UV-guided isolation of polyynes and polyenes from the roots of Codonopsis pilosula

Abstract

The UV-guided isolation of polyacetylenes from the crude extract of Codonopsis pilosula has successfully led to the characterization of five new polyynes, pilosulynes A–E (1–5), and two new polyenes, pilosulynes F and G (6 and 7), as well as five known analogues (8–12). Their structures were determined by spectroscopic methods, including ICD and 1D/2D NMR experiments. The absolute configurations of the 6,7-diol moiety of the isolated compounds were determined by the Snatzke's method, which revealed an induced circular dichroism after the addition of dimolybdenum tetraacetate in DMSO. Compound 6 exhibited anti-HCV activity in the HCVcc infection assay with an EC₅₀ value of 47.2 μM.

---

Introduction

Codonopsis pilosula (Franch.) Nannf. belongs to the family of Campanulaceae and its root is often utilized in traditional Chinese medicine (TCM). It has been used commonly in China to replenish qi, invigorate the spleen, and nourish the lungs, and it is sometimes used as a substitute for the much more costly Panax ginseng.¹ The roots of C. pilosula have a mild, sweet taste and are used in some Chinese food therapy recipes; for example, chicken soup may be prepared with it and other mild herbs to make a good flavor tonic.

Water-soluble polysaccharides from Radix Codonopsis pilosulae, the roots of Codonopsis pilosula, were reported to exhibit an immuno regulatory activity. Health supplementary products derived from Radix Codonopsis pilosulae have been developed due to their potent immunological activity, similar to the polysaccharides of Lucid Ganoderma (Ganoderma lucidum). Polyacetylene (polyyne) glycosides, such as lobetyolin (10) and lobetyolinin (11),^2,3 are also water-soluble constituents found in the water extract of C. pilosula. Thus, it is important to understand the biological activity of the polyynes obtained from C. pilosula. Its main constituents include sterols, triterpenes, glycosides, alkaloids, polyynes, and polysaccharides.^4–7 Recently, lobetyolin, a polyyne glycoside, was demonstrated to significantly stimulate the nuclear factor-kappa B (NF-κB),⁸ a transcriptional factor found in almost all types of animal cells and plays an important role in regulating the immune response to infection; it participates in inflammatory response, cancer, autoimmune disease, and synaptic plasticity.⁹ However, little is known about the biological properties and absolute configurations of the polyynes in C. pilosula. In an attempt to investigate the bioactive polyynes from a water extract, a series of extraction, isolation, and structural elucidation experiments were undertaken, which have resulted in the characterization of five new polyynes (1–5) and two new polyenes (6 and 7), as well as five known analogues (8–12).

Results and discussion

The H₂O extract of the roots of C. pilosula was subjected to Diaion HP-20 column chromatography. The fractions containing polyynes or polyenes were selected based on their characteristic UV spectra, which exhibited a palm-like or saw-tooth-like shape between 225 and 300 nm with a maximum absorption around 270 nm.¹⁰ The UV-guided fractionation successfully led to the isolation of five new polyynes (1–5) and two new polyenes (6 and 7), as well as five known analogues (8–12). The known compounds were identified as lobetyol (8),³ tetradeca-4E,8E,12E-triene-10-yne-1,6,7-triol (9),¹¹ lobetyolin (10),³ lobetyolinin (11),² and cordifolioidyne B (12)¹² by comparison of their physical and spectroscopic data with those reported in the literature.

The HRESIMS spectrum of pilosulyne A (1) exhibited a pseudomolecular ion peak at m/z 275.1261 [M + Na]⁺, which is consistent with a molecular formula of C₁₄H₂₀O₄ and 5 degrees of unsaturation. ¹³C NMR and DEPT spectroscopic data (Table 1) displayed 14 carbon signals, including 6 methylenes, 4 methines, and 4 quaternary carbons. UV (283, 267, 253, 240 nm) and IR (2164 cm⁻¹) spectra were closely correlated to those of known polyacetylenes, lobetyol (8), lobetyolin (10), and lobetyolinin (11).^2,3 In addition, proton resonances at δ 6.41 (1H, dt, J = 15.5, 4.6 Hz), and 5.81 (1H, dd, J = 15.5, 2.1 Hz) (Table 2) and carbon resonances at δ 82.5 (C), 70.7 (C), 74.2 (C), 77.4 (C), 108.5 (CH), and 148.2 (CH) (Table 1) corroborated the presence of two C≡C bonds and one C=C bond in conjugation.^2,3 This double bond was assigned as E owing to the large vicinal coupling constant (J = 15.5 Hz) between alkene protons. The ¹H–¹H COSY correlations between H-12/H-13 and H-13/H₂-14 and HMBC correlations from H-13 to C-11 and H-12 to C-10 indicated the attachment of a hydroxymethyl group at C-13 (Fig. 1). Two mutually coupled protons due to the hydroxyl-containing methines at δ 3.52 (1H, m) and 4.24 (1H, d, J = 6.2 Hz) were assigned to C-6 and C-7, respectively. This observation was corroborated by the presence of HMBC correlations from H-7 to C-8, C-9, and C-6, as well as from H-6 to C-8 (Fig. 1). Considering the molecular formula and degrees of unsaturation, it can be inferred that the remaining five sp³ carbons must be four aliphatic carbons (δ 33.4, CH₂; 26.9, CH₂; 26.6, CH₂; 33.6, CH₂), and one hydroxyl-containing methylene was assumed to be a hydroxypentyl functionality attached to C-6. The relevant NMR data in the literature suggest that threo and erythro vic-diols with similar partial structures have coupling constants of 6.0–7.0 Hz for threo diols and 3.0–4.0 Hz for erythro diols.^13–15 Accordingly, the vic-diol group of 1 was inferred to have a threo-configuration between H-6 and H-7 because of the value of its coupling constant (J = 6.2 Hz).

Table 1 ¹³C NMR spectroscopic data of compounds 1–3, 6, and 7

No.	1, δC, mult.^a	2, δC, mult.^b	3, δC, mult.^c	6, δC, mult.^a	7, δC, mult.^b
a Spectra were measured in CD3OD (125 MHz).b Spectra were measured in CD3OD (100 MHz).c Spectra were measured in CD3OD (75 MHz).
1	62.9, CH2	62.3, CH2	62.8, CH2	62.9, CH2	62.3, CH2
2	33.6, CH2	33.1, CH2	33.5, CH2	33.6, CH2	33.1, CH2
3	26.6, CH2	29.7, CH2	22.8, CH2	26.9, CH2	29.7, CH2
4	26.9, CH2	135.2, CH	37.7, CH2	26.8, CH2	134.5, CH
5	33.4, CH2	129.6, CH	72.5, CH	33.6, CH2	130.5, CH
6	75.3, CH	76.6, CH	150.6, CH	75.3, CH	76.7, CH
7	67.7, CH	67.8, CH	109.1, CH	76.2, CH	76.2, CH
8	82.5, C	83.8, C	74.1, C	143.9, CH	143.4, CH
9	70.7, C	70.8, C	75.3, C	110.7, CH	110.7, CH
10	74.2, C	74.7, C	66.1, C	88.9, C	88.9, C
11	77.4, C	79.4, C	84.2, C	88.8, C	88.8, C
12	108.5, CH	109.1, CH	16.6, CH2	112.1, CH	121.1, CH
13	148.2, CH	147.8, CH	32.3, CH2	143.4, CH	143.4, CH
14	62.6, CH2	61.1, CH	61.4, CH2	63.0, CH2	62.9, CH2

---

Table 2 ¹H NMR spectroscopic data of compounds 1–3, 6, and 7

No.	1, δH (J in Hz)^a	2, δH (J in Hz)^b	3, δH (J in Hz)^c	6, δH (J in Hz)^a	7, δH (J in Hz)^b
a Spectra were measured in CD3OD (500 MHz).b Spectra were measured in CD3OD (400 MHz).c Spectra were measured in CD3OD (300 MHz).
1	3.56, t (6.6)	3.58, t (6.6)	3.55, t (6.6)	3.55, t (6.5)	3.57, t (6.5)
2	1.56, m	1.65, quin (6.6)	1.53, m	1.56, m	1.62, quin (6.5)
3	1.57, m	2.16, dt (6.9, 6.6)	1.44, m	1.57, m	2.13, dt (7.0, 6.5)
4	1.38, m	5.82, dt (15.4, 6.9)	1.51, m	1.38, m	5.73, dt (15.4, 7.0)
5	1.68, m	5.57, dd (15.4, 6.6)	4.10, m	1.58, m	5.49, dd (15.4, 7.0)
	1.46, m			1.36, m
6	3.52, m	4.00, t (6.6)	6.25, dd (16.2, 5.7)	3.45, m	3.90, dd (7.0, 6.6)
7	4.24, d (6.2)	4.26, d (6.6)	5.72, d (16.2)	4.00, t (6.0)	4.00, dd (6.6, 6.0)
8				6.13, dd (15.6, 6.0)	6.09, dd (15.5, 6.0)
9				5.84, d (15.6)	5.85, d (15.5)
10	5.81, dd (15.5, 2.1)	5.64, br d (11.0)	2.42, t (7.2)	5.87 dd (16.0, 1.6)	5.85, dd (15.5, 1.6)
11	6.41, dt (15.5, 4.6)	6.24, dt (11.0, 6.4)	1.73, m	6.19, dt (16.0, 4.9)	6.18, dt (15.5, 5.0)
12	4.14, dd (4.6, 2.1)	4.31, dd (6.4, 1.2)	3.63, t (6.3)	4.12 dd (4.9, 1.6)	4.12, dd (5.0, 1.6)

---

Fig. 1 Selected ¹H–¹H COSY and HMBC correlations of compound 1.

The absolute configuration of the vic-diol moiety in 1 was assigned using the in situ dimolybdenum CD method developed by Snatzke and Frelek.^16,17 Briefly, this method involves the formation of chiral complexes in situ by mixing the chiral vic-diol with [Mo₂(OAc)₄], which generates a significant induced circular dichroism (ICD) spectrum. The sign of the Cotton effect around 310–320 nm in the ICD spectrum is related to the chirality of the diol moiety, which is expressed by the sign of the O–C–C–O torsion angle. The negative Cotton effect at 317 nm in the ICD spectrum of 1 led to the assignment of a 6R,7R configuration on the basis of the empirical rule (Fig. 2A).

Fig. 2 (A) Projection of the helicity rule for compound 1. (B) ¹H NMR chemical shift differences of MTPA esters of 3.

The HRESIMS analysis of pilosulyne B (2) provided a molecular formula of C₁₄H₁₈O₄, which is the same as that of a known compound, namely, threo-tetradeca-2,10-diene-4,6-diyne-1,8,9,14-tetrol.¹² Their NMR spectroscopic data were quite similar, but had differences in the coupling patterns of the C-12/C-13 double bond. The differences in NMR data were ascribed to the Z geometry of the C-12/C-13 double bond, which was corroborated by a coupling constant of 11.0 Hz between H-12 and H-13 and an NOE correlation between these two protons. The coupling constant (J = 6.6 Hz) between H-6 and H-7 suggested a threo-configuration for the vic-diol group of 2. Similarly, the negative Cotton effect at 320 nm observed in the ICD spectrum of 2 confirmed the 6R,7R configuration according to the Snatzke's empirical rule.

Analysis of the HRESIMS and ¹³C NMR spectroscopic data of pilosulyne C (3) suggested a molecular formula of C₁₄H₂₀O₃. The NMR spectroscopic data of 3 were similar to those of the aglycone moiety of cordifolioidyne B (12),¹² except that the NMR signals of the C-12/C-13 double bond in 12 disappeared and were replaced by a single bond in 3. This was confirmed by a detailed inspection of the 2D NMR experiments. The absolute configuration at C-5 was determined by the application of the Mosher's method.¹⁸ The (S)- and (R)-MTPA esters of 3 (3a and 3b, respectively) were prepared using the corresponding (R)- and (S)-MTPA chlorides. The determination of chemical shift differences for the protons neighboring C-5 led to the assignment of a 5S configuration in 3 (Fig. 2B).

Pilosulyne D (4) gave the same molecular formula as that of a known polyyne, lobetyolin (10),³ based on the interpretation of HRESIMS and ¹³C NMR spectroscopic data (Table 3). The ¹H NMR spectrum of 4 showed one anomeric proton at δ 4.35 (1H, d, J = 8.0 Hz), which is suggested to arise from one β-glucopyranose moiety resonating at δ_C 100.8 (CH, C-1′), 75.0 (CH, C-2′), 78.1 (CH, C-3′), 71.7 (CH, C-4′), 78.1 (CH, C-5′), and 62.8 (CH₂, C-6′).^3,12 The NMR spectra of the aglycone moiety of 4 were similar to those of 10, except for the coupling constants between H-12 and H-13. A coupling constant of 11.6 Hz between H-12 and H-13 suggested a Z-configuration of the C-12/C-13 double bond. This was confirmed by an NOE correlation between H-12 and H-13. The absolute configuration of the sugar moiety in 4 was determined by a reverse-phase HPLC analysis of its o-tolylthiocarbamate.¹⁹ The glucose liberated from the acid hydrolysis of 4 was treated with L-cysteine methyl ester followed by reaction with o-tolylisothiocyanate to afford the corresponding o-tolylthiocarbamate derivative. The retention time of the sugar derivative by HPLC analysis was found to be consistent with that of the standard D-glucose derivative. On the basis of biogenetic considerations, the absolute configurations of C-6 and C-7 were assigned to be the same as those of 1 and 2.

Table 3 ¹H and ¹³C NMR spectroscopic data of compounds 4 and 5

No.	4		5
	δH (J in Hz)^a	δC, mult.^b	δH (J in Hz)^a	δC, mult.^b
a Spectra were measured in CD3OD (400 MHz).b Spectra were measured in CD3OD (100 MHz).
1	3.61, t (6.4)	62.3, CH2	3.55, t (6.5)	62.8, CH2
2	1.68, m	33.1, CH2	1.55, m	33.5, CH2
3	2.22, q (6.8)	29.9, CH2	1.45, m	22.4, CH2
4	5.95, dt (15.2, 6.8)	139.0, CH	1.61, m	36.2, CH2
5	5.50, dd (15.2, 8.4)	126.7, CH	4.41, m	78.7, CH
6	4.31, dd (8.4, 6.0)	81.9, CH	6.27, dd (16.0, 6.0)	147.9, CH
7	4.48, d (6.0)	66.7, CH	6.12, d (16.0)	112.0, CH
8		82.7, C		80.4, C
9		71.1, C		74.8, C
10		78.6, C		75.3, C
11		76.1, C		80.7, C
12	5.57, dd (11.6, 1.6)	109.7, CH	5.85, br d (16.0)	108.9, CH
13	6.23, dq (11.6, 6.8)	144.3, CH	6.40, dt (16.0, 4.8)	147.9, CH
14	1.91, dd (6.8, 1.6)	16.6, CH3	4.15, dd (4.8, 2.0)	62.7, CH2
Glc
1′	4.35, d (8.0)	100.8, CH	4.37, d (7.6)	100.6, CH
2′	3.25, m	75.0, CH	3.47, dd (9.2, 7.6)	83.2, CH
3′	3.31, m	78.1, CH	3.55, dd (9.2, 8.6)	78.0, CH
4′	3.27, m	71.7, CH	3.30, m	71.7, CH
5′	3.18, m	78.1, CH	3.22, m	78.1, CH
6′	a: 3.88, dd (12.0, 2.4)	62.8, CH2	3.86, dd (12.0, 2.0)	63.0, CH2
	b: 3.68, dd (12.0, 5.9)		3.65, dd (12.0, 6.0)
Glc
1′′			4.60, d (7.6)	105.6, CH
2′′			3.26, dd (8.8, 7.6)	76.3, CH
3′′			3.39, dd (8.8, 8.8)	77.9, CH
4′′			3.34, dd (9.2, 8.8)	71.7, CH
5′′			3.31, m	78.4, CH
6′′			3.88, dd (12.0, 2.0)	63.0, CH2
			3.71, dd (12.0, 5.2)

---

The molecular formula of pilosulyne E (5) was found to be C₂₈H₃₆O₅, as deduced from the HRESIMS and ¹³C NMR data (Table 3). The ¹H NMR spectrum of 5 showed signals for two anomeric protons at δ 4.37 (1H, d, J = 7.6 Hz) and 4.60 (1H, d, J = 7.6 Hz), suggesting the presence of two sugar moieties. Except for the sugar residues, the NMR spectroscopic data of the aglycone moiety of 5 was found to be quite similar to those of a known monoglucoside (12),¹² suggesting that they shared the same aglycone moiety. By excluding the aglycone moiety, the remaining twelve oxygenated carbons in 5 were ascribed to the presence of two glucose residues (Table 3). The D-glucose was deduced to be the only sugar residue in 5 according to the RP HPLC analysis of the corresponding o-tolylthiocarbamate as described above. The carbon resonances assigned to C-2′ (83.2, CH) showed significant downfield shift, which clearly indicated the attachment of another glucose residue at this position. This was further confirmed by the HMBC correlations from H-1′′ to C-2′ and H-5 to C-1′. The large coupling constants (J = 7.6 Hz) of H-1′ and H-1′′ revealed the β-configuration of these two glucose residues. Biogenetically, the absolute configuration for the aglycone moiety of 5 is suggested to be the same as that of 3.

The HRESIMS spectrum of pilosulyne F (6) exhibited a pseudomolecular ion peak at m/z 277.1414 [M + Na]⁺, consistent with a molecular formula of C₁₄H₂₂O₄. Two pairs of trans double bonds [δ 6.19 (1H, dt, J = 16.0, 4.9 Hz), 5.87 (1H, dd, J = 15.5, 1.6 Hz) and 6.13 (1H, dd, J = 15.6, 6.0 Hz), 5.84 (1H, d, J = 15.6 Hz)] and carbon resonances at δ 88.8 (C), 88.9 (C), 110.7 (CH), 112.1 (CH), 112.1 (CH), and 143.4 (CH) (Tables 1 and 2) revealed the presence of one C≡C bonds and two C=C bonds in conjugation, which was confirmed by the UV absorption bands at 264 and 279 nm. A hydroxymethyl group (δ_H 4.12, 2H) was assigned at C-13 due to the presence of allylic coupling (J_12,14 = 1.6 Hz) and HMBC correlations from H₂-14 to C-12 and C-13 as well as H-13 to C-11. Two mutually coupled hydroxyl-containing methine protons at δ 3.45 (1H, m) and 4.00 (1H, t, J = 6.0 Hz) were assigned at C-6 and C-7, respectively, due to the presence of HMBC correlations from H-7 to C-8, C-9, and C-6, as well as H-8 to C-10. The remaining five sp³ carbons (δ 33.6, CH₂; 26.8, CH₂; 26.9, CH₂; 33.6, CH₂; 62.9, CH₂) were assigned to be a hydroxypentyl functionality attached at C-6. Similarly, the vic-diol group of 6 was determined to be in a threo-configuration by the coupling constant (J = 6.0 Hz) between H-6 and H-7. According to the Snatzke's empirical rule, the negative Cotton effect at 331 nm observed in the ICD of 6 confirmed the 6R,7R configuration.

The formula of pilosulyne G (7) was found to be C₁₄H₂₀O₄, 2 mass units less than that of 6, as deduced from the HRESIMS and NMR spectroscopic data. By comparison of the ¹H and ¹³C NMR spectroscopic data of 7 (Tables 1 and 2) with those of 6, it was found that 7 is the trans C-4/C-5 double bond (5.49, 1H, dd, J = 15.4, 7.0 Hz; 5.73, dt, J = 15.4, 7.0 Hz) derivative of 6. The coupling constant (J = 6.6 Hz) between H-6 and H-7 suggested a threo-configuration for the vic-diol group in 7. Similarly, the negative Cotton effect at 329 nm observed in the ICD of 7 confirmed the 6R,7R configuration.

Lobetyol (8), possessing a 6R,7R configuration, was the only polyyne of which the absolute configuration had been established in prior studies.^3,20 Except for 8, the absolute configurations of the other polyynes isolated from the genus Codonopsis have not been established before. Thus, we applied the aforementioned dimolybdenum CD method on compounds 7 and 9, and the resulting CD spectra showed negative Cotton effects at 323 and 315 nm, respectively, suggesting 6R,7R configurations for both compounds. Similarly, compound 8 also showed a negative Cotton effect at 317 nm, confirming the same 6R,7R configuration obtained by prior studies that established this by a series of chemical reactions.²⁰

Hepatitis C is an infectious disease caused by the hepatitis C virus (HCV) that primarily affects the liver. The current anti-HCV therapy utilizes interferons alone or pegylated interferons combined with ribavirin and it is usually accompanied by strong side effects and moderate success rate.²¹ Therefore, the discovery of small molecular inhibitors of HCV replication is an urgent need. Some polyynes, such as polyacetyleneginsenoside-Ro²² and minquartynoic acid, were reported to exhibit anti-virus activities.²³ Thus, the HCVcc infection assay was used to evaluate the anti-HCV activity of compounds 1–12. The results showed that compound 6 exhibited an anti-HCV activity with an EC₅₀ value of 47.2 μM, whereas the other compounds were found to be inactive in the HCVcc infection assay. In addition, 6 was found to be non-toxic toward the tested Huh7.5 cell lines with an IC₅₀ value of more than 100 μM. Compounds 3, 4 and 9–12 were also evaluated for their inhibitory effects on superoxide anion generation and elastase release by human neutrophils in response to FMLP/CB. However, they were not observed to exhibit inhibition effect in the abovementioned two assays.

Conclusion

The phytochemical studies of C. pilosula resulted in the characterization of five new polyynes (1–5) and two new polyenes (6 and 7), as well as five known analogues (8–12). To the best of our knowledge, this is the first report for the determination of the absolute configurations of vic-diols in polyynes and polyenes with only trace amounts of compound available. Considering its medicinal use and potential as a health supplementary product in the future, our work is clearly important because it reveals the chemical constituents in the water extract. In this study, we have also shown that compound 6 exhibited marginal anti-HCV activity in the HCVcc infection assay.

Experimental section

General experimental procedures

Melting points were determined on a Yanagimoto MP-S3 micromelting point apparatus. IR spectra were recorded on a Shimadzu FTIR spectrometer Prestige-21. Optical rotations were measured using a Jasco JASCO P-1020 polarimeter. UV spectra were obtained on a Hitachi UV-3210 spectrophotometer. ESI and HRESI mass spectra were recorded on a Bruker APEX II mass spectrometer. NMR spectra, including ¹H NMR, ¹³C NMR, COSY, NOESY, HMBC, HMQC, and 1D-TOCSY experiments, were recorded on Bruker AVANCE-500 and AMX-400 spectrometer0073. Silica gel (E. Merck 70–230, 230–400 mesh) and RP-18 gel (Merck, 230–400 mesh) were used for column chromatography. Pre-coated silica gel plates (Merck, Kieselgel 60 F₂₅₄, 0.25 mm) and pre-coated RP-18 F_254S plates (Merck) were used for TLC analysis. High-performance liquid chromatography (HPLC) was performed on a Shimadzu LC-10AT_VP series pumping system equipped with a Shimadzu SPD-M10A_VP diode array detector and a semi-preparative reversed-phase column (Merck, Hibar Purospher RP-18e, 5 μm, 250 × 10 mm).

Plant material

The roots of Codonopsis pilosula (Franch.) Nannf. were collected by Kaiser pharmaceutical company from Wen County, Gansu Province, People's Republic of China, in April 2008 and subsequently air-dried at room temperature without any heat-processing. We purchased this material from the company in August 2008 in Tainan, Taiwan, and the sample was authenticated by Prof. C. S. Kuoh (Department of Life Sciences, National Cheng Kung University, Tainan, Taiwan). The material A voucher specimen (CP-Wu 2008005) has been deposited in the Herbarium of National Cheng Kung University, Tainan, Taiwan.

Extraction and isolation

The roots of C. pilosula (5 kg, dry weight) were minced and extracted four times with H₂O under reflux for 5 h. The combined extract was concentrated and subjected to reverse-phase Diaion HP-20 column chromatography, eluting with H₂O, H₂O–MeOH (1 : 1), and MeOH to yield three subfractions, namely, CPW, CPWM, and CPM, respectively. The CPM fraction was selected for further purification because its UV spectra possessed characteristic palm-like shapes between 225 to 300 nm, which suggested the presence of polyynes or polyenes.¹⁰ The CPM fraction (90 g) was fractionated by open column chromatography on silica gel using CHCl₃–MeOH mixtures of increasing polarity to yield 8 fractions (CPM1 to CPM8). Among the eight fractions, CPM2–CPM5 were found to contain polyynes or polyenes, according to their UV spectra. CPM2 was separated by silica gel column chromatography with CH₂Cl₂–MeOH (9 : 1) to yield two subfractions (CPM2A and CPM2B). CPM2A was chromatographed on silica gel using CHCl₃–MeOH (8 : 1), followed by RP-18 HPLC (CH₃CN–H₂O, 30%) to obtain compounds 8 (1.0 mg) and 9 (1.6 mg). CPM2B was separated by RP-18 HPLC (CH₃CN–H₂O, 21%) to yield compounds 2 (0.4 mg) and 3 (1.7 mg). CPM3 was chromatographed on silica gel using EtOAc–MeOH (14 : 1), followed by column chromatography on RP-18 gel (MeOH–H₂O, 33%) to afford two subfractions CPM3A and CPM3B. Compound 7 (0.8 mg) was obtained from subfraction CPM3A by RP-18 HPLC (CH₃CN–H₂O, 13%). CPM3B was separated by RP-18 HPLC (CH₃CN–H₂O, 15%) to obtain compounds 1 (0.8 mg) and 6 (0.4 mg). CPM4 was separated by repeated column chromatography over RP-18 gel (MeOH–H₂O, 30%) and silica gel (CHCl₃–MeOH, 5 : 1), followed by RP-18 HPLC (CH₃CN–H₂O, 18%) to yield compounds 10 (37.0 mg) and 12 (6.0 mg). CPM5 was fractionated by RP-18 column chromatography (MeOH–H₂O, 30%) to yield two subfractions (CPM5A and CPM5B). Compound 4 (2.5 mg) was obtained from subfraction CPM5A by RP-18 HPLC (MeOH–H₂O, 39%). CPM5B was separated by repeated column chromatography over silica gel (CHCl₃–MeOH, 3 : 1) and RP-18 gel (MeOH–H₂O, 33%), followed by RP-18 HPLC (CH₃CN–H₂O, 15%) to yield compounds 5 (18 mg) and 11 (5.0 mg).

Pilosulyne A (1). Colorless oil; [α]²⁵_D +34.6 (c 0.07, MeOH); UV (MeOH) λ_max (log ε) 240 (3.37), 253 (3.58), 267 (3.73), 283 (3.64) nm; IR (KBr) v_max 3305, 2924, 2854, 2164, 1629, 1458, 1257, 1095, 1033 cm⁻¹; ¹³C NMR and ¹H NMR data, see Tables 1 and 2; ESIMS m/z 275 [M + Na]⁺; HRESIMS m/z 275.1261 [M + Na]⁺ (calcd for C₁₄H₂₀O₄Na, 275.1259).

Pilosulyne B (2). Colorless oil; [α]²⁵_D +20.0 (c 0.03, MeOH); UV (MeOH) λ_max (log ε) 241 (3.24), 254 (3.42), 268 (3.57), 284 (3.47) nm; IR (KBr) v_max 3313, 2924, 2854, 2164, 1678, 1632, 1261, 1026 cm⁻¹; ¹³C NMR and ¹H NMR data, see Tables 1 and 2; ESIMS m/z 273 [M + Na]⁺; HRESIMS m/z 273.1102 [M + Na]⁺ (calcd for C₁₄H₁₈O₄Na, 273.1103).

Pilosulyne C (3). Colorless oil; [α]²⁵_D +26.3 (c 0.14, MeOH); UV (MeOH) λ_max (log ε) 240 (3.49), 252 (2.69), 267 (3.86), 283 (3.76), 293 (3.03), 313 (2.84) nm; IR (KBr) v_max 3302, 2927, 2858, 2233, 2137, 1423, 1330, 1300, 1056, 952 cm⁻¹; ¹³C NMR and ¹H NMR data, see Tables 1 and 2; ESIMS m/z 259 [M + Na]⁺; HRESIMS m/z 259.1309 [M + Na]⁺ (calcd for C₁₄H₂₀O₃Na, 259.1310).

Pilosulyne D (4). Amorphous powder; [α]²⁵_D +23.0 (c 0.20, MeOH); UV (MeOH) λ_max (log ε) 240 (3.72), 253 (3.78), 267 (3.89), 283 (3.80) nm; IR (KBr) v_max 3321, 2927, 2877, 2229, 1654, 1573, 1415, 1076, 1045 cm⁻¹; ¹³C NMR and ¹H NMR data, see Table 3; ESIMS m/z 419 [M + Na]⁺; HRESIMS m/z 419.1680 [M + Na]⁺ (calcd for C₂₀H₂₈O₈Na, 419.1682).

Pilosulyne E (5). Amorphous powder; [α]²⁵_D −100.2 (c 0.1, MeOH); UV (MeOH) λ_max (log ε) 247 (4.29), 262 (3.89), 277 (4.13), 294 (4.27), 314 (4.18) nm; IR (KBr) v_max 3356, 2927, 2870, 2206, 1631, 1408, 1072, 1033 cm⁻¹; ¹³C NMR and ¹H NMR data, see Table 3; ESIMS m/z 581 [M + Na]⁺; HRESIMS m/z 581.2208 [M + Na]⁺ (calcd for C₂₆H₃₈O₁₃Na, 581.2210).

Pilosulyne F (6). Colorless oil; [α]²⁵_D +11.0 (c 0.03, MeOH); UV (MeOH) λ_max (log ε) 264 (3.93), 279 (3.83) nm; IR (KBr) v_max 3290, 2924, 2854, 2268, 1675, 1551, 1261, 1095, 1029, 952 cm⁻¹; ¹³C NMR and ¹H NMR data, see Tables 1 and 2; ESIMS m/z 277 [M + Na]⁺; HRESIMS m/z 277.1414 [M + Na]⁺ (calcd for C₁₄H₂₂O₄Na, 277.1416).

Pilosulyne G (7). Colorless oil; [α]²⁵_D +13.2 (c 0.04, MeOH); UV (MeOH) λ_max (log ε) 265 (3.71), 279 (3.62) nm; IR (KBr) v_max 3309, 2924, 2854, 2276, 1676, 1550, 1261, 1095, 1029, 952 cm⁻¹; ¹³C NMR and ¹H NMR data, see Tables 1 and 2; ESIMS m/z 275 [M + Na]⁺; HRESIMS m/z 275.1257 [M + Na]⁺ (calcd for C₁₄H₂₀O₄Na, 275.1259).

Determination of the absolute configuration of the 6,7-diol moieties in compounds 1, 2, and 6–9 by Snatzke's method

Dimolybdenum tetracetate and DMSO of spectroscopy grade (dried with 4 Å molecular sieves) were purchased from Acros. Following a reported procedure,^16,17 a solution of 6,7-diol in dry DMSO (5 × 10⁻⁴ M, 3 mL) was mixed with dimolybdenum tetraacetate (0.8 mg) in a ratio of ca. 1 : 1.2 diol/dimolybdenum tetraacetate. The first CD of the mixture was recorded immediately after mixing, and its time evolution was monitored until a stationary state was reached (about 30 min after mixing). The inherent CD was subtracted and the observed sign of the diagnostic band at around 310–320 nm in the induced CD spectrum was correlated to the absolute configuration of the 6,7-diol moiety.

Preparation of (S)- and (R)-MTPA esters of 3

To a solution of 3 (0.4 mg) in pyridine (0.4 mL), (R)-MTPA chloride (25 μL) was added, and the mixture was allowed to stand for 3 h at room temperature. The reaction was quenched by the addition of H₂O (1.0 mL), and the mixture was subsequently extracted with EtOAc (3 × 1.0 mL). The EtOAc-soluble layers were combined, dried over anhydrous MgSO₄, and evaporated. The residue was subjected to silica gel column chromatography using n-hexane–EtOAc (5 : 1) to yield the (S)-MTPA ester, 3a (0.3 mg). The same procedure was used to prepare the (R)-MTPA ester, 3b (0.4 mg from 0.4 mg of 3), with (S)-MTPA chloride. Selected ¹H NMR (CDCl₃, 300 MHz) of 3a: δ 7.400–7.500 (5H, m, Ph), 6.10 (1H, dd, J = 16.0, 7.1 Hz, H-6), 5.74 (1H, J = 16.0 Hz, H-7), 5.42 (1H, q, J = 7.0 Hz, H-5), 4.42 (2H, m, H₂-14), 4.20 (2H, m, H₂-1), 3.54 (3H, s, OMe), 3.53 (3H, s, OMe), 3.51 (3H, s, OMe), 2.38 (2H, t, J = 6.7 Hz, H₂-12), 1.94 (2H, m, H₂-13), 1.61 (2H, m, H₂-4), 1.58 (2H, m, H₂-2), 1.23 (2H, m, H₂-3). ¹H NMR (CDCl₃, 300 MHz) of 3b: δ 7.40–7.49 (5H, m, Ph), 6.01 (1H, dd, J = 16.0, 6.7 Hz, H-6), 5.63 (1H, J = 16.0 Hz, H-7), 5.40 (1H, q, J = 6.7 Hz, H-5), 4.42 (2H, m, H₂-14), 4.27 (2H, m, H₂-1), 3.54 (3H, s, OMe), 3.53 (3H, s, OMe), 3.51 (3H, s, OMe), 2.38 (2H, t, J = 6.7 Hz, H₂-12), 1.94 (2H, m, H₂-13), 1.69 (2H, m, H₂-4), 1.63 (2H, m, H₂-2), 1.33 (2H, m, H₂-3).

Determination of sugar configuration

Authentic samples of D-Glc (or L-Glc) and L-cysteine methyl ester hydrochloride (0.5 mg each) were dissolved in pyridine (0.1 mL) and heated at 60 °C for 1 h. Then, the o-tolylisothiocyanate (0.5 mg in 0.1 mL pyridine) was added to the mixture and heated at 60 °C for an additional 1 h. The reaction mixture was directly analyzed by reversed-phase HPLC (Cosmosil 5C18-AR II column; 4.6 × 250 nm; 25% CH₃CN in 50 mM H₃PO₄; 0.8 mL min⁻¹; 25 °C) and detected at 250 nm.¹⁹

The glycoside (0.3 mg, for each) was hydrolyzed by heating in 0.6 M HCl (0.1 mL) and neutralized with Amberlite IRA-400(OH). After the reaction mixture was dried in vacuo, the residue was dissolved in pyridine (0.1 mL) containing L-cysteine methyl ester (0.5 mg) and heated at 60 °C for 1 h. A 0.1 mL solution of o-tolylisothiocyanate (0.5 mg) in pyridine was added to the mixture, which was heated at 60 °C for an additional 1 h, to yield the corresponding o-tolylthiocarbamate derivative. The RP HPLC analysis of the o-tolylthiocarbamate derivative derived from the hydrolyte of the glycosides 4 and 5 showed peaks at 32.4 and 32.5 min, respectively, while the t_R values for standard L-Glc and D-Glc derivatives were observed at 28.8 and 32.5 min, respectively, suggesting the presence of a D-Glc residue in 4 and 5.

Measurement of superoxide anion generation

Human neutrophils from the venous blood of healthy, adult volunteers (18–32 years old) were isolated with the standard method of dextran sedimentation prior to centrifugation in a Ficoll-Hypaque gradient and hypotonic lysis of erythrocytes.²⁴ Neutrophil superoxide anion generation was determined using superoxide dismutase (SOD)-inhibitory cytochrome c reduction according to described procedures. Diphenyleneiodonium (DPI, an NADPH oxidase inhibitor) was used as the positive control in the generation of a superoxide anion.

Measurement of elastase release

The degranulation of azurophilic granules was determined by elastase release as described previously. The measurement of elastase release was carried out according to the described procedures.²⁵ The results are expressed as the percentage of elastase release in the (formyl-L-methionyl-L-leucyl-L-phenylalanine/cytochalasin B) FMLP/CB-activated, drug-free control system. Phenylmethylsulfonyl fluoride (PMSF, a serine protease inhibitor) was used as the positive control in the release of elastase.

Cell culture and cell viability assay

Human hepatoma Huh7.5 cells were maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% heat-inactivated fetal bovine serum (FBS) and 1% non-essential amino acid (NEAA). Viable cells were determined by the CellTiter 96 Aqueous One Solution Cell Proliferation Assay Kit.²⁶

Infectious HCV particles production and infection inhibition assay

The production of infectious JC1-Luc2A HCV particles (HCVcc) was performed as described previously.²⁷ For the infection inhibition assays, Huh7.5 cells were seeded in a 96-well plate at the density of 1 × 10⁴ cells per well. At 24 h after plating, JC1-Luc2A HCV reporter virus (0.1 MOI) was added into each well and incubated for 4 h. The virus-containing supernatant was removed and fresh medium with or without the tested compound was added and incubated for a total of 72 h, and the whole system was not kept away from light intentionally. The cell lysates were then collected for luciferase activity and cell viability assays. The luciferase activity assay was performed according to the manufacturer's instructions (Bright-Glo Promega). The relative firefly luciferase versus cell viability activity was reported as the mean plus or minus the standard deviation of three independent reactions. Honokiol, an anti-HCV compound,²⁸ was used as a positive control with EC₅₀ value of 9.4 μM (anti-HCV activity) and IC₅₀ value of 58.5 μM (cytotoxicity toward Huh 7.5 cells).

Acknowledgements

The work was supported by grants from the Ministry of Science and Technology of Republic of China awarded to T.-S.Wu and China Medical University (CMU103-N-12) awarded to C.-H.Chao.

Notes and references

1. W. L. Li, H. C. Zheng, J. Bukuru and N. De Kimpe, J. Ethnopharmacol., 2004, 92, 1.
2. K. Ishimaru, S. Sadoshima, S. Neera, K. Koyama, K. Takahashi and K. Shimomura, Phytochemistry, 1992, 31, 1577.
3. K. Ishimaru, H. Yonemitsu and K. Shimomura, Phytochemistry, 1991, 30, 2255.
4. G. H. Liu, Zhongguo Zhongxiyi Jiehe Zazhi, 1983, 3, 114.
5. H. Norr and H. Wagner, Planta Med., 1994, 60, 494.
6. Z. T. Wang, T. B. Ng, H. W. Yeung and G. J. Xu, Gen. Pharmacol., 1996, 27, 1347.
7. M. P. Wong, T. C. Chiang and H. M. Chang, Planta Med., 1983, 49, 60.
8. S. Hong, Y. Yong, K. Kang, S. Y. Shin, Y. H. Lee and Y. Lim, J. Microbiol. Biotechnol., 2009, 19, 556.
9. X. F. Sun and H. Zhang, Histol. Histopathol., 2007, 22, 1387.
10. C. F. Qiao, Z. D. He, Q. B. Han, H. X. Xu, R. W. Jiang, S. L. Li, Y. B. Zhang, P. P. H. But and P. C. Shaw, J. Food Drug Anal., 2007, 15, 258.
11. A. F. Magalhães, D. M. Vieira, E. G. Magalhães and G. J. Shepherd, Phytochemistry, 1988, 27, 3827.
12. R. Q. Mei, Q. Lu, Y. F. Hu, H. Y. Liu, F. K. Bao, Y. Zhang and Y. X. Cheng, Helv. Chim. Acta, 2008, 91, 90.
13. W. Aida, T. Ohtsuki, X. F. Li and M. Ishibashi, Tetrahedron, 2009, 65, 369.
14. Z. D. Jiang and W. H. Gerwick, Phytochemistry, 1990, 29, 1433.
15. M. S. Lehner, A. Steigel and R. Bauer, Phytochemistry, 1997, 46, 1023.
16. L. Di Bari, G. Pescitelli, C. Pratelli, D. Pini and P. Salvadori, J. Org. Chem., 2001, 66, 4819.
17. M. Gorecki, E. Jablonska, A. Kruszewska, A. Suszczynska, Z. Urbanczyk-Lipkowska, M. Gerards, J. W. Morzycki, W. J. Szczepek and J. Frelek, J. Org. Chem., 2007, 72, 2906.
18. I. Ohtani, T. Kusumi, Y. Kashman and H. Kakisawa, J. Am. Chem. Soc., 1991, 113, 4092.
19. T. Tanaka, T. Nakashima, T. Ueda, K. Tomii and I. Kouno, Chem. Pharm. Bull., 2007, 55, 899.
20. M. Yuda, K. Ohtani, K. Mizutani, R. Kasai, O. Tanaka, M. R. Jia, Y. R. Ling, X. F. Pu and Y. Saruwatari, Phytochemistry, 1990, 29, 1989.
21. M. Y. Lai, Intervirology, 2006, 49, 91.
22. H. Zhang, Z. Lu, G. T. Tan, S. Qiu, N. R. Farnsworth, J. M. Pezzuto and H. H. S. Fong, Tetrahedron Lett., 2002, 43, 973.
23. M. A. Rashid, K. R. Gustafson, J. H. Cardellina and M. R. Boyd, Nat. Prod. Lett., 2001, 15, 21.
24. T. L. Hwang, G. L. Li, Y. H. Lan, Y. C. Chia, P. W. Hsieh, Y. H. Wu and Y. C. Wu, Free Radical Biol. Med., 2009, 46, 520.
25. L. A. Sklar, V. M. McNeil, A. J. Jesaitis, R. G. Painter and C. G. Cochrane, J. Biol. Chem., 1982, 257, 5471.
26. J. C. Cheng, Y. J. Yeh, C. P. Tseng, S. D. Hsu, Y. L. Chang, N. Sakamoto and H. D. Huang, Cell. Mol. Life Sci., 2012, 69, 2621.
27. J. T. Huang, C. P. Tseng, M. H. Liao, S. C. Lu, W. Z. Yeh, N. Sakamoto, C. M. Chen and J. C. Cheng, J. Virol., 2013, 87, 4994.
28. N. Calland, J. Dubuisson, Y. Rouille and K. Seron, Viruses, 2012, 4, 2197.

---

Footnote

† Electronic supplementary information (ESI) available: NMR spectral data for new compounds 1–7 and the analytical profiles for the UV spectra of fractions. See DOI: 10.1039/c5ra02765a

---