684
PHYTOTHERAPY RESEARCH Phytother. Res. 19, 684–688 (2005) Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/ptr.1728 T. T. X. DONG ET AL.
Orthogonal Array Design in Optimizing the Extraction Efficiency of Active Constituents from Roots of Panax notoginseng T. T. X. Dong1, K. J. Zhao1,2, W. Z. Huang1, K. W. Leung1 and K. W. K. Tsim1* 1 2
Department of Biology, The Hong Kong University of Science and Technology, Clear Water Bay Road, Hong Kong, China Department of Pharmacy, Beijing Military Medical College of PLA, Beijing, China
The root of Panax notoginseng (Radix Notoginseng, Sanqi) is a commonly used traditional Chinese medicine, which is mainly cultivated in Wenshan of Yunnan China. The identified active constituents in Radix Notoginseng include saponin, flavonoid and polysaccharide; however, the levels of these active constituents vary greatly with different extraction processes. This variation causes a serious problem in standardizing the herbal extract. By using HPLC and spectrophotometry, the contents of notoginsenoside R1, ginsenoside Rg1, Rb1, Rd, and flavonoids were determined in the extracts of Radix Notoginseng that were derived from different processes of extraction according to an orthogonal array experimental design having three variable parameters: nature of extraction solvent, extraction volume and extraction time. The nature of extraction solvent and extraction volume were two distinct factors in obtaining those active constituents, while the time of extraction was a subordinate factor. The optimized condition of extraction therefore is considered to be 20 volumes of water and extracted for 24 h. In good agreement with the amount of active constituents, the activity of anti-platelet aggregation was found to be the highest in the extract that contained a better yield of the active constituents. The current results provide an optimized extraction method for the quality control of Radix Notoginseng. Copyright © 2005 John Wiley & Sons, Ltd. Keywords: Panax notoginseng; quality control; Chinese medicine; extraction.
INTRODUCTION Radix Notoginseng, the root of Panax notoginseng (Burk.) F. H. Chen, is a well-known traditional Chinese medicine named Sanqi belonging to the Ginseng genus. A wide variety of therapeutic uses of Radix Notoginseng have been reported including the promotion of blood circulation, removal of blood stasis, induction of blood clotting, relieving swelling and alleviating pain (Lei and Chiou, 1986; Cicero et al., 2003), which is also used for the treatment of coronary heart disease and cerebral vascular disease with favorable results (Zheng and Yang, 1994; Zheng, 2000; Chan et al., 2002). Radix Notoginseng is prescribed in numerous Chinese medicinal preparations including Yunnan Baiyao, which is used for the treatment of trauma and bleeding after internal and external injury, and Pianzai Huang, which is known for treating hepatitis. Different lines of evidence suggest that saponin (Gan and Zhen, 1992; Chen et al., 2001), flavonoid (Wei et al., 1980) and polysaccharide (Yang and Du, 2002) are the main active constituents in the roots of P. notoginseng. Saponins of Radix Notoginseng have been reported to increase
* Correspondence to: Dr Karl W. K. Tsim, Department of Biology, The Hong Kong University of Science and Technology, Clear Water Bay Road, Hong Kong SAR, China. E-mail:
[email protected] Contract/grant sponsor: University Grants Committee of the Hong Kong SAR, China; contract/grant number: AoE/B-10/01. Copyright © 2005 John Wiley & Sons, Ltd. Copyright © 2005 John Wiley & Sons, Ltd.
the blood flow of the coronary arteries (Huang et al., 1999), to prevent platelet aggregation (Wang et al., 2004) and to decrease the consumption of oxygen by heart muscles (Chen et al., 1983). Currently, over 20 different types of saponins have been identified in P. notoginseng root (Gan and Zhen, 1992; Li and Fitzloff, 2001) including ginsenosides, notoginsenosides and gypenosides. Among these saponins, ginsenoside Rg1, Rb1, Rd and notoginsenoside R1 are the major components found in Radix Notoginseng (Gan and Zhen, 1992; Dong et al., 2003). Thus, the contents of these four saponins are employed for the quality control of Radix Notoginseng. Similarly, flavonoid is known to increase coronary flow, to reduce myocardial oxygen consumption and to lower arterial pressure (Wei et al., 1980). While polysaccharide of P. notoginseng root is considered to be an active constituent having immunostimulating activities in vitro (Ohtani et al., 1987; Yang and Du, 2002). On the other hand, dencichine, an amino acid isolated from Radix Notoginseng, has a function on hemostasis (Zhao and Wang, 1985). Although different methods of extraction have been used in Radix Notoginseng, the optimized condition to achieve the best chemical composition and biological function in the herbal extract has not been determined. The nature of the solvent, the extraction volume and the time of extraction are crucial factors in determining the extraction efficiency of the active constituents in herbal medicines. In order to evaluate the quality of the extraction process of Radix Notoginseng, three levels of three different extraction parameters were considered by orthogonal array design. The experiment January(2005) 2005 Phytother.Received Res. 19,24 684–688 Accepted 24 May 2005
685
PANAX NOTOGINSENG EXTRACTION
of L9 (34) was adopted, and nine different groups of extractions were analysed. The amounts of notoginsenoside R1, ginsenoside Rg1, Rb1, Rd and flavonoids of Radix Notoginseng from different extractions were determined by reverse phase high-performance liquid chromatography (HPLC) and spectrophotometry. In addition, the anti-platelet aggregation activities of the extracts derived from different extraction protocols were also determined.
MATERIALS AND METHODS Plant materials. Fresh P. notoginseng plants were collected from Wenshan of Yunnan Province, China. The 3-year-old plants were collected from Wenshan during September to October in 2000. The botanical origins of all the materials in the form of whole plants were identified morphologically during the field collection. Roots of P. notoginseng were collected and dried under vacuum. A single batch of roots ground to powder of about 5 kg was obtained from ~100 plants of the same population. These grinding processes were done during the field collection before they were delivered to the laboratory. The collected powder was stored with silica gel that stabilized the chemical constituents. Their corresponding voucher specimens in the form of dry roots were deposited in the Department of Biology, The Hong Kong University of Science and Technology, Hong Kong, China. Extraction of chemical constituents. Three different parameters of extraction and three levels for each were studied by orthogonal array design of L9 (34) and nine combinations of extraction method having different parameters were established. For the determination of saponin and flavonoid, 10 g of ground powder (100 meshes) was soaked in different solvents (1/5 total volume) for different times and added onto a percolator, and nine different extracts were obtained. The samples were percolated with the designed volume and centrifuged. The supernatant was collected and concentrated. The final volume was adjusted to 25 mL with a volumetric flask. Samples were filtered through a Millipore filter unit. Then 20 µL of the sample was injected to reverse phase HPLC. Quantitative analysis. Notoginsenoside, ginsenoside and quercetin (a flavonoid standard) were purchased from National Institute for the Control of Pharmaceutical and Biological Products from Beijing, China. The identities of these chemical markers were confirmed by NMR and mass spectrometers, as determined by the suppliers. HPLC grade reagents were purchased from Fischer and Labscan (Dublin, Ireland). For the calibration of notoginsenoside R1, ginsenoside Rg1, Rb1, Rd, the standards were weighed and dissolved in 1 mL methanol to give serial concentrations. Three injections were performed for each dilution. The concentrations of these compounds in the samples were calculated according to the regression parameters derived from the standard curves. The HPLC system consisted of a Waters PC 800 integrator, a Waters 486 tunable absorbance detector and a Waters™ 600 pump. Chromatographic separations were carried out on a NOVA-PAK Copyright © 2005 John Wiley & Sons, Ltd.
C18 column (300 mm × 3.9 mm i.d., particle size 4 µm) with a guard column (NOVA-PAK C18, 20 mm × 3.9 mm i.d., particle size 4 µm), with CH3CN/50 mM KH2PO4 (20:80) as an eluent at a flow rate of 1.0 mL/min at room temperature and monitored at 203 nm. The running condition was developed as below: a linear gradient of CH3CN/50 mM KH2PO4 from a ratio of 20:80 to 30:70 from 2–6 min; a linear gradient of CH3CN/50 mM KH2PO4 from a ratio of 30:70 to 50:50 from 6–14 min; a linear gradient of CH3CN/50 mM KH2PO4 from a ratio of 50:50 to 30:70 from 14–16 min; an isocratic CH3CN/ 50 mM KH2PO4 (30:70) from 16–30 min; a linear gradient of CH3CN/50 mM KH2PO4 from a ratio of 30:70 to 20:80 from 30–35 min. This running condition was optimized to give the best resolution of all the HPLC peaks. The content of individual saponin was determined by HPLC using Rg1 as a reference standard. For flavonoid calibration, quercetin was weighed and dissolved in 50 mL 70% methanol to give serial concentrations. The absorbances of standard solution and samples, diluted 20 times in 70% methanol, were detected at 249 nm by UV spectrophotometry (Beckman DUR 650). Anti-platelet aggregation assay. Blood was collected from adult New Zealand white rabbits through a polyethylene cannula placed in the common carotid artery by a 10 mL syringe. The first few mL of blood were discarded, and the rest was diluted 10-fold with 3.8% tri-sodium citrate. The platelet-rich plasma was achieved by centrifugation at 800 rpm for 10 min. The platelet-poor plasma was achieved by centrifugation at 4000 rpm for 10 min; this plasma was used as the background reading in the assay (Ashida and Abiko, 1979). The extract was added 5 min before ADP (inducer; 10 µM final). The aggregation at 5 min (maximum; Amax) and at 1 min (A1’) were recorded by a Sanda-196 platelet aggregator (Shanghai, China). Ticlopidine was used as a positive control. The inhibition activity of platelet aggregation was calculated by formula: (ADP-induced Amax – sample-induced Amax) / (ADP-induced Amax) × 100%. Statistical analysis. In hierarchical clustering analysis of different samples, SPSS software (version 11.0 from Statistical Product and Service Solutions, Chicago, IL) was used. Data were evaluated for statistical significance on a minimum of six replicates using the unpaired t-test. Differences were considered to be significant when p ≤ 0.05. All data are expressed as mean ± SEM of triplicate determinations, where n was normally over 5.
RESULTS AND DISCUSSION Three different parameters of extraction and three levels for each were studied by orthogonal array design of L9 (34), and therefore nine extraction groups having different parameters were established. The factors and levels selected are displayed in Table 1. Notoginsenoside R1, ginsenoside Rg1, Rb1 and Rd are considered to be the key active constituents in roots of P. notoginseng. When the extract, derived from roots of P. notoginseng, was subjected to HPLC analysis under the gradient Phytother. Res. 19, 684–688 (2005)
686
T. T. X. DONG ET AL.
Table 1. Orthogonal array design to optimize the extraction of P. notoginseng roots Extraction parameters Groupa 1 2 3 4 5 6 7 8 9
Volume of solventb
Time (h)c
7 7 7 10 10 10 20 20 20
12 24 36 12 24 36 12 24 36
Nature of solventd H2O Ethanol Ethanol Ethanol Ethanol H2O 70% Ethanol H2O 30% Ethanol
30% 70% 30% 70%
a
Nine groups having different combinations of parameters during extraction. Each group contained 10 g of Radix Notoginseng powder as the starting material; five batches of preparations were done (n = 5). b Solvent volume in multiple times to total weight of herbs, i.e. 10 g × 7 = 70 mL in extraction volume, as listed in the group 1. c Extraction time. d The solvent used in the extraction.
Figure 1. Determinations of notoginsenoside R1, ginsenoside Rg1, Rb1, and Rd from roots of P. notoginseng. By using NOVAPAK C18 column in HPLC analyses, different saponins (at absorbance 203 nm) were separated; the running conditions were described in the Materials and Methods. Peaks corresponding to R1, Rg1, Rb1 and Rd are indicated. A typical chromatogram is shown.
elution, the peaks corresponding to different saponins were well separated. Figure 1 shows a typical chromatogram of ethanol extract from Radix Notoginseng. The individual peaks for notoginsenoside R1, ginsenoside Rg1, Rb1 and Rd were distinct. The peaks of these saponins were further identified by two means: (i) by
comparing the retention times of the unknown peaks with those of the standards eluted under the same conditions; and (ii) by spiking the sample with stock standard solutions of saponins, or dencichine. The quantitation was carried out by measuring the peak area according to the regression equation (Table 2). By using the established HPLC method, the amounts of notoginsenoside R1, ginsenoside Rg1, Rb1 and Rd were calculated from the calibration curves that were prepared with the standard solutions of each compound; these four saponins accounted for over 80% of total saponins in root of P. notoginseng (data not shown). The correlation coefficients of notoginsenoside R1, ginsenoside Rg1, Rb1 and Rd were from 0.9988 to 0.9998. The precision and repeatability of the tests were excellent, having a relative standard deviation (RSD) < 5%. The recovery experiment was carried out to evaluate the accuracy of the method. Known amounts of saponins were added to the sample and extracted accordingly. The recoveries of notoginsenoside R1, ginsenoside Rg1, Rb1 and Rd were from 96% to 103% (Table 2). The determination of flavonoid by using quercetin as a standard was done in a spectrophotometer, where the calibration curve and the recovery were as good as the analyses of saponins (Table 2). The amounts of five chemical markers within Radix Notoginseng showed significant variation under nine extractions, which are summarized in Fig. 2. The amount of notoginsenoside R1, at group 8 was higher than other groups, which showed ~2.7-fold difference when compared with the lowest group (e.g. group 3). Besides notoginsenoside R1, the amounts of ginsenoside Rg1, Rb1, Rd and the total flavonoids in group 8 also showed significantly higher levels than that of other groups; the difference was over 2-fold in general when compared with the lowest group. In contrast, the yields of saponins and flavonoids in the root extracts of groups 3 and 5 were significantly lower than that from others. Thus, the best extraction condition, chemically, should be considered to be 20 volume of water and extracted for 24 h as done in group 8. Different parameters were analysed statistically, and their results were summarized in Table 3. The nature of solvent and volume of extraction were two distinct factors, and the time of extraction was a subordinate factor. The extraction solvent and its volume (both at p < 0.05) are the two crucial parameters for a higher yield of notoginsenoside R1 as well as for ginsenoside Rb1. For the yield of ginsenoside Rg1 the extraction volume (p < 0.05) showed a more significant role than
Table 2. Calibration of notoginsenoside R1, ginsenoside Rg1, Rb1, Rd and flavonoids by HPLC analyses
Standard Notoginsenoside R1 Ginsenoside Rg1 Ginsenoside Rb1 Ginsenoside Rd Quercetin
Regression equation
y y y y y
= = = = =
547 904x 388 057x 240 491x 127 725x 0.07052x
− − − − +
240 765 678 882 175 137 65 269 0.05587
r2
Linearity (mg/mL)
Precision RSD (%)
Repeatability RSD (%)
Recovery (%)
0.9988 0.9991 0.9998 0.9993 0.9992
0.005~0.2 0.01~0.5 0.01~0.5 0.002~0.1 0.02~1.0
2.6 1.7 2.5 1.9 2.3
3.9 2.8 4.4 4.7 4.6
96 98 99 101 103
HPLC performance was described in Materials and Methods. Recovery was determined by adding known amounts of constituents into the root, where the amount of active constituents were known. These samples were subjected to HPLC analysis. The regression equation was used to calibrate the concentration of various active constituents. Quercetin is a flavonoid standard. The mean values are expressed here, and the SEM values of the five tested chemicals were less than 5% of the mean. The calibration was repeated five times (n = 5). Copyright © 2005 John Wiley & Sons, Ltd.
Phytother. Res. 19, 684–688 (2005)
PANAX NOTOGINSENG EXTRACTION
Figure 2. Comparison of five active constituents in the extracts from roots of P. notoginseng. The extraction conditions of different groups were described in Table 1. The amounts of notoginsenoside R1, ginsenoside Rg1, Rb1, and Rd were calibrated by HPLC as shown in Fig. 1, and the regression equations as listed in Table 2. Values are expressed in g/100 g of dry material and in mean ± SEM, where n = 5.
Table 3. Analyses of variances in extracting active constituents from roots of P. notoginseng Source of variationa
Fb
pc
Notoginsenoside R1
A B C
21.98 0.30 9.89
0.1 0.1), which could be because of the lower polarity of flavonoids compared with saponins. Besides the chemical analysis, the bioactivity of those P. notoginseng extracts from different groups was also tested here. Of the different proposed functions of Radix Notoginseng, the activity in preventing platelet aggregation is one of the known functions that could be mediated by saponin, and additionally that could be measured easily in vitro. Application of ADP in plasma induced platelet aggregation in a dose-dependent manner (Fig. 3A). The Radix Notoginseng extracts derived from different groups of extraction prevented the ADPinduced platelet aggregation. In line with the chemical analysis, group 8 contained the highest amount of saponins and flavonoids as well as activity in preventing the aggregation of platelets (Fig. 3B). The extracts derived from groups 4, 5, 6 and 9 also showed significant higher anti-platelet aggregation activity; indeed these groups contained higher contents of saponins and flavonoids than the others. Although the chemical and biological assays are in line to suggest that saponin could partly account for the anti-platelet aggregation activity in Radix Notoginseng as previously reported (Wang et al., 2004), the existence of other active factors with such activity should be not eliminated in the extracts. Indeed, recent studies from our laboratory have revealed small molecules from P. notoginseng root extract that have strong anti-platelet aggregation activity (Tsim et al., unpublished results), and these molecules are highly soluble in water, e.g. as the extraction condition in group 8. Phytother. Res. 19, 684–688 (2005)
688
T. T. X. DONG ET AL.
P. notoginseng belongs to the Ginseng genus and is a member of Araliaceae, and is an archaic plant originating 25000 years ago in southwest of Yunnan China (Zheng et al., 1994). Indeed, this region of China produces the most and the best quality of Radix Notoginseng. Not only because of the specific geographical properties, the farmers in Yunnan have experience of cultivating Radix Notoginseng of over a thousand years. Traditionally, the quality of traditional Chinese medicines is heavily dependent on the cultivated regions and the timing of the harvest, and in China it is called ‘Di Dao’ to describe the highest quality of herbs that are collected from the best region and at the best time (Dong et al., 2003). Although the source of Radix Notoginseng has been determined, the processes in extracting the best amounts of active constituents have not been revealed. In particular, the extracts (e.g. total saponins) of Radix Notoginseng are commonly used as health food products in China and South East Asia. The current results therefore provide an extraction methodology to obtain the highest amount of active constituents from Radix Notoginseng. Dencichine and total polysaccharides are also considered as active constituents in Radix Notoginseng; however, the levels of these two chemicals show very little variation in the quality control of the herb (Dong et al., 2003), and therefore have not been included in this analysis. The current extraction method could also be applied to the rhizome of P. notoginseng that normally is not used because of its irregular appearance. Because of its high content of active constituents, the rhizome is commonly used by manufacturers as the primary source of total saponin extraction. A single plant of P. notoginseng produces about 8 g of root per harvest, and the total yield of rhizome is about 25% of that. The contents of
notoginsenoside R1, ginsenoside Rg1, Rb1, Rd in rhizome are markedly higher than that in the roots (Dong et al., 2003). Besides, the content of total flavonoids and dencichine in the rhizome is slightly higher than the root. Having the aforementioned conditions in extracting active constituents, the rhizome of P. notoginseng has great economic value. Roots from Panax quinquefollius L. (American ginseng) and Panax ginseng C.A. Meyer (Korean ginseng) share a high degree of similarity in their chemical and genetic compositions compared with that of P. notoginseng. For instance, the spacer domains of 5SrRNA were sequenced and compared, which showed over 75% DNA identity among different members of the Panax family (Cui et al., 2003), even much higher sequence identity was revealed in other regions of the genome (Mihalov et al., 2000). Moreover, roots of P. notoginseng also show a close chemical resemblance to that of P. quinquefollius and P. ginseng. They all contain saponins as their active constituents, except for American and Korean ginsengs that contain predominantly Rb1 as the major ginsenoside (Mihalov et al., 2000), while Rg1 is the major component in Panax ginseng (Dong et al., 2003). Thus, it should be expected that the current described conditions in optimizing the extraction of active constituents could also be applied to American and Korean ginsengs. Acknowledgements The research was supported by grants from the Area of Excellence Scheme established under the University Grants Committee of the Hong Kong SAR, China (AoE/B-10/01 to KWKT). We are grateful to Dr Shao P. Li of Macau University for his fruitful discussion during the study.
REFERENCES Ashida SI, Abiko Y. 1979. Inhibition of platelet aggregation by a new agent, Ticlopidine. Thromb Haemost 40: 542–550. Chan P, Thomas GN, Tromlinson B. 2002. Protective effects of trilinolein extracted from Panax notoginseng against cardiovascular disease. Planta Med 68: 1024–1028. Chen JC, Chen LD, Tsauer W, Tsai CC, Chen BC, Chen YJ. 2001. Effects of ginsenoside Rb2 and Rc on inferior human sperm motility in vitro. Am J Chin Med 29: 155–160. Chen ZH, Wang DC, Li HL, Wei JX, Wang JF, Du YC. 1983. Hemodynamic effects of san chi (Panax notoginseng) root, leaf, flower and saponins on anesthetized dogs. Acta Pharm Sin 18: 818–822. Cicero AF, Vitale G, Savino G, Arlett R. 2003. Panax notoginseng (Burk.) effects on fibrinogen and lipid plasma level in rat fed on a high-fat diet. Phytother Res 17: 174–178. Cui XM, Lo CK, Yip KL, Dong TTX, Tsim KWK. 2003. Authentication of Panax notoginseng by 5S-rRNA spacer domain and random amplified polymorphic DNA (RAPD) analysis. Planta Med 69: 584–586. Dong TTX, Cui XM, Song ZH et al. 2003. Chemical assessment of roots of Panax notoginseng in China: Regional and seasonal variations in its active constituents. J Agric Food Chem 51: 4617–4623. Gan FY, Zhen GZ. 1992. Chemical composition studies of Panax notoginseng. Chin Pharm J 27: 138–143. Huang YS, Yang ZC, Yan BG, Hu XC, Li AN, Crowther RS. 1999. Improvement of early postburn cardiac function by use of Panax notoginseng and immediate total eschar excision in one operation. Burns 25: 35–41. Lei XL, Chiou GC. 1986. Cardiovascular pharmacology of Panax Copyright © 2005 John Wiley & Sons, Ltd.
notoginseng (Burk) F. H. Chen and Salvia miltiorrhiza. Am J Chin Med 14: 145–152. Li W, Fitzloff JF. 2001. A validated method for quantitative determination of saponins in notoginseng (Panax notoginseng) using high-performance liquid chromatography with evaporative light-scattering detection. J Pharm Pharmacol 53: 1637–1643. Mihalov JJ, Der Marderosian A, Pierce JC. 2000. DNA identification of commercial ginseng samples. J Agric Food Chem 48: 3744–3752. Ohtani K, Mizutani K, Hatono S et al. 1987. Sanchinan-A, a reticuloendothelial system activating arabinogalactan from sanchi-ginseng (roots of Panax notoginseng). Planta Med 53: 166–169. Wang J, Xu J, Zhong JB. 2004. Effect of Radix Notoginseng saponins on platelet activating molecule expression and aggregation in patients with blood hyperviscosity syndrome. Zhongguo Zhong Xi Yi Jie He Za Zhi 24: 312–316. Wei JX, Wang JF, Chang LY, Du YC. 1980. Chemical studies of san-chi Panax notoginseng I: Studies on the constituents of San-Chi root hairs. Acta Pharm Sin 15: 359–364. Yang F, Du Y. 2002. Synthesis of oligosaccharides derivatives related to those from sanqi, a Chinese herbal medicine from Panax notoginseng. Carbohydr Res 337: 485–491. Zhao GQ, Wang XX. 1985. Hemostasis constituent from Panax notoginseng- dencichine. Chin Trad Herb Drugs 16: 34–36. Zheng GZ, Yang CR. 1994. Biology of Panax notoginseng and Its Application. Science Press: Beijing, China, 9. Zheng XY (ed.). 2000. Pharmacopoeia of the People’s Republic of China. Chemistry Industry Press: Beijing, China, 180. Phytother. Res. 19, 684–688 (2005)
