Aqueous extract from the vegetative portion of

Aqueous extract from the vegetative portion of Spanish black radish (Raphanus sativus L.var.niger )induces detoxification enzyme expression in HepG2cellsPaul R.Hanlon *,Melissa G.Robbins,Lori D.Hammon,David M.BarnesStandard Process,Inc.,Department of Research and Development,1200West Royal Lee Dr.,Palmyra,WI 53156,United StatesA R T I C L E I N F O Article history:Received 14April 2009Received in revised form 28May 2009Accepted 3August 2009Available online 27September 2009Keywords:Glucosinolate Cruciferous Detoxification Antioxidant Raphasatin Raphanus sativusA B S T R A C TThe vegetative,or above ground,portion of radishes are not typically consumed and have not been characterized in terms of their phytochemical content or biological activity.The antioxidant capacity of the freeze-dried vegetative portion of mature Spanish black rad-ishes (harvested 2months after planting)was 3.7-,3.6-and 2.9-fold greater than freeze-dried roots as assayed by ferric ion reducing antioxidant power (FRAP),total phenolics and oxygen radical absorbance capacity (ORAC),respectively.No carotenoids were detected in the roots,however the vegetative portion contained 905.0and 421.2l g/g of lutein and b -carotene,respectively.Despite differences in glucosinolate composition,a class of com-pounds related to the induction of detoxification enzymes,the aqueous extracts of roots and the vegetative portion similarly induced quinone reductase activity and the expression of multiple phase I and II detoxification enzymes in the HepG2human hepatoma cell line.When juiced,the resulting freeze-dried juice and pulp of the vegetative portion maintained the ability to induce detoxification enzymes,although there were some differences in the phytochemical contents of these two dried materials.These data demonstrate that the ability of the vegetative portion of Spanish black radish,which has significantly greater antioxidant potential than roots,to induce detoxification enzyme expression was similar to that of the roots.Ó2009Elsevier Ltd.All rights reserved.1.IntroductionRadishes belong to the family Brassicaceae,also known as cru-ciferous vegetables,and are cultivated and consumed through-out the world.Consumption of cruciferous vegetables has been correlated with a decrease in risk for a number of chronic dis-eases,in particular cancer (Murillo and Mehta,2001),and data suggest that induction of detoxification enzymes is one of the primary mechanisms through which crucifer consumption de-creases risk (Keum et al.,2005;Yeh and Yen,2009).Antioxi-dants,which have also been associated with a reduction incancer risk,are also abundant in crucifer vegetables (Eberhardt et al.,2005;Gill et al.,2007;Tiku et al.,2008).A number of studies have also examined the biological activity of radishes and compounds from radishes in both in vitro and in vivo experiments.Much of this work has fo-cused on the activity of a class of compounds known as gluc-osinolates (Barillari et al.,2006;Hanlon et al.,2007;Nakamura et al.,2008;Papi et al.,2008).Upon tissue disruption (such as grinding or chewing)an enzyme also contained within the crucifer commonly called myrosinase (E.C.3.2.1.147)metabo-lizes the glucosinolates into numerous breakdown products1756-4646/$-see front matter Ó2009Elsevier Ltd.All rights reserved.doi:10.1016/j.jff.2009.08.001Abbreviations :VP ,vegetative portion;ITC,isothiocyanate;FRAP ,ferric ion reducing antioxidant power;ORAC,oxygen radical absorbance capacity;CYP,cytochrome P450;AA,ascorbic acid;GA,gallic acid;TE,trolox equivalents.*Corresponding author:Tel.:+12624956493;fax:+12624956399.E-mail address:phanlon@ (P .R.Hanlon).J O U R N A L O F F U N C T I O N A L F O O D S1(2009)356–365a v a i l ab l e a t w w w.sc i e n c ed i re c t.c o mjou r nal homep age:www.elsevi e r.co m /l ocat e/jf fsuch as isothiocyanates that are hypothesized to be the com-pounds that help prevent chronic disease(Keum et al.,2005). In vitro experiments have demonstrated that radish roots have potent antioxidant activity(Barillari et al.,2006),induce detoxification enzymes(Lee and Lee,2006;Hanlon et al.,2007) and inhibit proliferation of cancer cells(Kim et al.,2006; Barillari et al.,2008;Papi et al.,2008).In vivo experiments focusing on the activity of the radish root have also shown that radish consumption protects against chemical-induced injury(Baek et al.,2008)and reduces oxidant stress(Sipos et al.,2002;Lugasi et al.,2005).In the United States,consumption of radishes is much lower than the consumption of other crucifers,in particular broccoli(USDA,2001,2004),while consumption of radishes is much more common in countries such as Japan(Talalay and Fahey,2001;Suzuki et al.,2006).While the radish root is much more commonly consumed than the vegetative portion in the United States and most western countries,the vegeta-tive portion(or radish greens)are much more commonly con-sumed in Iran and some other Asian countries.With Spanish black radishes,which produce one of the largest roots of all the radish varieties,the vegetative,or above ground,portion grows to between14and18in.(35–45cm)tall.A rough esti-mation of yield from only the vegetative portion from an acre of Spanish black radish is 2500pounds( 1100kg)of dried material.If only the root of the radish is used,this results in a large amount of waste material.The purpose of this study was to determine whether the vegetative portion of Spanish black radishes have a phyto-chemical composition and biological activity that would pro-vide value.Since the vegetative portion is very infrequently consumed raw,especially in the United States,we addition-ally examined the effect of processing by analyzing the dry material resulting from juicing the vegetative portion.The dry material from the whole vegetative portion,dried juice material and the dried pulp(resulting from the juicing pro-cess)were compared to determine which would provide the most benefit as a dry material.2.Materials and methods2.1.MaterialsPlastic ware and general lab supplies used were purchased from Fisher Scientific(Hampton,NH).Chemicals and re-agents,including cell culture media,were purchased from Sigma(St.Louis,MO),unless otherwise indicated.Spanish black radishes,Raphanus sativus L.var.niger,were grown from seeds obtained from Johnny’s Selected Seeds(Winslow,ME) on the Standard Process farm(Palmyra,WI).2.2.Cell cultureHepG2cells at passage76were obtained from ATCC(Manas-sas,VA)and cultured in Modified Eagle Media(Sigma;St. Louis,MO)supplemented with1mM sodium pyruvate(Invit-rogen;Carlsbad,CA),100l M non-essential amino acid solu-tion(Invitrogen;Carlsbad,CA),100units of penicillin, 100l g/mL of streptomycin(Sigma;St.Louis,MO)and10%fe-tal bovine serum(Atlanta Biologicals;Norcross,GA).Cultured cells were maintained at37°C in a humidified5%CO2atmo-sphere.HepG2cells used in the experiments were between passage number80and84.2.3.Processing of Spanish black radish roots and vegetative portionSpanish black radishes were harvested from thefield 2months after seeds were planted,then the vegetative por-tion was separated from the roots.For the purposes of this study,all above ground material is referred to as the vegeta-tive portion and all below ground material is referred to as root.Roots were immediately frozen and then freeze dried.A portion of the vegetative portion was frozen intact and then freeze dried,resulting in‘‘Whole’’material.Another portion of vegetative portion was juiced using a Juicelady Pro(Salton Inc.;Miramar,FL)mechanical juicer.The juice and the pulp from the juicing were incubated at room temperature over-night to allow for complete hydrolysis of glucosinolates be-fore being frozen then freeze dried.These freeze-dried materials were then used for all phytochemical analyses. For cell culture experiments,5mL of water were added to 500mg of freeze-dried material and then the material was ro-tated at room temperature overnight to allow for complete myrosinase-catalyzed hydrolysis of glucosinolates.Material was then centrifuged to remove debris;the supernatant was subsequentlyfiltered through a0.45-l mfilter to produce solu-tions for treatment of cells.These extracts were then diluted appropriately to produce treatment media containing0.15, 0.5,1.5and5mg of dry radish material per mL of media.2.4.Antioxidant activity assaysFerric ion reducing antioxidant power(FRAP)assay was con-ducted using a previously published method(Benzie and Strain,1996).Samples were prepared by extracting300mg of material with10mL of water.In a clear96-well plate, 230l L of FRAP solution(0.3M acetate buffer,0.01M TPTZ and0.02M FeCl2)was added to10l L of sample.Triplicate extractions of the material were run in triplicate wells.The plate was incubated for8min at room temperature and then absorbance was measured at593nm.Data for the FRAP assay is expressed as ascorbic acid equivalents.Total phenolics were estimated by the Folin–Ciocalteu assay with gallic acid as a standard according to a previously described method (Singleton et al.,1999).Samples were prepared by extracting 100mg of material with2.5mL of70%acetone solution(70% acetone,29.5%water,and0.5%acetic acid).In a glass tube, 100l L of this extraction was incubated for2h at room tem-perature with5mL of diluted of Folin–Ciocalteu reagent and 4mL0.7M sodium carbonate.Absorbance was then mea-sured at650nm.Data for the total phenolics assay is ex-pressed as gallic acid equivalents.Oxygen radical absorbance capacity(ORAC)was performed according to a previously published method(Huang et al.,2002;Rasmussen et al.,2008).The250l L reaction mixture consisted of25l L of sample extract(at a concentration of25l g of material per mL of70%acetone solution)with50nMfluorescein and oxidation was initiated with20mM2,20-azobis(2-methylpro-pionamidine).Fluorescence was measured every2min overJ O U R N A L O F F U N C T I O N A L F O O D S1(2009)356–365357the course of90min.Data for the ORAC assay is expressed as trolox(a vitamin E analog)equivalents.2.5.Ascorbic acid quantitationAscorbic acid concentrations were measured using the official Association of Official Analytical Chemists(AOAC)method (45.1.14,from the16th edition).Three grams of sample were acid digested for2h with meta-phosphoric acid,thenfiltered through filter paper.Thefiltered sample was then titrated with2,6-dichlo-roindophenol and compared with an analytical standard.2.6.Quantitation of a-tocopherolVitamin E(a-tocopherol)concentrations in the samples was determined using GC/MS.For each sample,500mg of dry material was extracted four times with5mL of dichlorometh-ane.Solvent was removed using a rotary evaporator,then res-idue was resuspended in1mL of dicholoromethane.Samples were analyzed with an Agilent6890N GC(Palo Alto,CA) according to the following specifications:injector tempera-ture was220°C;oven temperature started at40°C for2m, then increased8°C/min from40to260°C,and remained at 260°C for a total run time of44.5min;columnflow was kept constant at1.3mL/min with a carrier gas of helium;injection volume was1l L.An HP-5MS capillary column(J&W Scien-tific;Folsom,CA)was used for separation(30m,0.25mm i.d.,0.25l mfilm thickness).Mass chromatography was per-formed at70eV using electron impact ionization.The identity of a-tocopherol was confirmed with mass spectrometry.2.7.HPLC quantitation of total isothiocyanate concentrationTotal isothiocyanates(ITCs)used in the cell culture experi-ments,as described above,were measured using a previously described method(Zhang et al.,1996;Hanlon et al.,2007).In an HPLC vial,50l L of the extract solution was combined with 450l L100mM potassium phosphate buffer(pH8.5)and 500l L of8mM1,2-benzenedithiol.Vials were then capped, given a nitrogen headspace,and incubated at65°C for2h. For analysis,50l L of samples were injected directly onto a RP-C18column(Vydac,201TP54,250·4.5mm i.d.,10l m; Grace Analytical,Columbia,MD).The mobile phase was 80:20MeOH:H2O with aflow rate of2.0mL/min.The cyclocon-densation product,1,3-benzodithiole-2-thione eluted from the column at approximately 2.7min with the detection wavelength set at365nm.An Agilent1100LC(Palo Alto,CA) was used for analysis.Serial dilutions of a10l M R-sulfora-phane(LKT Laboratories,St.Paul,MN)stock solution were used to create a linear standard;and total ITC units were ex-pressed as l mol ITC/mL of media.2.8.Quantitation of carotenoids by HPLCCarotenoid concentrations were determined using a previ-ously described method(Surles et al.,2004).To100mg of lyophilized radish material,2mL of ethanol+0.1%2,butyl-ated hydroxytoluene(BHT),150l L of80%potassium hydrox-ide and40l L of b-apo-80-carotenal(internal standard,25ng/mL concentration)were added and samples were heated at 70°C for15min.Next,3mL of water were added and samples were extractedfive times with5mL of hexane+0.1%BHT. Hexane was then removed by a centrifugal evaporator and residue was resuspended in1mL of mobile phase A+0.1% BHT.Samples were analyzed using an Agilent1100HPLC(Palo Alto,CA)using40l L of sample that were injected onto a Waters(Milford,MA)Resolve C18column,3.9·300mm with 5l m pore size.Mobile phase A consisted of95%acetonitrile, 5%water,10mM ammonium acetate and0.1%triethylamine. Mobile phase B consisted of85%acetonitrile,10%methanol, 5%dicloroethane,10mM ammonium acetate and0.1%trieth-ylamine.Carotenoids were then eluted from the column at a flow rate of2mL per min with a gradient procedure:100%A for3min;a12-min linear gradient to100%B;hold for3min at100%B;a1-min linear gradient back to100%A.The detec-tor was set to detect absorbance at444nm(lutein)and 453nm(b-carotene).2.9.Quinone reductase activityQuinone reductase activity was measured as previously de-scribed(Jiang et al.,2003;Hanlon et al.,2007).HepG2cells were cultured in12-well plates.Confluent cells were treated for two days with media containing treatment or water as a vehicle control.For each independent experiment,three wells of cells were exposed to each treatment,and each well then represented one replicate for the quinone reductase assay. After treatment was completed,media was removed and cells were washed once with1mL of ice-cold PBS.Cells were scraped into a cell harvest solution(25mM Tris–HCl pH7.4; 125mM sucrose)and then sonicated for10s.Protein concen-tration was determined using the Pierce BCA kit,following manufacturer’s protocol(Pierce Biotechnology;Rockford,IL). For measurement of quinone reductase activity,1l g of pro-tein was added to a96-well plate and200l L of quinone reduc-tase assay solution were added(25mM Tris–HCl pH7.4;60 l g/mL BSA;0.01%T ween20;5l Mflavin adenine dinucleotide; 80l M2,6-dichloroindophenol).Quinone reductase activity was calculated as the change in absorbance per min per l g of total protein.2.10.Cell viabilityHepG2cells were cultured in96-well plates.Confluent cells were then treated for two days with media containing crude radish extract or water as a vehicle control.Three replicate wells received100l L of each treatment media.After2days of treatment,assay was performed as per manufacturer’s pro-tocol(Promega CellTiter96ÒNon-Radioactive Cell Prolifera-tion Assay;Madison,WI).2.11.Real time quantitative PCRQuantitation of RNA expression was conducted as previously described(Hanlon et al.,2007).For each independent experi-ment three wells of cells were exposed to each treatment. For each sample,total RNA was collected from one well of a 6-well plate per manufacturer’s protocol using the Mag-MAX TM-96Total RNA Isolation Kit(Ambion;Austin,TX).RNA358J O U R N A L O F F U N C T I O N A L F O O D S1(2009)356–365concentrations were determined using a spectrophotometer (NanoDrop;Wilmington,DE).cDNA was generated with the Applied Biosystems High Capacity cDNA Archive Kit,follow-ing manufacturer’s protocol(Foster City,CA).Primer/probe kits for the genes measured were obtained from Applied Bio-systems(Foster City,CA).Each treatment was represented by triplicates and each PCR reaction was carried out in duplicate. PCR was performed on an ABI7300Sequence Detection Sys-tem.Each25l L reaction contained1l L of cDNA,12.5l L of Takara Premix Ex Taq TM(Takara Bio;Madison,WI)and1.25l L of the primer/probe reaction mixture(for the18s ribosomal RNA reaction the primer/probe mixture diluted1:6before 1.25l L was added).Fold changes between treatment groups were then determined using the reaction efficiency(Pfaffl, 2001)and18s ribosomal RNA as the reference gene: Ratio=(E target)D CP target(control-sample)/(E ref)D CP ref(control-sample).2.12.HPLC quantitation of glucosinolate concentrationGlucosinolate concentrations were measured using a previ-ously described method(Wathlelet and Wagstaffe,1991; Brown et al.,2002;Hanlon et al.,2007).T wo hundred milli-grams of dry radish material were extracted twice with2mL of boiling70%methanol.During extraction samples were incubated at95°C and were vortexed every minute.Before the second extraction,500l L of benzylglucosinolate(1mM) was added as an internal standard.After extractions,samples were centrifuged at1200g at10°C for15min,and the super-natant was transferred to a new tube.Next,1mL of the pooled supernatants was transferred to a new tube,and 150l L of0.5M barium acetate/0.5M lead acetate was added. The sample was vortexed and applied to a column with 100mg of A-25Sephadex that had been conditioned with 5mL of0.5M NaOH,5mL of0.5M pyridine acetate and 5mL of water.After the sample was allowed to drip through the column,the column was washed with3mL of0.02M pyr-idine acetate and then3mL of water.Then500l L of a20 U/mL sulfatase solution was allowed to penetrate the column, and the column was incubated overnight at room tempera-ture.The next day glucosinolates were eluted off the column with2mL of water,followed by an additional1mL of water. Samples were thenfiltered before40l L were injected for HPLC analysis.Samples were analyzed with an Agilent1100 HPLC(Palo Alto,CA)with a250mm Grace C18column(Grace Analytical;Columbia,MD),5l m pore size and4.6mm i.d. Desulfoglucosinolates were eluted from the column at aflow rate of0.8mL/min with a linear gradient of0–20%acetonitrile in water up to53min,then an additional7min with a linear gradient of20–0%acetonitrile.The amount of glucosinolates were determined using relative response factors(MacSharry, 1990;Wathlelet et al.,2004).2.13.GC/MS quantitation of individual ITCsIndividual ITCs in juice samples were quantitated using GC/ MS.The vegetative portion of Spanish black radishes were juiced with a Juicelady Pro mechanical juicer.At the indicated times after juicing,750l L of radish juice was transferred to a 1.5mL microfuge tube and immediately20l L of benzylisoth-iocyanate(BITC)internal standard(100mg/mL)and480l L of methylene chloride were added.Samples were vortexed immediately and then centrifuged to separate solvent and aqueous fractions.Solvent fractions werefiltered before anal-ysis using GC/MS,as previously described(Matusheski et al., 2004).Samples were analyzed with an Agilent6890N GC(Palo Alto,CA)according to the following specifications:injector and detector temperatures were200°C and280°C,respec-tively;oven temperature started at40°C for2m,then in-creased10°C/min from40to260°C,and remained at260°C for10min;head pressure was25psi;and injection volume was1l L.An HP-5MS capillary column(J&W Scientific;Fol-som,CA)was used for separation(30m,0.25mm i.d., 0.25l mfilm thickness).Concentrations of raphasatin(4-methylthio-3-butenyl isothiocyanate)were calculated using purified erucin(LKT Laboratories;St.Paul,MN)as a standard, since the structure of erucin is very similar to raphasatin(the only difference in structure is that raphasatin has a double bond that is absent in erucin)and these ITCs elute within one minute of each other under these chromatographic con-ditions.Concentrations of sulforaphene were calculated using purified sulforaphene(LKT Laboratories;St.Paul,MN). Mass chromatography was performed at70eV using electron impact ionization.The identity of raphasatin and sulforaph-ene were confirmed with mass spectrometry.2.14.Heavy metal concentrationsHeavy metal concentrations were determined by inductively coupled plasma mass spectroscopy(ICP-MS)by Bodycote Testing Group(Portland,OR).J O U R N A L O F F U N C T I O N A L F O O D S1(2009)356–3653592.15.Statistical analysisResults showing mean(±SD)values were tested for statistical differences between treatments by one-way ANOVA(Graph-Pad Prism,San Diego,CA)with T ukey’s multiple comparison posthoc test and p<0.05as minimum significance for Figs. 1,2and4;one-way ANOVA with Dunnet’s posthoc test and p<0.05as minimum significance for Figs.5and6;and one-way ANOVA with Bonferroni’s posthoc test and p<0.05as minimum significance for Fig.3.Non-normally distributed real-time PCR data(detailed in Fig.6)were transformed using a natural log function prior to statistical analysis.3.Results3.1.The vegetative portion of Spanish black radish has significantly more antioxidant activity and carotenoid content than the rootsSpanish black radishes were harvested and divided into the vegetative portion(VP)and root material before being freeze dried.Freeze-dried material was then assayed for antioxidant capacity using three methods that predominantly measure hydrophilic antioxidants:ferric ion reducing antioxidant power(FRAP),total phenolics and oxygen radical absorbance capacity(ORAC).In terms of antioxidant potential,the FRAP (l mol of ascorbic acid equivalents per gram of dry material), total phenolics(l g of gallic acid equivalents per mg of dry material)and ORAC(l mol trolox equivalents per gram of dry material)was significantly higher for the VP(141.0±4.0, 15.81±0.63and905.0±74.4,respectively)than for the root (38.1±2.8,4.42±0.05and315.3±32.1,respectively).The radish roots had27.0%,27.9%and34.8%of the antiox-idant capacity of the VP when measured with FRAP,total phenolics and ORAC,respectively(Fig.1).The VP also con-tained significantly more carotenoids,a class of hydrophobic compounds with antioxidant potential,than the roots(Fig.2). The VP contained905.0and421.2l g/g of lutein and b-caro-tene,respectively;however,there were no detectable carote-noids in the radish root as determined by HPLC analysis.The effect of processing on the antioxidant capacity of the VP was assessed in fresh material which was juiced and then the resulting juice and pulp was frozen and freeze dried after 24h of storage at room temperature.The results are reported as units per gram of dry material(for both the juice and the pulp)for the purpose that these experiments were conducted to evaluate the content of dried material,a processing step that would be required for their inclusion in other products. In terms of FRAP,total phenolics and ORAC,the VP juice (231.0±6.3,17.94±2.96and652.1±98.8,respectively)had360J O U R N A L O F F U N C T I O N A L F O O D S1(2009)356–365higher antioxidant capacity than the VP pulp(69.2±2.3, 11.02±0.18and748.7±38.1,respectively)in terms of FRAP and total phenolics,but not ORAC.Dried juice powder from the VP had significantly greater FRAP activity than the whole material(Fig.1).Both dried juice powder and dried pulp had significantly greater antioxidant activity than the roots in all three antioxidant assays.Processing significantly affected carotenoid concentrations(Fig.2),with dried juice and pulp materials containing significantly less lutein and b-carotene than the dried whole VP.3.2.Processing affects the glucosinolate and isothiocyanate content of the VP of radishesGlucosinolates are phytochemicals found in cruciferous vegetables that have been shown to induce detoxification enzymes after myrosinase-mediated metabolism to com-pounds such as isothiocyanates,including the extensively studied sulforaphane.A previous report demonstrated that glucoraphasatin is the predominant glucosinolate found inJ O U R N A L O F F U N C T I O N A L F O O D S1(2009)356–365361the root of Spanish black radish(Hanlon et al.,2007),as well as other varieties of radish such as daikon(Barillari et al., 2005b;Papi et al.,2008)and even the common table radish (data not shown).Analysis of the VP of Spanish black radish found significantly lower concentrations of glucoraphasatin, glucoerucin and gluconapin than in the root(Fig.3A),while the VP contained significantly higher concentrations than the root of glucoraphenin and glucobrassicin.Juicing of the VP had profound effects on the glucosinolate content of the resulting materials.Upon tissue disruption the glucosinolates are brought into contact with myrosinase en-zyme,leading to their metabolism.The effects of juicing on glucosinolate content were measured in VP frozen after allowing24h for myrosinase-mediated metabolism of gluco-sinolates.The pulp material had significantly lower concentra-tions of all four glucosinolates after24h of hydrolysis than was frozen and then dried intact VP(Fig.3B).No glucosinolates were detected in the juice material(data not shown).The production of the major isothiocyanates in the juice from the VP was then measured.Maximum concentrations of sulforaphene,the metabolite of glucoraphenin,and rapha-satin(4-methylthio-3-butenyl isothiocyanate),the metabolite of glucoraphasatin,were reached at30and15min after juic-ing,respectively(Fig.4).The concentration of sulforaphene declined slowly over24h,but even after24h significant levels remained.Raphasatin concentrations in the juice declined rapidly after15min and after90min raphasatin levels were no longer detectable by GC/MS.3.3.Aqueous extracts from the VP of Spanish black radishes induced detoxification enzyme activity and expression in HepG2human hepatoma cellsThe ability of cruciferous vegetable extracts to induce detoxi-fication enzyme activity and expression has been well estab-lished in the literature(Keck and Finley,2004;Juge et al.,2007). Prior to cell culture treatment,radish root,and the different processing products of the VP(whole,juice and pulp)were mixed with water overnight to allow for extensive myrosi-nase-mediated metabolism of glucosinolates into isothiocya-nates,as described in Section2.The total isothiocyanate concentration of the extract from each of these materials was significantly different from each other:3.54(±0.16),6.43 (±0.09),1.07(±0.03)and2.41(±0.15)nmol of ITC/mL for roots, VP whole,VP pulp and VP juice,respectively.HepG2,human hepatoma cells,were then treated with radish extracts atfinal concentrations of0.15,0.5,1.5and 5mg of dry radish material per mL of media.Extracts from all of the materials produced a dose-dependent induction of quinone reductase activity with significant induction at 1.5mg/mL(Fig.5A).A decrease in viable cells was only seen with the5mg/mL concentration of VP whole extract (Fig.5B);however,previous experiments had demonstrated that the5mg/mL concentration of root also produces a reduc-tion in viable cells(Hanlon et al.,2007).The effect of the radish extracts on the expression of var-ious detoxification enzyme RNA was subsequently deter-mined using real-time PCR.Cells were treated with the 1.5mg/mL concentration of all three extracts,which pro-duced significant induction of quinone reductase with all the extracts(Fig.5).The extracts from all four radish materi-als significantly induced the phase I detoxification enzymes cytochrome P450(CYP)1A1and CYP1B1,but not CYP1A2or CYP3A4(Fig.6A).The extracts from all four radish materials also significantly induced the expression of the phase II detoxification enzymes quinone reductase,heme oxygenase and thioredoxin reductase(Fig.6B).4.DiscussionMost cruciferous vegetable research has focused on com-monly consumed crucifers such as broccoli or Brussels sprouts and the phytochemicals they contain.The unpleasant taste and texture of the vegetative portion of radishes make it highly unlikely that this cruciferous material would be con-sumed fresh,and thus the most likely use this material would be through inclusion in feed or as a supplement.In a commer-cial setting,these uses would require processing including drying.Therefore,all of the experiments performed in this study(with the exception of ITC production in the juice in Fig.4)were conducted on dried material,and results were ex-pressed in terms of the amount of dried material.While a majority of the cruciferous vegetable research has focused on glucosinolates and isothiocyanates,most crucifer-ous vegetables are also a rich source of antioxidants(Plumb et al.,1996;Eberhardt et al.,2005).In comparison with other crucifers such as broccoli,Brussels sprouts,kale or watercress the radish is unique in that the portion of radishes that are typically consumed,the root,has significantly lower concen-trations of antioxidants than the portion that is not con-sumed as measured by several antioxidant assays(Fig.1). These assays measure different antioxidant properties (Huang et al.,2005),with the ORAC assay measuring peroxyl radical scavenging capacity and both the FRAP assay and total phenolics assay(Folin–Ciocalteu reagent)assessing ability to transfer an electron;however,in all assays the root had less antioxidant capacity than the VP.Additionally,carotenoids, a hydrophobic class of antioxidants(Sun et al.,2009)which have also been demonstrated to have chemopreventative properties(Zhang et al.,2007;Palozza et al.,2008),were not detectable in the roots but were present in the VP.The total antioxidant capacity of the root and vegetative portion of Spanish black radish is due to combinations of numerous compounds in these materials.In addition to the carotenoids(Fig.2),the concentration in radish roots,VP–whole,VP–pulp and VP–juice of ascorbic acid(1.24mg/g, 3.17,0.75and7.28,respectively)and a-tocopherol(undetect-able,177.9l g/g,75.3and75.6,respectively)were determined. Comparison of the total antioxidant capacity of these materi-als(FRAP,total phenolics or ORAC)with the individual compo-nents(lutein,b-carotene,ascorbic acid or a-tocopherol)using chi-square showed no associations.Since these individual compounds make up such a small portion of the materials, it is not surprising that the total antioxidant capacity of these materials can be linked to a single phytochemical.Further-more,a previous study also demonstrated that there was no correlation between the concentration of carotenoids, a-tocopherol or ascorbic acid with the total antioxidant capac-ity of a number of varieties of broccoli(Kurilich et al.,2002).362J O U R N A L O F F U N C T I O N A L F O O D S1(2009)356–365。