18 Soo-Jung Choi1 ∗ Janggyoo Choi2 ∗ Heejin Jeon1 Soo Kyung Bae1 Jaeyoung Ko3 Jinwoong Kim2 Kee Dong Yoon1 1 College
of Pharmacy and Integrated Research Institute of Pharmaceutical Sciences, The Catholic University of Korea, Bucheon, Republic of Korea 2 College of Pharmacy and Research Institute of Pharmaceutical Science, Seoul National University, Seoul, Republic of Korea 3 Materials Science Team, Medicinal Beauty Division, Amorepacific Corporation R&D Unit, Yongin, Republic of Korea Received September 13, 2014 Revised October 7, 2014 Accepted October 11, 2014
J. Sep. Sci. 2015, 38, 18–24
Research Article
Application of high-performance countercurrent chromatography for the isolation of steroidal saponins from Liriope plathyphylla High-performance countercurrent chromatography (HPCCC) with electrospray lightscattering detection was applied for the first time to isolate a spirostanol and a novel furostanol saponin from Liriope platyphylla. Due to the large differences in KD values between the two compounds, a two-step HPCCC method was applied in this study. The primary HPCCC employed methylene chloride/methanol/isopropanol/water (9:6:1:4 v/v, 4 mL/min, normal-phase mode) conditions to yield a spirostanol saponin (1). After the primary HPCCC run, the solute retained in the stationary phase (SP extract) in HPCCC column was recovered and subjected to the second HPCCC on the n-hexane/n-butanol/water system (1:9:10 v/v, 5 mL/min, reversed-phase mode) to yield a novel furostanol saponin (2). The isolated spirostanol saponin was determined to be 25(S)-ruscogenin 1-O--D-glucopyranosyl (1→2)[-D-xylopyranosyl (1→3)]--D-fucopyranoside (spicatoside A), and the novel furostanol saponin was elucidated to be 26-O--D-glucopyranosyl-25(S)-furost-5(6)-ene-1-3-22␣(1→2)-[-D-xylopyranosyl-(1→3)]--D-fucopyranoside 26-tetraol-1-O--D-glucopyranosyl (spicatoside D). Keywords: Electrospray light-scattering detection / Furostanol saponins / Highperformance countercurrent chromatography / Liriope platyphylla DOI 10.1002/jssc.201401007
1 Introduction The genus Liriope, belonging to the Liliaceae family, is abundantly distributed in subtropical and temperate regions and comprises about eight species globally [1]. In China, Korea, and Japan, roots of some Liriope species have been used traditionally to treat asthma, bronchial and lung inflammation, and sputum [2] and are added to various traditional medicinal formulas as expectorants and tonics. Recently, many studies have reported biological activities of Liriope species including anti-diabetic [3], anti-inflammatory, anti-asthmatic [4], laxative [5], anti-atopic dermatitis [6], and neuroprotective effects [7]. In addition, several reports have revealed that phenolic compounds and saponins from Liriope species showed antioxidant [8], anti-cancer [9, 10], estrogenic, anti-platelet [11], antiCorrespondence: Prof. Kee Dong Yoon, Integrated Research Institute of Pharmaceutical Sciences, College of Pharmacy, The Catholic University of Korea, 43 Jibong-ro, Wonmi-gu, Bucheonsi, Gyeonggi-do 420–743, Republic of Korea E-mail: [email protected] Fax: +82-2-2164-4059
Abbreviations: C/M/I/Wat, chloroform/methanol/isopropanol/water; E/B/W, ethyl acetate/n-butanol/water; ELSD, electrospray light-scattering detection; H/B/Wat, n-hexane/nbutanol/water; HPCCC, high-performance countercurrent chromatography; MC/M/I/Wat, methylene chloride/methanol/isopropanol/water; SP extract, stationary phase extract C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
thrombotic [12], and antiviral [13] activities. With regard to the secondary metabolites, steroidal saponins are characteristic components in Liriope species [14–18]. Among the known steroidal saponins, several saponins, such as spicatoside A, B, C, ophiopoginin B, prosapogenin II and III of spicatoside A, and -sitosterol glucoside, were isolated from Liriope spicata and Liriope plathyphylla in Korea [19–22], and several reports have demonstrated that spicatoside A stimulated growth hormone secretion [23], induced neurite outgrowth [24], enhanced memory consolidation [25], and increased mucin production and secretion [26]. Although the steroidal saponins are considered typical marker compounds of Liriope species, the polysaccharides are contained as major constituents and steroidal saponins are present in relatively low amounts, which complicate the isolation and determination of steroidal saponins. Therefore, efficient isolation methods are required to identify steroidal saponins in Liriope species. CCC is a type of liquid–liquid chromatography that employs only liquids for the mobile and stationary phases. Thus, there is no risk of chemical degradation and irreversible adsorption of target compounds onto a solid support, which is common in conventional column chromatography [27]. Among CCC instruments, high-performance countercurrent chromatography (HPCCC) is used to generate high g-levels up to 240 × g to retain a higher proportion of the ∗ These
authors contributed equally to this work.
www.jss-journal.com
J. Sep. Sci. 2015, 38, 18–24
Liquid Chromatography
19
Figure 1. Chemical structures of spicatoside A (1) and spicatoside D (2).
stationary phase even at a higher mobile-phase flow-rate than high-speed CCC. Therefore, faster and more efficient separation is possible compared with conventional CCC methods [28]. Many steroidal saponins of Liriope species have been isolated using solid support based conventional column chromatography, but there is no report concerning countercurrent separation of steroidal saponins from Liriope species. The present study describes HPCCC coupled with an electrospray light-scattering detection (ELSD) method for the isolation of spicatoside A and a novel furostanol saponin, spicatoside D (Fig. 1), from Liriope plathyphylla.
2 Materials and methods 2.1 Instrumentation The semipreparative HPCCC instrument employed in this study was a Spectrum (Dynamic Extractions, Berkshire, UK), which possesses two semipreparative coils of 70.5 mL with a 3.2 mm id. The -values of the semipreparative coils ranged from 0.52 to 0.86. The rotational speed was adjustable from 0 to 1600 rpm. The Spectrum was coupled to an IOTA S 300 pump (Ecom, Prague, Czech Republic), a Sedex 75 ELSD (Sedere, Olivet, France), a Foxy R2 fraction collector (Teledyne Isco, NE, USA), and a CCA-1111 circulatory temperature regulator (Eyela, Tokyo, Japan) to maintain the internal HPCCC temperature at 30⬚C. HPLC analyses were performed on an HPLC system equipped with a 1525 binary HPLC pump (Waters, MA, USA), a manual injection valve (Rheodyne, CA, USA) with a 20 L sample loop, and a Sedex 75 ELSD. HPCCC peak fractions were identified using 1 H and 13 C NMR spectroscopy and MS using an AVANCE 500 spectrometer (Bruker, Karlsruhe, Germany) and a 6530 QTOF mass spectrometer (Agilent Technologies, CA, USA), respectively.
a Millipore Milli-Q water purification system (Millipore, MA, USA). ACN and water used for HPLC analyses were of HPLC grade and purchased from Fisher Scientific Korea (Seoul, Korea). Dried tubers of L. plathyphylla, originating from Miryang province in Korea, were purchased from Humanherb (Deagusi, Korea). The voucher specimen (CU-140629) was deposited at the herbarium of the College of Pharmacy, The Catholic University of Korea.
2.3 Preparation of sample extract Dried tubers of L. plathyphylla (2 kg) were ground into rough powder and extracted with methanol (4 L × 90 min × three times) using an ultrasonic bath, and evaporated under reduced pressure to yield a methanol extract (74 g). The methanol extract was suspended in water and sequentially partitioned with organic solvents to yield ethyl acetate (2.3 g) and n-butanol (8.1 g) soluble extracts. The n-butanol-soluble extract was stored in a refrigerator (4⬚C) prior to HPLC analysis and HPCCC separation.
2.4 HPLC–ELSD analysis The n-butanol-soluble extract of the tubers of L. platyphylla and HPCCC peak fractions were analyzed using HPLC–ELSD with a Zorbax C18 column (250 × 4.6 mm, id, 5 m; Agilent Technologies, CA, USA). The mobile phases were acetonitrile for solvent A (0.1% formic acid) and water (0.1% formic acid) for solvent B. The mobile phase gradient was 25% A (0– 5 min), 25–65% A (5–35 min), and 65% A (35–40 min). The mobile phase flow-rate was 1.0 mL/min and the injection volume was 20 L.
2.5 Q-TOF-MS/MS analysis 2.2 Reagents and plant materials Organic solvents for HPCCC separation were of analytical grade and purchased from Daejung-Chemical and Metals (Kyeonggi-Do, Korea). Ultrapure water was obtained from C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
MS analysis was performed on 6530 Q-TOF mass spectrometer with ESI with the following acquisition parameters: drying gas (nitrogen) flow rate—8.0 L/min, drying gas temperature—300⬚C, nebulizer gas pressure—35 psig, capillary—3500 V, OCT RFV—750 V, fragmentor voltage— www.jss-journal.com
20
S.-J. Choi et al.
200 V. The collision energy was set to 60 V for negative MS/MS analysis. The MS system was operated by MassHunter Acquisition software, version B.05.00.
2.6 Evaluation of partition coefficient values (KD ) The partition coefficient (KD ) values of compounds 1 and 2 were determined by HPLC–ELSD analysis. The KD value of each target compound was calculated as the HPLC– ELSD peak area of the target compound in the stationary phase (upper phase) divided by that in the mobile phase (lower phase) under various ratios of two-phase solvent systems (chloroform/methanol/isopropanol/water (C/M/I/Wat, 7:6:1:4, 9:6:1:4, 11:6:1:4, v/v), methylene chloride/methanol/isopropanol/water (MC/M/I/Wat, 7:6:1:4, 9:6:1:4, 11:6:1:4, v/v), n-hexane/n-butanol/water (H/B/Wat, 3:7:10, 2:8:10, 1:9:10, v/v), and ethyl acetate/n-butanol/water (E/B/W, 1:9:10, v/v)). n-Butanol-soluble extract (5 mg) of L. plathyphylla was dissolved in a 1:1 mixture of upper and lower phase (each 2 mL) of the above two-phase solvent systems.
J. Sep. Sci. 2015, 38, 18–24
2.8 Second HPCCC procedure for compound 2 The second preparative HPCCC for the separation of 2 was performed after monitoring HPCCC chromatograms, similar to the primary HPCCC procedure. Two-phase solvent systems composed of H/B/Wat (3:7:10, 2:8:10, 1:9:10, v/v) and E/B/Wat (1:9:10, v/v) were used to check HPCCC separation patterns. Finally, the second preparative HPCCC was performed using the following conditions: H/B/Wat (1:9:10, v/v) two-phase solvent system, 5.0 mL/min flow rate, 1600 rpm rotational speed, and 100 mg of SP extract. The second preparative HPCCC was performed twice.
2.9 Structural determination of compounds 1 and 2 Structural determination of 1 and 2 in HPCCC peak fractions was performed based on Q–TOF-MS, 1 H, and 13 C NMR spectroscopic data, as well as comparison with previously published values.
3 Results and discussion 3.1 HPLC–ELSD analysis and evaluation of partition coefficient values
2.7 Primary HPCCC procedure for compound 1 To determine the optimal HPCCC pattern for 1, C/M/I/Wat (7:6:1:4, 9:6:1:4, 11:6:1:4, v/v) and MC/M/I/Wat (7:6:1:4, 9:6:1:4, 11:6:1:4, v/v) systems were used in the present study. The HPCCC coil was filled with upper aqueous phase at a flow rate of 10 mL/min. After filling stationary phase in the column, the HPCCC was rotated at 1600 rpm while the lower mobile phase was eluted in head-to-tail mode at a flow rate of 4 mL/min. Sample extract (20 mg) was dissolved in the lower organic phase and subjected to HPCCC after the mobile phase had emerged in the effluent. The HPCCC chromatograms were monitored using ELSD. The ELSD parameter was modulated with a tube temperature of 90⬚C, a gain level of 5, and 3.0 bar of nebulizer nitrogen gas pressure. Preparative HPCCC was conducted according to the results of the above HPCCC separation patterns. The HPCCC parameters were as follows: MC/M/I/Wat (11:6:1:4, v/v) twophase solvent system, 1600 rpm rotational speed, 4 mL/min flow rate, and 400 mg (200 mg × two times) of n-butanolsoluble extract. The tail outlet of the HPCCC was combined with ELSD through a split valve. The split ratio of the effluent was modulated using a spilt valve; one-tenth of the HPCCC effluent was subjected to ELSD and nine-tenths of the effluent was collected using a fraction collector for 1 min per fraction. After HPCCC separation, all solvents in the column were collected by displacement with N2 gas to measure the stationary phase retention, and it was evaporated under reduced pressure to yield the sample (stationary phase extract (SP extract)) for the second HPCCC. The primary preparative HPCCC for 1 was performed in duplicate. C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Previous reports have revealed that L. plathyphylla possessed furostanol- and spirostanol-type saponins as secondary metabolites. As shown in Fig. 2, the HPLC chromatogram of n-butanol-soluble extract of L. plathyphylla showed one peak (1) at 32.7 min and a relatively major peak (2) at 13.9 min, indicative of spirostanol and furostanol saponins, respectively. The KD values of target compounds should range from 0.2 to 2.0 for optimal CCC separation [29]. To select a suitable two-phase solvent system, C/M/I/Wat (7:6:1:4, 9:6:1:4 and 11:6:1:4, v/v), MC/M/I/Wat (7:6:1:4, 9:6:1:4 and 11:6:1:4, v/v), H/B/Wat (3:7:10, 2:8:10 and 1:9:10, v/v), and E/B/Wat (1:9:10, v/v) were tested to determine the partition coefficient of the two corresponding compounds because these two-phase solvent systems have been used previously to separate various saponins from medicinal plants. The KD values of 1 under various ratios of C/M/I/Wat and MC/M/I/Wat solvent systems ranged from 0.13 to 1.77, while those of 2 were over 15.5. The opposite results were observed under H/B/Wat and E/B/Wat systems; the KD values were over 17.8 for 1 and 0.24–1.33 for 2, respectively (Table 1). These results suggested that the difference in KD values between 1 and 2 is too large for separation using one-step HPCCC with a single two-phase solvent system. Therefore, the two-step HPCCC method was used to isolate 1 and 2.
3.2 HPCCC separation of compound 1 As described above, the KD values of 1 under C/M/I/Wat and MC/M/I/Wat systems ranged from 0.13 to 1.77. The www.jss-journal.com
Liquid Chromatography
J. Sep. Sci. 2015, 38, 18–24
21
Figure 2. HPLC–ELSD chromatograms of the n-butanol-soluble extract of L. plathyphylla (A), SP extract (B), and the HPCCC peak fraction of 1 (C) and 2 (D).
Table 1. Partition coefficient (KD ) values of 1 and 2
Solvent system (volume ratio, v/v)
KD value 1
Chloroform/methanol/isopropanol/water 7:6:1:4 (v/v) 0.13 9:6:1:4 (v/v) 0.18 11:6:1:4 (v/v) 0.32 Methylene chloride/methanol/isopropanol/water 7:6:1:4 (v/v) 1.42 9:6:1:4 (v/v) 1.53 11:6:1:4 (v/v) 1.77 n-Hexane/n-butanol/water 3:7:10 (v/v) 17.80 2:8:10 (v/v) >20 1:9:10 (v/v) >20 Ethyl acetate/n-butanol/water 1:9:10 (v/v) >20
2
15.5 19.3 >20 >20 >20 >20 0.24 0.59 1.33 0.92
optimum separation pattern was determined from the HPCCC separation using each two-phase solvent system. As expected based on their KD values, the elution times of 1 on C/M/I/Wat system (Fig. 3A–C) were shorter than those of MC/M/I/Wat system (Fig. 3D–F), but compound 1 was not separated from nearby impurity peaks. The MC/M/I/Wat system yielded better results; the peak corresponding to 1 was well resolved from impurity peaks and MC/M/I/Wat (11:6:1:4, v/v) gave the best separation. Methylene chloride is C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
similar to chloroform in terms of its physical and chemical properties, but methylene chloride is used instead of chloroform for reasons of safety [30]. However, methylene chloride is less polar than chloroform and the MC/M/I/Wat system yielded better separation than the C/M/I/Wat system in this study. Using the MC/M/I/Wat (11:6:1:4, v/v, 4 mL/min), 400 mg (200 mg × 2) of n-butanol-soluble L. platyphylla extract was introduced into HPCCC to yield 7.4 mg of 1 with a purity of 96.2% (Figs. 2C and 3G). The mean stationary phase retention was 71.1%.
3.3 HPCCC separation of compound 2 After primary HPCCC separation of 1, the stationary phase retained in the HPCCC coil was extruded by N2 gas and pooled to yield 236 mg of SP extract, which was analyzed using HPLC–ELSD. As shown in Fig. 2B, the peak corresponding to 1 disappeared and 2 was recovered in the SP extract. The SP extract was subjected to HPCCC under various H/B/Wat (3:7:10, 2:8:10, 1:9:10, v/v) and E/B/Wat (1:9:10, v/v) systems. On H/B/Wat systems, 2 eluted too rapidly to be separated from polar impurities under the 3:7:10 condition (v/v, 4 mL/min; Fig. 4A). The 2:8:10 (v/v, 4 mL/min) condition showed improved separation (Fig. 4B), but 2 was still eluted in proximity to impurity peaks (expressed as asterisk in the figure). The 1:9:10 (v/v, 4 mL/min) condition separated 2 from the impurity peaks (Fig. 4C), and increasing the mobile phase flow rate (5 mL/min) shortened the www.jss-journal.com
22
S.-J. Choi et al.
J. Sep. Sci. 2015, 38, 18–24 Figure 3. HPCCC–ELSD chromatograms of the n-butanolsoluble extract of L. plathyphylla. (A) C/M/I/Wat (7:6:1:4, v/v, 4 mL/min); (B) C/M/I/Wat (9:6:1:4, v/v, 4 mL/min); (C) C/M/I/Wat (11:6:1:4, v/v, 4 mL/min); (D) MC/M/I/Wat (7:6:1:4, v/v, 4 mL/min); (E) MC/M/I/Wat (9:6:1:4, v/v, 4 mL/min); (F) MC/M/I/Wat (11:6:1:4, v/v, 4 mL/min); (G) MC/M/I/Wat (11:6:1:4, v/v, 4 mL/min, sample amount: 400 mg (200 mg × two times)). C, chloroform; MC, methylene chloride; M: methanol; I: isopropanol; Wat: water.
Figure 4. HPLC–ELSD chromatograms of SP extract. (A) H/B/Wat (3:7:10, v/v, 4 mL/min); (B) H/B/Wat (2:8:10, v/v, 4 mL/min); (C) H/B/Wat (1:9:10, v/v, 4 mL/min); (D) H/B/Wat (1:9:10, v/v, 5 mL/min); (E) H/B/Wat (1:9:10, v/v, 4 mL/min); (F) H/B/Wat (1:9:10, v/v, 5 mL/min, sample amount: 200 mg (100 mg × two times)).H, n-hexane; chloroform; B, n-butanol; Wat: water; E: ethyl acetate.
elution time of 2 while maintaining identical separation patterns (Fig. 4D). The E/B/Wat (2:8:10, v/v, 4 mL/min) condition did not result in satisfactory separation compared with H/B/Wat (1:9:10, v/v, 5 mL/min) conditions (Fig. 4E). Therefore, H/B/Wat (1:9:10, v/v, 5 mL/min) conditions were selected for secondary HPCCC for the isolation of 2 from SP extract. A total of 200 mg (100 mg × 2) of SP extract subjected to HPCCC yielded 35.5 mg of 2, with a purity of over 98% (Figs. 2D and 4F). The mean stationary phase retention was 64.3%.
3.4 Structural determination of compounds 1 and 2 The structural elucidation of 1 and 2 was performed using 1 H and 13 C NMR spectroscopy and Q-TOF-MS, as well as published values of steroidal saponins from Liriope species. Consequently, 1 was determined to be 25(S)-ruscogenin 1O--D-glucopyranosyl (1→2)-[-D-xylopyranosyl (1→3)]--Dfucopyranoside (spicatoside A) [20] and 2 was identified as a novel compound. Compound 2 was obtained as a white amorphous powder with the molecular formula C50 H82 O23 from the negative ion Q-TOF-MS spectrum together with 1 H and 13 C NMR spectroscopy. Figure 5A shows the MS/MS of compound 2 C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
revealing [M–H]− (m/z 1049.5179) and its fragment ions at m/z 917.4747 [M–H–Xyl]− , 887.4648 [M–H–Glc]− , 755.4219 [M–H–Glc–Xyl]− , 609.3633 [M–H–2Glc–Xyl]− , and 447.3103 [M–H–2Glc–Xyl–Fucl]− . The 1 H NMR spectrum showed two singlet methyl resonances at ␦H = 1.37 (3H, s, H-19) and 0.94 ppm (3H, s, H-18), two doublet methyl signals at ␦H = 1.29 (3H, d, J = 6.9, H-21) and 1.03 ppm (3H, d, J = 6.7 Hz, H-27), and an olefinic proton resonance at ␦H = 5.91 ppm (1H, d, J = 5.9, H-6), which are attributable to a steroidal aglycone moiety. Furthermore, four anomeric protons arising from sugar moieties were observed at ␦H = 5.47 (1H, d, J = 7.8 Hz, H-1 ), 5.29 (1H, J = 7.7 Hz, H-1 ), 4.86 (1H, d, J = 7.5 Hz, H-1 ), and 4.83 ppm (1H, d, J = 7.8 Hz, H-1 ). The 13 C NMR spectrum showed 50 carbon resonances including two olefinic carbons (␦c = 140.1 (C-5) and 124.9 ppm (C-6)), one di-oxygenated carbon at ␦c = 111.0 ppm (C-22), and four anomeric carbon signals at ␦c = 106.6 (C-1 ), 105.5 (C-1 ), 105.3 (C-1 ) and 100.8 ppm (C-1 ), which are characteristic of furostanol saponins with ⌬5(6) [31]. The ␣-orientation of H-1 was determined based on its 1 H NMR resonance at ␦H = 3.89 ppm (1H, dd, J = 4.1, 11.2 Hz) [32]. The hydroxyl group at C-22 was assigned to the ␣-orientation from the hemiketal carbon resonance at ␦c = 111.0 ppm, which is higher than the -configuration by 3–4 ppm [31, 33, 34], and the 25S configuration was deduced from the chemical
www.jss-journal.com
J. Sep. Sci. 2015, 38, 18–24
Liquid Chromatography
23
Figure 5. Q-TOF-MS/MS (A) and key heteronuclear multiple bond correlation correlations of 2(B).
shift difference between two germinal-coupled protons of H26 (␦a −␦b = 0.60 > 0.57 ppm) [35]. Based on spectroscopic results, the aglycone of 2 was assigned as 25(S)-furost-5(6)ene-1-3-22␣-26-tetraol. Acid hydrolysis and 13 C NMR spectroscopy revealed a D-fucopyranose, a D-xylopyranose, and two D-glucopyranoses, and their -configuration were confirmed based on the coupling constant (>7.0 Hz) of each anomeric proton in the 1 H NMR spectrum. The interglycosidic linkages of sugar moieties were determined from the heteronuclear multiple bond correlation between ␦H = 4.86 ppm (H1 ) to ␦c = 83.2 ppm (C-1), ␦H = 5.29 ppm (H-1 ) to ␦c = 83.4 ppm (C-3 ), ␦H = 5.47 ppm (H-1 ) to ␦c = 79.2 ppm (C-2 ), and ␦H = 4.83 ppm (H-1 ) to ␦c = 75.7 ppm (Fig. 5B). Therefore, the chemical structure of 2 was determined to be 26-O--D-glucopyranosyl-25(S)-furost-5(6)-ene-1-3-22␣26-tetraol-1-O--D-glucopyranosyl (1→2)-[-D-xylopyranosyl(1→3)]--D-fucopyranoside (spicatoside D). 3.4.1 Compound 1 (spicatoside A) [␣]22 D : –64.1 (c = 0.05, MeOH); negative ion Q-TOF-MS: m/z 869.4526 [M–H]− (calcd 869.4535 for C44 H69 O17 ) and 905.4304 [M+Cl]− (calcd 905.4302 for C44 H70 O17 Cl); 1 H NMR (500 MHz, pyridine-d5 ): ␦H 5.56 (1H, d, J = 5.2 Hz, H-6), 0.86 (3H, s, H-18), 1.36 (3H, s, H-19), 1.11 (3H, d, J = 6.9 Hz, H-21), 1.07 (3H, d, J = 7.1 Hz, H-27), 4.85 (1H, d, J = 7.7 Hz, H-1 ), 1.52 (3H, d, J = 6.2 Hz, H-6 ), 5.29 (1H, d, J = 7.7 Hz, H1 ), 5.47 ppm (1H, d, J = 7.9 Hz, H-1 ); 13 C NMR (125 MHz, pyridine-d5 ): ␦c 83.2 (C-1), 37.7 (C-2), 68.6 (C-3), 44.0 (C-4), 140.1 (C-5), 124.8 (C-6), 32.7 (C-7), 33.3 (C-8), 50.7 (C-9), 43.2 C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
(C-10), 24.0 (C-11), 40.7 (C-12), 40.5 (C-13), 57.3 (C-14), 32.3 (C-15), 81.5 (C-16), 63.2 (C-17), 17.1 (C-18), 15.2 (C-19), 42.8 (C-20), 15.4 (C-21), 110.0 (C-22), 26.7 (C-23), 26.5 (C-24), 27.9 (C-25), 65.3 (C-26), 16.6 (C-27) 100.8 (C-1 ), 79.2 (C-2 ), 83.4 (C-3 ), 72.6 (C-4 ), 71.4 (C-5 ), 17.5 (C-6 ), 106.6 (C-1 ), 75.4 (C-2 ), 78.4 (C-3 ), 71.0 (C-4 ), 67.6 (C-5 ), 105.3 (C-1 ), 76.8 (C-2 ),79.0 (C-3 ), 72.3 (C-4 ), 78.7 (C-5 ), 63.6 ppm (C-6 ).
3.4.2 Compound 2 (spicatoside D) [␣]22 D : –74.2 (c = 0.07, MeOH); negative ion Q-TOF-MS: m/z 1049.5179 [M–H]− (calcd 1049.5169 for C50 H81 O23 ), 917.4747 [M–H–Xyl]− , 887.4648 [M–H–Glc]− , 755.4219 [M–H–Glc– Xyl]− , 609.3633 [M–H–2Glc–Xyl]− and 447.3103 [M–H–2Glc– Xy–Fucl]− ; 1 H NMR (500 MHz, pyridine-d5 ): ␦H 5.91 (1H, d, J = 5.9, H-6), 5.47 (1H, d, J = 7.8 Hz, H-1 ), 5.29 (1H, J = 7.7 Hz, H-1 ), 4.86 (1H, d, J = 7.5 Hz, H-1 ), 4.83 (1H, d, J = 7.8 Hz, H-1 ), 3.89 (1H, dd, J = 4.1, 11.2 Hz, H1), 1.37 (3H, s, H-19), 1.29 (3H, d, J = 6.9, H-21), 1.03 (3H, d, J = 6.7 Hz, H-27), 0.94 ppm (3 H, s, H-18); 13 C NMR (125 MHz, pyridine-d5 ): ␦c 83.2 (C-1), 37.6 (C-2), 68.6 (C-3), 44.0 (C-4), 140.1 (C-5), 124.8 (C-6), 32.4 (C-7), 33.3 (C-8), 50.7 (C-9), 43.2 (C-10), 24.0 (C-11), 40.8 (C-12), 41.1 (C13), 57.2 (C-14), 32.3 (C-15), 81.5 (C-16), 64.2 (C-17), 17.2 (C-18), 15.4 (C-19), 40.9 (C-20), 16.7 (C-21), 113.0 (C-22), 31.3 (C-23), 28.6 (C-24), 34.8 (C-25), 75.7 (C-26), 17.8 (C27), 100.8 (C-1 ), 79.2 (C-2 ), 83.4 (C-3 ), 72.6 (C-4 ), 71.4 (C-5 ), 17.5 (C-6 ), 106.6 (C-1 ), 75.4 (C-2 ), 78.7 (C-3 ), 71.0 (C-4 ), 67.6 (C-5 ), 105.5 (C-1 ), 76.8 (C-2 ), 78.9 (C-3 ), 72.3 www.jss-journal.com
24
S.-J. Choi et al.
J. Sep. Sci. 2015, 38, 18–24
(C-4 ), 78.9 (C-5 ), 63.6 (C-6 ), 105.3 (C-1 ), 75.6 (C-2 ), 78.8 (C-3 ), 72.0 (C-4 ), 78.4 (C-5 ), 63.1 ppm (C-6 ).
[10] Wang, H. C., Wu, C. C., Cheng, T. S., Kuo, C. Y., Tsai, Y. C., Chiang, S. Y., Wong, T. S., Wu, Y. C., Chang, F. R., Evid. Based Complement. Alternat. Med. 2013, doi: 10.1155/2013/857929.
4 Concluding remarks
[11] Tsai, Y. C., Chiang, S. Y., El-Shazly, M., Wu, C. C., Beerhues, L., Lai, W. C., Wu, S. F., Yen, M. H., Wu, Y. C., Chang, F. R., Food Chem. 2013, 140, 305–314.
The present work describes the first HPCCC–ELSD method for the isolation of two steroidal saponins (1 and 2) from L. platyphylla. On various two-phase solvent systems, the KD difference between 1 and 2 was too large to be separated in one-step run using a single two-phase solvent system, and thus a two-step HPCCC method was applied to yield respective target compounds within 50 min in each step. Based on 1 H and 13 C NMR spectroscopy and ESI-Q-TOF-MS/MS data, 1 was determined to be spicatoside A and 2 was identified as a novel furostanol saponin, 26-O--D-glucopyranosyl-25(S)furost-5(6)-ene-1-3-22␣-26-tetraol-1-O--D-glucopyranosyl (1→2)-[-D-xylopyranosyl-(1→3)]--D-fucopyranoside (spicatoside D). This study indicates that HPCCC is a simple, rapid, and efficient method for discovering natural products from medicinal plants. This work was supported by Basic Science Research Program through the National Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (2011-0011899); Cooperative Research Program for Agriculture Science & Technology Development program (Project no. PJ009830); Rural Development Administration, Republic of Korea; and Research Fund of the Catholic University of Korea (2011). The authors have declared no conflict of interest.
5 References [1] Li, G., Ra, W. H., Park, J. W., Kwon, S. W., Lee, J. H., Park, C. B., Park, Y. J., Biochim. Syst. Ecol. 2011, 39, 241– 252. [2] Kim, J. E., Hwnag, I. S., Choi, S. I., Lee, H. R., Lee, Y. J., Goo, J. S., Lee H. S., Son, H. J., Jang, M. J., Lee, S. H., Kang, B. C., Hwang, D. Y., Lab. Anim. Res. 2012, 28, 181–191. [3] Chen, X., Bai, X., Liu, Y., Tian, L., Zhou, J., Zhou, Q., J. Ethnopharmacol. 2009, 122, 205–209. [4] Lee, Y. C., Lee, J. C., Seo, Y. B., Kook, Y. B., J. Ethnopharmacol. 2005, 101, 144–152. [5] Kim J. E., Lee, J. Y., Kwak, M. H., Ko, J., Hong, J. T., Hwang, D. Y., BMC Complement. Altern. Med. 2013, 13, 333. [6] Kwak, M. H., Kim J. E., Hwang, I. S., Lee Y. J., An, B. S., Hong, J. T., Lee, S. H., Hwang, D. Y., J. Ethnopharmacol. 2013, 148, 880–889. [7] Lee, D. S., Choi, J., Kim, S. H., Kim, S., Biol. Pharm. Bull. 2014, 37, 954–960. [8] Li, W. J., Cheng, X. L., Liu, J., Lin, R. C., Wang, G. L., Du, S. S., Liu, Z. L., Molecules, 2012, 17, 1797–1808. [9] Zhao, R. P., Lin, S. S., Yuan, S. T., Yu, B. Y., Bai, X. S., Sun, L., Zhang, L. Y., Chin. J. Nat. Med. 2014 12, 24–29. C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
[12] Tian, Y., Ma, S., Lin, B., Kou, J., Yu, B., Indian J. Pharmacol. 2013, 45, 283–285. [13] Tsurng, T. J., Tsai, Y. C., Chiang, S. Y., Wang, G. J., Kuo, Y. C., Chang, Y. C., Wu, Y. Y., Wu, Y. C., Virus Res. 2014, 192, 16–24. [14] Li, Y. W., Qi, J., Zhang, W., Zhou, S. P., Wu, Y., Yu, B. Y., J. Sep. Sci. 2014, 37, 1762–1772. [15] Hu, Z., Jiang, C., Qin, M., Qi, J., Yu, B., Zhongguo Zhong Yao Za Zhi 2010, 35, 2508–2510. [16] Yu, B. Y., Hirai, Y., Shoji, J., Xu, G. J., Chem. Pharm. Bull. 1990, 38, 1931–1935. [17] Wang, K. W., Zhang, H., Shen, L. Q., Wang, W., Carbohydr. Res. 2011, 346, 253–258. [18] Do, J. C., Jung, K. Y., Song, Y. K., J. Nat. Prod. 1995, 58, 778–781. [19] Do, J. C., Sung, Y. K., Son, K. H., Kor. J. Pharmacogn. 1991, 22, 73–77. [20] Lee, D. Y., Son, K. H., Do, J. C., Kang, S. S., Arch. Pharm. Res. 1989, 12, 295–299. [21] Shin, J. S., Kor. J. Crop. Sci. 2002, 47, 236–239. [22] Kim, S. H., Kim, H. K., Yang, E. S., Lee, K. Y., Kim, S. D., Kim, Y. C., Sung, S. H., Sep. Purif. Technol. 2010, 71, 168–172. [23] Ha, H. G., Jung, D. Y., Kang, S. S., Kim, J. S., Ho, G., Repub. Korean Kongkae Taeho Kongbo 2003, KR2003070282. [24] Hur, J., Lee, P., Moon, E., Kang, I., Kim, S. H., Oh, M. S., Kim, S. Y., Eur. J. Pharmacol. 2009, 620, 9–15. [25] Kwon, G., Lee, H. E., Lee, D. H., Woo, H., Park, S. J., Gao, Q., Ahn, Y. J., Son, K. H., Ryu, J. H., Neurosci. Lett. 2014, 572, 58–62. [26] Park, S. H., Lee, H. J., Ryu, J., Son, K. H., Kwon, S. Y., Lee, S. K., Kim, Y. S., Hong, J. H., Seok, J. H., Lee, C. J., Phytomedicine 2014, 21, 172–176. [27] Ito, Y., J. Chromatogr. A 2005, 1065, 145–168. [28] Guzlek, H., Wood, P. L., Janaway, L., J. Chromatogr. A 2009, 1216, 4181–4186. [29] Ito, Y., J. Chromatogr. A. 2005, 1065, 145–188. [30] Chen, I. S., Shen, C. S. J., Sheppard, A. J., J. Am. Oil. Chem. Soc. 1981, 58, 599–601. [31] Zhang, T., Kang, L. P., Yu, H. S., Liu, Y. X., Zhao, Y., Xiong, C. Q., Zhang, J., Zou, P., Song, X. B., Liu, C., Ma, B. P., Steroids 2012, 77, 1298–1305. [32] Guo, Y., Liu, Y. X., Kang, L. P., Zhang, T., Yu, H. S., Zhao, Y., Xiong, C. Q., Ma, B. P., J. Asian Nat. Prod. Res. 2013, 15, 459–465. [33] Zou, K., Wang, J., Du, M., Li, Q., Tu, G., Chem. Pharm. Bull. 2006, 54, 1440–1442. [34] Dini, I., Tenore, G. C., Trimarco, E., Dini, A., Food Chem. 2005, 93, 205–214. [35] Agrawal, P. K., Steroids, 2005, 70, 715–724.
www.jss-journal.com
of Pharmacy and Integrated Research Institute of Pharmaceutical Sciences, The Catholic University of Korea, Bucheon, Republic of Korea 2 College of Pharmacy and Research Institute of Pharmaceutical Science, Seoul National University, Seoul, Republic of Korea 3 Materials Science Team, Medicinal Beauty Division, Amorepacific Corporation R&D Unit, Yongin, Republic of Korea Received September 13, 2014 Revised October 7, 2014 Accepted October 11, 2014
J. Sep. Sci. 2015, 38, 18–24
Research Article
Application of high-performance countercurrent chromatography for the isolation of steroidal saponins from Liriope plathyphylla High-performance countercurrent chromatography (HPCCC) with electrospray lightscattering detection was applied for the first time to isolate a spirostanol and a novel furostanol saponin from Liriope platyphylla. Due to the large differences in KD values between the two compounds, a two-step HPCCC method was applied in this study. The primary HPCCC employed methylene chloride/methanol/isopropanol/water (9:6:1:4 v/v, 4 mL/min, normal-phase mode) conditions to yield a spirostanol saponin (1). After the primary HPCCC run, the solute retained in the stationary phase (SP extract) in HPCCC column was recovered and subjected to the second HPCCC on the n-hexane/n-butanol/water system (1:9:10 v/v, 5 mL/min, reversed-phase mode) to yield a novel furostanol saponin (2). The isolated spirostanol saponin was determined to be 25(S)-ruscogenin 1-O--D-glucopyranosyl (1→2)[-D-xylopyranosyl (1→3)]--D-fucopyranoside (spicatoside A), and the novel furostanol saponin was elucidated to be 26-O--D-glucopyranosyl-25(S)-furost-5(6)-ene-1-3-22␣(1→2)-[-D-xylopyranosyl-(1→3)]--D-fucopyranoside 26-tetraol-1-O--D-glucopyranosyl (spicatoside D). Keywords: Electrospray light-scattering detection / Furostanol saponins / Highperformance countercurrent chromatography / Liriope platyphylla DOI 10.1002/jssc.201401007
1 Introduction The genus Liriope, belonging to the Liliaceae family, is abundantly distributed in subtropical and temperate regions and comprises about eight species globally [1]. In China, Korea, and Japan, roots of some Liriope species have been used traditionally to treat asthma, bronchial and lung inflammation, and sputum [2] and are added to various traditional medicinal formulas as expectorants and tonics. Recently, many studies have reported biological activities of Liriope species including anti-diabetic [3], anti-inflammatory, anti-asthmatic [4], laxative [5], anti-atopic dermatitis [6], and neuroprotective effects [7]. In addition, several reports have revealed that phenolic compounds and saponins from Liriope species showed antioxidant [8], anti-cancer [9, 10], estrogenic, anti-platelet [11], antiCorrespondence: Prof. Kee Dong Yoon, Integrated Research Institute of Pharmaceutical Sciences, College of Pharmacy, The Catholic University of Korea, 43 Jibong-ro, Wonmi-gu, Bucheonsi, Gyeonggi-do 420–743, Republic of Korea E-mail: [email protected] Fax: +82-2-2164-4059
Abbreviations: C/M/I/Wat, chloroform/methanol/isopropanol/water; E/B/W, ethyl acetate/n-butanol/water; ELSD, electrospray light-scattering detection; H/B/Wat, n-hexane/nbutanol/water; HPCCC, high-performance countercurrent chromatography; MC/M/I/Wat, methylene chloride/methanol/isopropanol/water; SP extract, stationary phase extract C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
thrombotic [12], and antiviral [13] activities. With regard to the secondary metabolites, steroidal saponins are characteristic components in Liriope species [14–18]. Among the known steroidal saponins, several saponins, such as spicatoside A, B, C, ophiopoginin B, prosapogenin II and III of spicatoside A, and -sitosterol glucoside, were isolated from Liriope spicata and Liriope plathyphylla in Korea [19–22], and several reports have demonstrated that spicatoside A stimulated growth hormone secretion [23], induced neurite outgrowth [24], enhanced memory consolidation [25], and increased mucin production and secretion [26]. Although the steroidal saponins are considered typical marker compounds of Liriope species, the polysaccharides are contained as major constituents and steroidal saponins are present in relatively low amounts, which complicate the isolation and determination of steroidal saponins. Therefore, efficient isolation methods are required to identify steroidal saponins in Liriope species. CCC is a type of liquid–liquid chromatography that employs only liquids for the mobile and stationary phases. Thus, there is no risk of chemical degradation and irreversible adsorption of target compounds onto a solid support, which is common in conventional column chromatography [27]. Among CCC instruments, high-performance countercurrent chromatography (HPCCC) is used to generate high g-levels up to 240 × g to retain a higher proportion of the ∗ These
authors contributed equally to this work.
www.jss-journal.com
J. Sep. Sci. 2015, 38, 18–24
Liquid Chromatography
19
Figure 1. Chemical structures of spicatoside A (1) and spicatoside D (2).
stationary phase even at a higher mobile-phase flow-rate than high-speed CCC. Therefore, faster and more efficient separation is possible compared with conventional CCC methods [28]. Many steroidal saponins of Liriope species have been isolated using solid support based conventional column chromatography, but there is no report concerning countercurrent separation of steroidal saponins from Liriope species. The present study describes HPCCC coupled with an electrospray light-scattering detection (ELSD) method for the isolation of spicatoside A and a novel furostanol saponin, spicatoside D (Fig. 1), from Liriope plathyphylla.
2 Materials and methods 2.1 Instrumentation The semipreparative HPCCC instrument employed in this study was a Spectrum (Dynamic Extractions, Berkshire, UK), which possesses two semipreparative coils of 70.5 mL with a 3.2 mm id. The -values of the semipreparative coils ranged from 0.52 to 0.86. The rotational speed was adjustable from 0 to 1600 rpm. The Spectrum was coupled to an IOTA S 300 pump (Ecom, Prague, Czech Republic), a Sedex 75 ELSD (Sedere, Olivet, France), a Foxy R2 fraction collector (Teledyne Isco, NE, USA), and a CCA-1111 circulatory temperature regulator (Eyela, Tokyo, Japan) to maintain the internal HPCCC temperature at 30⬚C. HPLC analyses were performed on an HPLC system equipped with a 1525 binary HPLC pump (Waters, MA, USA), a manual injection valve (Rheodyne, CA, USA) with a 20 L sample loop, and a Sedex 75 ELSD. HPCCC peak fractions were identified using 1 H and 13 C NMR spectroscopy and MS using an AVANCE 500 spectrometer (Bruker, Karlsruhe, Germany) and a 6530 QTOF mass spectrometer (Agilent Technologies, CA, USA), respectively.
a Millipore Milli-Q water purification system (Millipore, MA, USA). ACN and water used for HPLC analyses were of HPLC grade and purchased from Fisher Scientific Korea (Seoul, Korea). Dried tubers of L. plathyphylla, originating from Miryang province in Korea, were purchased from Humanherb (Deagusi, Korea). The voucher specimen (CU-140629) was deposited at the herbarium of the College of Pharmacy, The Catholic University of Korea.
2.3 Preparation of sample extract Dried tubers of L. plathyphylla (2 kg) were ground into rough powder and extracted with methanol (4 L × 90 min × three times) using an ultrasonic bath, and evaporated under reduced pressure to yield a methanol extract (74 g). The methanol extract was suspended in water and sequentially partitioned with organic solvents to yield ethyl acetate (2.3 g) and n-butanol (8.1 g) soluble extracts. The n-butanol-soluble extract was stored in a refrigerator (4⬚C) prior to HPLC analysis and HPCCC separation.
2.4 HPLC–ELSD analysis The n-butanol-soluble extract of the tubers of L. platyphylla and HPCCC peak fractions were analyzed using HPLC–ELSD with a Zorbax C18 column (250 × 4.6 mm, id, 5 m; Agilent Technologies, CA, USA). The mobile phases were acetonitrile for solvent A (0.1% formic acid) and water (0.1% formic acid) for solvent B. The mobile phase gradient was 25% A (0– 5 min), 25–65% A (5–35 min), and 65% A (35–40 min). The mobile phase flow-rate was 1.0 mL/min and the injection volume was 20 L.
2.5 Q-TOF-MS/MS analysis 2.2 Reagents and plant materials Organic solvents for HPCCC separation were of analytical grade and purchased from Daejung-Chemical and Metals (Kyeonggi-Do, Korea). Ultrapure water was obtained from C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
MS analysis was performed on 6530 Q-TOF mass spectrometer with ESI with the following acquisition parameters: drying gas (nitrogen) flow rate—8.0 L/min, drying gas temperature—300⬚C, nebulizer gas pressure—35 psig, capillary—3500 V, OCT RFV—750 V, fragmentor voltage— www.jss-journal.com
20
S.-J. Choi et al.
200 V. The collision energy was set to 60 V for negative MS/MS analysis. The MS system was operated by MassHunter Acquisition software, version B.05.00.
2.6 Evaluation of partition coefficient values (KD ) The partition coefficient (KD ) values of compounds 1 and 2 were determined by HPLC–ELSD analysis. The KD value of each target compound was calculated as the HPLC– ELSD peak area of the target compound in the stationary phase (upper phase) divided by that in the mobile phase (lower phase) under various ratios of two-phase solvent systems (chloroform/methanol/isopropanol/water (C/M/I/Wat, 7:6:1:4, 9:6:1:4, 11:6:1:4, v/v), methylene chloride/methanol/isopropanol/water (MC/M/I/Wat, 7:6:1:4, 9:6:1:4, 11:6:1:4, v/v), n-hexane/n-butanol/water (H/B/Wat, 3:7:10, 2:8:10, 1:9:10, v/v), and ethyl acetate/n-butanol/water (E/B/W, 1:9:10, v/v)). n-Butanol-soluble extract (5 mg) of L. plathyphylla was dissolved in a 1:1 mixture of upper and lower phase (each 2 mL) of the above two-phase solvent systems.
J. Sep. Sci. 2015, 38, 18–24
2.8 Second HPCCC procedure for compound 2 The second preparative HPCCC for the separation of 2 was performed after monitoring HPCCC chromatograms, similar to the primary HPCCC procedure. Two-phase solvent systems composed of H/B/Wat (3:7:10, 2:8:10, 1:9:10, v/v) and E/B/Wat (1:9:10, v/v) were used to check HPCCC separation patterns. Finally, the second preparative HPCCC was performed using the following conditions: H/B/Wat (1:9:10, v/v) two-phase solvent system, 5.0 mL/min flow rate, 1600 rpm rotational speed, and 100 mg of SP extract. The second preparative HPCCC was performed twice.
2.9 Structural determination of compounds 1 and 2 Structural determination of 1 and 2 in HPCCC peak fractions was performed based on Q–TOF-MS, 1 H, and 13 C NMR spectroscopic data, as well as comparison with previously published values.
3 Results and discussion 3.1 HPLC–ELSD analysis and evaluation of partition coefficient values
2.7 Primary HPCCC procedure for compound 1 To determine the optimal HPCCC pattern for 1, C/M/I/Wat (7:6:1:4, 9:6:1:4, 11:6:1:4, v/v) and MC/M/I/Wat (7:6:1:4, 9:6:1:4, 11:6:1:4, v/v) systems were used in the present study. The HPCCC coil was filled with upper aqueous phase at a flow rate of 10 mL/min. After filling stationary phase in the column, the HPCCC was rotated at 1600 rpm while the lower mobile phase was eluted in head-to-tail mode at a flow rate of 4 mL/min. Sample extract (20 mg) was dissolved in the lower organic phase and subjected to HPCCC after the mobile phase had emerged in the effluent. The HPCCC chromatograms were monitored using ELSD. The ELSD parameter was modulated with a tube temperature of 90⬚C, a gain level of 5, and 3.0 bar of nebulizer nitrogen gas pressure. Preparative HPCCC was conducted according to the results of the above HPCCC separation patterns. The HPCCC parameters were as follows: MC/M/I/Wat (11:6:1:4, v/v) twophase solvent system, 1600 rpm rotational speed, 4 mL/min flow rate, and 400 mg (200 mg × two times) of n-butanolsoluble extract. The tail outlet of the HPCCC was combined with ELSD through a split valve. The split ratio of the effluent was modulated using a spilt valve; one-tenth of the HPCCC effluent was subjected to ELSD and nine-tenths of the effluent was collected using a fraction collector for 1 min per fraction. After HPCCC separation, all solvents in the column were collected by displacement with N2 gas to measure the stationary phase retention, and it was evaporated under reduced pressure to yield the sample (stationary phase extract (SP extract)) for the second HPCCC. The primary preparative HPCCC for 1 was performed in duplicate. C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Previous reports have revealed that L. plathyphylla possessed furostanol- and spirostanol-type saponins as secondary metabolites. As shown in Fig. 2, the HPLC chromatogram of n-butanol-soluble extract of L. plathyphylla showed one peak (1) at 32.7 min and a relatively major peak (2) at 13.9 min, indicative of spirostanol and furostanol saponins, respectively. The KD values of target compounds should range from 0.2 to 2.0 for optimal CCC separation [29]. To select a suitable two-phase solvent system, C/M/I/Wat (7:6:1:4, 9:6:1:4 and 11:6:1:4, v/v), MC/M/I/Wat (7:6:1:4, 9:6:1:4 and 11:6:1:4, v/v), H/B/Wat (3:7:10, 2:8:10 and 1:9:10, v/v), and E/B/Wat (1:9:10, v/v) were tested to determine the partition coefficient of the two corresponding compounds because these two-phase solvent systems have been used previously to separate various saponins from medicinal plants. The KD values of 1 under various ratios of C/M/I/Wat and MC/M/I/Wat solvent systems ranged from 0.13 to 1.77, while those of 2 were over 15.5. The opposite results were observed under H/B/Wat and E/B/Wat systems; the KD values were over 17.8 for 1 and 0.24–1.33 for 2, respectively (Table 1). These results suggested that the difference in KD values between 1 and 2 is too large for separation using one-step HPCCC with a single two-phase solvent system. Therefore, the two-step HPCCC method was used to isolate 1 and 2.
3.2 HPCCC separation of compound 1 As described above, the KD values of 1 under C/M/I/Wat and MC/M/I/Wat systems ranged from 0.13 to 1.77. The www.jss-journal.com
Liquid Chromatography
J. Sep. Sci. 2015, 38, 18–24
21
Figure 2. HPLC–ELSD chromatograms of the n-butanol-soluble extract of L. plathyphylla (A), SP extract (B), and the HPCCC peak fraction of 1 (C) and 2 (D).
Table 1. Partition coefficient (KD ) values of 1 and 2
Solvent system (volume ratio, v/v)
KD value 1
Chloroform/methanol/isopropanol/water 7:6:1:4 (v/v) 0.13 9:6:1:4 (v/v) 0.18 11:6:1:4 (v/v) 0.32 Methylene chloride/methanol/isopropanol/water 7:6:1:4 (v/v) 1.42 9:6:1:4 (v/v) 1.53 11:6:1:4 (v/v) 1.77 n-Hexane/n-butanol/water 3:7:10 (v/v) 17.80 2:8:10 (v/v) >20 1:9:10 (v/v) >20 Ethyl acetate/n-butanol/water 1:9:10 (v/v) >20
2
15.5 19.3 >20 >20 >20 >20 0.24 0.59 1.33 0.92
optimum separation pattern was determined from the HPCCC separation using each two-phase solvent system. As expected based on their KD values, the elution times of 1 on C/M/I/Wat system (Fig. 3A–C) were shorter than those of MC/M/I/Wat system (Fig. 3D–F), but compound 1 was not separated from nearby impurity peaks. The MC/M/I/Wat system yielded better results; the peak corresponding to 1 was well resolved from impurity peaks and MC/M/I/Wat (11:6:1:4, v/v) gave the best separation. Methylene chloride is C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
similar to chloroform in terms of its physical and chemical properties, but methylene chloride is used instead of chloroform for reasons of safety [30]. However, methylene chloride is less polar than chloroform and the MC/M/I/Wat system yielded better separation than the C/M/I/Wat system in this study. Using the MC/M/I/Wat (11:6:1:4, v/v, 4 mL/min), 400 mg (200 mg × 2) of n-butanol-soluble L. platyphylla extract was introduced into HPCCC to yield 7.4 mg of 1 with a purity of 96.2% (Figs. 2C and 3G). The mean stationary phase retention was 71.1%.
3.3 HPCCC separation of compound 2 After primary HPCCC separation of 1, the stationary phase retained in the HPCCC coil was extruded by N2 gas and pooled to yield 236 mg of SP extract, which was analyzed using HPLC–ELSD. As shown in Fig. 2B, the peak corresponding to 1 disappeared and 2 was recovered in the SP extract. The SP extract was subjected to HPCCC under various H/B/Wat (3:7:10, 2:8:10, 1:9:10, v/v) and E/B/Wat (1:9:10, v/v) systems. On H/B/Wat systems, 2 eluted too rapidly to be separated from polar impurities under the 3:7:10 condition (v/v, 4 mL/min; Fig. 4A). The 2:8:10 (v/v, 4 mL/min) condition showed improved separation (Fig. 4B), but 2 was still eluted in proximity to impurity peaks (expressed as asterisk in the figure). The 1:9:10 (v/v, 4 mL/min) condition separated 2 from the impurity peaks (Fig. 4C), and increasing the mobile phase flow rate (5 mL/min) shortened the www.jss-journal.com
22
S.-J. Choi et al.
J. Sep. Sci. 2015, 38, 18–24 Figure 3. HPCCC–ELSD chromatograms of the n-butanolsoluble extract of L. plathyphylla. (A) C/M/I/Wat (7:6:1:4, v/v, 4 mL/min); (B) C/M/I/Wat (9:6:1:4, v/v, 4 mL/min); (C) C/M/I/Wat (11:6:1:4, v/v, 4 mL/min); (D) MC/M/I/Wat (7:6:1:4, v/v, 4 mL/min); (E) MC/M/I/Wat (9:6:1:4, v/v, 4 mL/min); (F) MC/M/I/Wat (11:6:1:4, v/v, 4 mL/min); (G) MC/M/I/Wat (11:6:1:4, v/v, 4 mL/min, sample amount: 400 mg (200 mg × two times)). C, chloroform; MC, methylene chloride; M: methanol; I: isopropanol; Wat: water.
Figure 4. HPLC–ELSD chromatograms of SP extract. (A) H/B/Wat (3:7:10, v/v, 4 mL/min); (B) H/B/Wat (2:8:10, v/v, 4 mL/min); (C) H/B/Wat (1:9:10, v/v, 4 mL/min); (D) H/B/Wat (1:9:10, v/v, 5 mL/min); (E) H/B/Wat (1:9:10, v/v, 4 mL/min); (F) H/B/Wat (1:9:10, v/v, 5 mL/min, sample amount: 200 mg (100 mg × two times)).H, n-hexane; chloroform; B, n-butanol; Wat: water; E: ethyl acetate.
elution time of 2 while maintaining identical separation patterns (Fig. 4D). The E/B/Wat (2:8:10, v/v, 4 mL/min) condition did not result in satisfactory separation compared with H/B/Wat (1:9:10, v/v, 5 mL/min) conditions (Fig. 4E). Therefore, H/B/Wat (1:9:10, v/v, 5 mL/min) conditions were selected for secondary HPCCC for the isolation of 2 from SP extract. A total of 200 mg (100 mg × 2) of SP extract subjected to HPCCC yielded 35.5 mg of 2, with a purity of over 98% (Figs. 2D and 4F). The mean stationary phase retention was 64.3%.
3.4 Structural determination of compounds 1 and 2 The structural elucidation of 1 and 2 was performed using 1 H and 13 C NMR spectroscopy and Q-TOF-MS, as well as published values of steroidal saponins from Liriope species. Consequently, 1 was determined to be 25(S)-ruscogenin 1O--D-glucopyranosyl (1→2)-[-D-xylopyranosyl (1→3)]--Dfucopyranoside (spicatoside A) [20] and 2 was identified as a novel compound. Compound 2 was obtained as a white amorphous powder with the molecular formula C50 H82 O23 from the negative ion Q-TOF-MS spectrum together with 1 H and 13 C NMR spectroscopy. Figure 5A shows the MS/MS of compound 2 C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
revealing [M–H]− (m/z 1049.5179) and its fragment ions at m/z 917.4747 [M–H–Xyl]− , 887.4648 [M–H–Glc]− , 755.4219 [M–H–Glc–Xyl]− , 609.3633 [M–H–2Glc–Xyl]− , and 447.3103 [M–H–2Glc–Xyl–Fucl]− . The 1 H NMR spectrum showed two singlet methyl resonances at ␦H = 1.37 (3H, s, H-19) and 0.94 ppm (3H, s, H-18), two doublet methyl signals at ␦H = 1.29 (3H, d, J = 6.9, H-21) and 1.03 ppm (3H, d, J = 6.7 Hz, H-27), and an olefinic proton resonance at ␦H = 5.91 ppm (1H, d, J = 5.9, H-6), which are attributable to a steroidal aglycone moiety. Furthermore, four anomeric protons arising from sugar moieties were observed at ␦H = 5.47 (1H, d, J = 7.8 Hz, H-1 ), 5.29 (1H, J = 7.7 Hz, H-1 ), 4.86 (1H, d, J = 7.5 Hz, H-1 ), and 4.83 ppm (1H, d, J = 7.8 Hz, H-1 ). The 13 C NMR spectrum showed 50 carbon resonances including two olefinic carbons (␦c = 140.1 (C-5) and 124.9 ppm (C-6)), one di-oxygenated carbon at ␦c = 111.0 ppm (C-22), and four anomeric carbon signals at ␦c = 106.6 (C-1 ), 105.5 (C-1 ), 105.3 (C-1 ) and 100.8 ppm (C-1 ), which are characteristic of furostanol saponins with ⌬5(6) [31]. The ␣-orientation of H-1 was determined based on its 1 H NMR resonance at ␦H = 3.89 ppm (1H, dd, J = 4.1, 11.2 Hz) [32]. The hydroxyl group at C-22 was assigned to the ␣-orientation from the hemiketal carbon resonance at ␦c = 111.0 ppm, which is higher than the -configuration by 3–4 ppm [31, 33, 34], and the 25S configuration was deduced from the chemical
www.jss-journal.com
J. Sep. Sci. 2015, 38, 18–24
Liquid Chromatography
23
Figure 5. Q-TOF-MS/MS (A) and key heteronuclear multiple bond correlation correlations of 2(B).
shift difference between two germinal-coupled protons of H26 (␦a −␦b = 0.60 > 0.57 ppm) [35]. Based on spectroscopic results, the aglycone of 2 was assigned as 25(S)-furost-5(6)ene-1-3-22␣-26-tetraol. Acid hydrolysis and 13 C NMR spectroscopy revealed a D-fucopyranose, a D-xylopyranose, and two D-glucopyranoses, and their -configuration were confirmed based on the coupling constant (>7.0 Hz) of each anomeric proton in the 1 H NMR spectrum. The interglycosidic linkages of sugar moieties were determined from the heteronuclear multiple bond correlation between ␦H = 4.86 ppm (H1 ) to ␦c = 83.2 ppm (C-1), ␦H = 5.29 ppm (H-1 ) to ␦c = 83.4 ppm (C-3 ), ␦H = 5.47 ppm (H-1 ) to ␦c = 79.2 ppm (C-2 ), and ␦H = 4.83 ppm (H-1 ) to ␦c = 75.7 ppm (Fig. 5B). Therefore, the chemical structure of 2 was determined to be 26-O--D-glucopyranosyl-25(S)-furost-5(6)-ene-1-3-22␣26-tetraol-1-O--D-glucopyranosyl (1→2)-[-D-xylopyranosyl(1→3)]--D-fucopyranoside (spicatoside D). 3.4.1 Compound 1 (spicatoside A) [␣]22 D : –64.1 (c = 0.05, MeOH); negative ion Q-TOF-MS: m/z 869.4526 [M–H]− (calcd 869.4535 for C44 H69 O17 ) and 905.4304 [M+Cl]− (calcd 905.4302 for C44 H70 O17 Cl); 1 H NMR (500 MHz, pyridine-d5 ): ␦H 5.56 (1H, d, J = 5.2 Hz, H-6), 0.86 (3H, s, H-18), 1.36 (3H, s, H-19), 1.11 (3H, d, J = 6.9 Hz, H-21), 1.07 (3H, d, J = 7.1 Hz, H-27), 4.85 (1H, d, J = 7.7 Hz, H-1 ), 1.52 (3H, d, J = 6.2 Hz, H-6 ), 5.29 (1H, d, J = 7.7 Hz, H1 ), 5.47 ppm (1H, d, J = 7.9 Hz, H-1 ); 13 C NMR (125 MHz, pyridine-d5 ): ␦c 83.2 (C-1), 37.7 (C-2), 68.6 (C-3), 44.0 (C-4), 140.1 (C-5), 124.8 (C-6), 32.7 (C-7), 33.3 (C-8), 50.7 (C-9), 43.2 C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
(C-10), 24.0 (C-11), 40.7 (C-12), 40.5 (C-13), 57.3 (C-14), 32.3 (C-15), 81.5 (C-16), 63.2 (C-17), 17.1 (C-18), 15.2 (C-19), 42.8 (C-20), 15.4 (C-21), 110.0 (C-22), 26.7 (C-23), 26.5 (C-24), 27.9 (C-25), 65.3 (C-26), 16.6 (C-27) 100.8 (C-1 ), 79.2 (C-2 ), 83.4 (C-3 ), 72.6 (C-4 ), 71.4 (C-5 ), 17.5 (C-6 ), 106.6 (C-1 ), 75.4 (C-2 ), 78.4 (C-3 ), 71.0 (C-4 ), 67.6 (C-5 ), 105.3 (C-1 ), 76.8 (C-2 ),79.0 (C-3 ), 72.3 (C-4 ), 78.7 (C-5 ), 63.6 ppm (C-6 ).
3.4.2 Compound 2 (spicatoside D) [␣]22 D : –74.2 (c = 0.07, MeOH); negative ion Q-TOF-MS: m/z 1049.5179 [M–H]− (calcd 1049.5169 for C50 H81 O23 ), 917.4747 [M–H–Xyl]− , 887.4648 [M–H–Glc]− , 755.4219 [M–H–Glc– Xyl]− , 609.3633 [M–H–2Glc–Xyl]− and 447.3103 [M–H–2Glc– Xy–Fucl]− ; 1 H NMR (500 MHz, pyridine-d5 ): ␦H 5.91 (1H, d, J = 5.9, H-6), 5.47 (1H, d, J = 7.8 Hz, H-1 ), 5.29 (1H, J = 7.7 Hz, H-1 ), 4.86 (1H, d, J = 7.5 Hz, H-1 ), 4.83 (1H, d, J = 7.8 Hz, H-1 ), 3.89 (1H, dd, J = 4.1, 11.2 Hz, H1), 1.37 (3H, s, H-19), 1.29 (3H, d, J = 6.9, H-21), 1.03 (3H, d, J = 6.7 Hz, H-27), 0.94 ppm (3 H, s, H-18); 13 C NMR (125 MHz, pyridine-d5 ): ␦c 83.2 (C-1), 37.6 (C-2), 68.6 (C-3), 44.0 (C-4), 140.1 (C-5), 124.8 (C-6), 32.4 (C-7), 33.3 (C-8), 50.7 (C-9), 43.2 (C-10), 24.0 (C-11), 40.8 (C-12), 41.1 (C13), 57.2 (C-14), 32.3 (C-15), 81.5 (C-16), 64.2 (C-17), 17.2 (C-18), 15.4 (C-19), 40.9 (C-20), 16.7 (C-21), 113.0 (C-22), 31.3 (C-23), 28.6 (C-24), 34.8 (C-25), 75.7 (C-26), 17.8 (C27), 100.8 (C-1 ), 79.2 (C-2 ), 83.4 (C-3 ), 72.6 (C-4 ), 71.4 (C-5 ), 17.5 (C-6 ), 106.6 (C-1 ), 75.4 (C-2 ), 78.7 (C-3 ), 71.0 (C-4 ), 67.6 (C-5 ), 105.5 (C-1 ), 76.8 (C-2 ), 78.9 (C-3 ), 72.3 www.jss-journal.com
24
S.-J. Choi et al.
J. Sep. Sci. 2015, 38, 18–24
(C-4 ), 78.9 (C-5 ), 63.6 (C-6 ), 105.3 (C-1 ), 75.6 (C-2 ), 78.8 (C-3 ), 72.0 (C-4 ), 78.4 (C-5 ), 63.1 ppm (C-6 ).
[10] Wang, H. C., Wu, C. C., Cheng, T. S., Kuo, C. Y., Tsai, Y. C., Chiang, S. Y., Wong, T. S., Wu, Y. C., Chang, F. R., Evid. Based Complement. Alternat. Med. 2013, doi: 10.1155/2013/857929.
4 Concluding remarks
[11] Tsai, Y. C., Chiang, S. Y., El-Shazly, M., Wu, C. C., Beerhues, L., Lai, W. C., Wu, S. F., Yen, M. H., Wu, Y. C., Chang, F. R., Food Chem. 2013, 140, 305–314.
The present work describes the first HPCCC–ELSD method for the isolation of two steroidal saponins (1 and 2) from L. platyphylla. On various two-phase solvent systems, the KD difference between 1 and 2 was too large to be separated in one-step run using a single two-phase solvent system, and thus a two-step HPCCC method was applied to yield respective target compounds within 50 min in each step. Based on 1 H and 13 C NMR spectroscopy and ESI-Q-TOF-MS/MS data, 1 was determined to be spicatoside A and 2 was identified as a novel furostanol saponin, 26-O--D-glucopyranosyl-25(S)furost-5(6)-ene-1-3-22␣-26-tetraol-1-O--D-glucopyranosyl (1→2)-[-D-xylopyranosyl-(1→3)]--D-fucopyranoside (spicatoside D). This study indicates that HPCCC is a simple, rapid, and efficient method for discovering natural products from medicinal plants. This work was supported by Basic Science Research Program through the National Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (2011-0011899); Cooperative Research Program for Agriculture Science & Technology Development program (Project no. PJ009830); Rural Development Administration, Republic of Korea; and Research Fund of the Catholic University of Korea (2011). The authors have declared no conflict of interest.
5 References [1] Li, G., Ra, W. H., Park, J. W., Kwon, S. W., Lee, J. H., Park, C. B., Park, Y. J., Biochim. Syst. Ecol. 2011, 39, 241– 252. [2] Kim, J. E., Hwnag, I. S., Choi, S. I., Lee, H. R., Lee, Y. J., Goo, J. S., Lee H. S., Son, H. J., Jang, M. J., Lee, S. H., Kang, B. C., Hwang, D. Y., Lab. Anim. Res. 2012, 28, 181–191. [3] Chen, X., Bai, X., Liu, Y., Tian, L., Zhou, J., Zhou, Q., J. Ethnopharmacol. 2009, 122, 205–209. [4] Lee, Y. C., Lee, J. C., Seo, Y. B., Kook, Y. B., J. Ethnopharmacol. 2005, 101, 144–152. [5] Kim J. E., Lee, J. Y., Kwak, M. H., Ko, J., Hong, J. T., Hwang, D. Y., BMC Complement. Altern. Med. 2013, 13, 333. [6] Kwak, M. H., Kim J. E., Hwang, I. S., Lee Y. J., An, B. S., Hong, J. T., Lee, S. H., Hwang, D. Y., J. Ethnopharmacol. 2013, 148, 880–889. [7] Lee, D. S., Choi, J., Kim, S. H., Kim, S., Biol. Pharm. Bull. 2014, 37, 954–960. [8] Li, W. J., Cheng, X. L., Liu, J., Lin, R. C., Wang, G. L., Du, S. S., Liu, Z. L., Molecules, 2012, 17, 1797–1808. [9] Zhao, R. P., Lin, S. S., Yuan, S. T., Yu, B. Y., Bai, X. S., Sun, L., Zhang, L. Y., Chin. J. Nat. Med. 2014 12, 24–29. C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
[12] Tian, Y., Ma, S., Lin, B., Kou, J., Yu, B., Indian J. Pharmacol. 2013, 45, 283–285. [13] Tsurng, T. J., Tsai, Y. C., Chiang, S. Y., Wang, G. J., Kuo, Y. C., Chang, Y. C., Wu, Y. Y., Wu, Y. C., Virus Res. 2014, 192, 16–24. [14] Li, Y. W., Qi, J., Zhang, W., Zhou, S. P., Wu, Y., Yu, B. Y., J. Sep. Sci. 2014, 37, 1762–1772. [15] Hu, Z., Jiang, C., Qin, M., Qi, J., Yu, B., Zhongguo Zhong Yao Za Zhi 2010, 35, 2508–2510. [16] Yu, B. Y., Hirai, Y., Shoji, J., Xu, G. J., Chem. Pharm. Bull. 1990, 38, 1931–1935. [17] Wang, K. W., Zhang, H., Shen, L. Q., Wang, W., Carbohydr. Res. 2011, 346, 253–258. [18] Do, J. C., Jung, K. Y., Song, Y. K., J. Nat. Prod. 1995, 58, 778–781. [19] Do, J. C., Sung, Y. K., Son, K. H., Kor. J. Pharmacogn. 1991, 22, 73–77. [20] Lee, D. Y., Son, K. H., Do, J. C., Kang, S. S., Arch. Pharm. Res. 1989, 12, 295–299. [21] Shin, J. S., Kor. J. Crop. Sci. 2002, 47, 236–239. [22] Kim, S. H., Kim, H. K., Yang, E. S., Lee, K. Y., Kim, S. D., Kim, Y. C., Sung, S. H., Sep. Purif. Technol. 2010, 71, 168–172. [23] Ha, H. G., Jung, D. Y., Kang, S. S., Kim, J. S., Ho, G., Repub. Korean Kongkae Taeho Kongbo 2003, KR2003070282. [24] Hur, J., Lee, P., Moon, E., Kang, I., Kim, S. H., Oh, M. S., Kim, S. Y., Eur. J. Pharmacol. 2009, 620, 9–15. [25] Kwon, G., Lee, H. E., Lee, D. H., Woo, H., Park, S. J., Gao, Q., Ahn, Y. J., Son, K. H., Ryu, J. H., Neurosci. Lett. 2014, 572, 58–62. [26] Park, S. H., Lee, H. J., Ryu, J., Son, K. H., Kwon, S. Y., Lee, S. K., Kim, Y. S., Hong, J. H., Seok, J. H., Lee, C. J., Phytomedicine 2014, 21, 172–176. [27] Ito, Y., J. Chromatogr. A 2005, 1065, 145–168. [28] Guzlek, H., Wood, P. L., Janaway, L., J. Chromatogr. A 2009, 1216, 4181–4186. [29] Ito, Y., J. Chromatogr. A. 2005, 1065, 145–188. [30] Chen, I. S., Shen, C. S. J., Sheppard, A. J., J. Am. Oil. Chem. Soc. 1981, 58, 599–601. [31] Zhang, T., Kang, L. P., Yu, H. S., Liu, Y. X., Zhao, Y., Xiong, C. Q., Zhang, J., Zou, P., Song, X. B., Liu, C., Ma, B. P., Steroids 2012, 77, 1298–1305. [32] Guo, Y., Liu, Y. X., Kang, L. P., Zhang, T., Yu, H. S., Zhao, Y., Xiong, C. Q., Ma, B. P., J. Asian Nat. Prod. Res. 2013, 15, 459–465. [33] Zou, K., Wang, J., Du, M., Li, Q., Tu, G., Chem. Pharm. Bull. 2006, 54, 1440–1442. [34] Dini, I., Tenore, G. C., Trimarco, E., Dini, A., Food Chem. 2005, 93, 205–214. [35] Agrawal, P. K., Steroids, 2005, 70, 715–724.
www.jss-journal.com
