Document 3QOnOvOwbRNEJw0mBZeoZK83x

AU6 @ CI Determinants of Benzene Metabolism and Dispositionu'* Paul M.Schlosser, Elaina M. Kenyon, Mark J. Seaton, and MicheleA. Medinsky The metabolism of benzene and the resulting dosimetry of benzene and Its mmolltes almost certainlyare primary factors In determiningthe health risks to humansfrom benzeneetpo8um. The mechanismsand pathwaysthat determinethe concentnth8 of benzeneand Itsmetaibolltos In various tissues have been the subject of consldomble debate. This artlcle ckrcribw Insights lnto benzene dosimetry and metabolism Obtrlned at CllT over tha past thtw yeam and subra quent lmpllcations for benzeneinduced myelotowicity and hmatotoxicity. The metaboik pathways and kinetics of benzenemetabolismhnnksne x p i d In w#w, pmvldlngsufficient datato developa mathematicalmodelof tha mctlonsthat occur. OifWunum In In VEtrObenamemotnbollsmamong mice, rats, and Individualhumnr have boon invmtlg8ted. Mice motabolkebenzene faster than rats, while the range of ratesoxhlbbd by human tluuo samples spans that of mka and rats. Some qualltatiw dlffemces hnnalso b#noboewad be!w8an In Mmand In vir0benzene metaboilsm. Thew dlfferenws can k axphinod, however, by Incorporating th8taglorurl distribution of liver enzymes Into a phymiologically baud phmrmacoklnetlc (PBPK) modd. By developing mathemaicai models capabb of descrlblng large data ntr and observing tha range of m@tabolic acthtlticw in samples of human I h ,w. expect to provide dmmetlcslly bettor tools for the assessment of human health risks from expoauros to bonzew. Benzene was used extensively in the humans. Themfactors leadto considerable metabolites at the site of toxicity, the bone past for productionof paints, resins, rubber. uncertaintyaboutthe riskto humans of low- marrow. Working under this hypothesis, re- inks, and dyes. It is currently 1%of gaso- level benzene exposure. search et CllT has focum Unckrdend- line and a feedstock for the manufacture of Benzeneis mydotoxicand hematotoxic ingthe V i r metabolismof behmand on synthetic organic chemicals (Ayres and Tay- in mice and rais (MacEachem et el., 1992; determininghowthii metaboliwninfiucHlc8s lor, 1089). Benzene is a constituentof gaso- Snyder etel., 1978). Etecwsemyebbxicity the dosimetry of benzene and its metabo- line fumes, automobile exhaust, and both and hematotoxidty have been observed lites in vim. In particular, our goal is to de- mainstreamand sid-am tobacco smoke prior to the oM8t of AML in humans, the velop a quantitative took8 ma!hematical (Wallace, 1990). It is listedas an air toxic in inddencoanddogreedtheseaaRedfects simulation model-that relates metabolic the Clean Air Act Amendment of 1990. Ex- in humans are presumed to correspond to rates in the her to the umcentrations of posure to high levels of benzene for ex- risk of human AML Further, because an benzene metabolites reaching the bone tended periods of time is associated with animalmodelforbenzeneinduwdAMLhas m a m . This tool is to be dmmbped using aplastic anemia and acute myelogenous not been identilied, the incidende of ben- data from rats and mice, for which exten- leukemia (AML) in humans (Infante et a/., zeneinducedmyelatoxicttyand hematotox- sive data sets exist or can be obtahd. In IS??). Workplaceexposurestobenzeneare icity inmice and rats is takento be a s u m addition to the collection and enatydr of now limited through safety regulations gate forAML when consideringthe relation- rodent data,we haw, h n measuring the (Runion and Scott,1Q85),but occupational ship betweten benzene exposun, levels, tis- rateofbenzenemetabolismb y h r lllunQb8 and environmental exposures to benzene sue metabolite-ntretlons, and disease from humandonon.The rodentdataww1be stilloccur (Hricko, 1994;Tmpa efal., 1994). response. Baaed on the observation by usedtodevetopandvalldetsthadmulation The effects of chronic exposure to low Sammett et a/.(1979) that partialhepatec- model.lhn. bychanglngihemodelparam- levelsof benzene are not known. Uncertain- tomy in rats is pmtech egainst the hema- sten to m f k t human-and mata- ties in exposure assessment in epidemio- totoxicityof benzene,theconversionof ben- bolk capacities, we will be ablrr to predict logical studies have made it difficult to ob- zene in the liver to one or more toxic me- the expo8uredose mlatbmhipfor benzene tain a reliablelow-doseextrapolation(Health tabolttes is belbved to play a significant meteboltto8in human bone marrow. Effects Institute, 1993). Also, a confirmed role in the indudknof hematotoxicity. If this The primary pathways for benzene animal model does not exist for benzene- is true, then the risk to humansdepends on inducedAML, the effect of most concern in the concentration, or -e, of these toxic (Continued on page 2) " ISSN 8755-4259 Publishedmonthlyby theChemkal IndustryInstituteofToxbkgy. Correspondence shouldbe directedtoWillannaGriffln, CllT informationServices, P.O. Box 12137,Research Triangb Park, NC 27709. Teiephone, (919)558-1200; FAX, (919) 5581300 Roger 0. McClellan,D.V.M. President SUPPORTING COMPANIES Air Products and Chemicals, Inc. Albemarle Corporation AliiedSignal, Inc. Amoco ChemlcalCompany ARC0 ChemicalCompany BASFCorporatlon Bayer Corporation BPAmerica, Inc. Chevron Corporation Ciba-GeigyCorporatlon The DowChemicalCompany Dow CornlngCorporation E, 1. du Pontde NemoursandCompany EastmanChemical Company EastmanKodak Company EthylCorporation Exxon Chemical Company FINA, Inc. FMCCorporation General ElectricCompany GeorgiaGulf Corporation The BFGoodrichCompany W. R. Grace& Co. Hoechst Celanese Corporation HBlsAmerica Inc. InternationalSpecialty Products The Lubrizol Corporation MallinckrodtChemical, Inc. Mobil Researchand Development Corporation MonsantoCompany Nalw ChemicalCompany NovacorChemicals Ltd. OccidentalChemicalCorporation Olln Corporation Phillips66 Company PPGIndustries,Inc. Quantum Chemical Corporation ReichhddChemicals, Inc. Rhh-Poulenc Inc. Rohmand Haas Company SandozAgro, Inc. Schuller International,Inc. Shell ChemicalCompany Texaw Inc. UnionCarbide Corporation UnocalCorporatlon Volkswagen of America Vuican MaterialsCompany BenzeneMetabdlsm (from page 1) metabolismintheliverare depictedin Figure 1. Oxidativereactionsleadto thb formationof benzene oxide (a reactive intermediate), phenol, hydroquinone, and smallamounts of catechol and trihydroxy benzene. These oxidized metabolites may be further transformedby-tion to compoundsthat are readily excreted in the urine. Benzene oxide can also be converted to an opened- ring metabolite,muamaldehyde, a precursor of muconicacid,Withisexcretedinthe urine. The conjugated metabolites of benzene are generally thought not to be responsible for its toxic effects, and a significant amount of researchhas focused on identifyingwhich of the intermediatemetabolites isthe badac- tor. Administmtion of either phenol or hydro- quinone alone to rodents fails to reproduce the hematotoxic and myelotoxic effects of benzene exposure (Cox, 1991; Kari et a/., 1992). These obsewations led investigators to the hypothssisthatneitherof thesemetabo- lites is responsiblefor benzenetoxicity andto the conjecture that muconaldehyde, the only semistable toxic metabolite that is not pro- duced from phend, might be responsible. However, as with phenol and hydroquinone, muconaldehydedonedoesnot reproducethe toxic effects of benzene (Guy et a/., 1991). Instead, benzene toxkity apparently resub from the combinedeffectsof severalmetabo- lites: muconaldehydeand hydroquinone act additively (Guy eld,1991), and phend and hydroquinonead synergistically(Barale eta/., 1990; Eastmond et el., 1987). The risk from benzene exposure may depend on the do- simetry of all three metabolites, with catechol and trihydroxybenzene playing lesser roles. To develop a simulation model for predicting in vivo dosimetry and metabolism inhumansbasedonin vitmdata, we first need to understandin vitmmetabolismand in vivo metabolism and dosimetry separately in a laboratory species such as the mouse. Only then can we merge the two levels of understanding,in viwandin vitm, andresolve any apparent differencesbetweenthetwo. In this article, we first describe the ongoing researchon livermetabolism, alongwithsome tentative physiological predictions based on in vitro results. Secondly, the development of whatiscumttyan'hdqmchtmathematical model of in vivodosimetry and metabolismis discussed. Finally, some implicstions of the current results and a synopsis of future directions are presented. Differences in the InWtno Metabolism ot Benoem,among Species and IndividualHumans The metabolismof benzene isfairly complicated in that it invdves a number of reac- tions (Figure 1). To predict the dosimetry of the various metabolites, we needto knowthe rateof each reactionas a functionof the rnetaboli concentrations. While these ratescan be determined in vivo for laboratory animals, our objective isto predicthumanmetabolism, where in dvoexpetimentsare precludedand the availability of tissue samples is limited. Differences exist among species as well as among indhridualhumans inthe ratesof m e tabolism. Since thesedifferences affect indi- vidual risk, we seek to understandandquantify the range of rates for benzene metaim lism in humans based on those measurements that can be obtainedwith human liver samples. While the rates of metabolism of a Glutathione conjugates A 0--* w-. Benzene Benzeneoxide &-Gul curondi e and sulfate H OH Trihydroxy benzene Muconaldehyde Muconicacid Hydroquinone Fig. 1 Primarypathwaysfor benzenemetabolismInthe liver. The oxldlzed metabolitesphenol, hydroquinone, catechol, and Mhydroxy benzene, along with the ringopen metabolite muconaldehyde, are all toxic to the bone marrow (myelotoxic)and affect the circulating bbod (hemstotoxic). The conjugated metabolites and muconic acid are readily excreted in the urlne. Whemical Industry Institute of Toxicology(CIIT), 1995. All rights reserved. xenobotic typically differ amongsoecies, the biochemical mechanisms of benzene biotransformation are expected to be nearly TABLE 1 ,K VALUES OF SEVERAL REACTIONS FOR MICE, RATS, AND HUMANS the same inall species. Therefore,our goal is f- to develop a mechanistically based mathematical model of metabolism in which all species and individual differences are explained by differences in parameter values such as enzyme activities. Since in vivo do- Species BO expoxidatlon @MY Reaction BMO hydrolysls (mW BMO glutathlone (mW Acrylonltrlle expoxidation @MIb simetty data in mice are also beingobtained and already exist forrats, these species can be used to test the extrapolation of in vifm Mouse 2.00 f 0.17 1.59 f 0.03 35.3 f 6.2 13fl rates to in vivo rates. Extrapolation among species or individualscan be tested by using Rat 3.75 f 0.20 0.26 f 0.10 13.8 f 0.3 11 f l a large in vitrr, data set for rat liver m e t a b lism in comparison with the mouse data. In particular,we candetermineif a modeldevel- Human 5.14 f 2.59 0.24f0.10to 1.65f0.12 10.4fl.O 1 2 f 1 to 18f 1 oped from mouse data, taken together with measurementsof rat and mournenzymeac- tivities, can be used to predict ratmetabolism. .From Csamldy ef al. (1992); bFromKedderis et al. (1993). Finally, a smaller data set obtained with hu- man tissues can be usedto determine model An initial mathematical simulation model saturation parameters such as Kmgeneral1 parameter values and their ranges for hu- of benzene oxidation by mouse and rat liver to have similar values in different species. mans. microsomes was described by Schlosser et The role of CYP 2E1 in benzen a/.(1993).This modelincorporatesa mecha- metabolismwas further exploredby Seato Oxidative Metabolism In wbo nism whereby benzene, phenol, hydro- et a/. (1994), who compared biotrans quinone, and catechol all interact through formationby liver samples from 10 human By usingdifferentfractions from livercells competitionfor am n reactionsiteon the as well as from mice and rats. Plots of th and supplying only the substratesand cofac- oxidizing enzyme, CYP. At least two CYP disappearance of benzene agd th tors necessaryfor specificmetabolicreactions, isozymesare responsiblefor benzeneoxida- formation of phenol and hydroquinone ar we can observe subsets of the reactionsde- tion, with CYP 2E1 beingpredominant (Koop shown in Figure 3. Benzene metabolisi picted in Figure land reducethe complexity efal., 1989; Nakajimaefal.,1993).The model correlated strongly with CYP 2E1 activi' of developing a modelfor all these reactions of Schlosser et a/. (1993) considers only a across speciesand individuals, suggestiv tomoremanag~pieces.onespecificgoal single enzyme, with effective parametersfor that a mathematical model could b is to develop quantitative descriptions or re- the total metaboiism of benzene to phenol, developed in which this activity is the on action rate equations for the oxidative reac- hydroquinone, catechol, and trihydroxy ben- difference among individuals or specie tions. BYperfohing incubationswith the mi- zene. The same model structure describes Sucha modelwas developedanddescribe crosomalfractionof lwerandsupplyingNADH metabolism by both mouse and rat mi- by Seaton et a/.(1994) basedon the mod andNADPHascofactors,wecanobserve the crosomes, but the parameter values for the of Schlosser et a/. (1993). This model al: set of oxidative reactions in isolation. As an twospecies are quite different. assumed that a single enzyme w i example, the disappearanceof benzene due The act*Myor maximal rate of re- responsible for oxidation, with benten to oxidationby liversamplesfromtwohumans, actionscatalyzedbyan enzymesuch asCYP phenol, and hydroquinonecompetingfor tl F-344 rats, and WC3F, mice is depicted in 2E1 mayvary significantly amonganimalspe- enzymatic reaction site. From Figure 3, v Figure 2. In these experiments, subcellular cies, but other kinetic parameters such as see that the modeldoes a fairly good job fractions of liversamplescontainingoxidation saturationor affinity constants (e.g.,K,,) may predicting data to which it has not been ' enzymes were incubated in vitro with ben- be very similar. For example, CSanady et a/. The ideaof auniversalmetabolismmod zene, and the amount of benzene oxidation (1992) examinedthe kineticsof 1,%butadiene in which CYP 2E1 activity abne explains was observed. As can be seen, mouse rni- (BD) epoxidation, butadiene monoepoxide dividualor species differencesinratesof b~ crosomes metabolizebenzenefaster than rat (BMO) hydrolysis, and BMOconjugationwith zene m e t a b d i , isquiteappealing. The dt microsomes,whichcorrelateswiththe fact that glutathione in vitro with liver samples from in Figure 3 not only suggest this possibil mice are more sensitive than rats to the mice, rats, and humans; apparent K,,, values but alsothat a firstgenerationmodelcapat hematotoxiceffectsof benzene. More impor- for these reactions are listed in Table 1. Val- of describingoxidativemetabolismby hum tantly, the data from the two human samples ues for acrylonitrile epoxidation obtained by liver might be developed almost mpletc illustratethe wide range in metabolic activity Kedderis et a/. (1993) are also listed inTable from mouse and rat data. Yet Schlosser ef for benzeneoxidationamonghumans, which 1.While differencesin Kmvaluesamongspe- (1993) obtained dissimilar values of satu spans the activity from mice and rats. Given cies and in some cases differences among tion and competitionparametersfor rats a that the same enzyme, cytochrome P450 individualhumansare observed, those differ- mice. How do we explain these conflicti (CYP) 2E1, is primarily responsiblefor ben- ences are often quite small and may not be resultsof Schlosser eta/.(1993),wheresa zene oxidation in mice and rats (Nakajima et statisticallysignificant. Further,thedifferences ration parameter values appear to be qL a/., 1993) and has a similar role in humans in K, values among species for a given reac- different betweenmiceandrats,andof Seal (Seaton et a/., 1994), understandingthe dif- tion are, in all cases, quite small compared et a/. (1994), where the kinetics of benze c ference in metabolic rates betweenmice and with the differencesin K,, values betweenre- oxidationappearvery similarinmice, rats, s rats should helpto explainthe range of rates actions. In the case of BMO hydrolysis and humans?The model of Seaton et a/. (191 observedfor humans. Inthis case, the rate of conjugation with glutathione, species differ- describes the resultsof benzeneincubatic m)a reaction may differ between two individu- ences of over an order of magnitude in the startingat a single concentration(4 a als, while the enzymatic mechanism is the respective Vmaxvalues were observed (val- same. ues not shown). Therefore we might expect (Continued on page izene Metabolism (from page 3) the data in Figure 3.If atwo-enzyme modelis among the rates of oxidation, sulfation, and abletodescribetheexistingmultiple-substrete glucuronidation for rats, mice, or individual Rionof CYP 2E1activity. When additional data sets,then inhibitionexperimentscanbe humans.Thereforewe ean assumethatthese iseand ratdata are considered,including performed to directly test and compare the rates vary independenttyin risk assessment insfrom benzeneandphenol incubations dependenceof benzeneandphenolmetabo- calculations, where the range of metabolism vo concentrations for each substrate, a lism on CYP 2E1 activity. rates among humans is important. le-enzymemodel is no longer capableof In addition to interindividual variability :ribing the entire dataset,evenfor a single ConjugationIn Mtto among human samples in the activities of :ies. In fact, more careful examinationof phenol sulfation, hydroquinone glucuroni- ire 3 reveals that there are different indials with nearty identical CYP 2El activibut significantly different extents of meIlism. If a single-enzyrnemodel is incapable of :ribing a larger dataset(onethat indudes iltsfrom incubationswith phenolasa sub:e),howcanthe correlationevidentinFig3 be explained and utilized? The rate of zene oxidation to phenol is determined sly by CYP 2E1 ectivity,which limits the labilityof phenol during benzene incuba- ;, and this limitedavailability may control rate at which benzenederived phenol is quently convertedto hydroquinone.On 3ther hand, when phenolis thesubstrate, vailability is not limitedby CYP 2E1 activndasecondCYPmaybeinvohred.There- As in the case of oxidation,the conjuga- tion reactionscan also be studied separately. Phenolsulfation andhydroquinoneglucuronidation are the primary routes for conjugation of benzenemetabolitesin rodents.Onceconjugated, the metaboliesare readily excreted inthe urine, 90 conjugationreactionsareconsideredto bedetoxicationpathways. The circulatinglevelsof the oxidizedmetabolitesare determinedbythebalance between oxidative ratesand conjugation rates.Thereforedeterminingthe ratesof conjugationreactionsrela- tive to the oxidative reactions and detennin- inghowthoseratesvary among individualhumare equally important asdeteminingthe rates of the oxidative reactions. The rates of conjugation reactions ean be measured using the cytosollc and microsomalsubcellular dation, andCYP2E1oxidation(Seatoneta/., IW),species dMerences observed among humans, mice, and rats are interesting and deservecomment.The mouseis moresensi- tive than the rat to the toxic effects of ben- Benzene 5 5- 40 t , the rate of conversion to hydroquinone I bemuchmoredependenton the activity CYP other than 2E1. Koop et a/. (1989) ~ e thdat several CYP isozymescanconphenolto hydroquinone. The hypothesis a CYPotherthan2E1contributesto phe- fractions, respectively,obtainedaftercentrifugation of liver homogenates. We examined differences inthe in VHrO ratesof conjugation among species or individuals using pooled samplesof rat andmouse liver and indiiual samplesof humanliver.We obsewedathree- xkWon is currently beingtested through development of a two-enzyme model, 're the activity of one enzyme is taken to hat of CYP 2E1. Candidates for the sec- enzymeindudeCYPlA1,CYP lA2,CYP 6, a Y C Y P X 1 1 (Nakajima eta/.,1993). :e parameter values have been set by chingthe modeltoa data set that includes I benzeneand phenolincubationsat mul3 concentrations, the dependence of fold variation among human samples in the rateof phenylsulfateformation (range, 0.3 to 0.9 nmoldm@min;Table 2). For laboratory animals, phenol sulfation was muchfaster in rat cytosol (1.2 nmole/mg/min)then inmouse cytosol (0.5 nmoldmglmin). Hydroquinone glucuronidation varied by almost threefold among human samples (range, 0.1 to 0.3 nmole/m@min;Table 2) and was more rapid in mouse liver microsomes (0.22 nmoldmgl -Model 01 tI30 Hydroquinone T 24i 1 WprediktedbenzenemetaWismonCYP min)comparedwith ratlivermicrosomes(0.08 activity will be tested and compared to nmoldmg/min). No correlationwas observed m CYP 2E1 activity (nmole/mVmin) 0 10 20 30 40 50 I. 2 Time (min) mparison of benzene metabolismwith liver microsomesfrom B6C3F1mice, F344 8, and humans (HL1 and HLlO). [Benzeney[Ben~enei]s~the proportion of ben- -ie remalning as a function of time relative to the initial benzene concentration. ubatlons were performedwith 4 pM benzene and 1mumi microsomal proteln, described in Schloeser et a/. (1993). Mouse and rat data were originally reported Schlosser et e/. (IQW); human data were reported in Seaton et a/. (1994). -Fig. 3 Metabolism of 4 pM benzene to phenol and hydroquinone from 16-min incubations wlth 1mglmi liver microsomal protein from a number of individual humans (HL 1-9), rats, and mice, shown as a function of cytochrome P450 (CYP) 2E1 activity. The amount of metabolite Is dlsplayed as a percentage of the total substrate and metabolite concentration (other metaboiltes not shown). Data (symbols) and model(solidline) are from Seeton eta/.(1994). (Adapted by permlsslon of Oxford University Press.) zene (Snyderand Kalf, 1994).Our results re- benzeneat the current permissibleexposure miainhumansisafundion of tissueresponse veal a greater capacity for activationand less limit of 1 ppm under the OccupationalSafety in addition to dosimetry. In particular, micc for detoxication in mice compared with rats, and HealthAct (Wallace, 1990). exhibit levels of the putative leukemogenic whereas human values include slow activa- The steady-state phenol and hydro- benzene metabolitesthat are at least as higt r'tion and detoxication (Table 2,HL2), moder- quinone concentrations for the individual hu- as is predictedin humans at exposure level: ate activation and rapiddetoxication(Table2, mans, rats, and mice are depicted in Figure assodated with AML in humans, but thesf HL 6),and rapid activation and slow detoxi- 4.According to modelcalculations, predicted levelsdo not give rise to AML in mice. There cation (Table2,HL 10).Observationsin mice steady-statephenolconcentrationsvariedsix- foremousebonemarrowapperenl)yrespond. and rats are in agreement with the sugges- fold among humans and correlated inversely to these agents in a m r that is qualita tion that susceptibility to benzene toxicity re- with measured microsomal activlty of CYP tivety different from humans. sults from a balance between oxidative and 2E1.This inverse relationship is most likely Variability in rates of conjugation plays i conjugativereactions.Moreover,theseobser- due to the fact that CYP 2E1 is also respon- definiteroleindeterminingsteady-stateblooc vations agreewith in viw data from Sabourin siMe for phenoloxidationto hydroquinone in concentrationsof phenol and hydroquinone et a/. (1987)and physiologicalmodel simula- this model. Inparticular,while the rateof phe- Although the in vivo rate constants for ben tions by Medkrsky eta/.(1989).Those stud- nol formation increases with CYP 2E1 activ- zene oxidation vary 13-foldamong humaq ies demonstratedgreater urinaryexcretionof ity, therateofTemovBJalsoincmaseswith CYP samples, the steady-state concentration c phenyl sulfate and less excretion of hydro- 2E1 activity, resulting in a net decrease in phenol in bloodvaries only 6-fold. Moreove quinone conjugates'in rats compared with concentration. Steady-state hydroquinone steady-stateconcentrationsof hydroquinon micefollowingbenzeneexposure, consistent concentrathsvariedfivefoklamonghumans. do not correlatewith rates of oxidation reac with Mgher CYP 2E1 activii and lower pha With laboratory animals, model simulations tions and appear to be influenced more b no1suifotransferaseactivity measuredin vifrr, predictedthat steadydate concentrationsof competition between phenol sulfation an in mouse liver. phenol and hydroquinonewould be higher in phenol oxidation and detoxicationdue to h) micethan in rats duringexposureto this very dmquinoneglucurod-dation. While the effec A PhyslologlcalCompartmental Model low benzene dose. These predictions are in of thehydroquinone:phendconcentrationri agreementwithhvivuobservationsfromstud- ti0inbloodon susceptibilityto benzenetoxic As discussed inthe introduction,benzene- ies with higher benzene concentrations ity is not dear, a syne-c relationship bt inducedhematotoxicttyandmyelotoxiciand (Sabourinetal.,1987).P r e d i isteady-state hveen these metabolites has been demor the riskofAMLinhumansareprobablyafunc- concentrations in mice were higher than in stratedk,m(Smithetd., 1989;lronseta tion of the circulating levels of oxidized me- humans, whereas rat values fell among the 1992;Kdachanaetal., 1993;Chapmaneta tabolitesthat occur duringbenzeneexposure. rangeof p r e d i i sfor humans.These simu- 1W). Gwen the range of activities for oxidation, lations suggest that the rat may be a good The results described above suggest sulfation, and glucuronidation indicated in model for h u m with respect to tissue do- significant interaction between Phase I (0) Table 2,what phenolandhydroquinoneblood simetry of benzene metabolites. dation)and Phase II (coniugaaion)pathwa) levels would OCCURWe have developed a Higher levels of oxidized metabolites in indeterminingbloodand tissue h l s of bel physiological cornmtmental model for esti- mice are due in part to the faster blood flow zene metabdies. The quality of these prc 1 mating steady-state blood concentrations of per unitbodyweight inthis smallspeciescom- dm mustbetestedaperimemtallyinlabc phenol and hydroquingne that might have paredwithhumans. More benzeneis brought ratMyanimalsinvivo beforeutilizingthem k been achieved durik continuous exposure to the l i r per unit time for metabolism. The h u m risk assessment Once this has bet of those 10individualsto 0.01 pMbenzene in model prediions for higher concentrations done,a modellikethat described here, takc blood (Seaton et a/., 1995).This Mood con- ofoxidizedmetabolitesinmouseblood, if cor- together withtheobserved rangesof enzyn centration is likely to result from exposure to rect, also suggestthat developmentof leuke- actMaiesamonghumans, can be usedto pr dict risk to individual humans with given e TABLE 2 INITIALRATES FOR BENZENE OXIDATION, PHENOLSULFATION, zyme activity profiles and sub?equently tt range of risk to humanpopulations. AND HYDROQUINONEGLUCURONIDATIONIN INDIVIDUAL HUMAN LIVER SAMPLES AND IN POOLED LIVER SAMPLES FROM MICEAND RATSa In Vivo Metabolism of Benzeneand Phend InMica Liver sample Human HL 1 HL 2 HL 6 HL 7 HL 10 \ I Mouse Rat Benzene oxidation (nmoldmglmln) 0.344 0.926 1.433 1.499 4.442 1.558 0.625 lnltial rate Phenol sulfatlon (nmoldmglmln) 0.309 0.581 0.867 0.919 0.485 0.485 1.195 Hydroqulnone glucuronidatlon (nmoldmglmln) ndb 0.117 0.281 0.167 0.106 0.218 0.077 The underlyinggoalof our in vivometab lism studies in mice has been the develo ment of aquantitative,mechanisticallybas description of the relationship between a phase benzene exposure concentratioi (ppm) and the amount of active metabolitc beingdeliveredto the target tissue. Oxidatit metabolismby CYP 2E1 isrequiredfor mar festation of the hemato-toxic and genoto) effects of benzene, but the dosimetry of a tive benzene metabolitesinthe target tissu bone marrow, is a net.resutt of competir toxification (oxidation) anddetoxification (00 jugation) reactions. Therefore we needto u derstandhowthese processesinteract in vi to determinetissue ummtmk*msofbenzet fTWtaboliteS. 'Results from Seaton et a/. (1995);bNotdetermineddue to insufficientsample. (Continued on page nzery Metabolism (from page 5) Phenol metabolism is of interest in this )textsince phenolis a majoroxidized me- olite of benzene, and there are distinct erences in both carcinogenic and lotoxic responses betweenbenzeneand molin mice. Forexample, while benzene :arcinogenic in a number of tissues in :e following both oral and inhalation ex- jure (Huff et a/., 1989), no increased in- 3nce of tumors was found in a two-year tional Cancer Institute(NCI) bioassay in ch male and female B6C3Fi mice were )osedto phenol in drinking water at levof up to 5000 ppm (NCI, 1980). Benle has been consistently foundto induce :ronuclei in mouse bone marrow cells ill, 1987). In contrast, both weakly posiand negativeresultshavebeenreported micronuclei induction in mouse bone rrow cells by phenol following intraperiea1 and oral administration (ATSDR, 19). These dramatic differences in carcinoiicity andgenotoxicity betweenbenzene i its major oxidized metabolite, phenol jure l ) , havepuzzledbenzeneresearchand led to different hypotheses about metabolitesresponsiblefor benzenetoxj. Insight into a possible metabolic basis the high potency of benzene as both an malcarcinogenandgenotoxicant andthe k of genotoxicity and carcinogenicity of mol can be gained by comparingthe uriy metabolite profiles after comparable I doses of benzene and phenol. This nparison is presented in Table 3, using data from CllT for urinary excretionof phenol metabolites following oral gavage of phenolin male B6C3Fi mice (Kenyoneta/., 1995) andsimilardata reportedby Sabourin eta/.(1989) following oral gavage with benzene. Differencesinthe excretionof hydroquinone glucuronideare quite marked, particularly at the lower dose (-1 mgkg). Hydroquinone glucuronide excretion in urine was sixfold and twofold greater after benzene comparedwith phenol at doses of approximately 1 and 10 mgkg, respectively. The in vitro studies described above indicate that hydroquinoneis produced by sequential oxidation of benzene via phenol. Howthen couldmore hydroquinonebe produced after administration of benzenethan after phenol? The differences in hydroquinoneformation after benzene and phenol administration are of interest for three reasons. First, hydroquinoneitself is a inducer of micronuclei and chromosomal aberrations in various test systems(IPCS, 1992).Thusgreater production of hydroquinone after benzene administrationcomparedwith phenoladministration, together with production of other putative toxic metabolites such as muconaldehydethat are uniqueto benzene, provide a possible metabolic basis for the carcinogenlc and genotoxic effects observed with benzene but not with phenol. Secondly, these data are in contrast to in vitro data (liver microsomes and isolated hepatocytes) that suggest greater production of hydroquinoneafter phenolcompared with benzene (Schlosser et a/., 1993; Schrenk and Bock, 1990). The implication io i Mouse ...................... ...................._.._....._....,.......... ....................A .................... h 5Y El i ...... ..................... ;AHL 2 j Ai 1 ",E. u- - g ............................................. j HL6 0A :;..................... : , 0 ........... .........-........ ...... .................... I..... ........... i HLl [Phenol] (nM) J- 4 eady-state blood concentrations of phenol and hydroqulnone for 10 lndlvldual imans, B6C3F1 mlce, and F344 rats when the blood concentratlon of benzene is Instant at 0.01 pM. Thls Mood concentratlon Is approximately what would be exoted for a contlnuous exposure to 1 ppm benzene. The steadyatate phenol and fdroquinone concentratlonr are calculatedfrom Invltn,oxidation and conjugatlon ,tlvitles, suchas those llstedInTable 1, uslng a physlologlcal compartmentmodel, , reported by Seaton et a/. (1995). (Adapted by permlsslon of Oxford Unlverslty e-.) here is that a mechanistic basis for this difference should be incorporated into our quantitative in vivo description of the relationship between benzene exposure and 3target tissue dosimetry. Finally, this differ- ence points out the need for caution when interpreting in vitro metabolism studies. We hypothesize that differences in the urinary metabolite profile of phenol compared with benzene following oral administration result from quantitative zonal differences in the distributionof metabolizingenzymes within the liver acinus, as illustrated in Figure 5. Phenol absorbed from the gut lumen has relatively greater opportunityfor conjugation than for oxidation, initially in gastrointestinalmucosalcells (sulfation and glucuronidation) and subsequently in the periportal hepatocytesof zone 1(sulfation) when absorbed into the portal circulation. The greater opportunity for phenol conjugationover oxidation is due to the relatively greater abundanceof conjugatingenzymes comparedwith oxidathm enzymesinthe areaswhere phenolwould be distributedwhen initially absorbed (Gebhardt, 1992; Morris and Pang, 1987). The net effect of this hypothetical scenario is that less phenol reacheshepaticzone 3,where oxidative enzyme activity is greatest, resulting in less hydroquinoneproduction. In contrast to phenol, benzene is only a E. 3substratefor CYP 2E1,and CYP2E1 activ- ity is much greater in the pericentral hepatocytes of zone 3 (Tsutsumi et el., 1989). i. Thus metabolismof benzenewould be minimal as benzene passes through periportal and midzonalhepatocytesprior to reaching hepatocytesin zone 3, where CYP 2El oxidation capacities are relatively greater (Gebhardt, 1992; Moms and Pang, 1987). Consequently, administration of benzene may result in the delivery of more free phenol to zone 3 oxidative enzymes than administration of phenol itself. The metabolic implicationsof these hypothetical scenarios are reflected in the observed urinary metabolite profiles presented in Table 3. Phe- nyl sulfate excretion is approximately two- fold higher and phenyl glucuronide excretion approximately four- to ninefold higher after phenol compared with benzene exposure, and oxidation products such as hydroquinone are higher after benzene exposure. Returning to the question of benzene toxicity, Smith et a/. (1989) hypothesized that benzoquinone, formed from the myeloperoxidase-mediatedoxidationof hydroquinone in the bone marrow, is responsible for benzene-induced hematotoxicity. The net effects of heterogeneous distribution of enzymes in the liver and the resulting preferential oxidation of benzene and conjugation of phenol are a higher blood levelof hydroquinoneafter benzeneadministrationcomparedwith phenoland more hy- - J \- f. TABLE 3 COMPARISON OF URINARY METABOLITES EXCRETEDAS PHENOL SULFATE, PHENOL GLUCURONIDE, AND HYDROQUlNONE GLUCURONlOEFOLLOWING GAVAGE ADMINISTRATION OF PHENOL OR BENZENE IN MALE B-F, MICE Total metabolites In urine (%p Chemical administered Dose (mghcg) Phenol sulfate Phenol glucuronide Hydroquinone glucuronide Benzeneb Phenol" 1 26 f 0.3 4.2 f 0.3 43 f 0.8 1.4 56 f 0.9 39 f 1.1 7.2 f 0.4 Benzeneb Phenol" 10 32 f 0.8 5.9 f 0.3 36 f 0.4 9.4 57 f 7.7 26 f 6.4 15 f 1.6 aValuesare means f SE. bunnarymetabolite data following benzene exposure are from Sabourin et a/.(1989). CData are from Kenyon et a/. (1995). (A) Benzene GI mucosa (6) Phenol Zone3 I Zone 2 4*;*i*:---- I* I Gut muGcoI sa lumen '' " W l PG Wl *,t *- 3 "I Fig. 5 Hypothesizedmajor relative hepatic zonal differences in metabolism of benzene (A) compared with phenol (B) foliowing gavage administration during an lnltlal pass through the liver. Symbols and abbreviations denote benzene(B), phenol b),hydroquinone (9, portal vein (PV), and the terminal hepatic venuk VI!).Benzene and phenol absorbed from the gastrointestinal (GI) tract are sequentially available for metabolism Inthe GI mucosaand periportal(zone 1)and prlcentral (zone 3)regions c of the liver. The overall capacity of GI mucosal metabolismIs low relative to the liver. Periportal locaiizatlon of the enzymes (sulfotransfemse) responsible for phenol sulfation suggests that less f r w phenol would reach the prlcentral hepatocytes where higher levels of CYP 2E1 activity are localized.The necessity for oxidation of benzene prior to conjugation suggeststhat more benrsnewill reach the perlcentral hepatocytes for oxidation to phenol and hydroqulnone. Phenol sulfate, PS; phenol glucuronide, PO. droquinone in bone marrowavailable foro dation to benzoquinone. Localization of z zymes inthe liver provides explanations differences both in observed toxicity and in vitrometabolismof benzene and phen Thus both the locationand the quantities benzene-metabolizingenzymes must be corporated into in vivo models to pred risks for humans exposed to benzene. The concept that the metabolicfate o metabolite derived from a precursor diffc from that of a preformed metabolite is r new. It was, in fact, proposed by Pang a Giilette in 1978for drugs. The kinetics o drug and its metabolite can differ due to ( ferences in diffusional barriersbetween dr and hepatocyte, the rate of communicati of the drugs between hepatocytes, and t distribution of drug-metabolizing enzyr systems within the liver parenchyma (Pa andTerrell, 1881). Usingthe drug-metat lite pair phenacetin and acetaminopht Pang and colleagues noted differences clearance of the chemicals from an isolat perfusedliverdependingon whether per sion was conducted in a normal (portal centrilobular) or retrograde(reversed) dirt tion. They attributed these differences preferential localization of P450 demethylation enzymes and sulfation 6 zymes indifferentzonesof theliver andst gested that existing models of well-stirr livers were oversimplifications and c o ~ only be used in limitingcaseswhm the d tributionof enzymes was viewed as ope tionally uniform. Using a combination of experimen and mathematical modeling approachc Pang and Terrell (lQ8l)determined tt substrate concentration at any point is fluenced.by preceding events during t metabolic processing of a substrate alo the lengthof the liversinusoid. Enzymes! tems such as sulfation, when present i stream along the sinusoidal flow pa modify residualsubstrateavailablefor me bolic activities such as oxidation witt downstream hepatocytes. In the case benzene and phenol, we propose that t residual phenolavailable for oxidationint centrilobutar region is very low when pt no1 is administered directly because mc of it is conjugated upstream of this regic In contrast, when benzene is administere the phenol is produced in the centrilobui region and thus is more available for su sequent oxidation to hydroquinone. Conclusions We cannot yet claim to have complett elucidated the factors that determine be zene metabolism and dosimetry, let aloi toxicity. Nevertheless, the research d scribed here does represent considerat (Continued on page Benzene Metabolism(from page 7) xogress in understandingand quantitating he factors that determine the distribution i f benzene and benzene metabolitesinthe mdy. In particular, we have seen that the luantity of a benzene metabolite produced s the result of subtle interplay among variIUS enzymes competing for substrates at a liven location, the distribution of those enymes in the liver, and relative ratesof perfu;ion in different species. Given the complex- ty of this system, itisnot surprisingthat many, Sometimes contradictory, mechanisms for )enzenemetabolismhavebeen proposed.At his point in time, many experimental obserrations must be explained by any proposed nodel. One example is the fact that in vivo idministration of benzene yields higher lev4s of hyclroquinonethan in vivo administraion of phenol but that in vitro metabolismof )heno1yields greater quantities of hydro- @nonethan h v i h metabolismof benzene. me mechanismsproposedhereare capable i f resolvingthesedisparateobservations and m be tested experimentally. In the case of iepaticenzyme localization, experimentswith solated hepatocytes(ina well-mixed suspen;ion) dbe compared with the resub of ~ormagl rade and retrograde perfused l i e r 3xperiments to determine if this Is indeed a significant factor. A second important point is that the indi- iidual pieces of research described here, aken together with further planned in vitn, 3xperiments, will allow ustoquantifythe relaionship between rates of biotransformation jetermined in vitta and benzene metabolite josimetry in vivo based on quantitative, nechanistic data. Once this has been done `or laboratoryanimals (miceand rats),we will lavea soundbasisfor predictingmetabolism and dosimetry in humans. These predictions Shouldthen leadto signicantty improvedrisk assessments for human exposures to benzene, since they will allow us to track the amount of critical benzene metabolites that reach the target tissue, bone marrow. Risk assessmentscanthen be basedontargettissue dosimetryratherthan on ambient air concentration.Giventhe differencesamongspecies and among individualhumans, such assessments should provide much better characterizations of risk. References AgencyforToxic SubstancesandDisease Registry (ATSDR) (1989). ToxicologEcalProfibfor Benzene. ATSDWP-88/03. Ayres, P. H. andTaylor,W. D. (1989). Solvents. In Principles and Methods of Toxicol- ~ ( 2 nedd.].(A. H.Hayes, ed.) RavenPress Ltd., NewYork, pP. 111-135. Barale, R., Marrazzini, A., Betti, C., Vangelisti, V., Lopneno, N., and Barrai, 1. (1990). Genotoxicity of two metabolites of benzene: Phenol and hydroquinone show strong synergistic effectsinvim. Mutat. Res, 244,1520. Chapman, D. E., Namkung, M. J., and Juchau, M. R. (1994). Benzeneandbenzene metabolitesasembryotoxicagents: Effectson cultured rat embryos. Toxicol. Appl. P h a m d . 128,129-137. Cox, L. A. (1991). Biological basis of chemical carcinogenesis: Insightsfrom benzene. RiskAnal. 11,453-464. CsanBdy, G. A., Guengerich, F. P., and Bond, J. A. (1992). Comparison of the biotransformation of 1,Sbutadiene and its metabolite, butadiene monoepoxide, by hepaticandpulmonarytissues from human, rats and mice. Carcinogenesis 13,1143-1153. Eastmond, D. A., Smith, M. T., and Irons, R. D. (1987). An interactionof benzene metabolites reproduces the myelotoxicity observed with benzene. Toxicoi. Appl. Phamacol. 91,85-95. Gebhardt, R. (1992). Metabolic zonation of the liver: Regulation and implitions for liverfunction. Pham~cdm. er. 53,275354. Guy, R. L., Hu, P., Witr, G.,Gddstein, B. D., and Snyder, R. (1991). Depressionof iron uptake intoerythrocytesinmiceby treatment with the combined benzene metabolites p benzoquinone, muconaldehyde and hydroquinone. J. M I . Toximl. 11,443-446. Health Effects Institute(1993). Research prioritiesto reduce uncertaintiesfor benzene. Research Prioritjes for Mobile Air Toxics. (Health Effects Institute, ed.) Health Effects Institute, Cambridge, MA, pp. 24-43. Hricko, A. (1994). Rings of controversy around benzene. Envim. Health Perspec$. 102,276-281. Huff, J. E., Haseman, J. K., DeMarini, D. M., Eustis, S., Maronpot, R. R., Peters,A. C., Persing, R. L., Chrisp, C. E., and Jacobs, A. C. (1989). Multiplesitecarcinogenicityof benzene in Fisher 344 rats and B6C3F1 mice. Environ. HealthPerspect. 82,125-163. Infante, P. E., Rinsky, R. A., Wagoner, J. K., andYoung, R. J. (1977). Leukemiain benzene workers. Lancet2, 76-78. International Programme on Chemical Safety (IPCS) (1992). EnvironmentalHealth Criteria for Hydroquinone. World Health Organization, Geneva, PCSEHCB2.8. Irons, R. D., Stillman, W. S., Colagiovanni, D. B., and Henry, V. A. (1992). Synergistic action of benzene metabolite hydroquinone on myelopoietic stimulating activity of granulocytedmacrophagecdony-stimulatinginv&o. Pm. Nafl.Acad. Sci. U SA 89,3691-3695. Kalf,G. F. (1987). Recentadvances inthe metabolismandtoxicity of benzene. Crit. Rev. TOXiCOl. 18, 141-159. Kari, F. W., Bucher, J., Eustis, S. L., Haseman,J. K., and Huff, J. E. (1992).Toxicity and carcinogenicity of hydroquinone in F344hl rats and B6C3F1 mice. FoodChem. TOxiwl.30,737-747. Kedderis, G. L., Batra, R., and Koop, D. R. (1993). Epoxidation of acrylonitrile by rat and human cytochrome P450. Chem, Res, Toxid. 6,866-871. Kenyon, E. M., Seeley, M. E., Janszen, 1D., and Medinsky, M. A. (1995). Dose, route and sexdependent urinary excretion of phe- no1 metabolites in W 3 F 1 mice. J. Toxiwi. Environ. Health44,219-233. Kolachana, P., Subrahmanyam, V. V., Meyer, K., Zhang, L., and Smith, M. (1993). Benzeneand its phenolii metabolitesproduce oxidative DNA damage in HL-60cells in vitro and in the bone marrow in vivo. CancerRes. 53,1023-1026. Koop, D. R., Laethem, C. L., andSchnier, G. G. (1989). Identificationof ethanol-inducible P450 isozyme 3a (P45011El) as a benzene and phenol hydroxylase. Toxicol. Appl. Pha/Tnacol. 98,278-288. MacEachem, L., Snyder, R., and Laskin, D. L (1992).Alterationsinthemorphdogyand functionalactivityof bonemarrowphagocytes followingbenzenetreatmentof mice. To&/. &pi. Phamcd. 117,147-154. Medinsky, M. A., Sabourin, P. J., Luder, G., Bimbaum, L. S., and Henderson, R. F. (1989). A physiological model for simulation m.of benzene metabolism by rats and mice. Toid. P~WEICO~9.9,i93-206. MonIs,M. E. andPang, K. S. (1987). Com- petition between two enzymes for substrate removalinliver:Modulatingeffectsdueto sub- strate recruitment of hepatocyte activii. J. `3,Phamawkinet. Biophann. 15,473-496. Nakajima, TIWang, R.-S., Elovaara, E., Park, S. S., Gelboin, H. V., and Vainio, H. (1993). CytochromeP450-relateddifferences between rats and mice in the metabolism of benzene,toluene andtrichloroethylenein liver microsomes. Biochem. Phannacol.45,1079- 1085. National Cancer Institute (NCI) (1980). BioassayofPhenol for Possible Carcinoge- nrcify Departmentof Healthand Human Ser- vices, Technical Report Series No. NCI-CG- TR-203. Pang, K. S. and Gillette, J. R. (1978). Ki- netics of metabolism and elimination in the perfusedrat liver preparation:Differencesbe- tween the elimination of preformed acetominophen and acetominophen formed from phenacetin. J, Phannawl. Exp. Thei 207, 178-194. Pang, K. S. andTerrell, J. .A. (1981). Ret- rogradeperfusionto probethe heterogeneous distribution of hepatic drug metabolizing en- zymes in rats. J. PhamawL f i p . 7heE 216, 339-346. Runion, H. E. andScott,L. M. (1985). Ben- zene exposure in the United States 1978- 1983: An overview. Am. J. Ind. Med. 7,385- )*393. Sabourin, P. J., Bechtold, W. E., Griffith, W. C., Birnbaum, L. S., Lucier, G., and Henderson, R. F.(1989). Effectof exposure concentration, exposure rate and route of administrationon metabolismof benzene by 8 , F344 rats and B6C3F1mice. Toxicol.Appl. Pharmacol.99, 421-444. Sabourin, P. J., Chen, B. T., Lucier, G., Birnbaum, L. S., Fisher, E.,and Henderson, R. F. (1987). Effect of dose on the absorp- tion and excretion of [14C]benzene admin- istered orally or by inhalation in rats and mice. Toxicol.Appl. Pharmacol. 87, 1-12. Sammett, D., Lee, E. W., Kocsis, J. J., and Snyder, R. (1979). Partial hepatectomy reduces both metabolismandtoxicityof ben- zene. J. Toxicol.Environ. Health 5,785-792. Schlosser, P. M., Bond, J. A., and Med- insky, M. A. (1993). Benzene and phenol metabolism by mouse and rat liver mi- crosomes. Carcinogenesis 14, 2477-2486. Schrenk, D. and Bock, K. W. (1990). Me- tabolism of benzenein rat hepatocytes: Influ- ence of inducers of phenol glucuronidation. Drug Metab. Dispos. 18,720-725. Seaton, M. J., Schlosser, P. M., Bond, J. A., and Medinsky, M. A. (1994). Benzene metabolism by human liver microsomes in Left to right: Drs. Medinsky, Kenyon, Seaton,and Schlmsser relation to cytochrome P450 2E1 activity. Carcinogenesis 15, 1799-1806. Seaton, M. J., Schlosser, P. M., Bond, J. The Authors A., and Medinsky, M.A. (1995). In vitroconju- gationof benzenemetabolitesby humanliver: Paul Schiosser received his Ph.D. in chemicalengineering from the University Potentialinfluenceof interindividualvariability Rochester in 1988 and subsequently performedpostdoctoral researchat the Califc on benzene toxicity. Carcinogenesis 16(7), nla Institute of Technology. Dr. Schlosser JoinedCllT In 1991, with his initial foc 1519-1527. primarily on the benzene project. He has an adjunct faculty position at North Cai Smith, M. T., Yager, J. W., Steinmetz, K. iina State Universlty and is on the editorial review board of Non//near T/mes a L., and Eastmond, D.A. (1989). Peroxidase- U/gest He has a general Interwt In the modeling of complex bloioglcalchemlc ' dependent metabolism of benzene's phe- physical systems. Other projects at CllT in which he Is involved are dosimetry nolic metabolites and its potential role in inhaled gases, endocrine toxicology, and genetic susceptlbiilty. benzene toxicity and carcinogenicity. el el^ M. Kenyon, who received her Ph.D. In toxicology from the Universtty Environ. Health Perspect. 82, 23-29. Massachusetts at Amhent, was a postdoctoralfellow at CllT from 1992to early 1% Snyder, C. A., Goldstein, B. D., While at CilT, she conducted In VIVOdosimetry researchon benzene and extendm Sellakumar, A., Wolman, S. R., Bromberg, physiologically based pharmacokinetlc(PBPK) model for benzeneto predict the c I., and Laskin, S. (1978). Hematotoxicity of simetry of benzene, phenol, and hydroquinone Inbone marrow. This work was s i inhaled benzene to Sprague-Dawiey rats ported in part by a National Research Service Award (NRSA) from the NationalIns and AKR mice at 300 ppm. J. Toxicol. tute of Environmental Health Sciences. Dr. Kenyon Is currently with the Health I Environ. Health 4, 605-618. fects ResearchLaboratory, U.S. Environmental ProtectionAgency, ResearchTrians Snyder, R. and Kalf, G. F. (1994).A per- Park, NC, where she is conducting research in the area of pharmacokineticsto c spective on benzene leukemogenesis. Crit. velop PBPK models for both volatile organics and metailolds such as arsenic. Rev. Toxicol.24, 177-209. Mark J. Seaton received his Ph.D. in pharmacologyfrom PennsylvaniaState U Tompa, A., Major, J., and Jakab, M. G. versity in 1993 and then came to CllT as a postdoctoral fellow to study In v/rrotv (1994). Monitoring of benzene-exposed tabolism of butadtene by human liver. Dr. Seaton has been more recently involv workers for genotoxic effects of benzene: with In vltro studies of butadiene metabolism by hepatic and pulmonary cells. Tt Improved-working-condition-related de- work Is supported In part by an NRSA. He Is currently performing research on t crease in the frequencies of chromosomal metabolic and toxic effects of coexposures to benzene and unleaded garroline. aberrationsin peripheralbloodlymphocytes. Michele A. Medinsky is a senior scientist at CllT who has had a long-standi Mutaf. Res. 304, 159-165. interest in assessing the health risks of airborne materials. She conducted her d Tsutsumi, M., Lasker,J. M., Shimizu, M., sertation research at Lovelace inhalation Toxicology Research institute (ITRI) in I Rosman,A. S., and Lieber,C. S. (1989). The buquerque, NM, and receivedher Ph.D. Inbiologyfrom the Universityof New Mexic intralobulardistributionof ethanol-inducible Followinga postdoctoralappointment at CIIT, Dr. Medinsky accepted a staff posit11 P45011E1in rat and human liver. Hepatology at iTRi and then returned to CiiT in 1989 as a staff scientist, where she serves IO,437-446. coordinator for the benzene prolect and participates in several other projects cc Wallace L. (1990). Major sources of ex- cerned with evaluating factors influencing the disposition of inhaled chemlcais. I posureto benzene andother volatile organic Medinsky is a Diplomate of the American Board of Toxicology. She currently serv chemicals. RiskAnal. IO,59-64. as 8 member of the NAS-NRC Committee of the Society of Toxicology.