Effective Ingredients Of Cistanche: Separation And Purification Method Of Phenylethanoid Glycosides

Contact: joanna.jia@wecistanche.com

Isolation And Purification Of Phenylethanoid Glycosides From Cistanche Deserticola By High-speed Counter-current Chromatography

Li a,b, Rong Tsao b,*, Raymond Yang b, Chunming Liu a,

J. Christopher Young b, Honghui Zhu b

a Department of Chemistry, Changchun Normal University, Changchun 130032, China

b Guelph Food Research Centre, Agriculture and Agri-Food Canada, 93 Stone Road West, Guelph, Ontario, Canada N1G 5C9

Received 31 May 2007; received in revised form 16 October 2007; accepted 29 October 2007

Abstract:

Five phenylethanoid glycosides (PhGs), echinacoside, cistanoside A, acteoside, isoacteoside, and 20-acetylacteoside, were isolated and purified from Cistanche deserticola for the first time by high-speed counter-current chromatography (HSCCC) using two biphasic systems, one consisting of ethyl acetate–ethanol-water (5:0.5:4.5, v/v/v) and another of ethyl acetate–n-butanol–ethanol-water (0.5:0.5:0.1:1, v/v/v/v). A total of 28.5 mg of echinacoside, 18.4 mg of cistanoside A, 14.6 mg of acteoside, 30.1 mg of isoacteoside and 25.2 mg of 20-acetylacteoside were purified from 1412 mg of the n-butanol extract of Cistanche deserticola, each at over 92.5% purity as determined by HPLC. The structures were identified by their retention time, UV, LC–ESI-MS in the negative ion mode, and confirmed by NMR experiments. The characteristic LC–ESI-MSn fragmentation pattern of the five compounds is discussed, and found to be a very specific and useful tool for the structural identification of PhGs from this important medicinal plant.

Crown Copyright © 2007 Published by Elsevier Ltd. All rights reserved.

Keywords: Cistanche deserticola; Phenylethanoid; Echinacoside; Cistanoside A; Acteoside; Isoacteoside; 20-Acetylacteoside; HSCCC; LC–ESI-MS; NMR

1. Introduction

Cistanche deserticola Y.C. Ma is a well-known Chinese herbal medicine and has been widely used in traditional preparations similar to today’s functional food ingredients or food supplements (Wong, Li, Cheng, & Chen, 2006). The major bioactive constituents in C. deserticola are phenylethanoid glycosides (PhGs), including echinacoside, acteoside, isoacteoside, 20-acetylacteoside, and cistanoside A (Kobayashi, Karasawa, & Miyase, 1984; Kobayashi et al., 1987). PhGs are widely distributed in the plant kingdom and have been extensively studied for their various biological functions such as hepatoprotective (Xiong et al., 1998), anti-inflammatory, antinociceptive activity (Schapoval et al., 1998), and antioxidant activities (Cheng, Wei, Guo, Ni, & Liu, 2005; He, Lau, Xu, Li, & But, 2000; Li, Wang, & Wang, 1997; Li, Wang, Zheng, Liu, & Jia, 1993; Wang, Jiang, Wu, & Wang, 2001; Xiong, Kadota, Tani, & Namba, 1996), improving sexual function (Xie & Wu, 1993; Zong, He, Wu, & Chen, 1996), and sedative effect (Lu, 1998).

Owing to the above significant bioactivities, large quantities of pure compounds are urgently needed as reference standards and for various in vitro and in vivo studies related to the use of traditional Chinese medicines. Effective methods for the isolation, purification, and structural characterization of PhGs, therefore become necessary. However, such work usually requires the use of multiple chromato- graphic steps for sample clean up and isolation (Gross, Lahloub, Anklin, Schulten, & Sticher, 1988; Nishimura, Sasaki, Inagaki, Chin, & Mitsuhashi, 1991; Ravn, Nishibe, Sasahara, & Li, 1990; Shoyama, Matsumoto, & Nishioka, 1987), which usually results in low recovery rates due to irreversible adsorptions of PhGs onto the solid support during separation (Lei et al., 2001). In contrast, high-speed counter-current chromatography (HSCCC) has become an effective alternative to the conventional chromatographic techniques for the separation of some PhGs from plant extracts (Lei et al., 2001; Li et al., 2005). Lei et al. successfully separated acteoside and 20-acetylacteoside from Cistanches salsa (C.A. Mey.) G. Beck by using HSCCC (Lei et al., 2001). The authors of this paper have previously reported the separation of acteoside and isoacteoside from Plantago psyllium L. by HSCCC (Li et al., 2005). However, no report has been published on the separation and purifi- cation of multiple PhGs from Cistanche deserticola using HSCCC. Due to the lack of standards, LC-MS methods have been developed and used as a powerful analytical tool for rapid characterization and identification of some PhGs in plant extract (Li et al., 2005; Wang et al., 2000).

In this paper, we report a HSCCC method developed for the preparation of echinacoside, cistanoside A, acteoside, isoacteoside and 20-acetylacteoside from Cistanche deserticola. Characterization and analysis of the five PhGs separated were accomplished by the use of LC coupled with in-line ESI mass spectrometry and NMR experiments. The retention time, molecular weight, and the characteristic fragment ions of the five PhGs are presented and discussed in this paper. The structures of the five PhGs identified in this investigation are shown in Fig. 1.

echinacoside in cistanche

2. Experimental

2.1. Chemicals and reagents

Acteoside was purchased from Sigma–Aldrich (Oak- Ville, ON), echinacoside was purchased from ChromaDex (Santa Ana, CA). Isoacteoside was isolated from P. psyllium L. (Li et al., 2005). Cistanche deserticola was purchased from Beijing TongRenTang Medicinal Store (China). All solvents were of HPLC grade and purchased from Caledon Laboratories Ltd. (Georgetown, ON).

2.2. Sample preparation

Cistanche deserticola (20 g) was extracted five times at room temperature for 12 h each with 100 mL of 80% aqueous ethanol. Each time the extraction mixture was filtered through a Whatman No.1 filter paper (Whatman International Ltd., Maidstone, England). The combined filtrate was concentrated to 100 mL in vacuo at < 40 °C. The resulting aqueous solution was defatted twice, each with 100 mL of hexane, and then extracted successively for five times, each with 100 mL n-butanol. The n-butanol layers were combined and concentrated to dryness in vacuo at < 40 °C, which yielded 2.2 g of n-butanol extract. The extract was stored at -20 °C before HSCCC separation.

2.3. HSCCC separation procedure

The preparative HSCCC was carried out in a Model CCC-1000 high-speed counter-current chromatography (Pharma-Tech Research, Baltimore, Maryland USA). This apparatus had three preparative coils, connected in series (total volume, 325 mL). The revolution speed of the apparatus could be regulated between 0 and 2000 rpm. The HSCCC system was equipped with an HPLC pump (Pharma-Tech Research, Baltimore, Maryland, USA), a Model 450 UV detector (Alltech, USA), a Model L 120 E flat-bed recorder (Linseis Inc., Princeton Jct, USA), a fraction collector (Advantec MFS Inc., USA) and a sample injection valve with a 10-mL sample loop.

A mixture of ethyl acetate–ethanol-water (5:0.5:4.5, v/v/v) was shaken vigorously in a separatory funnel and let stand and separate at room temperature. The two phases were used in the HSCCC after they reached equilibrium. The entire coiled column was first filled with the upper layer, which serves as the stationary phase. The lower layer (mobile phase) was then pumped into the head-end of the column at a flow-rate of 1.5 mL/min. The rotation speed was set at 1050 rpm. A sample (ca. 230 mg each time) dis- solved in 8 ml of the mixture of ethyl acetate–ethanol-water (5:0.5:4.5, v/v/v) was loaded into the injection valve after the system reached hydrodynamic equilibrium. This biphasic solvent system was selected based on the partition coefficient (K), which was 0.87, 1.11, and 1.32 for acteoside, isoacteoside, and 20 acetylacteoside, respectively. The K value was the ratio of the concentrations in the top and bottom layers of the same compound as determined by HPLC (Fig. 2). The effluent from the outlet of the column was continuously monitored by a UV detector at 254 nm and collected into test tubes with a fraction collector set at 4 min for each tube.

2.4. LC conditions

static auto-sampler and a photodiode array detector (DAD) were used for the analysis of PhGs in the n-butanol extract of Cistanche deserticola and fractions collected from the HSCCC separation. The separation was carried out in a Phenomenex ODS-C18 column (250 × 4.6 mm, 5 lm) with a C18 guard column. The binary mobile phase consisted of acetonitrile (solvent A) and water containing 2% acetic acid (solvent B). All solvents were filtered through a

0.45 lm filter prior to use. The flow-rate was kept constant at 1.0 mL/min for a total run time of 25 min. The system was run with a gradient program: 0–20 min: 90% B to 60% B; 20–22 min: 60% B to 0% B; and 22– 25 min, 0% B to 90% B. The sample injection volume was 10 lL. Peaks of interest were monitored at 320 nm by a DAD detector.

2.5. LC–ESI-MS experiments

LC-MS experiments were carried out using a Finnigan LCQ DECA ion trap mass spectrometer (Thermo Finnigan, San Jose, CA, USA) equipped with an electrospray ionization (ESI) source. The samples were analyzed in the same chromatographic condition. A negative model was used for data collection. The sheath gas and auxiliary flow-rates were set at 96 and 7 (arbitrary unit), respectively. The capillary voltage was set at 29 V and its temperature was controlled at 350 °C. The entrance lens voltage was fixed at 40 V and the multipole RF amplitude was set at 540 V. The ESI needle voltage was controlled at 4.5 kV. The tube lens offset was 16 V, the multipole lens 1 offset was 8.20 V and the multipole lens 2 offset was 10.5 V. The electron multiplier voltage was set at 980 V for ion detection.

2.6. NMR for Identification

NMR spectra were recorded on a Bruker Avance-600 spectrometer (Bruker BioSpin Ltd., Milton, Canada). Only compounds 2 and 5 (no standards available) were subjected to NMR experiments. Samples were dissolved in CD3OD.

3. Results and discussion

3.1. HSCCC separation

The n-butanol extract of Cistanche deserticola and the fractions corresponding to each peak isolated by HSCCC were analyzed by HPLC, and the results are given in Fig. 2. Five major compounds (peaks 1–5) were separated and detected with retention times at 9.2 min, 10.7 min, 12.3 min,

13.3 min and 16.2 min, respectively.

A successful HSCCC separation depends largely on a suitable two-phase solvent system that provides an ideal partition coefficient (K) around 1 for the desired compound. Such a biphasic system should also yield a reasonably short settling time (Chen, Games, & Jones, 2003; Foucault & Chevolot, 1998; Oka, Oka, & Ito, 1991). In our experiment, we selected four series of solvent systems according to the solubility of the target compounds in Cistanche deserticola. HPLC was used to measure the concentration in each phase, from which the K values of the target compounds were calculated. Two systems, ethyl acetate– n-butanol–ethanol-water (4:0.6:0.6:5, v/v/v/v) and ethyl acetate–water (1:1, v/v), have been previously used in HSCCC to separate acteoside and 20-acetylacteoside fromC. salsa, acteoside, and isoacteoside from P. Psyllium, respectively (Lei et al., 2001; Li et al., 2005). Although the first system had a relatively short settling time, it had poor performance in separating the PhGs of Cistanche deserticola, due to the low K values for compounds 1 and 2, and high K values for compounds 3–5. The K values were very low for compounds 1–4 in the second system, but very high for compound 5 (Table 1). A modified system containing ethyl acetate–ethanol-water (5:0.5:4.5, v/v/v), gave an ideal K value for compounds 3–5 at 0.87, 1.11, and 1.32, respectively, and resulted in the good separation of these three compounds (Fig. 3A and B). This system, however, produced

too small a K value for compounds 1 and 2, causing the two compounds to co-elute near the solvent front (fraction 1 in Fig. 3A). A further modification to the system (ethyl acetate–n-butanol–ethanol-water (0.5:0.5:0.1:1, v/v/v/v) elevated the K values for compound 1 and 2 to 0.52 and 0.92, respectively, and led to complete separation (Fig. 3C). Fig. 3A shows the HSCCC separation of a sample containing 230 mg of the n-butanol extract of Cistanche deserticola using ethyl acetate–ethanol-water (5:0.5:4.5, v/v/v). Fractions that were confirmed by HPLC to contain only compounds 3, 4, or 5 were combined separately, and those containing compounds 3 and 4 were pooled, freeze-dried, and re-subjected to the HSCCC for further separation (Fig. 3B). The two-step HSCCC separation described above yielded a total of 14.6 mg, 30.1 mg, and 25.2 mg of compounds 3–5 from 1412 mg n-butanol extract. Fig. 3C shows the HSCCC separation of a sample containing compounds 1 and 2 (fraction 1 in the first separation) using ethyl acetate–n-butanol–ethanol-water (0.5:0.5:0.1:1, v/v/ v/v). A total of 28.5 and 18.4 mg of compounds 1 and 2 were obtained. The chromatographic purities of the freeze-dried compounds 1–5 were over 92.5%, which were directly used for LC–ESI-MS and NMR analyses.

3.2. Structural identification by LC–ESI-MS and NMR

Tentative identification of compounds 1, 3, and 4 was achieved by congruent retention times and UV spectral data with those of the authentic echinacoside, acteoside, and isoacteoside (Fig. 2). Compounds 2 and 5 were unknown, however, the UV spectra of all five compounds were highly similar, indicating similar structural features.

To further investigate the structures of these five compounds, LC–ESI-MSn experiments were attempted and the results are shown in Fig. 4 and Table 2. Compounds related to the peaks (1–5) in Fig. 2 exhibited intense deprotonated molecular ions [M H]— at m/z 785, 799, 623, 623, and 665, respectively, in the negative mode. Dimeric [2M H]— ions were also observed for peaks 1–4 in Fig. 2. These confirmed the molecular weights of peaks 1–5 to be 786, 800, 624, 624, and 666, respectively. The LC–MSN data (Table 2) provided highly useful structural information for the five PhGs, such as the neutral loss of

an Experimental procedure: approximately 1 mg of each sample was weighed in a 10 mL test tube into which 1 mL of each phase of the pre-equilibrated two-phase solvent system was added. The test tube was capped and shaken vigorously for 1 min, and allowed to stand until it separated completely. An aliquot of 100 lL of each layer was taken out and evaporated separately to dryness in vacuo at <40 °C. The residue was dissolved in 10 lL methanol and analyzed by HPLC for determining the partition coefficient (K) of compounds 1–5. The K value was expressed as the peak area of the target compound in the upper phase divided by that in the lower phase.

the caffeoyl moiety (162), the glucose moiety (162), the rhamnose moiety (146), a CH2 radical (14), and the COCH2 group (42).

The LC–ESI-MS of peak 1 is shown in Fig. 4A. The deprotonated molecular ion [M H]— at m/z 785 with a high abundance and a dimeric deprotonated molecular

ion [2M H]— at m/z 1571 was observed in the negative mode, suggestive of a molecular weight of 786, which was the same as that of echinacoside. Further investigation in the LC–MS2 experiment of the m/z 785 ions yielded one main daughter ion at m/z 623 (Fig. 4B) produced directly from m/z 785 by loss of a caffeoyl moiety or a hexose moiety as [M 162 H]—. The LC–MS3 spectrum of m/z 623 exhibited two major ions at m/z 477 and 461, and two minor ions at m/z 315 and 179 (Fig. 3C, Table 2). The mass differences between m/z 623 and the fragment ions m/z 477 and 461 were 146 and 162, respectively, corresponding to the loss of a rhamnose unit and a glucose moiety or a caf-feel moiety [M 162 H]—. The m/z 623 ions also lost a caffeoyl moiety and a rhamnose moiety to produce the m/z 315 ions. The ion at m/z 179 was produced from the cleavage of the caffeoyl moiety, with the negative charge remaining- ing on the part of the caffeoyl moiety. LC–ESI-MSN experiment on the authentic echinacoside showed the same fragmentation pattern. Peak 1 was therefore confirmed to be echinacoside.

For peak 2, the LC–ESI-MS showed m/z 799 as the deprotonated molecular ion [M H]— and m/z 1599 as its dimeric ion, suggestive of a molecular mass of 800. During MS2 experiments, the m/z 799 ions formed three daughter

ions at m/z 637, 623, and 475 (Table 1). The ion at m/z 637 was produced directly from the parent ion of m/z 799 again due to the neutral loss of the caffeoyl [M—162—H]— or a glucose moiety [M—162—H]—. The ion at m/z 623 resulted from the loss of a CH2 radical. The ion at m/z 475 was formed from the neutral loss of both the caffeoyl moiety [M 162 H]— and the glucose moiety from the parent ion. The MS3 experiment on m/z 637 produced three ions at m/z 619, 491, and 475, corresponding to the losses of one water, a rhamnose unit, and a glucose moiety, respectively. The m/z 623 daughter ion produced m/z 461 and 315 in the MS3 study, which followed the same fragmentation pathways as echinacoside as discussed above. The LC– ESI-MSN data supported a tentative identification for peak 2 as cistanoside A.

LC–ESI-MS experiments were also conducted for peaks

3 and 4 (tR 12.49 and 13.46 min in Fig. 2). Both peaks showed the same [M H]— ion at m/z 623 and dimer at m/z 1247 in the negative mode (Table 1), indicating they are possibly isomers with the same molecular weight of

624, the same as acteoside and isoacteoside. The MS2 spectra of the [M H]— ions also showed one same daughter ion of m/z 461, which indicated the loss of the caffeoyl moiety from the parent ion m/z 623 (Table 1). Similar MS3 spectra were obtained for the two compounds. For peak 3, the MS3 spectrum of the ion at m/z 461 formed three ions at m/z 315, 161, and 135. The m/z 315 is formed after losing rhamnose as discussed earlier. The ion m/z 161 was produced from the cleavage of the caffeoyl moiety, followed by a further loss of one water; the charge remained on the part of the caffeoyl moiety. The ion at m/z 135 [agly- cone 18 H]— arose from the cleavage of the glycosidic bond at C1 position with an additional loss of one water, leaving the charge to be on the part of the aglycone moiety. The MS3 spectrum of peak 4 followed the same fragmentation pathway as peak 3 except for the missing ion at m/z 135. The molecular ion and the fragmentation patterns of these two compounds are consistent with the literature data on acteoside and isoacteoside (Wang et al., 2000), although an additional ion m/z 153 was found and

assigned as [aglycone H]— by Wang et al. (Wang et al., 2000). This ion may have been unstable and lost water to give m/z 135 in our experiment. Based on the MS data and the congruent retention times of peaks 3 and 4 with the standards, they are therefore identified as acteoside and isoacteoside, respectively.

The LC–ESI-MS data of peak 5 are shown in Table 2. A deprotonated molecular ion [M H]— (m/z 665) was the only ion found in the negative mode, implying a molecular mass of 666. Three daughter ions were observed at m/z 623, 503, and 461 in the MS2 experiment (Table 2). The daughter ions at m/z 623 and 503 were formed directly from the parent ion by loss of a COCH2 group and a caffeoyl moiety, respectively. The ion at m/z 461 came from the loss of both the caffeoyl and the COCH2 moiety [M 162 42 H]— from the parent ion. In the MS experiment, m/z 623 produced m/z 461, and m/z 503 yielded three ions at m/z 485, 461, and 315. The MS3 spectrum of the daughter ion m/z 461 gave two ions at 443 and 315. By comparing the LC– MSN fragmentation pattern of peak 5 with other compounds reported in this study and with other reported (Li et al., 2005; Wang et al., 2000), we concluded that peak 5 was structurally highly related to acteoside with the only difference being the COCH3 unit on the R3 position. A tentative identity was therefore given to peak 5 as 20-acetylacteoside (Fig. 1).

The structures of the two tentatively identified compounds, peak 2 (as cistanoside A) and peak 5 (as 20-acetyl- acteoside) were confirmed of their structures by 1H NMR. The chemical shifts and coupling constants of all protons in compounds 2 and 5, as shown in Table 3., matched with the reported NMR data for cistanoside A and 20-acetylacteoside, respectively (Kobayashi et al., 1984, 1987). 2D NMR experiments (long-range COSY, ROESY, and CH correlation) were also conducted in the present study and they further confirmed the identification (data not shown).

phenylethanoid glycosides in cistanche

4. Conclusions

In the present paper, HSCCC was successfully used for the isolation and purification of echinacoside, cistanoside A, acteoside, isoacteoside, and 20-acetylacteoside from the n-butanol extract of C. deserticola. It is therefore a proven means for semi-preparative separation of bioactive. Meanwhile, the structures of the five PhGs in Cistanche deserticola have been investigated by means of LC–ESI-MSN; some characteristic features of PhGs were found, which allowed us to determine the functional groups in the structures. The LC–ESI-MSN method is therefore a powerful tool for rapid identification of phenylethanoids and their glycosides in Cistanche deserticola extracts, especially when substantiated by NMR data.

Acknowledgment

The authors would like to thank Jun Gu of the Nuclear Magnetic Resonance Centre, University of Guelph, Ontario, Canada for her assistance in NMR experiments. This project was partially supported by funding from the Jilin Province, China (No. 20060904).

References

Chen, L. J., Games, D. E., & Jones, J. (2003). Isolation and identification of four flavonoid constituents from the seeds of Oroxylum Indicum by high-speed counter-current chromatography. Journal of Chromatography A, 988, 95–105.

Cheng, X. Y., Wei, T., Guo, B., Ni, W., & Liu, C. Z. (2005). Cistanche deserticola cell suspension cultures: Phenylethanoid glycosides bio-synthesis and antioxidant activity. Process Biochemistry, 40, 3119– 3124.

Foucault, A. P., & Chevolot, L. (1998). Counter-current chromatography: instrumentation, solvent selection, and some recent applications to natural product purification. Journal of Chromatography A, 808, 3–22.

Gross, G.-A., Lahloub, M. F., Anklin, C., Schulten, H.-R., & Sticher, O. (1988). Teucrioside, a phenylpropanoid glycoside from Teucrium Chamaedrys. Phytochemistry, 27, 1459–1463.

He, Z. D., Lau, K. M., Xu, H. X., Li, P. C., & But, P. P. H. (2000).

Antioxidant activity of phenylethanoid glycosides from Brandisia chance. Journal of Ethnopharmacology, 71, 483–486.

Kobayashi, H., Karasawa, H., & Miyase, T. (1984). Studies on the constituents of Cistanchis Herba. III. Isolation and structures of new phenylpropanoid glycosides, Cistanoside A and B. Chemical & Pharmaceutical Bulletin, 32, 3009–3014.

Kobayashi, H., Oguchi, H., Takizawa, N., Miyase, T., Ueno, A., Usmanghani, K., et al. (1987). New phenylethanoid glycosides from Cistanche tubulosa (Schrenk) Hook. F.I. Chemical & Pharmaceutical Bulletin, 35, 3309–3314.

Lei, L., Yang, F. Q., Zhang, T. Y., Tu, P. F., Wu, L. J., & Ito, Y. (2001).

Preparative isolation and purification of acteoside and 2’-acetyl acteoside from Cistanches salsa (C.A. Mey) G. Beck by counter-current chromatography. Journal of Chromatography A, 912, 181–185.

Li, L., Tsao, R., Liu, Z. Q., Liu, S. Y., Yang, R., Young, J. C., et al. (2005). Isolation and purification of acteoside and isoacteoside from Plantago psyllium L. by high-speed counter-current chromatography. Journal of Chromatography A, 1063, 161–169.

Li, L. L., Wang, X. W., & Wang, X. F. (1997). Antilipid peroxidation and antiradical action of glycosides in herba Cistanches. China Journal of Chinese Materia Medica, 22, 364–367.

Li, J., Wang, P. F., Zheng, R. L., Liu, Z. M., & Jia, Z. J. (1993). Protection of phenylpropanoid glycosides from Pedicularis against oxidative hemolysis in vitro. Planta Medica, 59, 315–317.

Lu, M. C. (1998). Studies on the sedative effect of Cistanche deserticola.Journal of Ethnopharmacology, 59, 161–165.

Nishimura, H., Sasaki, H., Inagaki, N., Chin, M., & Mitsuhashi, H. (1991). Nine phenethyl alcohol glycosides from Stachys szeboldzz. Phytochemistry, 30, 965–969.

Oka, F., Oka, H., & Ito, Y. (1991). A systematic search for suitable two-phase solvent systems for high-speed counter-current chromatography. Journal of Chromatography, 538, 99–108.

Ravn, H., Nishibe, S., Sasahara, M., & Li, X. (1990). Phenolic compounds from Plantago Asiatica. Phytochemistry, 29, 3627–3631.

Schapoval, E. E. S., Winter de Vargas, M. R., Chaves, C. G., Bridi, R., Zuanazzi, J. A., & Henriques, A. T. (1998). Antiinflammatory and antinociceptive activities of extracts and isolated compounds from Stachytarpheta cayennensis. Journal of Ethnopharmacology, 60, 53–59. Shoyama, Y., Matsumoto, M., & Nishioka, I. (1987). Phenolic glycosides from diseased roots of Rehmannia glutinosa Var. Purpurea. Phytochemistry, 26, 983–986.

Wang, X. W., Jiang, X. Y., Wu, L. Y., & Wang, X. F. (2001). Scavenging effect of glycosides of Cistanche deserticola on free radicals and its protection against OH-induced DNA damage in vitro. Journal of Chinese Pharmacology, 36, 29–31.

Wang, Y. M., Zhang, S. J., Luo, G. A., Hu, Y. N., Hu, J. P., Liu, L., Zhu,

Y., & Wang, H. J. (2000). Analysis of phenylethanoid glycosides in the extract of herbs Cistanchis by LC/ESI-MS/MS. Acta Pharmaceutica Sinica, 35, 839–842.

Wong, C.-C., Li, H.-B., Cheng, K.-W., & Chen, F. (2006). A systematic survey of antioxidant activity of 30 Chinese medicinal plants using the ferric reducing antioxidant power assay. Food Chemistry, 97, 705–711. Xie, J. H., & Wu, C. F. (1993). Effect of ethanolic extract of Cistanche deserticola on the contents of monoamine neurotransmitters in rat

brain. Chinese Traditional and Herbal Drugs, 24, 417–419.

Xiong, Q., Hase, K., Tezuka, Y., Tani, T., Namba, T., & Kadota, S. (1998). Hepatoprotective activity of phenylethanoids from Cistanche deserticola. Planta Medica, 64, 120–125.

Xiong, Q. B., Kadota, S., Tani, T., & Namba, T. (1996). Antioxidative effects of phenylethanoids from Cistanche deserticola. Biological & Pharmaceutical Bulletin, 19, 1580–1585.

Zong, G., He, W., Wu, G. L., & Chen, L. H. (1996). Comparisons between Cistanche deserticola YC Ma and Cistanche tubular (Schenk) Wight on some pharmacological activities. Journal of Traditional Chinese Medicine, 21, 436–438.