含鼠李糖之黃酮醇苷在大鼠之代謝研究

Abstract

含鼠李糖之黃酮醇苷廣佈於植物界，其中芸香苷已被廣泛報導具多種生物活性，但這類化合物在體內之代謝仍鮮有報告。研究顯示薑黃(Curcuma longa L.)地上部及守宮木(Sauropus androgynus (L.) Merr.)具有這類成分，守宮木的新鮮汁液曾被使用於減重，曾於1995年台灣造成23例之阻塞性細支氣管炎(bronchiolitis obliterans)，據研究其成因可能與這類成分有關，因此本研究擬探討其體內吸收與代謝物，以供釐清這些副作用機制之用。 本實驗從薑黃地上部正丁醇可溶部分，藉已建立之分離方法，使用凝膠管柱層析及逆相管柱層析，得到五個各近200 mg的含鼠李糖槲皮素苷：quercetin 3-O- α-L-rhamnosyl-(1→2)-[α-L-rhamnosyl-(1→6)]-β-D-galactoside (2)、quercetin 3-O- [α-L-rhamnosyl-(1→2)]-β-D-galactoside (3)、quercetin 3-O-[α-L-rhamnosyl-(1→2)]-α-L-rhamnoside (7)、quercetin 3-O-α-L-rhamnoside (8)、quercetin 3-O-[α-L-rhamnosyl- (1→2)]-β-D-glucuronide (11)。這五個成分連同地上部水煎萃取物及正丁醇可溶部分，分別進行大鼠胃內灌食，以代謝籠分別收集48小時之糞便與尿液，經液相-液相分配、凝膠管柱及逆相管柱層析集中代謝物，並利用液相層析-質譜儀(LC-MS)及液相層析-固相萃取-微樣細管轉移-核磁共振儀(LC-SPE-TT-NMR)技術，與未餵食樣品之控制組糞便及尿液比對以鑑定代謝物。液相層析分析發現餵食2 及7之糞便含有較多代謝物，與文獻報導槲皮素主要經肝膽排除相符。本實驗於2組糞便樣品中鑑定4個代謝物(M1–M4)、7組尿液樣品中鑑定2個代謝物(M5及M6)及糞便樣品中鑑定6個代謝物(M7–M12)。其中，以LC-SPE-TT-NMR共鑑定10個代謝物(M1–M10)，以LC-MS推定2個代謝物(M11–M12)。 三醣苷2組糞便樣品中鑑定之醣苷鍵部分水解代謝物(M1及M4)，顯示其可能經腸道菌叢去除醣基，然而M2及M3為黃酮(flavones)應非源自黃酮醇苷2，疑來自飼料成分。雙醣苷7組尿液樣品中發現疑內生性之脂肪酸類代謝物M5及M6，而於糞便樣品分離原形藥(7) 8.4 mg (回復率為4.8%)，並發現小分子芳香性代謝物M7–M9，後者顯示黃酮醇苷可能經phase I之環斷裂(ring fission)代謝。此外，於7組糞便樣品中鑑定到的M10，屬葡萄糖接合之代謝物，此代謝路徑僅有少數報導，且該葡萄糖基接在Rha 2-OH而非酚基此現象尚屬首見。另又以LC-MS發現與黃酮醇有紫外光吸收相符之代謝物M11及M12，經與原形藥7的分子離子峰比對，推測M11屬葡萄醣醛酸接合代謝物，而M12則為去酚基及甲基化代謝物。 動物實驗期間觀察到一大鼠呼吸有聲且體重有減輕之現象，此與守宮木造成之副作用相似，唯仍需進一步進行病理檢驗確認。

Rhamnosyl containing flavonol glycosides are widely distributed in the plant kingdom. Among them, rutin has been comprehensively studied including versatile bioactivities. The metabolic study, especially the metabolite identification of such group, however, is rarely reported. The aerial part of Curcuma longa L. and Sauropus androgynus (L.) Merr. have been reported to be rich in such flavonols rhamnoside. Ground juice from fresh S. androgynus was used as a diet for weight loss. In Taiwan, outbreak of 23-case bronchiolitis obliterans in 1995 after chronic digest of the juice and later on research indicated that flavonol rhamnosides might cause the pulmonary toxicity. Hence, this study was aimed to investigate the metabolites of this type compounds for further clarification of their potential side effect. Following the developed procedure, five quercetin rhamnosides with amount around 200 mg were isolated from the n-BuOH-soluble fraction of C. longa. These five compounds are quercetin 3-O– α-L-rhamnosyl-(1→2)-[α-L-rhamnosyl-(1→6)]-β-D- galactoside (2), -[α-L-rhamnosyl-(1→2)]-β-D-galactoside (3), -[α-L-rhamnosyl-(1→2)]-α-L-rhamnoside (7), -α-L-rhamnoside (8), and -[α-L-rhamnosyl-(1→2)]-β-D- glucuronide (11). Along with the water extract and n-BuOH-soluble fraction, they were administered individually to four rats in metabolic cages via gavage, and both urine and feces were collected with a period of 48 hours. Metabolites in both samples were focused by liquid-liquid partitioning, gel filtration, and reverse-phase chromatography in sequence. The metabolites were picked up via comparison of the HPLC profile of the treated and the control group. Hyphenation techniques, including LC-MS and LC-SPE-TT-NMR, were applied for metabolite identification. HPLC analysis indicated the feces samples, obtained from the 2 and 7 groups, to contain more metabolites in accordance with the previous reports that quercetin rhamnoside excreted predominantly through biliary duct. Metabolites M1–M4 were identified from the feces sample of the 2-treated group. Two (M5 & M6) and six metabolites (M7–M12) were identified from urine and feces sample of the 7-treated group, respectively. Of those, M1–M10 were identified by LC-SPE-TT-NMR and M11 & M12 by LC-MS. In the trioside 2-treated feces sample, partial hydrolysis of the glycosidic bond led to the corresponding metabolite M1 and M4, indicating that gut microflora might involve in deglycosylation. The identified flavone metabolites M2 and M3 are apparently not derived from 2 but from the fodder. In the dioside 7-treated urines, two endogenous fatty acid metabolites, M5 and M6, were identified. In 7-treated feces, however, the parent compound 7, up to 4.8% recovery, and simple aromatic metabolites, M7–M9, were identified. The latter three indicates that phase I biotransformation like ring fission may play a role in the metabolism of flavonol glycosides. Glucose conjugated metabolite M10 was identified from the 7-treated feces sample. Such conjugation was rarely reported. In addition, this is the first report that such conjugation occurs at Rha 2-OH rather than the phenolic group. Analysis of the MS of the LC-peaks with UV absorption consistent with flavonols led to the identification of M11 and M12. By comparing the m/z values with the parent 7, M11 was elucidated as a glucuronide conjugated 7 while M12 as a mono-dephenolated and mono-O-methylated 7. During the animal experiment of the 7 group, a rat suffering wheezes and body weight decrease, similar to those happened in S. androgynous incidence, was observed. Further pathological examination is required to clarify whether rhamnosyl containing flavonol glycosides could cause these symptoms.