Novel phthalide derivatives from the rhizomes of Ligusticum chuanxiong and their inhibitory effect against lipopolysaccharide-induced nitric oxide production in RAW 264.7 macrophage cells

Abstract

Five new phthalide derivatives, chuanxiongnolides L1–L5 (1–5), together with three known phthalide dimers (6–8), were isolated from the rhizomes of Ligusticum chuanxiong Hort.. Compound 1 represents a novel type of phthalide derivative, biogenetically derived from coniferyl alcohol and ligustilide. Compound 2 is the first example of an E-ring expanded phthalide dimer with a 5/6/6/6/6 fused-ring system, which is different from the 5/6/6/6/5 fused-ring system in normal phthalide dimers. Their structures were established using spectroscopic data, and the absolute configurations were determined by a circular dichroism (CD) exciton chirality method. To confirm the absolute configuration of compound 1, electronic circular dichroism (ECD) calculations were also conducted. Compounds 1, 2, 6, and 7 exhibited inhibitory effects against lipopolysaccharide-induced nitric oxide production in RAW 264.7 macrophage cells with half maximal inhibitory concentration (IC₅₀) values ranging from 3.0 to 12.6 μM. Furthermore, plausible biosynthetic routes for 1 and 2 were also proposed.

---

Introduction

Chuanxiong rhizoma (CXR), the dried rhizomes of Ligusticum chuanxiong Hort. (family Umbelliferae), is an authentic herbal medicine of the Sichuan province of China, was first recorded in Shennong's Classic of Materia Medica (Shen Nong Ben Cao Jing), published during the period of the Han Dynasty (200–300 A.D.), and has been used in traditional Chinese medicine for many centuries. CXR possesses various functions including promoting the circulation of blood, expelling wind, and alleviating pain.¹ Due to its beneficial effects and reputation for facilitating blood circulation and dispersing blood stasis, CXR has been widely used in China for the treatment of various cardiovascular and cerebrovascular diseases in clinics² such as hypertension,³ stroke,⁴ thrombus formation,⁵ and atherosclerosis.⁶ Previous phytochemical investigations have reported the isolation and identification of various different structural types of its components,^7–12 among which, several constituents belonging to four types, namely alkaloids, phthalides, phenols, and organic acids, have been found to be pharmacologically active with properties including antioxidant,¹³ anti-inflammatory,^10,14 neuroprotective,¹⁵ vasorelaxing,¹⁶ and antiproliferative¹⁷ effects.

Phthalides and their derivatives have attracted researchers' attention since 1991 (ref. 18) due to their structural diversities and extensive biological activities.^8,16,19–22 These compounds can structurally be divided into two types, monomeric phthalides and phthalide dimers. In the following decades, the structural diversity of phthalides from CXR has been extensively studied, leading to the isolation and identification of various compounds.^{14,18,23–26} The main and bioactive monomeric phthalides contained in CXR, such as Z-ligustilide, E-ligustilide, senkyunolide A, 3-butylidenephthalide, etc., are considered to be precursors in the generation of phthalide dimers, which involves a [2 + 2] or [4 + 2] cycloaddition (such as the Diels–Alder reaction)²⁷ of two phthalide units. For example, the formation of senkyunolide O,¹⁸ senkyunolide P,¹⁸ levistolide A,¹⁸ tokinolide B,¹⁸ chuanxiongdiolide R1,¹⁴ and chuanxiongdiolide R2.¹⁴

Inflammation is the complex biological response of tissues to harmful stimulus. Mostly produced by inducible nitric oxide synthase (iNOS), nitric oxide (NO) is a gaseous short-lived signalling molecule with several physiological and pathophysiological effects. In inflammation, high and prolonged production of NO may lead to cytotoxic and pro-inflammatory effects,²⁸ which indicate that the inhibition of NO release might be an important and attractive therapeutic target in the treatment of inflammatory diseases.

A previous report suggested that the aqueous extract of CXR exhibited anti-inflammatory effects.²⁹ Moreover, two monomeric phthalides, Z-ligustilide and senkyunolide A, were demonstrated to have potential applications in the treatment of inflammation.³⁰ As for the anti-inflammatory effects of phthalide dimers, some phthalide dimers were reported to have significantly inhibitory effects against lipopolysaccharide (LPS)-induced NO production.¹⁴ To obtain a better understanding of the structural diversity of these important secondary metabolisms and their anti-inflammatory effect, a detailed chemical investigation of CXR was carried out. As a result, five new phthalide derivatives (1–5) and three known phthalide dimers (6–8) were obtained. Usually, the biogenesis of phthalide dimers from the Ligustrum genus could be explained by the coupling of two monomeric phthalide units; here, compound 1 represents a novel type of phthalide derivative, biogenetically derived from coniferyl alcohol and ligustilide. Compound 2 is the first example of an E-ring expanded phthalide dimer with a 5/6/6/6/6 fused-ring system, possibly formed by levistolide A^7,18 (6), which is the most abundant phthalide dimer in CXR.⁷ Herein, details of the isolation, structural elucidation, inhibitory effect on the LPS-induced NO production in RAW 264.7 macrophages, and the biogenetic origins of five new phthalide derivatives, chuanxiongnolides L1–L5 (1–5) and three known phthalide dimers (6–8) are described.

Results and discussion

A 95% EtOH extract of CXR was subjected to silica gel Sephadex LH-20 column chromatography and reverse-phase semi-preparative high-performance liquid chromatography (RP-SP-HPLC) to afford compounds 1–8 (Fig. 1). The known compounds were identified as levistolide A^18,27 (6), ansaspirolide³¹ (7), and chuanxiongdiolide R1 (ref. 8) (8) by comparison of their spectroscopic data with literature values.

Fig. 1 Chemical structures of compounds 1–8.

Compound 1 was obtained as a pale yellow oil. Its molecular formula was established to be C₂₀H₂₂O₆ by high-resolution electrospray ionization mass spectrometry (HRESIMS) with a protonated molecular ion at a m/z of 359.1491 [M + H]⁺ (calcd for C₂₀H₂₃O₆, 359.1495), corresponding to ten degrees of unsaturation. Its infrared (IR) spectrum indicated the presence of hydroxyl (3385 cm⁻¹), methine (2931 cm⁻¹), lactone (1772 cm⁻¹), and aromatic ring (1712, 1516, and 1367 cm⁻¹) functionalities. The ¹H nuclear magnetic resonance (NMR) data (Table 1) of 1 showed signals for one methyl group at δ_H 1.32 (t, J = 7.1 Hz, H-4′′); two methylenes at δ_H 1.54 (m, H-9b), 1.94 (t, J = 10.5 Hz, H-9a), 1.62 (m, H-8b), and 2.01 (m, H-8a); two oxygen-bearing methylenes at δ_H 4.28 (m, H-3′′), 4.17 (dd, J = 8.9, 11.2 Hz, H-3b), and 4.53 (dd, J = 7.6, 8.9 Hz, H-3a); three methines at δ_H 2.45 (m, H-3a), 2.68 (d, J = 8.5 Hz, H-4), and 2.96 (m, H-5); three methines assignable to a 1,3,4-trisubstituted benzene moiety (ABX spin system) at δ_H 6.49 (d, J = 2.0 Hz, H-2′), 6.51 (dd, J = 2.0, 8.0 Hz, H-6′), and 6.82 (d, J = 8.0 Hz, H-5′); an olefinic methine at δ_H 7.18 (d, J = 6.5 Hz, H-6); one methoxyl group at δ_H 3.85 (s, 3′-OCH₃); and one hydroxyl proton at δ_H 5.52 (brs, 4′-OH). The ¹³C NMR spectroscopic data with the aid of a distortionless enhancement of polarization transfer (DEPT) experiment suggested the presence of 20 carbon resonances, including two ester carbonyl groups at C-1 (δ_C 174.8), and C-1′′ (δ_C 163.7); three quaternary aromatic carbons at C-1′, C-3′, and C-4′ (δ_C 135.9, 146.7, and 144.7); three aromatic methines at C-2′, C-5′, and C-6′ (δ_C 109.5, 114.8, and 119.4); one olefinic methine at C-6 (δ_C 142.4); one olefinic quaternary carbon at C-7 (δ_C 138.2); two oxygen-bearing methylenes at C-3′′ (δ_C 61.2) and C-3 (δ_C 70.4); one quaternary carbon at C-7a (δ_C 46.0); two methylenes at C-9 (δ_C 27.6) and C-8 (δ_C 22.1); three methines at C-5 (δ_C 40.1), C-4 (δ_C 46.1), and C-3a (δ_C 52.8); one methyl group at C-4′′ (δ_C 14.2); and one methoxyl group (δ_C 56.0). The aforementioned information indicated seven degrees of unsaturation occupied by one benzene ring, one double bond, and two carbonyls, suggesting that the remaining three required the presence of a tricyclic system in 1.

Table 1 ¹H NMR spectroscopic data (400 MHz, CHCl₃) for compounds 1–5 (δ in ppm, J in Hz)

No.	1	2	3	4	5
3	4.17 dd (8.9, 11.2); 4.53 dd (7.6, 8.9)
3a	2.45 m
3-OH				5.31 s
4	2.68 d (8.5)	1.99 m; 2.30 m	1.99 m; 2.36 m	1.80 m; 2.13 m	1.90 m; 2.25 m
5	2.96 m	1.72 m; 1.96 m	1.54 m; 2.01 m	1.55 m; 1.94 m	1.54 m; 1.91 m
6	7.18 d (6.5)	2.55 m	2.61 t (7.5)	2.57 t (7.6)	2.53 t (7.6)
7		3.38 d (8.6)	3.28 d (8.3)	3.28 brd (8.8)	3.17 d (8.7)
8	1.62 m; 2.01 m	5.17 t (7.9)			3.54 dd (2.2, 10.4)
9	1.54 m; 1.94 t (10.5)	2.30 m		2.06 m; 2.49 m	1.17 m; 1.31 m
10		1.28 m; 1.45 m		1.65 m	1.35 m; 1.55 m
11		0.92 t (7.3)		0.95 t (7.2)	0.96 t (7.2)
2′	6.49 d (2.0)
3′		4.85 dd (4.6, 9.2)
3′-OCH3	3.85 s
4′			1.43 m; 2.06 m	1.44 m; 2.08 m	1.38 m; 2.01 m
4′-OH	5.52 brs
5′	6.82 d (8.0)	1.40 m; 2.16 m	1.32 m; 1.91 m	1.31 m; 1.90 m	1.29 m; 1.87 m
6′	6.51 dd (2.0, 8.0)	1.25 m; 1.96 m	3.03 m	3.03 m	2.99 m
7′		2.81 m	7.36 d (6.6)	7.38 d (6.6)	7.36 d (6.6)
8′		7.72 d (6.8)	4.97 t (7.5)	5.06 t (7.5)	4.96 t (7.5)
9′		1.81 m	2.14 q (7.5)	2.18 q (7.4)	2.13 q (7.3)
10′		1.53 m	1.42 m	1.45 m	1.42 m
11′		0.93 t (7.5)	0.91 t (7.4)	0.93 t (7.2)	0.90 t (7.3)
3′′	4.28 m
4′′	1.32 t (7.1)

---

The planar structure of 1 was elucidated by analysis of its 2D NMR spectroscopic data including ¹H–¹H correlation spectroscopy (COSY), heteronuclear single-quantum correlation spectroscopy (HSQC), heteronuclear multiple-bond correlation spectroscopy (HMBC), and nuclear Overhauser effect spectroscopy (NOESY) spectra. Its ¹H–¹H COSY spectrum, in combination with its HSQC spectroscopic data, established three independent spin systems (I: H-3′′/4′′; II: H-3/3a/4/5/6 and H-5/9/8; and III: H-2′/5′/6′), as shown in Fig. 2. The HMBC spectrum (Fig. 2) showed correlations from H-5′ to C-4′ and C-3′, from H-6′ to C-4′, and from H-2′ to C-3′ and C-4′, which confirmed that the OCH₃ and OH groups were located at C-3′ and C-4′, respectively. Furthermore, HMBC correlations from H-2′ to C-4, from H-6′ to C-4, and from H-4 to C-1′, C-2′, and C-6′ suggested that the 1,3,4-trisubstituted benzene ring was located at C-4. HMBC correlations from H-3a to C-1, from H-3 to C-1 and C-7a, and the corresponding shifts (H-3 at δ_H 4.17 and 4.53; C-3 at δ_C 70.4; and C-1 at δ_C 174.8) and the ¹H–¹H COSY correlations (H-3/H-3a) indicated a five-membered lactone ring in 1. Besides, HMBC correlations from H-3 to C-4, H-4 to C-3, H-5 to C-3a, and from H-6 to C-7a, together with analysis of the spin system II (Fig. 2), demonstrated the presence of a cyclohexenyl ring fused to positions C-3a and C-7a on the five-membered ring. Specifically, the long range HMBC correlation (Fig. 2, blue arrow) from H-6 to C-1 confirmed the above proposal.

Fig. 2 Key ¹H–¹H COSY and HMBC correlations of compounds 1–5.

The aforementioned information only accounted for eight degrees of unsaturation, which suggested the existence of one more ring in 1. The ¹H–¹H COSY correlations of H-5/H9/H8 and the HMBC correlations from H-8 to C-7a, C-1, C-3a, C-5, and C-7 and from H-9 to C-7a, C-4, and C-6 suggested that C-5 and C-7a were linked by a –CH₂CH₂– bridge. Thus, the planar structure of 1 was finally established as shown (Fig. 1).

The relative configuration of 1 was deduced from the NOESY spectrum. Key NOE interactions observed between H-3/H-4, H-4/H-8, and H-4/H-9 suggested that these protons were positioned on the same side of the cyclohexenyl ring. The abovementioned information, together with the observations of the NOE interactions of H-4/H-6′, H-2′/H-5, and H-2′/H-3a, established the structure of 1 as shown (Fig. 3), which is highly consistent with the lowest energy conformer generated from the theoretical conformational analysis.

Fig. 3 Key NOESY correlations of compounds 1–5.

The absolute configuration of 1 was determined by the Harada–Nakanishi nonempirical rule for exciton chirality circular dichroism (CD).^32,33 Compound 1 shows a positive Cotton effect at 233 nm and a negative Cotton effect at 218 nm. This was assigned to exciton coupling between the π–π* transitions of the two chromophores: the phenyl ring and the α,β-unsaturated ketone chromophores (Fig. 4).

Fig. 4 The positive exciton chirality of compounds 1–5.

To confirm the absolute configuration of 1, the electronic circular dichroism (ECD) spectra of (3aR, 4S, 5R, and 7aS) and (3aS, 4R, 5S, and 7aR) were calculated using the time-dependent density functional theory (TDDFT) method at the B3LYP/6-31+G(d) level. The calculated ECD spectrum of (3aR, 4S, 5R, and 7aS) showed the same pattern as the experimental ECD spectrum of 1 and was generally opposite to that of (3aS, 4R, 5S, and 7aR) (Fig. 6). Thus, the absolute configuration of 1 was confirmed to be 3aR, 4S, 5R, and 7aS and 1 was named chuanxiongnolide L1.

Fig. 5 CD and UV spectra of compounds 1–5 in MeOH.

Fig. 6 Comparison of the experimental ECD spectrum of 1 in MeOH (red) with calculated ECD spectra for 3aS, 4R, 5S, and 7aR (black) and 3aR, 4S, 5R, and 7aS (blue).

Compound 2 was obtained as a pale yellow oil. Its molecular formula was determined to be C₂₄H₂₈O₅ by HRESIMS (m/z 397.2002, [M + H]⁺, calcd for C₂₄H₂₉O₅, 397.2015) and ¹³C NMR data, indicating 11 degrees of unsaturation. Its IR spectrum exhibited absorption bands for conjugated carbonyl (1767 cm⁻¹) and double bond (1719 cm⁻¹) functionalities. Detail analysis of the ¹H NMR spectroscopic data and ¹H–¹H COSY spectrum indicated that 2 had a butylidene as a side chain, which was linked to a tetrasubstituted carbon, as confirmed by its HMBC correlations from H-9 (δ_H 2.30, m) to C-3 (δ_C 148.2). Besides, its ¹H NMR spectrum shows another olefinic proton at δ_H 7.72 (d, J = 6.8 Hz, H-8′), which was attached to a methylene carbon (C-8′, δ_C 150.4) and correlated to the ester carbon at δ_C 160.8 (C-1′) in its HMBC spectrum. These signals indicated the presence of one α,β-unsaturated γ-lactone. With the aid of DEPT and HSQC spectral data, additional signals assignable to one methyl group, four methines (one oxygenated), and six methylenes were identified in its ¹H and ¹³C NMR spectra and their connectivity was confirmed by analysis of the ¹H–¹H COSY and HMBC spectra (Fig. 2). The aforementioned information, in combination with the biogenetic considerations and MS data, indicated that 2 was a phthalide dimer. The NMR spectroscopic data of 2 were similar to those of levistolide A^18,27 (6), a normal phthalide dimer in CXR, except for the absence of resonances of the Δ^3′,8′ double bond at the butylidene side chain and the presence of resonances assignable to one oxygenated methine at 3′ (δ_H 4.85/δ_C 84.9) and one ketone carbonyl group at C-4′ (δ_C 205.6). The downfield chemical shift of C-4a′ (δ_C 49.2), HMBC correlations from H-3′ to C-4′ and H-9′ to C-4′, and a weak long range correlation from H-8′ to C-4′ (Fig. 2, blue arrow) located the carbonyl group at C-4′, which implied 2 was an E-ring expanded phthalide dimer derived from levistolide A (6) (Scheme 1).

Scheme 1 Proposed biogenetic pathway to compounds 1 and 2.

Similarly, the relative configuration of 2 was determined through NOE correlations and vicinal coupling constants. The small coupling constant (J = 8.3 Hz) of H-6 and H-7 indicated that they were in the same orientation. NOE correlations of H-5′/H-7, H-6/H-6′, and H-3′/H-5′ observed in the NOESY spectrum indicated that only the configuration in Fig. 3 could give such stereostructure correlations. The absolute configuration of 2 was determined by the same exciton chirality CD method as 1. Compound 2 showed a positive Cotton effect at 242 nm and a negative Cotton effect at 210 nm. This was assigned to exciton coupling between the π–π* transitions of the two conjugated ester chromophores. The positive first Cotton effect indicated the positive chirality between the two axes of electric transition moments (Fig. 4). Thus, the absolute configuration of 2 was confirmed as 3′R, 4a′R, 7′R, 6R, and 7R and 2 was named chuanxiongnolide L2. Besides, the NOE correlation (Fig. 3) of H4/H8 indicated the Z configuration of the butylidene side chain.

Compound 3 was obtained as a pale yellow oil. The HRESIMS of 3 gave a [M + H]⁺ ion peak at m/z 341.1378 (calcd for C₂₀H₂₁O₅, 341.1389), corresponding to a molecular formula of C₂₀H₂₀O₅, which suggested 11 degrees of unsaturation. Comparison of the ¹H and ¹³C NMR data of 3 with those of the known compound, levistolide A (6), revealed that the ¹H NMR and ¹³C NMR resonances of another butylidene side chain had disappeared, and the replacement of an ester carbonyl group at C-3 (δ_C 163.8) in 3 was observed, indicating that an oxidative cleavage of the Δ^3,8 double in the butylidene side chain of levistolide A (6) was responsible for this structural change. The HMBC correlation from H-4 (δ_H 1.99 and 2.36) to C-3 (δ_C 163.8) confirmed the above proposal (Fig. 2). The relative configuration of 3 was elucidated on the basis of the analysis of the NOE correlations, as well as the coupling constant values. The NOE correlations between H-7 and H-4′ and H-6 and H-5′ observed in the NOE spectra, together with the small coupling constant (J = 8.8 Hz) of H-6 and H-7, established the relative configuration, as shown in Fig. 3. The Z configuration of the butylidene side chain was confirmed by the NOE correlation of H-4′ and H-8′. The absolute configuration of 3 was assigned using the CD exciton chirality method. Compound 3 exhibited a split CD curve with the positive first Cotton effect at 246 nm and the negative second Cotton effect at 223 nm, which were caused by the transition reaction from the two chromophores (Fig. 4 and 5). Therefore, the absolute configuration of 3a′R, 6′R, 6R, and 7R was assigned for 3 and it was named chuanxiongnolide L3.

Compound 4 was obtained as a pale yellow oil and was assigned the molecular formula C₂₄H₂₈O₆ for its HRESIMS [M + Na]⁺ at m/z 435.1778 (calcd for C₂₄H₂₈O₆Na, 435.1784). Taking the MS data into consideration, 4 was also inferred to be a dimeric phthalide. Its ¹H NMR and ¹³C NMR spectroscopic data were similar to those of levistolide A (6), except for some signals assignable to the butylidene side chain. The absence of the proton signal assignable to H-3 in the ¹H NMR spectrum of 3 and the observation of the down field chemical shift of the carbon signals of C-3 in its ¹³C NMR spectrum, in combination with it MS data, indicated that the hydroxyl group was attached to C-3. Besides, the replacement of the methylene at C-8 by a ketone group was confirmed by the observation of an additional carbonyl resonance (C-8, δ_C 202.5) in its ¹³C NMR spectrum and corresponding HMBC correlations from 3-OH to C-3 and C-8, from H-9 to C-3, and from H-10 to C-8. Thus, the planar structure of 4 was established as shown (Fig. 1). It could be deduced that 4 was generated by the oxidation of the Δ^3,8 double bond of levistolide A (6). The relative configuration of 4 was inferred by a NOESY experiment and the ¹H NMR spectroscopic data. The cross-peaks of H-7/H-4′ and H-6/H-5′ observed in the NOESY spectrum suggested that H-6, H-7, H-4′, and H-5′ were on the same face. Due to the deshielding effect of the ketone carbonyl group, the downfield chemical shifts of H-4 were observed in its ¹H NMR spectrum (Table 1), which helped to establish the relative configuration as shown (Fig. 3). The applied CD exciton chirality method determined the absolute configuration of 4 as 3a′R, 6′R, 3R, 6R, and 7R and 4 was named chuanxiongnolide L4.

Compound 5 was obtained as a pale yellow oil. The HRESIMS of 5 gave a [M + H]⁺ ion peak at m/z 415.2121 (calcd for C₂₄H₃₁O₆, 415.2121), corresponding to a molecular formula of C₂₄H₃₀O₆. The ¹H NMR and ¹³C NMR spectroscopic data resembled closely those of 4 (Tables 1 and 2), except for the absence of the ¹³C NMR signal for the carbonyl group and the presence of a signal for an oxygenated-methine at δ_C 73.7, which was supported by HMBC correlations from H-8 (δ_H 3.54, dd, J = 2.2, 10.4 Hz) to C-3 (δ_C 106.3), C-9 (δ_C 32.5), and C-10 (δ_C 19.0) and from H-9 (δ_H 1.13 and 1.71, m) to C-8 (δ_C 73.7). The aforementioned information indicated that 5 was also generated by the oxidation of the Δ^3,8 double bond of levistolide A (6). The relative configuration of 5 was deduced by the NOE correlations and its absolute configuration was determined by the CD method (Fig. 5) as 3a′R, 6′R, 6R, and 7R and 5 was named chuanxiongnolide L5. As for the absolute configurations of the chirality centers of C-3 and C-8, further X-ray diffraction analysis was needed.

Table 2 ¹³C NMR spectroscopic data (400 MHz, CHCl₃) for compounds 1–5 (δ in ppm)

No.	1	2	3	4	5
1	174.8, C	169.0, C	163.3, C	169.7, C	170.2, C
3	70.4, CH2	148.2, C	163.8, C	102.0	106.3, C
3a	52.8, CH	156.0, C	149.3, C	164.2	165.4, C
4	46.1, CH	19.1, CH2	19.5, CH2	20.3, CH2	21.4, CH2
5	40.1, CH	27.8, CH2	28.5, CH2	29.4, CH2	29.3, CH2
6	142.4, CH	37.9, CH	38.5, CH	38.5, CH	38.6, CH
7	138.2, C	41.1, CH	42.4, CH	41.3, CH	41.0, CH
7a	46.0, C	125.2, C	145.0, C	133.2, C	131.1, C
8	22.1, CH2	113.6, CH		202.5, C	73.7, CH
9	27.6, CH2	28.3, CH2		36.2, CH2	32.5, CH2
10		22.5, CH2		16.9, CH2	19.0, CH2
11		13.7, CH3		13.3, CH3	14.1, CH3
1′	135.9, C	160.8, C	164.2, C	164.7, C	165.0, C
2′	109.5, CH
3′	146.7, C	84.9, CH	149.8, C	150.3, C	150.6, C
3′-OCH3	56.0, CH3
3′a			47.2, C	47.7, C	47.5, C
4′	144.7, C	205.6, C	31.0, CH2	31.7, CH2	31.5, CH2
4′-OH	5.52 brs
4a′		49.2, C
5′	114.8, CH	29.2, CH2	26.0, CH2	25.7, CH2	25.7, CH2
6′	119.4, CH	27.2, CH2	41.3, CH	41.6, CH	41.7, CH
7′		40.0, CH	142.9, CH	143.0, CH	142.8, CH
7′a			133.4, C	134.3, C	134.3, C
8′		150.4, CH	109.3, CH	109.3, CH	108.8, CH
8a′		129.8, C
9′		37.0, CH2	27.5, CH2	27.6, CH2	27.5, CH2
10′		18.7, CH2	22.3, CH2	22.5, CH2	22.4, CH2
11′		14.1, CH3	14.0, CH3	14.1, CH3	13.6, CH3
1′′	163.7, C
3′′	61.2, CH2
4′′	14.2, CH3

---

Structurally, compound 1 was determined to possess an unprecedented carbon skeleton. As mentioned, phthalides and phenols are two types of normal constituents in CXR and thus they were considered responsible for the formation of 1 (Scheme 1). 4,5-Dihydro-1,3-isobenzofurandione³⁴ (A), a monomeric phthalide isolated from the same genus (Angelica sinensis), underwent the “endo” Diels–Alder addition with a normal phenol, coniferyl alcohol, which was formed by hydrolysis of a coniferyl ferulate³⁵ contained in CXR, to yield 1a. Then, 1a underwent hydrolysis to produce 1b. Ultimately, an esterification reaction was the key step to convert 1b to 1. Different from those normal phthalide dimers featuring a 5/6/6/6/5 fused-ring system, compound 2 possesses the first example of an E-ring expanded phthalide skeleton with a 5/6/6/6/6 fused-ring system. As shown in Scheme 1, oxidation, hydrolysis, and intramolecular esterification reactions were possibly responsible for the formation of 2.

All the isolates were evaluated for their inhibitory activity against LPS-induced NO production in RAW 264.7 macrophage cells. Indomethacin (IND), a nonselective cyclooxygenase inhibitor, and L-N⁶-(1-iminoethyl)-lysine (L-NIL), a selective inhibitor of iNOS, were selected as positive controls (half maximal inhibitory concentration (IC₅₀) were 9.4 and 55.8 μM, respectively). Cell viability was measured using a 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazoliumbromide (MTT) method. The cell viability of all tested compounds should be above 95% at the treated concentration during incubation for 24 h. Thus, in this experiment, the concentrations used in the bioassay were determined to be 3.13–50 μM for compounds 1–5 and 0.625–10 μM for compounds 6–8, respectively. Compounds 1–2 show inhibitory effects on LPS-induced NO production in RAW 264.7 macrophages with IC₅₀ values of 12.0 and 12.6 μM (Table 3), respectively. In addition, the inhibitory potentials of compounds 6 and 7 were much stronger than those of L-NIL. Comparison of the IC₅₀ values of compounds 2 (12.6 μM) and 6 (6.6 μM) with those of compounds 3–5 (IC₅₀ > 50 μM) indicated that the Δ^3,8 double bond was important to the NO inhibitory effect.

Table 3 Inhibitory effect of compounds 1–8 on LPS-activated NO production in RAW 264.7 macrophage cells^a

Compound	IC50 (μM)
a L-NIL: L-N^6-(1-iminoethyl)-lysine. IND: indomethacin. Results were obtained from three independent experiments and expressed as mean ± SD.
1	12.0 ± 3.7
2	12.6 ± 1.4
3	>50
4	>50
5	>50
6	6.6 ± 0.6
7	3.0 ± 0.5
8	>10
L-NIL	9.4 ± 1.6
IND	55.8 ± 1.2

---

Conclusions

In summary, five new phthalide derivatives, chuanxiongnolide L1–L5 (1–5), and three known phthalide dimers (6–8) were isolated from the rhizomes of L. chuanxiong. Their structures were determined by spectroscopic data and a CD exciton chirality method. To confirm the absolute configuration of compound 1, ECD calculations were also conducted. Compound 1 represents a novel type of phthalide derivative. Compound 2 is the first example of an E-ring expanded phthalide dimer with a 5/6/6/6/6 fused-ring system. Compounds 1, 2, 6, and 7 exhibited potential inhibitory effects on LPS-induced NO production in RAW 264.7 macrophages with IC₅₀ values ranging from 3.0 to 12.6 μM. Plausible biosynthetic routes of 1 and 2 were proposed and the preliminary structure–activity relationship of compounds 2–6 was also discussed.

Experimental section

General experimental procedures

Optical rotations were obtained on an Autopol III polarimeter (Rudolph Research Analytical, Flanders, NJ, USA). Ultraviolet (UV) spectra were measured with a Cary 300 UV-visible (Vis) spectrophotometer (Varian, Inc., Palo, Alto, CA, USA). IR spectra were acquired on a Nexus 470 Fourier transform IR (FT-IR) spectrometer (Thermo Nicolet, Inc., Madison, WI, USA) using KBr disks. CD spectra were recorded on a JASCO J-810 spectropolarimeter (Jasco, Hachioji, Tokyo, Japan). HRESIMS was recorded on a Waters Xevo G2 Q-TOF mass spectrometer (Waters, Milford, MA, USA). 1D and 2D NMR spectra were recorded on a Bruker AV 400 spectrometer (Bruker, Karlsruhe, Baden-Wuerttemberg, Germany) with tetramethylsilane (TMS) as an internal standard. Column chromatographic (CC) separations were carried out using silica gel (200–300 mesh; Qingdao Marine Chemical Factory, China), and thin-layer chromatography (TLC) was conducted on silica gel GF₂₅₄ plates from the same company. TLC spots were viewed under UV light at 254 or 365 nm. A Phenomenex Prodigy C₁₈ column (250 × 21.2 mm i.d., 10 μm; Phenomenex, Torrance, CA, USA) was used for RP-SP-HPLC and the separations were performed on a RP-SP-HPLC system consisting of a LabTech P600 pump, a LC3000 UV detector, and a 7125 Rheodyne injector (Rheodyne, Cotati, CA, USA) with a loop of 1 mL. The LC Workstation was Labtech Chromsoftware (LabTech Co., Beijing, China). Reagents of analytical grade were purchased from Beijing Chemical Works (Beijing, China). Acetonitrile (MeCN) of chromatographic grade was purchased from Tianjin Xihua Special Type Reagent Factory (Tianjin, China). The murine macrophage cell line RAW 264.7 was obtained from the Cell Resource Center, IBMS, CAMS/PUMC (Beijing, China). Dulbecco's modified Eagle's medium (DMEM), phosphate buffered saline (PBS), and fetal bovine serum (FBS) were purchased from Gibco™ (Grand Island, NY, USA). 96-Well plates were purchased from Corning Costar (Corning Inc., Cambridge, MA, USA). Dimethylsulphoxide (DMSO), LPS, Griess reagent, MTT, IND, and L-NIL were purchased from Sigma-Aldrich Co. (St. Louis, MO, USA).

Plant material

The rhizomes of L. chuanxiong were collected from the Xingquan village of Aoping town in Pengzhou city of the Sichuan Province of the People's Republic of China in May 2015 and were identified by Prof. Xiu-Wei Yang of the School of Pharmaceutical Sciences, Peking University Health Science Center, Peking University. A voucher specimen (accession number: 20150520CXR) was deposited at the State Key Laboratory of Natural and Biomimetic Drugs (Peking University, Beijing, China).

Extraction and isolation

The air-dried rhizomes of L. chuanxiong (10.5 kg) were powdered and extracted with 95% EtOH : aqueous solution (20 L × 72 h × 8) at room temperature. The extract was evaporated under reduced pressure to obtain a residue (1.5 kg). The residue was fractionated by a silica gel column eluted with a gradient of petroleum ether–acetone (100 : 1, 80 : 1, 50 : 1, 30 : 1, 15 : 1, 5 : 1, and 1 : 1, v/v) to give seven fractions (Fr1–Fr7) based on TLC analysis. Fr3 (13 g) was subjected to an open octadecylsilane (ODS) column eluted with a step gradient of 40–90% MeCN in H₂O to give six subfractions (Fr3.1–6). Fr3.4 (1.5 g) was purified by Sephadex LH-20 CC to yield compounds 1 (9.0 mg) and 6 (10.0 mg). Fr4 (9.1 g) was subjected to ODS CC eluting with a step gradient of 30–100% MeOH in H₂O to yield eight subfractions (Fr4.1–8). Fr4.2 (500.0 mg) was first separated by Sephadex LH-20 CC using CH₂Cl₂–MeOH (1 : 1, v/v) for elution and then purified by RP-SP-HPLC eluted with MeOH–H₂O (35 : 65, v/v, 8 mL min⁻¹, detector UV_max 230 nm) to afford compounds 5 (25.1 mg, retention time (t_R) 121 min) and 7 (7.3 mg, t_R 133 min). Fr4.4 (10.0 mg) was first separated by RP-SP-HPLC eluted with MeCN–H₂O (65 : 35, v/v, 5 mL min⁻¹, detector UV_max 230 nm) and then purified by a Sephadex LH-20 CC using CHCl₃–MeOH (1 : 1, v/v) to give compound 2 (3.0 mg). Fr5 (301.4 mg) was first purified by Sephadex LH-20 CC using CHCl₃–MeOH (1 : 1, v/v) for elution and then subjected to a silica gel column eluting with CH₂Cl₂–EtOAc (60 : 1, v/v) to give compound 8 (15.5 mg). Fr6 (300.6 mg) was subjected to a silica gel column eluted with CH₂Cl₂–acetone (40 : 1, v/v), followed by RP-SP-HPLC eluted with MeCN–H₂O (55 : 45, v/v, 8 mL min⁻¹, detector UV_max 254 nm) to give compounds 3 (60.4 mg, t_R 49 min) and 4 (75.5 mg, t_R 66 min).

Chuanxiongnolide L1 (1). pale yellow oil; [α]²⁰_D + 23.1 (c 0.1, MeOH); UV (MeOH) λ_max 233 nm (sh); IR (KBr) ν_max cm⁻¹: 3385, 2931, 1772, 1712, 1516, 1260 cm⁻¹; ¹H and ¹³C NMR data, see Tables 1 and 2; HRESIMS m/z 359.1491 [M + H]⁺ (calcd for C₂₀H₂₃O₆, 359.1495).

Chuanxiongnolide L2 (2). pale yellow oil; [α]²⁰_D + 16.3 (c 0.1, MeOH); UV (MeOH) λ_max 243 nm (sh); IR (KBr) ν_max cm⁻¹: 3417, 2958, 2935, 1719, 1364, 1242 cm⁻¹; ¹H and ¹³C NMR data, see Tables 1 and 2; HRESIMS m/z 397.2002 [M + H]⁺ (calcd for C₂₄H₂₉O₅, 397.2015).

Chuanxiongnolide L3 (3). pale yellow oil; [α]²⁰_D + 13.9 (c 0.1, MeOH); UV (MeOH) λ_max 240 nm (sh); IR (KBr) ν_max cm⁻¹: 2955, 1775, 1711, 1434, 1361, 1269 cm⁻¹; ¹H and ¹³C NMR data, see Tables 1 and 2; HRESIMS m/z 341.1378 [M + H]⁺ (calcd for C₂₀H₂₁O₅, 341.1389).

Chuanxiongnolide L4 (4). pale yellow oil; [α]²⁰_D + 26.1 (c 0.1, MeOH); UV (MeOH) λ_max 237 nm (sh); IR (KBr) ν_max cm⁻¹: 3381, 2957, 2932, 1773, 1723, 1268, 1217, 1068 cm⁻¹; ¹H and ¹³C NMR data, see Tables 1 and 2; HRESIMS m/z 435.1778 [M + Na]⁺ (calcd for C₂₄H₂₈O₆Na, 435.1784).

Chuanxiongnolide L5 (5). pale yellow oil; [α]²⁰_D + 9.7 (c 0.1, MeOH); UV (MeOH) λ_max 235 nm (sh): 3400, 2956, 2935, 1770, 1709, 1269 cm⁻¹; ¹H and ¹³C NMR data, see Tables 1 and 2; HRESIMS m/z 415.2121 [M + H]⁺ (calcd for C₂₄H₃₁O₆, 415.2121).

ECD calculations

The relative configurations of compound 1 were initially established according to the NOESY correlations. A SYBYL-X V1.1.2 software package was used for conformational analysis using MMFF94S force field with an energy cut off of 10 kcal mol⁻¹. In the Gaussian 09 software, the conformers obtained were used for geometry reoptimizations at the B3LYP/6-31G(d) level and the PCM solvent model was set as methanol. At the B3LYP/6-31+G(d) level in methanol, TDDFT ECD calculations for the optimized conformers were conducted. Using SpecDis v1.51 software, ECD curves were obtained based on their rotatory strengths with a half-band of 0.3 eV, and the final ECD spectra were obtained according to the Boltzmann-calculated contribution of each conformer.

Cell viability

Cell viability were evaluated by a conventional MTT assay.³⁶ RAW 264.7 cells were cultured in DMEM, supplemented with 10% FBS in a 37 °C humidified incubator with a 5% CO₂/95% air atmosphere prior to seeding of 96-well plates for the designated experiments. Briefly, the cells were seeded at a density of 3 × 10⁵ cells per mL in 100 μL aliquots in a 96-well culture plate. After overnight incubation, the cells were treated with LPS (1 μg mL⁻¹) in the absence or presence of the test compounds at different concentrations. After 24 h incubation, the supernatant (100 μL) was removed and MTT stock solution (20 μL, 5 mg mL⁻¹) was added to each well again to incubate for 4 h. Then, a dissolving solution containing 10% sodium dodecyl sulphate (100 μL), 5% isopropanol, and HCl (0.012 M) was added to each well to incubate for 8 h, and the absorbance was determined at 492 nm.

NO inhibition assay

Using the Griess method,^37,38 the nitrite concentration was determined in the culture supernatants. Briefly, after 24 h incubation with tested compounds, the culture supernatant (100 μL) was removed into another 96-well plate, where nitrite reacted with the standard Griess reagent (100 μL) for 15 min. The optical density of each well was measured at 540 nm by a Microplate Reader (Multiskan MK3, Thermo, USA). Fresh culture media were used as blanks and the IC₅₀ values were obtained using the software origin 7.5. All the assays were conducted in triplicate, and the results were expressed as the mean ± SD of the three independent experiments. The nitrite concentration was determined by using a standard curve of sodium nitrite made up in DMEM. NO inhibitory rate (%) = [C_NO(LPS) − C_NO(compound)]/C_NO(LPS) × 100, where C is the concentration.

Acknowledgements

This research was partly supported by the National Natural Science Foundation of China (81473321), the National Key Technology R & D Program of China (2011BAI07B08), and the Beijing Municipal Natural Science Foundation (7152086).

References

1. Chinese Pharmacopoeia Commission, Pharmacopoeia of the People's Republic of China, China Medical Science and Technology Press, Beijing, 2010, vol. I, p. 38.
2. W. X. Li, Y. P. Tang, Y. Y. Chen and J. A. Duan, Molecules, 2012, 17, 10614–10651.
3. M. Zhang, S. L. Chen, S. W. Seto, Y. W. Kwan and S. W. Chan, Pharm. Biol., 2009, 47, 366–371.
4. K. J. Chen and K. Chen, China Med. J., 1992, 105, 870–873.
5. M. Li, S. Handa, Y. Ikeda and S. Goto, Thromb. Res., 2001, 104, 15–28.
6. L. Zhang, Y. R. Jiang, C. Y. Guo, C. F. Wu, K. J. Chen and H. J. Yin, Chin. J. Integr. Med., 2009, 15, 359–364.
7. Q. Liu, J. D. Zhou, J. G. Yu, Y. X. Xie, X. Y. Jiang, H. Yang and X. Q. Chen, J. Chromatogr. A, 2014, 1364, 204–213.
8. J. Huang, X. Q. Lu, C. Zhang, J. Lu, G. Y. Li, R. C. Lin and J. H. Wang, Fitoterapia, 2013, 91, 21–27.
9. Q. H. Yang, Y. Liang, Q. Xu, Y. Zhang, L. Xiao and L. Y. Si, Phytomedicine, 2011, 18, 1148–1152.
10. T. C. T. Or, C. L. H. Yang, A. H. Y. Law, J. C. B. Li and A. S. Y. Lau, Neuropharmacology, 2011, 60, 823–831.
11. C. P. Miao, S. H. Wu, B. Z. Luo, J. Wang and Y. W. Chen, Fitoterapia, 2010, 81, 1088–1090.
12. R. Yan, S. L. Li, H. S. Chung, Y. K. Tam and G. Lin, J. Pharm. Biomed. Anal., 2005, 37, 87–95.
13. L. Y. Wang, Y. P. Tang, X. Liu, M. Zhu, W. W. Tao, W. X. Li and J. A. Duan, Chin. J. Nat. Med., 2015, 13, 401–408.
14. J. Huang, X. Q. Lu, C. Zhang, J. Lu, G. Y. Li, R. C. Lin and J. H. Wang, Fitoterapia, 2013, 91, 21–27.
15. Y. L. Lin, G. J. Wang, C. L. Huang, Y. C. Lee, W. C. Liao, W. L. Lai, Y. J. Lin and N. K. Huang, J. Ethnopharmacol., 2009, 122, 417–423.
16. S. S. K. Chan, A. O. K. Choi, R. L. Jones and G. Lin, Eur. J. Pharmacol., 2006, 537, 111–117.
17. T. F. Lee, Y. L. Lin and Y. T. Huang, Planta Med., 2007, 73, 527–534.
18. T. Naito, T. Katsuhara, K. Niitsu, Y. Ikeya, M. Okada and H. Mitsuhashi, Heterocycles, 1991, 32, 2433–2442.
19. M. J. Liang, L. C. He and G. D. Yang, Life Sci., 2005, 78, 128–133.
20. H. Y. Peng, J. R. Du, G. Y. Zhang, X. Kuang, Y. X. Liu, Z. M. Qian and C. Y. Wang, Biol. Pharm. Bull., 2007, 30, 309–312.
21. X. Kuang, Y. Yao, J. R. Du, Y. X. Liu, C. Y. Wang and Z. M. Qian, Brain Res., 2006, 1102, 145–153.
22. J. Wang, J. R. Du, Y. Wang, X. Kuang and C. Y. Wang, Acta Pharmacol. Sin., 2010, 31, 791–797.
23. T. Naito, T. Katsuhara, K. Niitsu, Y. Ikeya, M. Okada and H. Mitsuhashi, Phytochemistry, 1992, 31, 639–642.
24. T. Naito, Y. Ikeya, M. Okada, H. Mistuhashi and M. Maruno, Phytochemistry, 1996, 41, 233–236.
25. L. S. Lim, P. Shen, Y. H. Gong and E. L. Yong, Phytochemistry, 2006, 67, 728–734.
26. Y. H. Li, S. L. Peng, Y. Zhou, K. B. Yu and L. S. Ding, Planta Med., 2006, 72, 652–656.
27. M. Kaouadji, F. De Pachtere, C. Pouget, A. J. Chulia and S. Lavaitte, J. Nat. Prod., 1986, 49, 872–877.
28. C. Bogdan, Nat. Immunol., 2001, 2, 907–916.
29. Y. Dai, P. P. H. But, Y. P. Chan, H. Matsuda and M. Kubo, Biol. Pharm. Bull., 2002, 25, 1175–1178.
30. L. Liu, Z. Q. Ning, S. Shan, K. Zhang, T. Deng, X. P. Lu and Y. Y. Cheng, Planta Med., 2005, 71, 808–813.
31. S. x. Deng, S. N. Chen, P. Yao, D. Nikolic, R. B. van Breemen, J. L. Bolton, H. H. S. Fong, N. R. Farnsworth and G. F. Pauli, J. Nat. Prod., 2006, 69, 536–541.
32. M. Koreeda, N. Harada and K. Nakanishi, J. Am. Chem. Soc., 1974, 96, 266–268.
33. N. Harada and K. Nakanishi, Circular dichroic spectroscopy-exciton coupling in organic stereochemistry, University Science Books/Oxford University Press, Mill Valley, CA/Oxford, 1983.
34. Y. Pei, C. B. Tan, W. R. Xu, P. Liu, B. N. Liu, W. Liu, C. L. Chen and L. D. Tang, Chin. Herb. Med., 2010, 2, 236–241.
35. T. Yi, K. S. Y. Leung, G. H. Lu, K. Chan and H. Zhang, Chem. Pharm. Bull., 2006, 54, 255–259.
36. T. Mosmann, J. Immunol. Methods, 1983, 65, 55–63.
37. L. C. Green, D. A. Wagner, J. Glogowski, P. L. Skipper, J. S. Wishnok and S. R. Tannenbaum, Anal. Biochem., 1982, 126, 131–138.
38. W. Wei, X. W. Wu, G. G. Deng and X. W. Yang, Phytochemistry, 2016, 123, 58–68.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra10023f

---