. Each hydrolysate was extracted using CHCl₃, and then the aqueous layers were evaporated to remove the HCl and to afford sugar from each hydrolysate. Each H₂O-soluble fraction was added to anhydrous pyridine (400 µL) containing l-cysteine methyl ester hydrochloride (1.1 mg), which were heated at 60 °C for 1 h. Then, O-tolyl isothiocyanate (30 µL) was added to the reaction mixture and heated at 60 °C for another 1 h. After the reaction, each reaction mixture was analyzed by LC-MS (0.7 mL/min; 25% aqueous CH₃CN with 0.1% formic acid for 40 min). The authentic samples of l-rhamnopyranose and d-glucopyranose were derivatized and analyzed by the same method mentioned above, and these monosaccharides were also derivatized with d-cysteine methyl ester hydrochloride instead of its l-form to prepare their enantiomeric derivatives. The hydrolysate derivatives of 9 and 11 were 25.4 min (l-rhamnopyranose) and 14.9 min (d-glucopyranose) which matched well with those of authentic l-rhamnopyranose (25.3 min) and d-glucopyranose (14.9 min), respectively, rather than their enantiomeric derivatives (13.2 min (d-rhamnopyranose) and 13.7 min (l-glucopyranose)) (Figures S15 and S22).

2.6. NO Production Assays

Analogous as described in [18]. The BV-2 cells, developed by Dr. V. Bocchini at the University of Perugia (Perugia, Italy), were used for this study [19,20]. The cells were seeded in a 96-well plate (4 × 10⁴ cells/well) and incubated in the presence or absence of various doses of tested compounds (1–24). Lipopolysaccharide (LPS) (100 ng/mL) was added to all the pre-treated wells except the control one and grown for 1 d. The produced levels of nitrite (NO₂), a soluble oxidized product of NO, was evaluated with 0.1% N-1-napthylethylenediamine dihydrochloride and 1% sulfanilamide in 5% phosphoric acid, aka the Griess reagent. The supernatant (50 μL) was mixed with the Griess reagent (50 μL). After 10 min the absorbance was gauged at 570 nm. For a positive control, the reported nitric oxide synthase (NOS) inhibitor N^G-monomethyl-l-arginine (l-NMMA) was employed. Graded sodium nitrite solution was utilized to determine nitrite concentrations. A 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) assay was used for the cell viability assay.

2.7. Assays for Nerve Growth Factor NGF Release from C6 Cells

Analogous as described in [21]. C6 glioma cell lines were used to measure the nerve growth factor (NGF) of the culture medium, which were fixed with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin (PS) in an incubator filled with 5% CO₂. The cells were seeded in a 24-well culture plate (1 × 10⁵ cells/well) and incubated for 24 h. The cells were treated with or without 20 µM of the compounds (1–24), together with serum-free Dulbecco’s modified Eagle’s medium (DMEM) for another 24 h. Released NGF levels from the supernatants from each cell were measured using an ELISA development kit (R&D System, Minneapolis, MN, USA). Besides, the cell viability was evaluated by 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) assay by comparison with 6-shogaol as a positive control and the results are expressed as percentage of the control group.

2.8. Cytotoxicity Assessment

The cytotoxicity of purified compounds (1–24) were tested against the A-549 (non-small cell lung adenocarcinoma), SK-OV-3 (ovary malignant ascites), SK-MEL-2 (skin melanoma), and MKN-1 (adenosquamous carcinoma), utilizing the sulforhodamine B colorimetric (SRB) method [22]. Cisplatin (≥98%; Sigma–Aldrich) served as a positive control.

3. Results and Discussion

3.1. Structure Elucidation of Compounds 1–24

Compound 1 was purified as a colorless gum and its molecular formula was determined as C₂₂H₂₈O₁₀ according to the deprotonated molecular ion at m/z at 451.1600 [M – H]^– in the HRFABMS data (calcd. for C₂₂H₂₇O₁₀, 451.1599). Analysis of the ¹H NMR spectrum of 1 indicated the presence of two 1,3,4,5-tetrasubstituted benzene rings (δ_H 7.29 (2H, s, H-2′,6′) and 6.58 (2H, s, H-2,6)), two oxygenated methines (δ_H 4.99 (1H, brs, H-7) and 4.23 (1H, ddd, J = 6.6, 3.5, 3.0 Hz, H-8)), two oxygenated methylenes (δ_H 4.06 (2H, t, J = 5.3 Hz, H-9′), 3.93 (1H, overlap, H-9a), and 3.54 (1H, m, H-9b)), a methylene (δ_H 3.23 (2H, t, J = 5.4 Hz, H-8′)), and four methoxy groups (δ_H 3.96 (6H, s, 3′,5′-OCH₃) and 3.88 (6H, s, 3,5-OCH₃)). The ¹³C spectrum of 1 showed 22 carbon signals for two 1,3,4,5-tetrasubstituted benzene rings (δ_C 153.6 (×2), 147.2 (×2), 139.9, 134.2, 133.1, 130.3, 105.6 (×2), and 102.5 (×2)), a carbonyl carbon (δ_C 199.1), four methoxy carbons (δ_C 56.6 (×2) and 56.5 (×2)), four oxygenated carbons (δ_C 87.5, 73.1, 60.7, and 58.3), and a methylene carbon (δ_C 40.4). These NMR spectra of 1 were similar to those of 3 (Table 1) [23], except for the presence of the carbonyl carbon signal at δ_C 199.1 (C-7′) in 1 rather than the methylene carbon signal at δ_C 34.4 (C-7′) in 3. The HMBC correlations from H-2′,6′/H-8′/H-9′ to C-7′ were observed, suggesting that carbonyl carbon is located at C-7′. The planar structure of 1 was established through 2D NMR spectra analysis including COSY, HSQC, and HMBC spectra (Figure 2A). In 8-O-4′-oxyneolignan derivatives, the relatively small (~4 Hz) and large (~8 Hz) coupling constants between H-7 and H-8 indicate erythro- or threo-configuration, respectively, in chloroform-d [24,25,26]. When the H-7 peak was analyzed to measure the ³J_H-7,H-8 value, however, its shape was found to be a broad singlet. Alternatively, the H-8 peak showed the splitting pattern of doublet of doublet of doublet (ddd) with coupling constants of 6.7, 3.5, and 3.0 Hz. In order to deduce the ³J_H-7,H-8 value among them, the H-7 peak was simulated with coupling constants of 3.5 and 6.7 Hz. As shown in Figure 2B (middle) the simulated peak with the 3.5 Hz coupling constant was more similar to that of experimental peak rather than that with 6.7 Hz coupling constant. Therefore, the ³J_H-7,H-8 value was assigned as 3.0 or 3.5 Hz which suggested erythro-form. The difference of ¹³C NMR chemical shift between C-7 and C-8 can be also used for its relative configuration assignment. The Δδ_C-8,C-7 value was ~14.6 ppm in the erythro-form whereas ~15.5 ppm was observed in the threo-form [24]. By observing 14.4 ppm of Δδ_C-8,C-7 value in the ¹³C NMR data of 1 (Figure 2C), the relative configuration between C-7 and C-8 was deduced as erythro-form. The 8R configuration of 1 was confirmed by means of analysis of its ECD spectrum displaying a characteristic negative Cotton effect at 244 nm (Figure 2D) [26,27]. Thus, the structure of 1 was established as (7S,8R)-7,9,9′-trihydroxy-3,5,3′,5′-tetramethoxy-7′-oxo-8-O-4′-neolignan.

Compounds 2 and 3 were previously reported from Bursera tonkinensis [23], and Epimedium sagittatum [28] and Fraxinus mandshurica [29], respectively, without determination of the absolute configuration of C-7 and C-8 (Figure S23–S26). The relative configuration between C-7 and C-8 were assigned as erythro by the same empirical rule applied for 1; the relatively small ³J_H-7,H-8 value was observed from both 2 (3.5 Hz) and 3 (3.2 Hz) (Table 1). A positive Cotton effect at 241 nm from ECD spectrum of 2 indicated an 8S configuration of 2 and the absence of any characteristic Cotton effect from the ECD spectrum of 3, suggesting 3 as a racemic mixture (Figure 2D).

Compound 9 was isolated as a colorless gum with a molecular formula of C₂₈H₃₈O₁₃. The ¹H NMR spectrum of 9 showed two 1,3,4,5-tetrasubstituted benzene rings (δ_H 6.76 (2H, s, H-2′,6′) and 6.75 (2H, s, H-2,6)), two oxygenated methines (δ_H 5.06 (1H, d, J = 8.4 Hz, H-7′) and 5.03 (1H, d, J = 8.4 Hz, H-7)), two oxygenated methylenes (δ_H 3.82 (1H, m, H-9a), 3.78 (1H, dd, J = 11.2, 5.2 Hz, H-9′a), 3.69 (1H, dd, J = 11.2, 4.4 Hz, H-9′b), and 3.57 (1H, dd, J = 9.9, 4.9 Hz, H-9′b)), two methines (2.51 (1H, m, H-8) and 2.38 (1H, m, H-8′)), four methoxy groups (δ_H 3.90 (12H, s, 3,5,3′,5′-OCH₃)), and an α-rhamnopyranosyl unit (δ_H 4.70 (1H, d, J = 1.3 Hz, H-1′′), 3.81 (1H, m, H-2′′), 3.65 (1H, dd, J = 4.5, 3.2 Hz, H-3′′), 3.61 (1H, dq, J = 9.5, 6.5 Hz, H-5′′), 3.40 (1H, m, H-4′′), and 1.27 (3H, d, J = 6.5 Hz, H-6′′)). The ¹³C NMR spectrum of 2 revealed 28 resonances, comprising 12 benzene ring carbons [δ_C 149.5 (×4), 136.4, 136.3, 134.7, 134.3, 105.1 (×2), and 104.8 (×2)], four methoxy carbons (δ_C 57.0 (×4)), four oxygenated carbons (δ_C 85.2, 84.5, 67.8, and 61.6), two methine carbons (δ_C 54.6 and 51.8), and a rhamnopyranosyl unit (δ_C 102.3, 74.1, 72.7, 72.3, 70.3, and 18.2). These spectroscopic data were almost identical with those of 7S,7′S,8R,8′R-icariol A2 9′-O-α-l-rhamnopyranoside (25) [30]. Intensive 2D NMR data analysis (Figure 3A) confirmed that 9 and 25 have the same the planar structures. The absolute configuration of the known compound 25 was determined as 7S,7′S,8R,8′R by observing the negative Cotton effect around 240–250 nm in the ECD spectrum (Figure S9) [30], however, the new compound 9 showed the positive Cotton effect around that area (246 nm, Figure 3B), indicating 7R,7′R,8S,8′S-configuration. In addition, the opposite sign of the specific rotation values of 9, [α]${}_{\mathrm{D}}^{25}$+ 30 (c 0.1, MeOH), and 25, [α]${}_{\mathrm{D}}^{25}$− 51.4 (c 0.1, MeOH) [30], supported the initial assignment. Finally, the l-form of rhamnopyranosyl moiety was assigned by LC-MS analysis on the chiral derivatized monosaccharide obtained by hydrolysis of 9 and following chemical reaction [17]. The retention time of derivatized rhamnopyranose (25.4 min) was matched with that of standard l-rhamnopyranose (25.3 min) rather than d-rhamnopyranose (13.2 min) (Figure 2C). Thus, the structure of 9 was determined as 7R,7′R,8S,8′S-icariol A2 9′-O-α-l-rhamnopyranoside.

Compound 11 was obtained as a colorless gum, and its molecular formula was deduced as C₁₇H₂₄O₁₀ based upon the HRFABMS data. The ¹H and ¹³C NMR spectra of 11 were comparable to those of 3-hydroxyl-1-(4-hydroxy-3,5-dimethoxyphenyl)-1-propanone (or β-hydroxypropiosyringone, 12) [31], except for the presence of resonances for a β-glucopyranosyl unit (δ_H 5.00 (1H, d, J = 7.5 Hz, H-1′), 3.67 (1H, dd, J = 11.9, 2.3 Hz, H-6′a), 3.55 (1H, dd, J = 11.9, 5.3 Hz, H-6′b), 3.40 (1H, dd, J = 8.8, 7.5 Hz, H-2′), 3.33 (1H, t, J = 8.8 Hz, H-3′), 3.30 (1H, t, J = 8.8 Hz, H-4′), 3.12 (1H, ddd, J = 8.8, 5.3, 2.3 Hz, H-5′); δ_C 104.6, 78.6, 78.0, 75.9, 71.5, and 62.7). On the basis of the HMBC cross peak of H-1′ with C-4, the location of the β-glucopyranosyl unit was confirmed to be connected at C-4 by an ether bond (Figure 4A). Acid hydrolysis of 11 afforded glucopyranose, which was identified as d-form by the same method as described above (Figure 4B). Thus, the structure of 11 was elucidated as β-hydroxypropiosyringone 4′-O-β-d-glucopyranoside.

The other known compounds were identified as (7S,8R)-1-[4-(O-β-d-glucopyranosyl)-3-methoxyphenyl]-2-[4-(3-hydroxypropyl)-2,6-dimethoxyphenoxyl]-1,3-propanediol (4) [32], 3-[4-[(1R,2S)-2-hydroxy-2-(4-hydroxy-3-methoxyphenyl)-1-(hydroxymethyl)ethoxy]-3-methoxyphenyl]propyl β-d-glucopyranoside (5) [27], syringatesinol (6) [33], pinoresinol-4-O-glucoside (7) [34], 8-hydroxypinoresinol-4′-O-β-d-glucose (8) [35], (–)-lyoniresinol 3α-O-β-d-glucopyranoside (10) [36], 3-hydroxyl-1-(4-hydroxy-3,5-dimethoxyphenyl)-1-propanone (12) [31], (2R)-O-[4′-(3″-hydroxypropyl)-2′,6′-dimethoxyphenyl]-3-O-β-d-glucopyranosyl-sn-glycerol (13) [37], 4-allyl-2,6-dimethoxyphenyl 1-O-β-d-xylopyranosyl-(1→6)-β-d-glucopyranoside (14) [38], 4-allyl-2-hydroxyphenyl 1-O-β-d-apiosyl-(1→6)-β-d-glucoside (15) [39], 4-allyl-2-methoxyphenyl 6-O-β-d-apiosyl-(1→6)-β-d-glucoside (16) [40], 3,5-dimethoxy-1-hydroxy-4-O-β-d-glucoside (17) [41], 3,4,5-trimethoxyphenol 1-O-β-d-glucopyranoside (18) [42], 3,4,5-trimethoxyphenyl 1-O-β-d-apiofuranosyl-(1′′→6′)-β-d-glucoside (19) [43], tachioside (20), isotachioside (21) [44], vanillyl alcohol-4-O-β-d-glucopyranoside (22) [45], benzyl β-d-xylopyranosyl-(1→6)-β-d-glucopyranoside (23) [46], and 4-hydroxy-2,6-dimethoxyphenyl 6′-O-vanilloyl-β-d-glucopyranoside (24) [47] by comparing their spectroscopic data with reported data.

3.2. Anti-Neuroinflammatory Activity of the Isolated Compounds (1–24)

Alzheimer’s disease, a neurodegenerative disease, is an unsolved social and medical problem and is strongly associated with neurotrophins (e.g., NGF, brain-derived neurotrophic factor (BDNF), neurotrohpin-3 (NT-3), and NT-4/5) and neuroinflammatory mediators (e.g., nitric oxide (NO), tumor necrosis factor alpha (TNF-α), interleukin (IL)-1β, nuclear factor kappa B (NF-κB), and prostaglandin (PG)E₂) [48]. Therefore, many phytochemicals have been tested to discover novel anti-neurodegenerative agents, which revealed that phenolic constituents are one of the major neurotrophic and anti-neuroinflammatory compounds [48,49]. In line with this finding, the isolated phenolic compounds in this study (1−24) were assessed for their anti-neuroinflammatory activities by measuring NO levels in LPS-activated BV-2 cells (

Table 3. Inhibitory effects of selected compounds on nitric oxide (NO) production in lipopolysaccharide (LPS)-activated BV-2 cells.

Compound	IC50 (μM) 1	Cell viability (%) 2
1	18.66	113.97 ± 1.36
2	19.75	105.99 ± 7.78
3	18.63	94.33 ± 3.84
6	12.69	101.33 ± 8.15
9	97.50	97.67 ± 3.20
12	72.50	114.97 ± 5.17
17	68.77	106.59 ± 1.27
20	41.67	103.50 ± 2.62
21	15.00	103.50 ± 2.62
24	70.08	94.46 ± 10.89
l-NMMA 3	21.35	104.56 ± 4.20

¹ IC₅₀ value of each compound was defined as the concentration (μM) that caused 50% inhibition of NO production in LPS-activated BV-2 cells. ² Cell viability after treatment with 20 μM of each compound was determined by 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) assay and is expressed in percentage (%). The results are averages of three independent experiments, and the data are expressed as mean ± SD. ³ N^G-monomethyl-l-arginine (l-NMMA) as positive control.

) and for their neurotrophic effects evaluated by assessing the secretion of NGF from C6 cells (

Table 4. Effects of selected compounds on nerve growth factor (NGF) secretion in C6 cells.

Compound	NGF Secretion 1	Cell Viability (%) 2
1	110.27 ± 2.18	96.59 ± 2.78
2	127.98 ± 3.95	101.75 ± 2.40
3	149.89 ± 1.97	105.00 ± 3.21
6	120.09 ± 2.72	109.25 ± 8.03
12	135.10 ± 5.59	94.41 ± 1.83
20	157.73 ± 1.50	11.83 ± 8.24
21	125.00 ± 1.50	117.97 ± 8.50
6-shogaol 3	149.53 ± 5.36	97.00 ± 0.17

¹ C6 cells were treated with 20 μM of compounds. After 24 h, the content of NGF secretion in C6-conditioned media was measured by ELISA. The level of secreted NGF cells is expressed as percentage of the untreated control. The data shown represent the means ± SD of three independent experiments performed in triplicate. ² Cell viability after treatment with 20 μM of each compound was determined by MTT assay and is expressed in percentage (%). The results are averages of three independent experiments, and the data are expressed as mean ± SD. ³ 6-shogaol as positive control.

). Compounds 1−3, 8–O–4′ type neolignans, significantly reduced NO levels with IC₅₀ values of 18.63–19.75 μM, which showed better activity than a well-known inhibitor of NO synthase (NOS), N^G-monomethyl-l-arginine (l-NMMA, IC₅₀ 21.35 μM) [50]. These data were consistent with the previous studies in which diverse 8–O–4′ type neolignans exhibited potent NO inhibitory activity [49]. However, compounds 4 and 5, the other 8–O–4′ type neolignans, were not active (>100 μM) in this study, which might be attributed to the different configuration at C-7 and/or the glucopyranosyl unit attached to the molecules. The most potent compound in this study was 6, a furanofuran-type lignan, with IC₅₀ value of 12.69 μM, whereas the other glucosylated furanofuran-type lignans 7 and 8 did not display NO inhibitory activity. The simple phenolic glucoside 21 was a potent NO inhibitor (IC₅₀ 15.00 μM) but transfer of the methoxy group at C-3 to C-2 decreased the potency (IC₅₀ 41.67 μM, 20). The other compounds showed weak or no activity.

3.3. Neurotrophic Activity of the Isolated Compounds (1–24)

Similar to the NO inhibitory activity profile of the 8–O–4′ type neolignans, only 1–3 displayed effect on NGF release with stimulation levels of 110.27 ± 2.18%, 127.98 ± 3.95%, and 149.89 ± 1.97%, respectively, while the latter was comparable to that of positive control, 6-shogaol (149.53 ± 5.36%) [51]. The addition of a ketone group at C-7′ (1) and removal of the methoxy group at C-5 (2) in 3 decreased NGF secretion activity. Compound 20 showed powerful NGF release effect with stimulation levels of 157.73 ± 1.50% but this compound was cytotoxic to the tested C6 cells (cell viability: 11.83 ± 8.24%). Compounds 6, 12, and 21 were mild neurotrophic agents (120.09 ± 2.72%, 135.10 ± 5.59%, and 125.00 ± 1.50%, respectively) and the other compounds were inactive (<100%).

3.4. Cytotoxicity of the Isolated Compounds (1–24)

Podophyllotoxin is a representative anticancer natural product with a lignan scaffold and its synthetic derivative etoposide (VePesid^®) is chemotherapy medication used for the treatments of diverse types of cancer [2]. Since most of the isolated phytochemicals (1–24) are derivatives of lignan or its monomer, phenylpropanoid, their cytotoxic activity against four human cancer cell lines, A549 (non-small-cell lung adenocarcinoma), SK-OV-3 (ovary malignant ascites), SK-MEL-2 (skin melanoma), and MKN-1 (adenosquamous carcinoma) were evaluated via the SRB bioassay method with etoposide as a positive control substance (IC₅₀ 0.98, 2.15, 1.80, and 3.47 μM, respectively). As a result, four 8–O–4′ type neolignans, 1−4, and a phenolic glucoside 24 showed an inhibitory effect on MKN-1 cancer cell with IC₅₀ values of 11.15, 14.03, 14.16, 7.96, and 19.04 μM, respectively (

Table 5. Cytotoxicity of selected compounds against four cultured human cell lines.

Compound	IC50 (μM) 1
	A-549	SK-OV-3	SK-MEL-2	MKN-1
1	>30.0	>30.0	>30.0	11.15
2	>30.0	>30.0	>30.0	14.03
3	>30.0	>30.0	>30.0	14.16
4	>30.0	>30.0	>30.0	7.96
24	>30.0	>30.0	>30.0	19.04
Etoposide 2	0.98	2.15	1.80	3.47

¹ 50% inhibitory concentration; the concentration of compound that caused a 50% inhibition in cell growth. ² Etoposide as positive control.

).

4. Conclusions

A total of 24 phenolic compounds (1−24) including three new compounds (1, 9, and 11) were isolated and characterized from the twigs of C. sinensis. All isolates were evaluated for their anti-neuroinflammatory, neurotrophic, and cytotoxic activities. As a result, compounds 1−3, 8–O–4′ type neolignans, displayed significant results in all activity evaluations. Compounds 6, 12 and 21 showed anti-neuroinflammatory effects and/or neurotrophic activity. Compounds 24 displayed inhibitory effect on MKN-1 cancer cell. Through these results, it was suggested that the phenolic compounds in C. sinensis would be responsible for one of the various activities of this plant which could be the basis for the traditional use of C. sinensis for inflammatory diseases. Especially, 8–O–4′ type neolignans showed promising multi-target potentials as a drug against neurodegenerative diseases.