Production and Characterisation of Taraxacum officinale Extracts Prepared by Supercritical Fluid and Solvent Extractions

There are many reports on biological activities of pentacyclic triterpenoids, which could be relevant to the pharmacological effects including anti-inflammatory properties. Dandelion (Taraxacum officinale Wiggers et Weber, Asteraceae) is one of the best known European medicinal plants, rich in triterpenoids, which has been used for the treatment of various inflammatory diseases such as rheumatoid arthritis and also for many infectious disorders. The aim of this work was to investigate the supercritical fluid extraction (SFE) of dandelion crude drugs (Taraxaci radix and T. folium) with carbon dioxide, to study the extraction of triterpenoids and phytosterols and to compare supercritical CO2 extracted products and extracts made by traditional solvent extractions (n-hexane and ethanol 96%). Solvent extractions were carried out using a Soxhlet extractor. To define the effect of temperature and pressure on the yield of supercritical fluid extraction, a 2 factorial 3 level experiment chain was performed. The content of triterpenes and phytosterols was determined, after saponification, by thin layer chromatographydensitometry. The products gained by SFE were different from the traditional ones concerning their apparence and composition; triterpenes and their esters could be extracted quantitatively by supercritical fluid extraction using CO2 as solvent; the extraction dynamic for β-amyrin and β-sitosterol was different; triterpenes have a higher concentration in the SFE product then in traditional ones. By means of supercritical fluid extraction of Taraxacum crude drugs, in function of the selectivity of the solvent, temperature, pressure and accompanying constituents, qualitatively new products can be gained. These may serve as prospective raw materials for phytopharmaceuticals. INTRODUCTION There is an increasing demand for natural products as curative agents and foods. Regulations in the food, pharmaceuticaland cosmetic industries are getting more rigorous so the traditional solvent extraction could cause problems. In this study we examined the extraction possibilities of the active materials of Taraxacum officinale. The root (Taraxaci radix) and herb (Taraxaci folium and herba) are not only traditional medicines but may serve as prospective raw materials for modern pharmaceuticals. The anti-inflammatory activity of dandelion extracts has been recently confirmed in animal studies (Mascolo, 1987) and aqueous extracts seem to have anti-tumour activity (Nevall et al., 1996). Based on pharmacological studies, dandelion is one of the components of phytomedicines used in therapy for hepatitis and the drug also has diuretic and choleretic actions (Bisset, 1994; Bradley, 1992). There are numerous studies about chemical composition of Taraxacum officinale (Czygan 1990; Hegnauer 1964; Hegnauer 1989; Komissarenco and Derkach 1981). Roots contain several triterpenes, including taraxol, taraxerol, taraxasterol, ψtaraxasterol, and β-amyrin; sterols (stigmasterols, β-sitosterol); inulin (ca. 25%) (Rutherford 1972); sugars (fructose, glucose, sucrose); pectin; glucosides; choline; phenolic acids e.g., caffeic and p-hydroxyphenylacetic acids; gum; vitamins; and others. The drug Taraxaci radix cum herba contains: sesquiterpenlactones, triterpenes, Proc. Int. Conf. on MAP Eds. J. Bernáth et al. Acta Hort. 597, ISHS 2003 58 phytosterols, carbohydrates, phenolic acids, flavonoids, etc (Bissett 1994; Bradley 1992; Newall et al, 1996; Williams et al., 1996). From their characteristic principles our attention has been directed to triterpenes and phytosterols with anti-inflammatory activity (Kashiwada et al., 1998; Safayhi and Sailer, 1997). The yield of active substances during supercritical fluid (SFE) and Soxhlet extraction were compared how to produce active materials under soft conditions. MATERIALS AND METHODS All reagents were of reagent quality (Reanal RT, Hungary; Linde RT, Hungary). Silica gel TLC plates were from Merck KGaA (Darmstadt, Germany). The dandelion root and leaves were from Rózsahegyi Kft (Erdőkeresztes, Hungary). A herbarium specimen is deposited in the Department of Pharmacognosy, Semmelweis University. Soxhlet extraction. The dry drug was humected with ethanol (or n-hexane) at room temperature for 30 min before extraction. The extraction went until exhaution. The drug weight was 1200 g and solvent volume 2650 ml. The solvent flow rate (7 kg/h) was controlled with steam pressure adjusted by a valve and the extractor temperature (4045C) was controlled by a thermostat. During the extraction the solvent flow rate, the vessel temperatures, and the dry content of the extracts were measured every half hour. SFE process. One kg raw material was used for extraction. The CO2 mass flow was kept to 7 kg/h. The extraction was considered finished when the deposit between two checks was less than 0.1 w/w % calculated on drug. Assuming that the yield is not a linear function of the extraction temperature and pressure, we set the parameters to three different levels to be able to fit a non-linear function of the yield. In the middle of the design (300 bar, 50 °C) we completed 4 experiments to determine the standard deviation. Phytoanalytical methods. Measurements of the triterpene and phytosterol content of the extracts were carried out. Saponification was made according to our earlier reports (Kristó et al., 2000). Thin-layer chromatography densitometry. Standards of β-amyrin (0.05 g/mL) and β-sitosterol (0.025 g/mL), samples of SFE products and Soxhlet extracts were separated using n-hexane-ethyl acetate (6:2 w/w) as mobile phase on silica gel layers (Kieselgel 60 F254 MERCK 20 x 10 cm). Cerium sulphate reagent in acidic medium was selected as detection reagent. After heating for 10 min at 100C, measurements were performed at 600 nm, in the zig-zag mode (y = 0.2 mm, 0.2 x 1.2 mm) using a Shimadzu CS-930 densitometer. The calculation was based on calibration graphs. RESULTS AND DISCUSSION To evaluate the efficacy of various extraction techniques for preparing triterpene and phytosterol rich extracts from dandelion, β-amyrin and β-sitosterol were used as key compounds. Their separation was excellent (Rf values of β-amyrin and β-sitosterol were 0.58 and 0.43 respectively) in the saponified extracts. The calibration graphs used for densitometric analysis were in the range of 2.5 5.5 ng/mL, where r was always greater than 0.9777 (Fig. 1). When the quantities of triterpenes and phytosterols were measured three times in 10 different root and leaf samples we could conclude that Taraxaci folium is a better source for triterpenes while Taraxaci radix is superior in phytosterol content (Table 1). From Taraxaci radix the most β-amyrin was extracted by the Soxhlet method, using ethanol as solvent (889 mg/ 100 g), followed by supercritical CO2: 450 bar, 35C (424 mg/ 100 g) and the Soxhlet method using n-hexane (215 mg/ 100 g). The same data for Taraxaci folium were: 677 mg/ 100 g (Soxhlet method, ethanol), 446 mg/ 100 g (supercritical CO2: 450 bar, 65C), 496 mg/ 100 g (Soxhlet method, n-hexane) (Fig. 2). A possible answer for the greater quantity extracted by ethanol could be that triterpenes may be present in the roots in glycosidic form and therefore ethanol is a better solvent to extract them than the apolar supercritical CO2. N-hexane extracted almost the same amount as the supercritical CO2, because they have almost the same polarity. However, the SFE is superior to n-hexane extraction due to its better selectivity.