Medicine, Science & Technology

A general method for the dereplication of flavonoid glycosides utilizing high performance liquid chromatography/mass spectrometric analysis

A general method for the dereplication of flavonoid glycosides utilizing high performance liquid chromatography/mass spectrometric analysis
of 5
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
    A General Method for the Dereplication of Flavonoid Glycosides Utilizing High PerformanceLiquid Chromatography/Mass SpectrometricAnalysis Howard L. Constant, 1 Karla Slowing, 2 James G. Graham, 1 John M. Pezzuto, 1 Geoffrey A. Cordell 1 and Christopher W. W. Beecher 1  1 Program for Collaborative Research in the Pharmaceutical Sciences, Department of Medicinal Chemistry and Pharmacognosy,College of Pharmacy, University of Illinois at Chicago, 833 S. Wood St., Chicago, IL 60612, USA 2 Departmento de Farmacologia, Facultad de Farmacia, Universidad Complutense de Madrid, Avda. Complutense s/n, 28040, Madrid,Spain Using high performance liquid chromatography–electrospray mass spectrometry a method was developed toseparate flavonoids and flavonoid glycosides and to obtain molecular weight information. Collision-induceddissociation was then used to obtain the molecular masses of the flavonoid aglycones. Thus, the mass of aflavonoid glycoside and its aglycone can now be determined in a mixture, and it is not necessary to isolate thecomponents. An aqueous extract of  Eugenia jambos L. was used to demonstrate the utility of the technique inanalysing extracts for flavonoid glycosides. © 1997 by John Wiley & Sons, Ltd. Phytochem. Anal. 8 , 176–180 (1997)No. of Figures: 7.No. of Tables: 1.No. of References: 28.) Keywords: collision-induced dissociation; dereplication; electrospray mass spectrometry; flavonoids; high performance liquidchromatography. INTRODUCTION Flavonoids and their glycosides are biologically activecompounds commonly found in plants, and constitute oneof the largest classes of natural products known. Theirchemotaxonomic distribution ranges from the simpleflavones in the primitive liverworts to the complex, highlyevolved rotenoids in the legumes (Harborne et al. , 1975;Harborne and Mabry, 1982; Harborne, 1988).These natural products display broad biological activity(Cody et al. , 1986, 1988). Examples include anti-oxidant(Gy¨orgy et al. , 1964), radical scavenging (Husain et al. ,1987; Robak and Gryglewski, 1987; Hatano et al. , 1988),and hypolipidemic activities (Sharma, 1979), and inhibitionof 5-lipoxygenases and cyclo-oxygenases (Takahama, 1985;Laughton et al. , 1991). Flavonoids also inhibit the oxidativemodification of low density lipoproteins (LDL) by macro-phages and damage caused by ultra violet (UV) irradiation(DeWhalley, 1990; N`egre-Salvayre and Salvayre, 1992).The oxidized form of LDL has been shown to be cytotoxicand has been implicated in the formation of atheroscleroticplaques (Steinberg et al. , 1989). The isoflavonoid genisteinhas also been found to induce cell differentiation and toinhibit angiogenesis (Constantinou et al. , 1990; Fotsis et al. ,1993). Some studies have found that flavonoids mediate acancer chemopreventive effect (Deschner et al. , 1991;Messina et al. , 1994).Although the flavonoids are biologically active and are of substantial current commercial and clinical interest, they areoften regarded as nuisance compounds since they arewidespread in nature. For this reason, it is important to beable to characterize the active elements of phenolic mixturesas quickly as possible. With this information, it can bedetermined if an active agent merits the resources requiredfor its isolation and structure determination.A variety of mass spectrometric techniques have beenused to analyse natural products, especially flavonoid andflavonoid glycosides (Wolfender et al. , 1992, 1994).Recently, one important technique that has been used withsuccess is thermospray ionization mass spectrometry (TSP-MS) (Schr¨oder and Merfort, 1991; Wolfender et al. , 1992;Slimestad and Hostettmann, 1996). Standard glycosideswere introduced by loop injection into the TSP interface andthe flavonoid glycosides and aglycones were monitored.Using TSP-MS,  Arnica montana and  A. chamissonis wereanalysed for adulteration with  Heterotheca inuloides whichcontains rutin in significant quantities (Schr¨oder andMerfort, 1991). Also, phenolic glycosides from Picea abies were analysed by TSP-MS. The chromatography afforded aseparation of the phenolic glycosides and the MS of theglycoside and aglycone, along with UV information, wereused to assess 32 phenolics in the needles of P. abies (Slimestad and Hostettmann, 1996). The present techniqueoffers a significant advance over these reported techniquesnot only because of increased stability and efficiency of ionization of the electrospray interface over the thermosprayinterface but also because it directly ties the biologicalactivity to both the molecular weight of the active flavonoidglycoside and its aglycone portion.The technique of rapidly characterizing the biologicallyactive constituents in a mixture to establish if they havebeen previously identified is known as dereplication (Suff-  Correspondence to: C. W. W. Beecher. (E-mail: grant sponsor: National Institute of Health: Contract grantnumber: P01 CA48112.CCC 0958–0344/97/040176–05 $17.50  Received 19 February 1996  © 1997 by John Wiley & Sons, Ltd.  Revised 15 July 1996  Accepted 27 August 1996  PHYTOCHEMICAL ANALYSIS, VOL. 8 , 176–180 (1997)  ness, 1987). We have previously described an efficientmethod for dereplicating moderately lipophilic extracts(Constant and Beecher, 1995). We presently describe ageneral method for the dereplication of flavonoids andflavonoid glycosides and the dereplication of an aqueousextract of  Eugenia jambos L. (Myrtaceae) using electro-spray mass spectrometry. EXPERIMENTAL Materials. Flavonoid standards were purchased fromIndofine chemicals (Somerville, NJ, USA) or from SigmaChemical Co. (St. Louis, MO, USA). Each standard wasprepared as a 1mg/mL stock solution. A mixture of 13flavonoids was prepared for optimization of the chromato-graphic conditions. Triethylamine and acetonitrile werepurchased from Fisher Scientific (Pittsburgh, PA, USA).Water was drawn from a Nanopure system. Chromatography. Chromatography was performed with aHewlett Packard (Wilmington, DE, USA) electrospraysystem comprised of a 1090 Series II L high performanceliquid chromatograph (HPLC) equipped with a photodiodearray detector, a 59987A electrospray, and a 5989B massspectrometer. The splitter used was a 1:50/1:125 streamsplitter from LCPackings (San Francisco, CA, USA), andthe postcolumn pump was an ABI MicroGradient pump.The HPLC separation was performed with a KromasilC-18 reverse phase column (250  3.2mm i.d.) with 5   mpacking material (Technikrom, Wilmette, IL, USA). Thegradient began at 20% acetonitrile in water and was held atthis concentration for the first 10min. This was followed bya linear gradient to 40% acetonitrile over the next 8min, alinear increase to 75% acetonitrile over the following10min and then a sharp transition to 100% acetonitrile overthe next 2min. The column was washed with 100%acetonitrile for 7min and was then returned to startingconditions and the column re-equilibrated for 10minkeeping the flow rate constant at 0.75mL/min. The diodearray detector (DAD) was set to record between 210 and450nm. The eluent was split 1:50 after the DAD, and thesmaller stream mixed with postcolumn solvent (metha-nol:water:triethylamine; 90:9.8:0.2) prior to introductioninto the mass spectrometer (MS). The larger stream wasrecovered in 96-well bioassay plates on a Gilson F-204series fraction collector (Middleton, WI, USA) and thentested for cyclo-oxygenase inhibition (Kulmacz and Lands,1987). Electrospray mass spectrometry. The protocol required twoseparate injections for each sample. For the first injection,the electrospray (ES) lenses and skimmers were adjusted todetect [M-H]  ions in a ‘normal’ negative ion mode. Forthe second injection, the MS was tuned so that ions wouldbe subjected to collision-induced dissociation (CID) condi-tions, specifically to cleave the sugar moieties from theglycosylated flavonoids and thus yield information concern-ing the aglycone (non-glycosylated) portion of themolecule. In order to obtain the necessary voltages for theskimmers and lenses for CID, naringin was infused into theESMS and the lenses and skimmers were adjusted to obtainmaximum abundance for the aglycone (naringenin). Themajor differences between the ‘normal’ tune file and theCID tune file were that the entrance lens voltage and thecapillary exit voltage were decreased ( V  ent =90,V cap =  200), and the first skimmer voltage was slightlyincreased (SKIM1=  41) for the CID tune file. Voltages forthe entrance lens were ramped (50 to 170) and V  cap andSKIM1 were set at  120 and  45, respectively. The effectof these changes was to increase the voltage differencebetween the capillary exit voltage and the first skimmervoltage. This increased the potential energy of the individ-ual ions and accelerated the ions into collision withmolecules of drying gas. The resulting fragments werefocused and measured by the spectrometer. In ‘normal’operation, such fragments are rarely seen. RESULTS AND DISCUSSION As can be seen in Fig. 1 and Table 1 the chromatographicsystem reported here separates a variety of flavonoids andtheir glycosides. This solvent system elutes the tanninstested within the first 5min (data not shown) whilstglycoside flavonoids were eluted over the next 10min, andnon-glycosylated flavonoids over the next 15min.Fig. 2A shows the total ion chromatograph (TIC)recorded by the MS in the normal mode, while Fig. 2Bshows the TIC in the CID mode. The glycoside ionabundances recorded in the CID mode are as large as thoserecorded in the normal mode, whereas the ion currents forthe non-glycosylated flavonoids are diminished in the CIDmode. The additional ions seen in the glycosidic flavonoidpeaks are the ions that are generated by CID.The spectra recorded in the CID mode of all theaglycones tested were similar, if not identical, to thoserecorded in the normal mode. However, the spectra Figure 1. HPLC separation of the flavonoid mixture. (For detailsof the HPLC system see Experimental section). Key to peaknumbering: 1  —rutin; 2  —naringin; 3  —daidzein; 4  —luteolin; 5  —quercetin; 6  —apigenin; 7  —genistein; 8  —naringenin; 9  —Hesperetin; 10  —formononetin; 11  —chyrsin; 12  —biochanin-A; 13  —tangeretin. Table1.Retention times (by HPLC-UV) of flavonoids used inthis study CompoundRetention TimenumberFlavonoid( R  t ; min) 1 Rutin4.44 2 Naringin11.05 3 Daidzein17.02 4 Luteolin18.15 5 Quercetin18.48 6,7,8 Apigenin/Genistein/Naringenin20.45 9 Hesperetin21.18 10 Formononetin22.70 11 Chrysin24.75 12 Biochanin A25.40 13 Tangeretin25.90 DEREPLICATION OF FLAVONOID GLYCOSIDES177 Phytochem. Anal. VOL. 8, 176–180 (1997)© 1997 by John Wiley & Sons, Ltd.  recorded in the CID Mode of the glycosylated flavonoidswere significantly different from those recorded in thenormal mode. In every glycosylated flavonoid tested, asignificant second ion was seen in the spectra recorded inthe CID mode. This ion was representative of the agly-cone.In Fig. 3 the spectra of naringin (MW 580) yielded anappropriate ion at m/z 579 in the normal (Fig. 3A) mode andshowed this ion and an additional ion at m/z 271 in the CIDmode (Fig. 3B). The aglycone from naringin is naringenin(MW 272) and would be expected to yield an ion at m/z 271.Another example is shown in Fig. 4: the compound rutin isquercetin-3-rutinoside (MW 610) and thus is expected toyield an ion at m/z 609 (Fig. 4A), whilst quercetin (theaglycone) would be predicted to show an ion at m/z 301. Inthis case, the ion is not at m/z 301 but rather at m/z 300 (Fig.4B) and could be due to formation of the quinone anion.This ion is diagnostic for quercetin glycosides and was alsoseen in the spectra of two other quercetin glycosides tested,as well as in the spectrum of quercetin itself (data notshown). All of the glycosides tested yielded an appropriateion for their aglycone, consequently, the method appears tobe of very general applicability.The aqueous extract of  Eugenia jambos L. (Myrtaceae)was analysed using this HPLC/MS protocol. Fig. 5A and Bshows the negative ion TIC and HPLC chromatogram of theaqueous extract of  E. jambos using the normal tune file, andFig. 5C shows the negative ion TIC adjusting conditions forglycoside cleavage (TIC-CID). There are two distinct areasof biological activity (cyclo-oxygenease inhibition). Thefirst area is between 5 and 7min and the second area isbetween 9.8 and 11min. Analysing the first area in the TIC(Fig. 6A) shows an ion at m/z 595 and the TIC-CID showsions at m/z 316 and 595, corresponding to the aglycone andthe glycoside, respectively. Both ions are found in the areaof 6.2min and follow the same chromatographic profile of each other and of the biological response (Fig. 6B). In thearea of 9.8 to 11min two quercetin-based glycosides yieldan ion at m/z 300 in addition to their parent ions at m/z 579and 447 (Fig. 7A and B). The three compounds identified bythe TIC and TIC-CID correspond to three compounds whichhave been isolated from  E. jambos namely myricetin-3- O -xylose-rhamnose, quercetin-3- O -xylose-rhamnose andquercetrin (Slowing et al. , 1994). The retention times forthese three compounds in the TIC are 6.2, 10.2 and 10.7minfor myricetin-3- O -xylose-rhamnose (MW 596), quercetin-3- O -xylose-rhamnose (MW 580), and quercetrin (MW448), respectively. These three compounds have also beenreported active in the cyclooxygenase inhibition assay Figure 2. TIC of the flavonoid mixture in the normal mode ( A )and in the CID mode ( B ). (Protocol for the HPLC analysis was thesame as for Fig. 1. For key to peak numbering see legend to Fig.1). Figure 3. Mass spectra of naringin in the normal mode ( A ) andin the CID mode ( B ). Figure 4. Mass spectra of rutin in the normal mode ( A ) and inthe CID Mode ( B ). H. L. CONSTANT  ET AL. 178© 1997 by John Wiley & Sons, Ltd.PHYTOCHEMICAL ANALYSIS, VOL. 8, 176–180 (1997)  (Slowing et al. , 1995).As noted in our earlier publication (Constant andBeecher, 1995), the basic approach to dereplication takenhere involves the HPLC separation of an active extract withthe concurrent monitoring of the UV/VIS spectra, electro-spray mass spectra and the biological activity of the effluentin an information gathering phase. This is followed by ananalytical phase in which these data are examined relative tothe published chemotaxonomic and pharmacological data inorder to ascertain the probability that an active agent is apreviously known compound. This method worked well aslong as the separation afforded by the HPLC was sufficient,but it was not diagnostic where there was less than optimalseparation.In these experiments we have gone beyond the simpleenhancement of the chromatography by using the techniqueof collision-induced dissociation to measure both the massof the glycoside and the mass signature of the aglycone.When these data are combined in the analytical stage withthe previously known aglycones from closely related plants,they provide a powerful tool for the dereplication of aparticular sample.Using the sample of  E. jambos , the method reported hereidentifies both the weight of the parent flavonoid glycosideand its aglycone, thereby significantly increasing thestrength of the identification. The above method is nowbeing used on a number of plant samples that have beenobserved to be active in a radical scavenging assay. Itshould be widely applicable for the determination of  Figure 5. Negative ion TIC ( A ) HPLC chromatogram ( B ) andnegative ion TIC using CID mode ( C ) of an aqueous extract of Eugenia jambos  . Figure 6. Mass spectrum in the normal mode ( A ) and theextracted ion chromatograms of selected ions ( m/z  316 and 595)in the cleavage mode ( B ) in the time range of 5 to 7 min of anaqueous extract of Eugenia jambos  . Figure 7. Mass spectrum in the normal mode ( A ) and theextracted ion chromatograms of selected ions ( m/z  300, 447 and579) in the cleavage mode ( B ) in the time range of 9.8 to 12minof an aqueous extract of Eugenia jambos  . DEREPLICATION OF FLAVONOID GLYCOSIDES179 Phytochem. Anal. VOL. 8, 176–180 (1997)© 1997 by John Wiley & Sons, Ltd.  flavonoids and their glycosides over a wide variety of bioassay conditions. Furthermore, the stability andefficiency of electrospray ionization means that the sensitiv-ity and applicability of this dereplication technique isextremely broad. Acknowledgements The work reported in this paper was supported, in part, by a grant from theNational Institute of Health (P01 CA48112). The authors would also like toacknowledge the technical and equipment support of Hewlett Packard Inc.(Wilmington, DE, USA). REFERENCES Cody, V., Middleton, E. Jr. and Harborne, J. B., eds. (1986). Plant Flavonoids in Biology and Medicine: Biochemical, Pharma- cological and Structure-Activity Relationships. Alan R. LissInc., New York.Cody, V., Middleton, E. Jr., Harborne, J. B. and Beretz, A., eds.(1988). Plant Flavonoids in Biology and Medicine II: Bio- chemical, Cellular, and Medicinal Properties. Alan R. LissInc., New York.Constant, H. L. and Beecher, C. W. W. (1995). A method for thedereplication of natural product extracts using electrosprayHPLC/MS. Nat. Prod. Lett. 6 , 193–196.Constantinou, A., Kiguchi, K. and Huberman, E. (1990). Induc-tion of differentiation and DNA strand breakage in humanHL-60 and K-562 leukemia cells by genistein. Cancer Res. 50 ,2618–2624.Deschner, E. E., Ruperto, J., Wong, G. and Newmark, H. L.(1991). Quercetin and rutin as inhibitors of azoxymethanol-induced colonic neoplasia. Carcinogenesis 12 , 1193–1196.DeWhalley, C. V., Rankin, S. M., Hoult, J. R. S., Jessup, W. andLeake, D. S. (1990). Flavonoids inhibit the oxidative mod-ification of low density lipoproteins by macrophages. Biochem. Pharmacol. 39 , 1743–1750.Fotsis, T., Pepper, M., Adlercreutz, H., Fleischmann, G., Hase, T.,Montesano, R. and Schweigerer, L. (1993). Genistein, adietary-derived inhibitor of in vitro  angiogenesis. Proc. Natl.Acad. Sci. USA 90 , 2690–2694.Gy¨orgy, P., Murata, K. and Ikehata, H. (1964). Antioxidantsisolated from fermented soybeans (Tempeh). Nature 203 ,870–872.Harborne, J. B., ed. (1988). The Flavonoids: Advances in Research Since 1980  . Chapman and Hall, New York.Harborne, J. B. and Mabry, T. J., eds. (1982). The Flavonoids: Advances in Research 1975–1981. Methuen, New York.Harborne, J. B., Mabry, T. J. and Mabry, H., eds. (1975). The Flavonoids. Academic Press, New York.Hatano, T., Kagawa, H., Yasuhara, T. and Okuda, T. (1988). Twonew flavonoids and other constituents in Licorice root: Theirrelative astringency and radical scavenging effects. Chem.Pharm. Bull. 36 , 2090–2097.Husain, S. R., Cillard, J. and Cillard, P. (1987). Hydroxyl radicalscavenging activity of flavonoids. Phytochemistry 26 ,2489–2491.Kulmacz, R. J. and Lands, E. M. (1987). Cyclooxygenase:measurement, purification and properties. In Prostaglan- dins and Related Substances: A Practical Approach. (Benedetto, T., McDonald-Gibson, R. G., Nigram, S. andSlater, T. F., eds.), pp. 209–227. IRL Press, Oxford.Laughton, M. J., Evans, P. J., Moroney, M. A., Hoult, J. R. S. andHalliwell, B. (1991). Inhibition of mammalian 5-lipoxygenaseand cyclo-oxygenase by flavonoids and phenolic dietaryadditives: Relationship to antioxidant activity and to iron ionreduction ability. Biochem. Pharmacol. 42 , 1673–1681.Messina, M. J., Persky, V., Setchell, K. D. R. and Barnes, S.(1994). Soy intake and cancer risk: A review of the in vitro  and in vivo  data. Nutr. Cancer 21 , 113–131.N`egre-Salvayre, A. and Salvayre, R. (1992). Quercetin preventsthe cytotoxicity of oxidized LDL on lymphoid cell lines. Free Radic. Biol. Med. 12 , 101–106.Robak, J. and Gryglewski, R. J. (1987). Flavonoids are scaven-gers of superoxide anions. Biochem. Pharmacol. 37 ,837–841.Schr¨oder, E. and Merfort, I. (1991). Thermospray liquid chroma-tography/mass spectrometric studies of flavonoidglycosides from Arnica montana  and Arnica chamissonis  extracts. Biol. Mass Spectrom. 20 , 11–20.Sharma, R. D. (1979). Isoflavones and hypercholesterolemia inrats. Lipids 14 , 535–540.Slimestad, R. and Hostettman, K. (1996). Characterization ofphenolic constituents from juvenile and mature needles ofNorway Spruce by means of high performance liquidchromatography-mass spectrometry. Phytochem. Anal. 7 ,42–48.Slowing, K. V., Constant, H. L., Beecher, C. W. W. (1995). Theanti-inflammatory activity of Eugenia jambos  . AmericanSociety of Pharmacognosy 36th Annual Meeting, Oxford,Mississippi, USA.Slowing, K., S¨ollhuber, M., Carretero, E. and Villar, A. (1994).Flavonoid glycosides from Eugenia jambos. Phytochemistry  37 , 255–258.Steinberg, D., Parthasarathy, S., Carew, T. E., Khoo, J. C. andWitztum, J. L. (1989). Beyond cholesterol: Modifications oflow-density lipoprotein that increase its atherogenicity. N.Engl. J. Med. 320 , 915–924.Suffness, M. (1987). New approaches to the discovery ofantitumour agents. In Biologically Active Natural Products  .(Hostettmann, K. and Lea, P. Eds.), pp. 85–104. OxfordUniversity Press, New York.Takahama, U. (1985). Inhibition of lipoxygenase-dependent lipidperoxidation by quercetin: Mechanism of antioxidativefunction. Phytochemistry 24 , 1443–1446.Wolfender, J. L., Maillard, M., Marston, A. and Hostettmann, K.(1992). Mass spectrometry of underivatized naturally occur-ing glycosides. Phytochem. Anal. 3 , 193–214.Wolfender, J. L., Maillard, M. and Hostettmann, K. (1994).Thermospray liquid chromatography–mass spectrometry inphytochemical analysis. Phytochem. Anal. 5 , 153–182. H. L. CONSTANT  ET AL. 180© 1997 by John Wiley & Sons, Ltd.PHYTOCHEMICAL ANALYSIS, VOL. 8, 176–180 (1997)
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks