Preparative separation of alkaloids from Litsea cubeba using combined applications of pH-zone-refining and high-speed counter-current chromatography

Abstract

Litsea cubeba is characterized by the presence of aporphine alkaloids. But few recent reports about the preparative separation of alkaloids from L. cubeba are found. The traditional separation method is time consuming and solvent consuming and irreversible adsorption is inevitable. In this research, pH-zone-refining counter-current chromatography and high-speed counter-current chromatography are applied to separate the alkaloids from a chloroform extract of L. cubeba. The crude extract was fractionated using the solvent system: chloroform–methanol–water (4 : 3 : 3, v/v) with different concentrations of hydrochloric acid (retainer) in the aqueous stationary phase and triethylamine (eluter) in the organic mobile phase to determine the ideal conditions for screening for the aporphine alkaloids. Using 1.5 g of the chloroform extract, 68.1 mg of norisocorydine (93.5% purity), 215.5 mg of isoboldine (96.3% purity), 612.3 mg of the mixture of boldine and laurotetanine, 108.8 mg of reticuline (97.4% purity) and 92.6 mg of laurolitsine (97.6% purity) were obtained with the selected conditions where 60 mM of hydrochloric acid was added to the stationary phase and 10 mM of triethylamine was used in the mobile phase. The mixture of boldine and laurotetanine was further separated using high-speed counter-current chromatography with a two-phase solvent system composed of ethyl acetate–methanol–water (4 : 1 : 5, v/v). Two alkaloids, laurotetanine (285.7 mg) and boldine (112.3 mg), were obtained from 500 mg of the mixture, in a one-step separation, with the relative purity of 94.8% and 96.2%, respectively. The purities of the isolated alkaloids were determined using high performance liquid chromatography and the chemical structures were confirmed using electrospray ionization-mass spectrometry, proton nuclear magnetic resonance (¹H-NMR) and carbon-13 (¹³C)-NMR.

---

1. Introduction

High-speed counter-current chromatography (HSCCC) is a continuous liquid–liquid partition chromatography without the solid support matrix. The liquid stationary phase is retained in the separation columns by gravity and the centrifugal force field avoids the disadvantages arising from the interaction between the sample and the solid support, such as irreversible absorption and denaturation of the target components. High recovery, large sample injection, high efficiency and the ease of scaling-up are the unique features of HSCCC. It is an excellent chromatographic technique, and it has been widely used in the separation and purification of natural products, antibiotics and rare elements with various solvent systems.¹

Another technique, pH-zone-refining counter-current chromatography (pH-zone-refining CCC), which is derived from the conventional HSCCC, enables the separation of organic acids and alkaloids into a succession of highly concentrated rectangular peaks that elute according to their pK_a values and hydrophobicities.^2–4 Besides having the advantages of HSCCC, pH-zone-refining CCC extends the capability of HSCCC in various ways including an over 10-fold increase in sample loading capacity, high concentration and high purity of fractions, and concentration of minor impurities. In addition, pH-zone-refining CCC enables the detection of components by their pH, which is especially useful if the analyte has low or no ultraviolet absorption.^3–5 Because of the advantages described previously, it has been successfully used as a large scale preparative technique for separating ionizable compounds which include alkaloids,^6–8 synthetic colors,^9,10 isomers,^11,12 peptide derivatives,^13,14 and so on.

Litsea cubeba (Lour) Pers. is mainly distributed in Anhui, Jiangsu, Zhejiang, Jiangxi and Fujian provinces of China and belongs to the Lauraceae family and is regarded as an important traditional Chinese medicine (TCM). This herbal medicine possesses the characteristics of dispelling wind and eliminating dampness and regulating qi to alleviate effects of pain. In TCM, wind is one of the six pathogenic factors, which include wind, cold, summer-heat, dampness, dryness and fire. Qi, blood and body fluids are the fundamental substances which make up the human body and are responsible for the vital activities of the body. Therefore, dispelling the pathogenic factors and regulating qi is of benefit to physical fitness. In the treatment of rheumatic arthralgia, traumatic injury, cold, abdominal pains, vomiting and diarrhoea, this herbal medicine plays an important role.¹⁵ The representative chemical components in L. cubeba are aporphine alkaloids. Pharmacological research has demonstrated that L. cubeba has anticancer, anti-inflammatory, anti-anaphylaxis, antioxidation and antimicrobial biological activities and it can also relieve asthma.¹⁵ But the pharmacodynamic substance of this herbal medicine is still not known. Therefore, the separation and identification of the aporphine alkaloids from L. cubeba will give the basis for determining the pharmacodynamic substance and establishing a scientific method of quality evaluation. It is necessary to develop a rapid and efficient method to separate these aporphine alkaloids. However, the chemical structures of these alkaloids are very similar which may lead to many difficulties for the separation and purification by pH-zone-refining CCC or HSCCC. In the present study, combining the applications of pH-zone-refining CCC and HSCCC was used to separate the alkaloids from the chloroform extract of L. cubeba and six aporphine alkaloids were successfully obtained using the optimized conditions.

2. Experimental

2.1. Materials and regents

The solvents used in this experiment, including petroleum ether (Pet), chloroform (CHCl₃), ethyl acetate (EtAc), hydrochloric acid (HCl), triethylamine (TEA), ammonia water (NH₃·H₂O), dimethylsulfoxide (DMSO), and tetramethylsilane (TMS), were of analytical grade (Nanjing Regent Factory, Nanjing, China). Methanol (MeOH) and acetonitrile (MeCN) used for high performance liquid chromatography (HPLC) analysis were of chromatographic grade (Fisher Scientific, Fair Lawn, NJ, USA). Double distilled water (Nihon Millipore Kogyo KK, Japan) was used for all aspects of the experimental work.

The material, L. cubeba (Lour) Pers., was collected in Fujian province. The species was identified by Dr Jia Li (Shandong University of Traditional Chinese Medicine, Jinan, China).

2.2. Apparatus

The pH-zone-refining CCC and HSCCC measurements in this study were made using a TBE-300B HSCCC apparatus (Tauto Biotech, Shanghai, China) with three preparative coils connected in series (diameter of 2.6 mm, total volume 300 mL) and a 20 mL sample loop. The revolution speed was regulated in the range of 0–1000 rpm with a speeder. The HSCCC system was equipped with an HX-1050 constant temperature circulating device (Beijing Boyikang Laboratory Instrument Co., Ltd, Beijing, China) to stabilize the separation temperature. A constant flow pump (Beijing Shengyitong Technology Co., Ltd., Beijing, China) was used to pump the stationary and mobile phase. Continuous monitoring of the effluent was achieved using a Waters 8823B-UV detector (Beijing BINTA Instrument Technology, Beijing, China) operating at 254 nm and a TP110 pH meter (Beijing Time Power Measure and Control Equipment Co., Ltd, Beijing, China). An R1200 recorder (Hangzhou Meacon Automation Technology Co., Ltd, Hangzhou, China) was used to record the pH-zone-refining CCC and HSCCC chromatograms.

The analysis of the CHCl₃ extract and the purified alkaloids was performed using a Waters Alliance system including a Waters 2998 Photodiode Array Detection (DAD) system, a Waters 2695 MultiSolvent delivery system, a Waters 2695 system controller, a Waters 2695 pump, and an Empower 3 Workstation (all from Waters, Milford, USA).

2.3. Extraction of crude sample

A portion of dried L. cubeba powder (5.0 kg) was extracted with 50 L of 95% aqueous ethanol under reflux for 2 h in a round bottomed flask. This procedure was repeated twice and the extracts were combined. After filtration, the extracts were concentrated on a Rotavapor at 55 °C. The residue was redissolved in 1000 mL of water (H₂O) containing 0.2% HCl in a beaker and partitioned three times, in a separator funnel, with equal volumes of Pet and CHCl₃ to remove the nonbasic substances. Then, the acidic solution was made alkaline with NH₃·H₂O to pH 10.0 and then partitioned three times with an equal volume of CHCl₃. The CHCl₃ layers were combined and evaporated under reduced pressure at 55 °C. Finally, 5.3 g of the CHCl₃ extract was obtained and used for further separation.

2.4. Preparation of the solvent system and sample solution

2.4.1. Preparation of solvent system and sample solution for the pH-zone-refining CCC. The two-phase solvent system of CHCl₃–MeOH–H₂O (4 : 3 : 3, v/v) was equilibrated in a separator funnel and separated before use. The upper aqueous phase (the stationary phase) was acidified with 60 mM HCl. The lower organic phase (the mobile phase) was made basic by adding TEA, resulting in a 10 mM solution. Both phases were degassed by ultrasonic mixing prior to use.

The sample solution was prepared by dissolving 1.5 g of the CHCl₃ extract in a mixture of 10 mL of stationary phase with 60 mM HCl and 10 mL of mobile phase without TEA.

2.4.2. Preparation of the solvent system and sample solution for the HSCCC. The two-phase solvent system EtAc–MeOH–H₂O (4 : 1 : 5, v/v) was equilibrated in a separator funnel. The upper organic phase was separated before use as the stationary phase and the lower aqueous phase was used as the mobile phase. The two phases were degassed by ultrasonic mixing before being pumped into the coils.

A mixture of boldine and laurotetanine (500 mg) was dissolved in a mixture of 5 mL of the aqueous phase and 5 mL of the organic phase.

2.5. Separation procedure

2.5.1. Separation procedure of pH-zone-refining CCC. The separation of the pH-zone-refining CCC was carried out by filling the whole column with the stationary phase at 15 mL min⁻¹ and then the sample solution containing 1.5 g of the CHCl₃ extract was injected. After setting and controlling the rotation speed at 850 rpm, the mobile phase was pumped into the column at 2.0 mL min⁻¹. The ultraviolet (UV) absorption of the effluent was continuously monitored at a wavelength of 254 nm using the UV monitor and each 10 mL fraction was collected in a test tube. At the same time, the pH value of each fraction was measured using a pH meter. When the separation was completed, the retention of the stationary phase as one of the important parameters could be measured by forcing the column contents with pressurized air, into a measuring vessel.

2.5.2. Separation procedure of the HSCCC. The separation procedure of the HSCCC was similar to that of the pH-zone-refining CCC. The major differences between them are as follows: (1) the sequence of sample injection and establishing the hydrodynamic equilibrium. In the separation of HSCCC, the sample dissolved in a mixture of stationary and mobile phases was injected when the hydrodynamic equilibrium was established; (2) the measurement of pH values is not necessary in the HSCCC separation. Besides these two points, there are no other differences between these two techniques in the separation procedure.

2.6. Analysis and identification

The purified alkaloids from the pH-zone-refining CCC and HSCCC as well as the CHCl₃ extract were analyzed using HPLC with an Agilent Zorbax SB-C₁₈ column (250 mm × 4.6 mm, id: 5 μm) at 310 nm and at a column temperature of 25 °C. The mobile phase was MeOH–0.2% NH₃·H₂O. The effluent was monitored using a DAD and the injection volume was 10 μL.

Electrospray ionization-mass spectrometry (ESI-MS) was performed with an Agilent 5973N mass selective detector (MSD) and a Pentium 4 computer with MSD Productivity ChemStation Software (Agilent Technologies Co. Ltd) was used for the identification of the purified compounds. The mass spectrometer scanned over the range 29–400 mat scan 1 s with an ionizing voltage of 70 eV and an ionization current of 150 mA. The NMR spectrum was recorded with a Varian 400 spectrometer (Varian, Palo Alto, CA, USA) with DMSO as the solvent and TMS as the internal standard.

3. Results and discussion

3.1. HPLC analysis of the CHCl₃ extract

The CHCl₃ extract of L. cubeba was analyzed using analytical HPLC and several mobile phases were tested, including MeOH–H₂O, MeOH–H₂O (0.2% TEA), MeCN–H₂O, and MeCN–H₂O (0.2% TEA), and so on. Finally, baseline separation of the target compounds was obtained when MeOH (A)–0.2% NH₃·H₂O (B) (0–20 min, 40%–75% A; 20–21 min, 75%–100% A; 21–30 min, 100%–100% A) was used as mobile phase and the flow rate was set at 1.0 mL min⁻¹. The HPLC chromatogram of the CHCl₃ extract is shown in Fig. 1. This chromatogram contained six major peaks (peak 1–6), which correspond to laurolitsine (1), boldine (2), laurotetanine (3), reticuline (4), isoboldine (5) and norisocorydine (6) in sequence, based on the peak area normalization method at the optimized detective wavelength of 310 nm. In the HPLC chromatogram of the CHCl₃ extract, peak 1 represented 8.5% of the total sample and peaks 2, 3, 4, 5 and 6 were 17.1%, 37.6%, 10.1%, 20.5% and 6.3% of the total sample, respectively.

Fig. 1 HPLC chromatogram for the CHCl₃ extract from L. cubeba. Experimental conditions: chromatography column: Agilent ZORBAX SB-C₁₈ column (250 mm × 4.6 mm, id: 5 μm); mobile phase: MeOH (A)–0.2% NH₃·H₂O (B) (0–20 min, 40%–75% A; 20–21 min, 75%–100% A; 21–30 min, 100%-100% A); column temperature: 25 °C; flow rate: 1.0 mL min⁻¹; UV detection wavelength: 310 nm; injection volume: 10 μL.

3.2. Optimization of the two-phase solvent system

3.2.1. Optimization of the two-phase solvent system for pH-zone-refining CCC. A suitable two-phase solvent system which could provide ideal dissociation (K_D) values in both acidic (K_acid ≪ 1) and basic (K_base ≫ 1) conditions as well as good solubility of the crude sample are required to obtain a successful separation using pH-zone-refining CCC.³ The K_D values could be determined using analytical HPLC and calculated using the following formula where C_s and C_m of each compound could be replaced by the corresponding peak areas in the stationary and mobile phase:

K_D = C_s/C_m

C_s: the mass concentration of solute in the stationary phase; C_m: the mass concentration of solute in the mobile phase.

From the experience obtained of separating alkaloids using the pH-zone-refining CCC, the solvent systems composed of Pet–EtAc–MeOH–H₂O with different volume ratios were investigated. However, none of them could provide a suitable K_D value. Considering the polarity of the aporphine alkaloids, the crude sample must have good solubility in CHCl₃, which was often used for the extraction of alkaloids. Therefore, the solvent system of CHCl₃–MeOH–H₂O was examined. All the alkaloids dissolved in the lower CHCl₃ phase using this biphasic solvent system at the common volume ratio of 4 : 3 : 3 with 10 mM HCl in the upper phase and 10 mM TEA in the lower phase. All the alkaloids were eluted in a short time but the target compounds were not separated completely. However, the result was still helpful because the crude sample was separated into several fractions as shown in Fig. 2(a). The chromatogram obtained with TLC indicated that the alkaloids were eluted in a short time and separated incompletely which meant that the retention capability of this solvent system was poor. In this context, the question was how to improve the retention capability. On the one hand, the volume ratio of the solvent system could be adjusted to an appropriate value. On the other hand, the concentration of the retainer acid in the upper stationary phase could be increased to improve the retention capability.¹⁶ The concentration of the retainer in the stationary phase mainly determined the concentration and retention time of the analyte. In order to improve the retention capability and achieve efficient resolution of the target compounds, a proper amount of the retainer acid should be added to the aqueous stationary phase.¹⁷ The resolution of the sample was radically improved by adding 60 mM HCl to the aqueous stationary phase and 10 mM TEA to the organic mobile phase as shown in Fig. 2(b).

Fig. 2 pH-zone-refining CCC and TLC chromatograms for the separation of the CHCl₃ extract from L. cubeba. Experimental conditions: solvent system: (a) CHCl₃–MeOH–H₂O (4 : 3 : 3, v/v); 10 mM HCl in the upper aqueous phase and 10 mM TEA in the lower organic phase; (b) CHCl₃–MeOH–H₂O (4 : 3 : 3, v/v); 60 mM HCl in the upper aqueous phase and 10 mM TEA in the lower organic phase; revolution speed: 850 rpm; flow rate: 2 mL min⁻¹; sample size: 1.5 g; UV detection wavelength: 254 nm; ^a crude sample.

In general, Pet–EtAc–MeOH–H₂O (5 : 5 : x : 10 − x, 1 ≤ x ≤ 9, v/v) and CHCl₃–MeOH–H₂O (4 : 3 : 3, v/v) are two common solvent systems used in the separation of natural products by pH-zone-refining CCC.^17–20 Pet–EtAc–MeOH–H₂O was eliminated by calculating the K_D values and then CHCl₃–MeOH–H₂O was examined by optimizing the concentration of the retainer. Finally, the two-phase solvent system, CHCl₃–MeOH–H₂O (4 : 3 : 3, v/v), with 60 mM HCl in the aqueous stationary phase and 10 mM TEA in the organic mobile phase, was selected.

3.2.2. Optimization of the two-phase solvent system for HSCCC. Successful separation by HSCCC largely depends upon use of a suitable two-phase solvent system which provides an ideal partition coefficient (0.5 < K < 2.5) for the target compound.^21,22 A small K value usually results in a poor peak resolution, whereas a large one tends to produce excessive sample band broadening.^23–25 After partitioning the sample between the two phases, the K values were determined using analytical HPLC. From the HPLC chromatograms, the K value of each component was determined by determining the ratio of the corresponding peak areas as mentioned previously. Several two-phase solvent systems were tested and the K values were measured and are summarized in Table 1:

K = C_s/C_m

C_s: the mass concentration of solute in the stationary phase; C_m: the mass concentration of solute in the mobile phase.

Table 1 The partition coefficient of compound (2) and compound (3) in several solvent systems

Pet–EtAc–MeOH–H2O	Partition coefficient (K)
	Compound (2)	Compound (3)
1 : 9 : 1 : 9	21.6	5.9
3 : 7 : 2 : 8	18.1	3.9
0 : 4 : 1 : 5	12.0	1.7

---

As can be seen from Table 1, the two-phase solvent systems of Pet–EtAc–MeOH–H₂O (1 : 9 : 1 : 9, v/v) and (3 : 7 : 2 : 8, v/v) were tested and the K values were larger than 2.0. Therefore, it took a long time for the target compounds to be eluted and this resulted in poor resolution. Then, the two-phase solvent system EtAc–MeOH–H₂O (4 : 1 : 5, v/v) was investigated, which could provide a suitable K value for laurotetanine (3) and the separation results are shown in Fig. 3.

Fig. 3 HSCCC chromatogram for the separation of the mixture of compound (2) and compound (3). Experimental conditions: solvent system: EtAc–MeOH–H₂O (4 : 1 : 5, v/v); the upper organic phase as the stationary phase; revolution speed: 850 rpm; flow rate: 2 mL min⁻¹; sample size: 500 mg; UV detection wavelength: 254 nm.

The selection of the two-phase solvent system for HSCCC entirely depends on the measurement of the K values of the solvent system: Pet–EtAc–MeOH–H₂O with different volume ratios. After testing, the two-phase solvent system composed of EtAc–MeOH–H₂O (4 : 1 : 5, v/v) was used for the separation of the mixture.

3.3. Preparative separation of aporphine alkaloids using pH-zone-refining CCC and HSCCC

Fig. 2(b) shows a typical chromatogram obtained with pH-zone-refining CCC for the separation of 1.5 g of the CHCl₃ extract using the two-phase solvent system CHCl₃–MeOH–H₂O (4 : 3 : 3, v/v) with 60 mM HCl in the aqueous stationary phase and 10 mM TEA in the organic mobile phase. The total separation time was about five hours. These aporphine alkaloids were eluted as five irregular rectangular peaks. The measurement of the pH value of each fraction also presented five pH zones, which indicated that the separation of the target compounds was successful. This separation yielded 68.1 mg of norisocorydine (6), 215.5 mg of isoboldine (5), 612.3 mg of the mixture of boldine (2) and laurotetanine (3), 108.8 mg of reticuline (4) and 92.6 mg of laurolitsine (1) from 1.5 g of CHCl₃ extract.

Fig. 3 shows the separation of the mixture of boldine (2) and laurotetanine (3) by HSCCC using the two-phase solvent system EtAc–MeOH–H₂O (4 : 1 : 5, v/v). While boldine (2) was retained in the stationary phase, laurotetanine (3) was eluted in 2.5 hours because of its appropriate K value (1.7). The separation yielded 285.7 mg of laurotetanine (3) and 112.3 mg of boldine (2) from 500 mg of the mixture.

The purified alkaloids were determined by HPLC with the selected analytical condition of the CHCl₃ extract at the optimized detective wavelength of 310 nm. Fig. 4 displays the HPLC chromatograms of the pure fractions separated by pH-zone-refining CCC and HSCCC, which illustrates that these aporphine alkaloids were obtained with the purities of all being over 93%.

Fig. 4 HPLC chromatograms and chemical structures of the purified aporphine alkaloids from L. cubeba. Experimental conditions: Agilent Zorbax SB-C₁₈ column (250 mm × 4.6 mm, id: 5 μm); mobile phase: MeOH (A)–0.2% NH₃·H₂O (B) (0–20 min, 40%–75% A; 20–21 min, 75%–100% A; 21–30 min, 100%–100% A); column temperature: 25 °C; flow rate: 1.0 mL min⁻¹; UV detection wavelength: 310 nm; injection volume: 10 μL.

3.4. Structural identification

Identification of the alkaloids purified by pH-zone-refining CCC and HSCCC were carried out using UV, ESI-MS, proton nuclear magnetic resonance (¹H-NMR) and carbon-13 (¹³C)-NMR. The identification of each separated alkaloid was as follows:

Compound (1) (peak 1 in Fig. 1). Brown powder in CHCl₃; UV (λ_max, MeOH): 217.2, 275.1 and 308.3 nm; positive ESI-MS, m/z 314 [M + H]⁺; ¹H-NMR (deuterated DMSO (DMSO-d₆), 400 MHz) δ ppm: 7.92 (1H, s, H-11), 6.70 (1H, s, H-8), 6.53 (1H, s, H-3), 3.80 (3H, s, 10-OCH₃), 3.59 (3H, s, 1-OCH₃), 3.55 (1H, m, H-6a), 3.15 (1H, d, J = 4.0 Hz, H_e-5), 2.80 (1H, d, J = 4.0 Hz, H_a-5), 2.76 (1H, m, H_e-4), 2.62 (1H, d, J = 12.0 Hz, H_e-7), 2.49 (2H, m, H_a-4 and H_a-7); for ¹³C-NMR (DMSO-d₆) data, see Table 2. The ¹H- and ¹³C-NMR data were consistent with those previously reported^26,27 and compound (1) was identified as laurolitsine.

Table 2 ¹³C-NMR spectroscopic data (400 MHz) for laurolitsine (1), boldine (2), laurotatanine (3), isoboldine (5), norisocorydine (6)

Position	δC^a
	(1)	(2)	(3)	(5)	(6)
a The experiments were performed in DMSO and the NMR chemical shift are given in ppm related to TMS at 0.0 ppm as the internal reference.
1	143.1	143.5	144.2	141.4	143.2
1a	127.1	126.9	126.8	120.4	125.1
1b	126.3	126.9	126.5	123.4	129.2
2	149.6	150.4	152.4	147.3	152.0
3	115.1	114.6	111.6	109.9	112.0
3a	129.7	128.0	129.0	126.7	129.2
4	28.9	27.2	28.2	28.7	27.5
5	43.1	52.1	42.2	53.4	41.7
6a	53.9	62.0	53.3	62.9	53.7
7	36.5	32.5	35.4	33.9	36.4
7a	130.1	128.7	129.5	129.5	127.8
8	115.6	115.8	115.6	115.6	119.0
9	146.5	146.6	146.5	145.9	112.4
10	146.3	146.8	146.6	146.0	149.1
11	112.6	112.5	112.9	114.2	144.4
11a	123.5	122.9	122.9	124.0	120.0
N-CH3		42.3		43.9
1-OCH3	59.7	59.8	59.9		61.3
2-OCH3			56.2	56.4	56.3
10-OCH3	56.2	56.2	56.2	56.4	56.3

---

Compound (2) (peak 2 in Fig. 1). Brown powder in CHCl₃; UV (λ_max, MeOH): 217.2 and 316.6 nm; positive ESI-MS, m/z 328 [M + H]⁺; ¹H-NMR (DMSO-d₆, 400 MHz) δ ppm: 7.87 (1H, s, H-11), 6.78 (1H, s, H-8), 6.59 (1H, s, H-3), 3.79 (3H, s, 10-OCH₃), 3.59 (3H, s, 1-OCH₃), 3.24 (1H, m, H_e-7), 3.08 (1H, m, H_e-5), 3.05 (1H, m, H_e-4), 2.81 (1H, m, H-6a), 2.67 (1H, m, H_a-5), 2.67 (3H, s, N-CH₃), 2.54 (2H, m, H_a-4 and H_a-7). Compound (2) was identified as boldine supported by the ¹H- and ¹³C-NMR (DMSO-d₆) data (see Table 2) which were in good agreement with those found in the literature.^26,28

Compound (3) (peak 3 in Fig. 1). Brown powder in CHCl₃; UV (λ_max, MeOH): 212.5 and 315.4 nm; positive ESI-MS, m/z 328 [M + H]⁺; ¹H-NMR (DMSO-d₆, 400 MHz) δ ppm: 7.89 (1H, s, H-11), 6.71 (1H, s, H-8), 6.71 (1H, s, H-3), 3.80 (3H, s, 10-OCH₃), 3.78 (3H, s, 2-OCH₃), 3.74 (1H, m, H-6a), 3.61 (3H, s, 1-OCH₃), 3.28 (1H, d, J = 4.0 Hz, H_e-5), 2.92 (2H, m, H_a-5 and H_e-4), 2.68 (2H, m, H_a-4 and H_e-7), 2.54 (1H, m, H_a-7); for ¹³C-NMR (DMSO-d₆) data, see Table 2. Compared with the ¹H- and ¹³C-NMR data given in the literature,^27–30 compound (3) was identified as laurotetanine.

Compound (4) (peak 4 in Fig. 1). Yellow powder in CHCl₃; UV (λ_max, MeOH): 212.5 and 283.3 nm; positive ESI-MS, m/z 330 [M + H]⁺; ¹H-NMR (DMSO-d₆, 400 MHz) δ ppm: 6.74 (1H, d, J = 8.0 Hz, H-5′), 6.63 (1H, s, H-5), 6.56 (1H, s, H-2′), 6.52 (1H, d, J = 8.0 Hz, H-6′), 6.42 (1H, s, H-8), 3.71 (3H, s, 6-OCH₃), 3.71 (3H, s, 4′-OCH₃), 3.50 (1H, m, H-1), 3.04 (1H, m, H_e-3), 2.82 (1H, m, H_e-9), 2.67 (2H, m, H_e-4 and H_a-9), 2.50 (2H, m, H_a-3 and H_a-4), 2.30 (3H, s, N-CH₃); ¹³C-NMR (DMSO-d₆, 400 MHz) δ ppm: 64.5 (C-1), 47.3 (C-3), 25.1 (C-4), 124.9 (C-4a), 112.3 (C-5), 146.3 (C-6), 144.7 (C-7), 114.8 (C-8), 130.5 (C-8a), 42.9 (C-9), 133.3 (C-1′), 117.3 (C-2′), 146.1 (C-3′), 146.3 (C-4′), 112.3 (C-5′), 120.4 (C-6′), 56.1 (6-OCH₃), 55.9 (4′-OCH₃), 42.9 (N-CH₃). ¹H- and ¹³C-NMR data were in good agreement with those previously reported^28,29 and compound (4) was identified as reticuline.

Compound (5) (peak 5 in Fig. 1). Colorless crystals in CHCl₃; UV (λ_max, MeOH): 212.5, 279.8 and 304.7 nm; positive ESI-MS, m/z 328 [M + H]⁺; ¹H-NMR (DMSO-d₆, 400 MHz) δ ppm: 7.99 (1H, s, H-11), 6.72 (1H, s, H-8), 6.60 (1H, s, H-3), 3.79 (3H, s, 2-OCH₃), 3.75 (3H, s, 10-OCH₃), 2.94 (4H, m, H_e-7, H_e-4, H_e-5 and H-6a), 2.57 (1H, m, H_a-4), 2.45 (3H, s, N-CH₃), 2.30 (2H, m, H_a-5 and H_a-7). Compound (5) was identified as isoboldine supported by the ¹H- and ¹³C-NMR (DMSO-d₆) data (see Table 2) which were consistent with those found in the literature.²⁸

Compound (6) (peak 6 in Fig. 1). Brown powder in CHCl₃; UV (λ_max, MeOH): 217.2, 266.8 and 302.4 nm; positive ESI-MS, m/z 328 [M + H]⁺; ¹H-NMR (DMSO-d₆, 400 MHz) δ ppm: 6.92 (1H, d, J = 4.0 Hz, H-9), 6.91 (1H, d, J = 4.0 Hz, H-8), 6.81 (1H, s, H-3), 3.85 (3H, s, 10-OCH₃), 3.78 (3H, s, 2-OCH₃), 3.69 (1H, d, J = 12.0 Hz, H-6a), 3.62 (3H, s, 1-OCH₃), 3.33 (1H, d, J = 8.0 Hz, H_e-5), 2.97 (2H, m, H_e-4 and H_a-5), 2.79 (2H, m, H_a-4 and H_e-7), 2.51 (1H, m, H_a-7); for ¹³C-NMR (DMSO-d₆) data, see Table 2. Compared with the data given in the literature,^29–31 compound (6) was identified as norisocorydine.

4. Conclusions

The chemical structures of alkaloids in L. cubeba are very similar and there is no significant difference in both polarity and alkalinity, which leads to many difficulties in separating these alkaloids using pH-zone-refining CCC or HSCCC. The efficient separation and purification of six aporphine alkaloids from the CHCl₃ extract of L. cubeba was developed using the combined application of pH-zone-refining CCC and HSCCC. From 1.5 g of the crude sample, four purified alkaloids including norisocorydine (68.1 mg), isoboldine (215.5 mg), reticuline (108.8 mg), and laurolitsine (92.6 mg) and a mixture were obtained after separation using pH-zone-refining CCC. The mixture was further separated using HSCCC and another two alkaloids, laurotetanine (285.7 mg) and boldine (112.3 mg), were obtained. The results of the study demonstrated that the combined application of pH-zone-refining CCC and HSCCC is a rapid and efficient method for the separation and purification of natural products.

Acknowledgements

Financial support from the National Natural Science Foundation of China (20872083) and the Key Science and Technology Program of Shandong Province (2014GZX219003) are gratefully acknowledged.

References

1. Y. Ito and R. L. Bowman, Science, 1970, 167, 281–283.
2. Y. Ito and Y. Ma, J. Chromatogr. A, 1996, 753, 1–36.
3. Y. Ito, J. Chromatogr. A, 2005, 1065, 145–168.
4. Y. Ito, J. Chromatogr. A, 2013, 1271, 71–85.
5. Y. J. Pan and Y. B. Lu, J. Liq. Chromatogr. Relat. Technol., 2007, 30, 649–679.
6. R. L. Hu, X. J. Dai, Y. B. Lu and Y. J. Pan, J. Chromatogr. B: Anal. Technol. Biomed. Life Sci., 2010, 878, 1881–1884.
7. Q. Yu, S. Q. Tong, J. Z. Yan, C. Q. Hong, W. F. Zhai and Y. Q. Li, J. Sep. Sci., 2011, 34, 278–285.
8. X. Wang, J. H. Liu, Y. L. Geng, D. J. Wang, H. J. Dong and T. Y. Zhang, J. Sep. Sci., 2010, 33, 539–544.
9. A. Weisz, E. P. Mazzola and Y. Ito, J. Chromatogr. A, 2011, 1218, 8249–8254.
10. A. Weisz, E. P. Mazzola and Y. Ito, J. Chromatogr. A, 2009, 1216, 4161–4168.
11. S. Q. Tong, J. Z. Yan and Y. X. Guan, J. Chromatogr. A, 2008, 1212, 48–53.
12. A. Weisz, A. Idina, J. Ben-Ari, M. Karni, A. Mandelbaum and Y. Ito, J. Chromatogr. A, 2007, 1151, 82–90.
13. Y. Ma and Y. Ito, J. Chromatogr. A, 1995, 702, 197–206.
14. Y. Ma and Y. Ito, J. Chromatogr. A, 1997, 771, 81–88.
15. B. Lin, H. Zhang and X. X. Zhao, J. Ethnopharmacol., 2013, 147, 327–334.
16. L. Fang, Y. Q. Liu, B. Yang, X. Wang and L. Q. Huang, J. Sep. Sci., 2011, 34, 2545–2558.
17. C. L. Sun, J. Li, X. Wang, W. J. Duan, T. Y. Zhang and Y. Ito, J. Chromatogr. A, 2014, 1370, 156–161.
18. X. Wang, H. J. Dong, X. K. Shu, Z. J. Zheng, B. Yang and L. Q. Huang, Molecules, 2012, 17, 14968–14974.
19. X. K. Shu, J. Li, F. Liu, X. J. Lin, X. Wang and C. X. Song, J. Sep. Sci., 2013, 36, 2680–2685.
20. X. Y. Wang, X. K. Shu, X. Wang, J. Q. Yu and F. Jing, Molecules, 2014, 19, 12619–12629.
21. H. T. Lu, Y. Jiang and F. Chen, J. Chromatogr. A, 2003, 1017, 117–123.
22. High-Speed Countercurrent Chromatography, Chemical Analysis, ed. Y. Ito and W. D. Conway, Wiley/Interscience, New York, 1996.
23. X. Wang, H. J. Dong, Y. Q. Liu, B. Yang, X. Wang and L. Q. Huang, J. Chromatogr. B: Anal. Technol. Biomed. Life Sci., 2011, 879, 811–814.
24. H. T. Lu, Y. Jiang and F. Chen, J. Chromatogr. A, 2004, 1023, 159–163.
25. Y. Ito, Counter-current chromatography, in Countercurrent Chromatography: Theory and Practice, ed. N. B. Mandava and Y. Ito, Marcel Dekker, New York, 1988.
26. X. H. Yan, X. Y. Wei, H. H. Xie, M. F. Liu and F. X. Zhang, J. Trop. Subtrop. Bot., 2000, 8, 324–328.
27. E. V. Costa, L. M. Dutra, A. Nepel and A. Barison, Biochem. Syst. Ecol., 2013, 51, 331–334.
28. S. Y. Zhang, Q. Guo, Y. Cao, Y. Zhang, X. L. Gao, P. F. Tu and X. Y. Chai, China J. Chin. Mater. Med., 2014, 39, 3964–3968.
29. R. Wang, S. G. Tang, H. Y. Zhai and H. Q. Duan, China J. Chin. Mater. Med., 2011, 36, 1032–1036.
30. Y. H. Qu, M. Jiang, L. Y. Wang, Y. C. Li and Y. N. Wang, China J. Chin. Mater. Med., 2015, 40, 94–97.
31. M. L. R. Ferreira, I. C. de Pascoli, I. R. Nascimento, J. Zukerman-Schpector and L. M. X. Lopes, Phytochemistry, 2010, 71, 469–478.

---