differentiation of herbs linked to “chinese herb nephropathy” from the liquid chromatographic...

Download Differentiation of herbs linked to “Chinese herb nephropathy” from the liquid chromatographic determination of aristolochic acids

If you can't read please download the document

Upload: wan-chan

Post on 07-Aug-2016

220 views

Category:

Documents


6 download

TRANSCRIPT

  • Analytica Chimica Acta 576 (2006) 112116

    Differentiation of herbs linked to Chineseionm C, Kow

    bruary2006

    Abstract

    A HPLC and Aextracted by %. Exeluting unde acidand 7.29 ng traditdifferentiatio causan example urity 2006 Else

    Keywords: A

    1. Introdu

    Aristolochic acids, methylenedioxy-8-methoxy-10-nitro-phenanthrene-1-carboxylic acid (AA-I) and its demethoxylatedderivative (AA-II) (Fig. 1), are derived mainly from Aristolochiaspecies [1some plantplants havesnake bitesetc. The Phthat they caand promo

    Recentistration offibrosis [7,8in lymphoithe prolongbeen consitheir metabdetected inurinary bla

    CorresponE-mail ad

    ingabout 100 cases of renal disease were reported because ofAAs-containing slimming drugs, some of them died and morethan half require renal replacement. DNAAAs adducts weredetected in the kidney and ureter of the patients suffered from

    0003-2670/$doi:10.1016/j4]. Low levels of AAs have also been reported ins from the genus Asarum [4,5]. The AAs-containinglong been used in herbal medicines to treat tumours,, obstetrics, rheumatism, small pox and pneumonia,armacopoeia of the Peoples Republic of China statesn be used to remove heat, relieve pain, dispel wind

    te diuresis [6].publications, however, have pointed out that admin-

    AAs-containing drug may cause renal interstitial] and carcinogen in rats [9,10]. Tumours were found

    d organs, lung, forestomach and uterus of mice afterdosing of AAs [9]. The carcinogenicity of AAs has

    dered to be caused from DNA binding of AAs afterolic activation [1013]. The DNAAAs adducts wereforestomach, glandular stomach, liver, kidney and

    dder of rats [10,13].

    ding author. Tel.: +852 34117070; fax: +852 34117348.dress: [email protected] (Z. Cai).

    the disease [14,15]. The renal disorder caused has been termed asaristolochic acid nephropathy (AAN) or Chinese herb nephropa-thy (CHN) [16]. The AAN has also been reported in France,United Kingdom, United States, Japan and China.

    The use of Aristolochia species in herbal medicines is cur-rently no longer permitted in the United States, Canada, Aus-tralia, as well as most European and Asian countries. Further-more, US-FDA has advised customers to stop the use of anyherbal products-containing AAs in 2001. However, it has beenfound that AAs-containing herbs were mistakenly used becausemany herbs share similar Chinese folk name and some of themhave very similar appearance. For example, Stephania tetran-dra (Hanfangji or Fenfangji) had been inadvertently substitutedwith Aristolochia fangchi (Guangfangi), because they are bothnamed as Fangji in Chinese. Aristolochia manshuriensis (Guan-mutong), Clematis species (Chuanmutong) and Akebia species(Baimutong) are all Mutong in Chinese, and they have beenmisused in some traditional Chinese medicines. Therefore, it isvery important to have a readily available analytical method fordetecting AAs and for rapidly differentiating the herbs.

    see front matter 2006 Elsevier B.V. All rights reserved..aca.2006.03.008liquid chromatographic determinatWan Chan, Ka Man Hui, Wing Tat Poon, Ki

    Department of Chemistry, Hong Kong Baptist UniversityReceived 28 October 2005; received in revised form 18 Fe

    Available online 10 March

    method was developed and applied to analyze aristolochic acids (AA-Iusing ultrasonication with the extraction efficiency of better than 82

    r a gradient program using methanol and water-containing 0.5% aceticfor AA-II per injection was successfully applied for the analysis ofn of Chinese medicinal herbs that have previously been misused and

    that analysis of selected component markers could serve for health secvier B.V. All rights reserved.

    ristolochic acids; Chinese herbal medicine; HPLC; Chinese herb nephropathy

    ction Durherb nephropathy from theof aristolochic acids

    hung Lee, Zongwei Cai loon, Hong Kong SAR, China

    2006; accepted 2 March 2006

    A-II) in Chinese medicinal herbs. The herb samples weretracts were then filtered and injected onto a C18 column

    . The method with the detection limits of 1.33 ng for AA-Iional Chinese medicine (TCM) and related products anded toxicological effects. The developed protocol providedand quality control of TCM consumption.

    a slimming regimen in Belgium in the early 1990s,

  • W. Chan et al. / Analytica Chimica Acta 576 (2006) 112116 113

    Fig. 1. Chemical structures of AA-I and AA-II.

    HPLC with UV detection has been used for the quantitativeanalysis of AAs in herbal products [1,3,4,17]. Other analyti-cal approaches include cyclic voltammetry [18] and LC/MS[5,17]. The use of soxhlet extraction and cyclic voltammetryprovided a detection limit of 1 108 M [18]. However, thesoxhlet extraction is solvent and time-consuming. LC/MS anal-ysis using trifluoroacetic acid as the mobile phase modifier gavean on-column detection limit of 12.5 ng of AAI in methano-lic extract [5]. Sample extracted by boiling methanol followedby sample clean-up using SPE give similar detection limit of2 ng for both LC/UV and LC/MS detections [17]. This paperdescribes a sensitive LC/UV method for the quantization of AAsin Chinese herbal and the use of chromatographic fingerprint forthe differentiation of the easily confusing herbs.

    2. Experim

    2.1. Chem

    Standarpurchasedfrom Panrepurchased

    from a Milli-Q ultra-pure water system with the water outletoperating at 18.2 M (Millipore, Billerica, USA).

    2.2. Herb materials

    Reference herbs of four Aristolochia species were purchasedfrom Chinese National Institute for the Control of Pharma-ceutical and Biological Products. Other medicinal herbs andproprietary Chinese medicines were purchased from local herbalstores in March 2004 before the banning of AAs-containingproducts by the Department of Health, Hong Kong. The namesof the targeted herbs are listed in Table 1.

    2.3. Standard solutions

    Stock solution at 500g/ml of total AAs was prepared bydissolving 2.5 mg of the standard mixture in 5 ml methanol, giv-ing 190 and 290g/ml of AA-I and AA-II, respectively. Thestock solution was diluted 10-fold with methanol to give AA-Iconcentration at 19g/ml and AA-II at 29g/ml. With differentspiked levels of AAs in a blank herb Mutong, five-point calibra-tion curves were established with injection amount ranging from0 to 570 ng for AA-I and from 0 to 870 ng for AA-II.

    2.4. Sample preparation

    bal sr andg apsul

    y 1 gof mion o

    Table 1Levels of aris roduc

    Chinese name

    Guanmutong ensisGuangfangjiMadoulingQingmuxiangXungufengXixinHuaitongb nchTianxiantengFangjidMutongNanmuxiang chZhushagenJinerhuanJinguolanLong Dan XieQingshen XiaBi Ying Ching

    a Not detecb This herbc This herbd The identie Chinese mental

    icals

    d mixture-containing 38% AA-I and 58% AA-II wasfrom Sigma (St. Louis, MO, USA). Acetic acid wasac (Barcelona, Spain). HPLC-grade methanol was

    from Lab-scan (Dublin, Ireland). Water was obtained

    Hergrindeby usinuse. Caimatel10 mlmentat

    tolochic acids (g/g, dry weight) determined in herbs and Chinese medicinal pPharmaceutical name

    Caulis Aristolochiae ManshuriRadix Aristolochiae FangchiFructus AristolochiaeRadix AristolochiaeAristolochia mollissima HanceHerba AsariAristolochia moupinensis Fra

    c

    Herba AristolochiaeRadix Stephania tetrandraCaulis Clematidis ArmandiAristolochia yunnanensis FranAristolochia versicolorHerba Asarl Inslgnls

    Aristolochia tuberosa

    Gan Wan (Capsule)e opang Pille (Capsule)e

    ted.was identified and confirmed as Guanmutong.was identified and confirmed as Qingmuxiang.ty of this herb was not known.edicinal products.amples were ground into fine powder by a high-speedsieved through a 0.6 mm sieve. Pills were grounded

    mortar and pestle. They were stored at 20 C untiles were decapsulated immediately prior use. Approx-of powder was accurately weighed and extracted withethanol in an ultrasonic bath for 30 min. After sedi-f the solid material, the clear suspension was filtered

    ts

    AA-I AA-II

    1985.0 458.6838.1 58.8484.7 55.4820.6 122.2139.2 NDa17.2 ND787.0 107.3260.7 37.020.3 NDND NDND NDND NDND NDND ND150.9 ND0.17 0.13ND ND

  • 114 W. Chan et al. / Analytica Chimica Acta 576 (2006) 112116

    through a 0.2m PTFE filter disk. The filtrate was concentratedto 1 ml before injected into the HPLC system described below.

    2.5. HPLC

    HPLC aHPLC systor, a binaMA, USA)ume was 3(4.6 mm i.dscanned rawhile quan254 nm.

    The HPmethanol (dient progrof 1 ml/miwere 15.6 a

    3. Results

    3.1. HPLC

    AA-I anionized whfor the revmized acidand organiby using thof AA-I anchromatogple (Guanmshown in F

    The HPmatographUV absorpode arrayat 224.7, 2250.7, 299the UV abthose of AA

    The calithan 0.9995coefficienty = 6186.4xmated fromdetection linjection, rknown amoThe recovewere compthe relative(1.16g/g)At high spibetter thanparable rec

    Chromatograms of AAs standard (A), Guanmutong (B) and Long DanWan Capsule (C).

    ion [3,4], while having advantages consuming less timeganic solvent.trix spiked samples were prepared by spiking standarde into 1 g of AAs-free Mutong. The spiked samples wereed and analyzed in the same way as described for the reals such that each injection contained 19.0 ng AA-I andAA-II. Average detected levels (n = 7) were 16.2 ng for

    nd 28.3 ng for AA-II, given the relative analytical error ord accuracy of better than 15%. Relative standard deviation.) representing the method precision was 6.2% and 4.7%-I and AA-II, respectively.developed method was used to analyze AAs in several

    e herbs and medicinal products. The analytical results-containing products are listed in Table 1. Among thed herbs, Guanmutong (Caulis AristolochiaeManshurien-uangfangji (Radix Aristolochiae Fangchi), Madoulingus Aristolochiae) and Qingmuxiang (Radix Aristolochiae)ound to contain both AA-I and AA-II, while only AA-detected in Xungufeng (Aristolochia mollissima Hance)xin (Herba Asari). Guanmutong was found to have thet levels of total AAs at 2.44 mg/g (dry weight). In agree-with previously reported results [14], the concentra-f AA-I was much higher than that of AA-II in all thes analyzed. Low levels of AA-I and AA-II were alsod in a slimming product Qingshen Xiaopang Pills atconditions

    nalysis was carried out on a Waters Alliance 2695tem equipped with a 996 photodiode array detec-ry gradient pump and a column oven (Milford,. The column oven was set at 30 C. Injected vol-0l, and was made onto an Agilent C18 column. 150 mm, 5m particle size). UV absorption wasnging from 200 to 450 nm for qualitative analysis,titation was performed with single wavelength at

    LC mobile phases consisted of aqueous (A) andB), each containing 0.5% acetic acid. A solvent gra-am from 40% B and to 70% in 10 min at a flow raten was used. The retention time of AA-I and AA-IInd 13.8 min, respectively.

    and discussion

    analysis of the AAs

    d AA-II contain moiety of carboxylic acid. They areen alkaline buffer is used, resulting in peak tailingersed-phase HPLC analysis. In this study, an opti-ic buffer with 0.5% of acetic acid in both aqueousc phase was used. The analytes were well separatede solvent gradient program, with the retention timesd AA-II at 15.6 and 13.8 min, respectively. Typicalrams of AA standards, an AA reference herbal sam-utong) and an AA-containing capsule medicine are

    ig. 2.LC detection of AA-I and AA-II based on the chro-ic retention time was confirmed by using diagnostiction spectral patterns obtained from the photodi-

    detection. The maximum absorption was observed49.5, 320.6 and 393.4 nm for AA-I as well as 217.7,.2 and 357.6 nm for AA-II. During sample analysis,sorbance of the targeted peaks was compared with

    standards for confirmation.bration curves gave correlation coefficients of betterfor both AA-I and AA-II, with the linear regression

    for AA-I and AA-II being y = 4623.0x 10172.4 and 45470.6, respectively. Detection limits were esti-the intercept of the calibration curve [19,20]. The

    imits of AA-I and AA-II were 1.3 and 7.3 ng perespectively. Recovery test was performed by spikingunt of AA-I and AA-II into 1 g of AAs-free Mutong.ry samples were analyzed and the AAs peak areasared with those obtained from the standard runs. Atly low concentrations of AA-I (0.76g/g) and AA-II, the recoveries were 113% and 82%, respectively.ked levels of AAs (500g/g), the recovery was also85%. The sonication extraction method yield com-overy to other AA extraction methods such as liquid

    Fig. 2.Xie Gan

    extractand or

    Mamixturextractsample29.0 ngAA-I ametho(R.S.Dfor AA

    TheChinesof AAtargetesis), G(Fructwere fI wasand Xihighesmenttion osampledetecte

  • W. Chan et al. / Analytica Chimica Acta 576 (2006) 112116 115

    0.17 and 0.13g/g, respectively. The Chinese medicinal prod-uct Long Dan Xie Gan Wan Capsule used as health supplementof fatty livtain AA-Iin herbs M(Aristolochversicolor)(AristolochQing LongLong CapsAAs analytain Xixin (the concenweight) innot detectesule and Blimits.

    3.2. Differcomparison

    Chinesebetween 19accidentalfangchi [7,in France,Although ttries, it hasin herbal millegal AAbecause mahave the sbelong tofangchi (Gdra (HanfaFangji in Cpeople [8].tis speciesall Mutongtraditionalwere purchance, whicincidents.

    The chrdifferentiatthe developfrom localherb storethe purchather investilar HPLCthe determthan thosesequent inthat Guanminal prescrthe similar

    HPLC chromatogram comparison of Guanmutong (A), herbal slicesed as Huaitong (B), Qingmuxiang (C) and herbal slices purchased asnteng (D).

    r study is needed to find out the reason for the differ-f AA-I and AA-II levels and their concentration ratiosn the misused Guanmutong (Huaitong) and referenceutong.

    m the comparison of HPLC chromatograms, we have alsothat Qingmuxiang (Radix Aristolochiae) (Fig. 3C) mayeen misused as Tianxianteng (Herba Aristolochiae)D). Tianxianteng purchased from the herb store wased and found to contain both AA-I and AA-II. How-

    carefully investigation of the obtained HPLC profilesfalse Tianxianteng and that of the reference Qingmuxi-ig. 3C and D) reveal very similar fingerprint patterns. Thed concentration ratio of AA-I to AA-II in Tianxiantengas also close to that in Qingmuxiang (6.7), although

    al AAs contents were different. The difference in AAts might be due to discrepancy of sources such as theand date of the herb production as well as the storageatment of herbs after their harvest. This again demon-the strong need of standardization and quality control ofer, cholecystitis and hepatitis B was found to con-at 150.9g/g (dry weight). No AAs were detectedutong (Caulis Clematidis Armandi), Nanmuxiangia yunnanensis Franch), Zhushagen (Aristolochia, Jinerhuan (Herba Asarl Inslgnls) and Jinguolania tuberosa) as well as the medicinal products XiaoCapsule and Bi Ying Ching Capsule. Xiao Qing

    ule and Bi Ying Ching Capsule were chosen forsis in this study because they were listed to con-Herba Asari) as one of their ingredients. Althoughtration of AA-I was determined at 17.2g/g (drythe individual analysis of reference Xixin, AAs wered in the medicinal products Xiao Qing Long Cap-i Ying Ching Capsule with the current detection

    entiation of confusing herbs through ngerprintand detection of AAs

    herb nephropathy (CHN) case reported in Belgium90 and 1992 was generally believed to relate to the

    replacement of Stephania tetrandra by Aristolochia8,11,14,15]. Similar cases have also been reportedUnited Kingdom, United States, Japan and China.he uses of AAs have been banned in many coun-been found that AAs are sometimes mistakenly usededicines and medicinal products, in addition to the

    s-adducted slimming drugs. The misuse may occurny herbs share similar Chinese folk name, i.e. they

    ame ending in their Chinese names although theydifferent plant families. For example, Aristolochiauangfangi) has been misused as Stephania tetran-ngji or Fenfangji) because they are both called ashinese, which has caused kidney disease for manyAristolochia manshuriensis (Guanmutong), Clema-

    (Chuanmutong) and Akebia species (Baimutong) arein Chinese, and they have been misused in some

    Chinese medicines. Furthermore, some herbs thatased as shredded slices possess very similar appear-h have resulted in herb misuse and thus medicinal

    omatographic fingerprint may serve for the rapidion of the easily confusing herbs. We have applieded analytical method for investigating some herbsstores. A Chinese herb sold as Huaitong in a

    was analyzed. The analytical results indicated thatsed Huaitong contained AAs (Fig. 3B). Our fur-gation showed that the false Huaitong has simi-profiles as that of Guanmutong (Fig. 3A), althoughined levels of AAs in Huaitong were much lowerdetermined in reference herb of Guanmutong. Con-vestigation by local medicinal authority indicatedutong has been misused as Huaitong for medic-

    iptions in this store and the reason of misuse wasChinese end name for Huaitong and Guanmutong.

    Fig. 3.purchasTianxia

    Furtheence o

    betweeGuanm

    Frofoundhave b(Fig. 3analyzever, a

    of theang (Fdetecte(7.0) wthe totcontenplaceand trestratesherbs.

  • 116 W. Chan et al. / Analytica Chimica Acta 576 (2006) 112116

    Another example of herb misuse was that AA-I was detectedwith concentration of 20.3g/g (dry weight) in a Fangji (RadixStephania tetrandra) herb purchased from one of the localherb stores. But it is known that Fangji contains no AAs[4,7,8,11,14,15], the identity of the sold Fangji that we havepurchased and analyzed was of highly suspicious. Although theidentity of this doubtful herb so far has not been revealed, ourAAs analysis has stopped the further sale of this AAs-containingFangji.

    4. Conclusion

    The developed HPLC method for the analysis of AAs inChinese herbs and medicines provided good sensitivity andreproducibility. AAs peak confirmation was achieved by thecomparison of retention time and UV spectra with those obtainedfrom AA standards. The method was applied to the AA analysisin some herbs and slimming medicines purchased from localherb stores. Guanmutong (Caulis Aristolochiae Manshuriensis)was found to have the highest AA concentrations, and the levelsof AA-II was founded to be much lower than that of AA-I. Theanalytical results have also revealed that some AA-containingherbs have been misused for Chinese medicinal prescription.Thus, the AAs analysis may provide an efficient approach forthe differenappearancemisuse.

    Acknowled

    The autfrom the Hthe ResearHong Kongthis study.

    References

    [1] E.S. Ong, S.O. Woo, Y.L. Yong, J. Chromatogr. A 904 (2000) 57.[2] E.S. Ong, S.O. Woo, Electrophoresis 22 (2001) 2236.[3] T.Y. Lee, M.L. Wu, J.F. Deng, D.F. Hwang, J. Chromatogr. B 766 (2002)

    169.[4] K. Hashimoto, M. Higuchi, B. Mankino, I. Sakakibara, M. Kubo,

    Y. Komatsu, M. Maruno, M. Okada, J. Ethnopharmacol. 64 (1999)185.

    [5] T.T. Jong, M.R. Lee, S.S. Hsiao, J.L. Hsai, T.S. Wu, S.T. Chiang, S.Q.Cai, J. Pharm. Biomed. Anal. 33 (2003) 831.

    [6] The State Pharmacopoeia Commission of P.R. China, Pharmacopoeia ofthe Peoples Republic of China, vol.1, Chemical Industry Press, Beijing,2000.

    [7] F.D. Debelle, J.L. Nortier, E.G. De Prez, C.H. Garbar, A.R. Vienne, I.J.Salmon, M.M. Deschodt-Lanckman, J.L. Vanherweghem, J. Am. Soc.,Nephrol. 13 (2002) 431.

    [8] J.L. Vanherweghem, M. Depierreux, C. Tielemans, D. Abramowicz, M.Dratwa, M. Jadoul, C. Richard, D. Vandervelde, D. Verbeelen, R. Van-haelenfastre, M. Vanhaelen, Lancet 341 (1993) 387.

    [9] U. Mengs, Arch. Toxicol. 61 (1988) 504.[10] W. Pfau, H.H. Schmeiser, M. Wiessler, Carcinogenesis 11 (1990) 313.[11] M. Stiborova, E. Frei, M. Wiessler, H.H. Schmeiser, Chem. Res. Toxicol.

    14 (2001) 1128.[12] H.H. Schmeiser, E. Frei, M. Wiessler, M. Stiborova, Carcinogenesis 18

    (1997) 1055.[13] M. Stiborova, E. Frei, B. Sopko, K. Sopkova, V. Markova, M. Lankova,

    T. Kumstyrova, M. Wiessler, H.H. Schmeiser, Carcinogenesis 24 (2003)1695.

    Stiborova, E. Frei, B. Sopko, M. Wiessler, H.H. Schmeiser, Carcino-esis. Bi. Sc

    . Cosy. Ios

    03) 2. Sunanal.. Sk

    s, 5th.H. Htiation of Chinese herbs that have similar name orand thus avoid the incident resulted from the herb

    gements

    hors would like to thank the Faculty Research Grantong Kong Baptist University (FRG04-05/II-35) andch Grant Council, University Grants Committee of

    (HKBU2154/04M) for their financial supports for

    [14] M.gen

    [15] C.AH.H

    [16] J.P[17] J.-R

    (20[18] Z.Y

    Bio[19] D.A

    ysi[20] S.S23 (2002) 617.eler, M. Stiborova, M. Wiessler, J.P. Cosyns, C.V. deStrihou,hmeiser, Carcinogenesis 18 (1997) 1063.ns, Drug Saf. 26 (2003) 33.

    et, G.E. Raoelison, K.H. Hostettmann, Food Chem. Toxicol. 419., L.F. Liu, X.F. Zheng, C.H. Fan, Q. Wang, G.X. Li, Anal.Chem. 378 (2004) 388.

    oog, E.J. Holler, T.A. Nieman, Principles of Instrumental Anal-ed., Harcourt Brace, Orlando, FL, 1998, p. 13.o, J.Z. Yu, Environ. Sci. Technol. 38 (2004) 862.

    Differentiation of herbs linked to "Chinese herb nephropathy" from the liquid chromatographic determination of aristolochic acidsIntroductionExperimentalChemicalsHerb materialsStandard solutionsSample preparationHPLC conditions

    Results and discussionHPLC analysis of the AAsDifferentiation of confusing herbs through fingerprint comparison and detection of AAs

    ConclusionAcknowledgementsReferences