J. Chil. Chem. Soc., 58, Nº 4 (2013) 2089 ARISTOLOCHIC ACIDS IN THE ROOTS OF ARISTOLOCHIA CHILENSIS, A DANGEROUS CHILEAN MEDICINAL PLANT ALEJANDRO URZÚA* 1 , ANGEL OLGUÍN, ROCÍO SANTANDER 1 Laboratory of Chemical Ecology, Department of Environmental Sciences, Faculty of Chemistry and Biology, University of Santiago de Chile, Av. Libertador Bernardo O’Higgins 3363, Santiago, Chile (Received: July 30, 2013 - Accepted: November 21, 2013) ABSTRACT The chemical composition of aristolochic acids (AAs) from the roots of A. chilensis was determined by high-performance liquid chromatography with diode- array detection (HPLC-DAD), a technique widely used for the detection and quantiication of AAs in herbal medicines. The roots contained a mixture of AA-I (1), AA-II (2), AA-III (3), AA-IV (4), AA-Ia (5), AA-IIIa (6), AA-IVa (7), and aristoloside (8), indicating that A. chilensis is not suitable for use as a medicinal plant due to the harmful effects of the aristolochic acids. INTRODUCTION Species of the genus Aristolochia (Aristolochiaceae) have been used in folk medicine throughout the world to treat various diseases 1 . Aristolochia can be characterized by their levels of aristolochic acids (AAs), a group of 10-nitrophenanthrene-1-carboxylic acids that normally include a 3,4-methylenedioxy moiety and substitutions at C-6 and/or C-8 with –OCH 3 and –OH groups; AAs with the former substitution are known as phenolic AAs. Structures with other types of substituents have also been identiied, including N-glycosides and O-glycosides of phenolic AAs. 2-4 AA-I (1) and AA-II (2) are powerful carcinogens in mice, rats and humans. Studies have shown that these AAs are genotoxic, mutagenic, and nephrotoxic. 5-10 The mechanism underlying the toxicity of AAs AA-I (1) and AA-II (2) involves a reduction of the nitro group (which is catalyzed by an enzyme) to the very reactive cyclic nitrenium ion, which then binds covalently to DNA and/or proteins. DNA adducts with the reduction products of AAs have been found in the kidney and ureter tissues of rats and humans that had consumed AAs. 5,6,10,11 In a recent review of the ethnopharmacological properties of this group, 99 species of Aristolochia were included 1 , but the authors obtained phytochemical data for only 24 of these species; for the other species, the part of the plant that was studied was not related to the part used in ethnomedicine. The authors of the review indicated the importance of obtaining new ethnopharmacological, phytochemical, and epidemiological data for Aristolochia used in folk medicine, particularly in Continents where such information is scarce. The roots of Aristolochia chilensis Bridges ex Lindl. were used in Central Chile during the XIX and XX centuries as an anti-hemorrhagic agent and to expel the residual placenta after childbirth. 12 The plant was consumed in two ways: as an infusion of approximately 2 g of powdered roots in a cup of water and as a decoction by boiling 30-40 g in 250 ml of water. 13 Although the risk of using A. chilensis in folk medicine is evident, the plant continues to be utilized and even promoted on the Internet (http://www. losmedicamentos.net/planta/oreja-de-zorro-hierba-de-la-virgen-maria-clon). This promotion may occur, at least in part, because the content and composition of AAs from the roots of A. chilensis have not been studied. In this study, high-performance liquid chromatography with diode- array detection (HPLC-DAD), a technique widely used for the detection and quantiication of AAs in herbal medicine 14-24 was used to determine the chemical composition of AAs from the roots of A. chilensis. The indings reveal that the roots contain a signiicant amount of AAs, making it dangerous to use A. chilensis as a medicinal plant because of the harmful effects of AAs. MATERIALS AND METHODS Plant material Representative samples of the roots of A chilensis Bridges ex Lindl. were collected during the lowering season at Cuesta Lo Prado (15 km west of Santiago, 33º 28’ S, 70º 56’ W, 750 m above sea level) in September 2012. Voucher specimens (SGO-152461) were deposited in the Herbarium of the National Natural History Museum in Santiago, Chile. Extraction of AAs Oven-dried and powdered roots of A. chilensis (40 g) were extracted with light petroleum ether (35-65º) in a Soxhlet apparatus, and the defatted plant material was then extracted with MeOH. The methanolic extract was evaporated in vacuo. The syrupy residue was agitated with 100 mL of 3% NaHCO 3 for 6 h, allowed to stand for 24 h at 10 ºC, and iltered. The clear iltrate was washed with CHCl 3 (5 x 50 mL). Upon evaporation, the washings with CHCl 3 yielded a brown gum that contained no acids and was not further investigated. The aqueous phase was adjusted to pH 2 with HCl and extracted with CHCl 3 (6 x 50 mL). Evaporation of the combined extracts in vacuo yielded a fraction of crude non-phenolic AAs (43.4 mg) 22 . The acid solution was then extracted with AcOEt (6 x 50 mL). Evaporation of the combined extracts in vacuo yielded a second fraction of crude phenolic AAs (103 mg). The fraction of crude phenolic AAs was re-suspended in 10 mL of 3% NaHCO 3 and iltered. The clear iltrate was adjusted to pH 2 with HCl and extracted with AcOEt (6 x 10 mL). Evaporation of the combined extracts in vacuo yielded a puriied fraction of phenolic AAs (45.5 mg) 22 . This procedure was repeated using ive samples (40 g) of A.chilensis roots obtained from ive different plants 22 . The fractions of both the non-phenolic and phenolic AAs were subjected to preparative TLC on pre-coated plates of silica gel 60 F254 Merck (1.0 mm thickness, 20 x 20 cm) in CHCl 3 -MeOH (95:5). Mixtures of AAs were detected under UV irradiation at 365 nm and eluted from the plates with CHCl 3 -MeOH (70:30). The fractions obtained from each band were then analyzed by HPLC- DAD. HPLC-DAD analysis of AAs The AA fractions in MeOH were directly injected (20 µl) into an analytical HPLC (Waters 600) with a reverse-phase Symmetry column (5 μm particle size; 25 x 0.46 cm). Gradient elution was performed using a mobile phase of 0.1% acetic acid in water (solution A) and 0.1% acetic acid in acetonitrile (solution B) in the following manner: 0-5 min, isocratic elution with 70% A / 30% B; 5-45 min, linear gradient from 70% A / 30% B to 55% A / 45% B. A Waters 2996 diode-array-detector (DAD) was used for detection, and spectra were recorded at wavelengths between 200 and 800 nm 22,24 . The AA fractions from the ive collected samples (40 g each), each of which was obtained from a different plant, were analyzed independently. Acid hydrolysis of compound 8. Compound 8 (4 mg) was dissolved in 2 mL of MeOH and 2 mL of 10% HCl. The mixture was stirred at room temperature for 5 hr; the solvent was then evaporated, and MeOH was added and evaporated several times until the HCl was eliminated. The remaining residue was diluted with H 2 O (2 mL) and extracted with AcOEt; the AcOEt extract yielded AA-IVa (7). The H 2 O layer was evaporated to dryness, yielding a solid (1.0 mg). Final puriication was performed via preparative PC using the system n-BuOH-EtOH-H 2 O (4:9:1), yielding pure D-(+)-glucose. e-mail: alejandro.urzua@usach.cl