BIOINORGANIC CHEMISTRYBIOINORGANIC CHEMISTRYIVANO BERTINIUniversity of FlorenceHARRYB. GRAYCalifornia Institute of TechnologySTEPHENJ. LIPPARDMassachusettsInstitute of TechnologyJOANSELVERSTONE VALENTINEUniversity of California, Los Angeles~.University Science BooksMill Valley, CaliforniaUniversity Science Books20 Edgehill RoadMill Valley, CA94941Fax (415) 383-3167This book is printed on acid-free paper.Copyright 1994 by University Science BooksReproduction or translation of any part of this work beyond thatpermitted by Section 107 or 108 of the 1976 United States CopyrightAct without the permission of the copyright holder is unlawful.Requests for permission or further information should be addressed tothe Permissions Department, University Science Books.Library of Congress Cataloging-in-Publication DataBioinorganic chemistry / authors/editors Ivano Bertini, Harry B. Gray,Stephen Lippard, Joan Valentine.p. em.Includes bibliographical references.ISBN 0-935702-57-1: $58.001. Bioinorganic chemistry. I. Bertini, Ivano.QP531.B543 1994574. 19'214-dc20 91-67870CIPPrinted in the United States of America10 9 8 7 6 5 4 3 2 1ContentsList of Contributors viPreface viiAcknowledgments viii1Transition-Metal Storage, Transport, andBiomineralization 1ELIZABETH C. THEIL and KENNETHN.RAYMOND2 TheReactionPathways of Zinc Enzymes andRelatedBiological Catalysts 37IVANO BERTINI and CLAUDIO LUCHINAT3 Calciumin Biological Systems 107

4 Biologicaland Synthetic Dioxygen Carriers 167GEOFFREY B. JAMESON and JAMES A. IBERS5 DioxygenReactions 253JOAN SELVERSTONEVALENTINE6 Electron Transfer 315HARRY B. GRAY and WALTHERR. ELLIS, JR.7Ferredoxins, Hydrogenases, andNitrogenases: Metal-SulfideProteins 365EDWARDI. STIEFEL and GRAHAM N. GEORGE8 Metal/Nucleic-AcidInteractions 455JACQUELINEK. BARTON9Metals in Medicine 505STEPHEN J.LIPPARDSuggestedReadings 585Index 593viList of ContributorsNumbers in parentheses indicate the pages on whichthe authors' contributions begin.JACQUELINE K. BARTON (455), Division of Chemistry and Chemical Engineering,California Institute of Technology, Pasadena, California 91125.IVANO BERTINI (37), Department of Chemistry, University of Florence, Via GinoCapponi 7, 50121 Florence, Italy.WALTHER R. ELLIS, JR. (315), Department of Chemistry, University of Utah, SaltLake City, Utah 84112.STURE FORSEN (107), Physical Chemistry 2, Chemical Centre, University of Lund,P. O. Box 124, S-22100, Lund, Sweden.GRAHAM N. GEORGE (365), Stanford Synchrotron Radiation Laboratory, P. O. Box4349, Bin 69, Stanford, California 94309.HARRY B. GRAY (315), Beckman Institute, California Institute of Technology,Pasadena, California 91125.JAMES A. lEERS (167), Department of Chemistry, Northwestern University,Evanston, Illinois 60208.GEOFFREY B. JAMESON (167), Department of Chemistry, Georgetown University,Washington, D.C. 20057.JOHAN KORDEL (107), Physical Chemistry 2, Chemical Centre, University of Lund,P. O. Box 124, S-221 00 Lund, Sweden.STEPHEN J. LIPPARD (505), Department of Chemistry, Massachusetts Institute ofTechnology, Cambridge, Massachusetts 02139.CLAUDIO LUCHINAT (37), Institute of Agricultural Chemistry, University ofBologna, Via1e Berti Pichat 10,40127 Bologna, Italy.KENNETH N. RAYMOND (1), Department of Chemistry, University of California,Berkeley, California 94720.EDWARD I. STIEFEL (365), Exxon Research and Engineering Company, ClintonTownship, Rt. 22 East, Annandale, New Jersey 08801.ELIZABETH C. THEIL (1), Department of Biochemistry, North Carolina StateUniversity, Raleigh, North Carolina 27695-7622.JOAN SELVERSTONE VALENTINE (253), Department of Chemistry andBiochemistry, University of California, Los Angeles, California 90024.PrefaceThisbookcoversmaterial that couldbeincludedinaone-quarterorone-semestercourseinbioinorganicchemistryforgraduatestudentsandadvancedundergraduatestudentsinchemistryorbiochemistry. Webelievethat suchacourseshouldprovidestudentswiththebackgroundrequiredtofollowtheresearch literature in the field. The topics were chosen to represent those areas ofbioinorganicchemistrythat arematureenoughfor textbookpresentation.Althougheachchapter presentsmaterial at amoreadvancedlevel thanthat ofbioinorganictextbookspublishedpreviously, thechaptersarenot specializedreviewarticles. Whatwehaveattemptedtodoineachchapteristoteachtheunderlyingprinciples of bioinorganicchemistry aswellasoutlining thestate ofknowledge in selected areas.Wehavechosennot toincludeabbreviatedsummariesof theinorganicchemistry, biochemistry, and spectroscopy that students may need as backgroundin order tomaster thematerialpresented. Weinstead assumethatthe instructorusingthisbook will assignreadingfromrelevantsourcesthat isappropriatetothe background of the students taking the course.Fortheconvenienceof theinstructors, students, andotherreadersof thisbook, wehaveincludedanappendixthat listsreferences toreviews of theresearch literature that we have found to be particularly useful in our courses onbioinorganic chemistry.viiviiiAcknowledgmentsThe idea of preparinga bioinorganicchemistry textbook wasconceived byoneof us(IB)ata"MetalsinBiology"GordonConferenceinJanuary, 1986. Thecontributingauthorswere recruited tothe projectshortly thereafter. The projectevolved as a group effort, with substantial communication among the authors atall stagesof planningandexecution. Bothfirst andreviseddraftsof thebookwereclass-testedat UCLA, Caltech, andtheUniversityof Wisconsinandmodifiedinresponsetothereviewsofstudents andteachers. ParticularlyvaluablesuggestionsweremadebyProfessor JudithN. Burstyn(UniversityofWisconsin);Ken Addess, Raymond Ro, Kathy Kinnear, Clinton Nishida, RogerPak, MarleneSisemore(UCLA); andDeborahWuttke(Caltech) duringthereview process. We thank them for their contributions.Even with all this help, the book would never have seen the light of day hadit not beenfor thededicationandhardworkof DebbieWuttke. WithRBG,Debbiecheckedeverylinethroughfourroundsof galleysandpages. Graziemille, Debbie!Ivana BertiniHarry B. GrayStephen J. LippardJoan Selverstone ValentineBIOINORGANIC CHEMISTRY1Transition-Metal Storage, Transport,andBiomineralizationELIZABETHC. THEILDepartment ofBiochemistryNorthCarolina StateUniversityKENNETHN. RAYMONDDepartment of ChemistryUniversity of California atBerkeleyI. GENERALPRINCIPLESA. Biological Significanceof Iron, Zinc, Copper, Molybdenum, Cobalt,Chromium, Vanadium, andNickelLivingorganisms storeandtransport transitionmetals bothtoprovide appro-priate concentrations of them for useinmetalloproteinsor cofactorsandtopro-tect themselvesagainst thetoxiceffectsof metal excesses; metalloproteinsandmetal cofactorsarefoundinplants, animals, andmicroorganisms. Thenormalconcentrationrangefor eachmetal inbiological systemsisnarrow, withbothdeficienciesandexcessescausingpathological changes. Inmulticellularorgan-isms, composedof avarietyofspecializedcell types, thestorageof transitionmetalsandthesynthesisof thetransportermoleculesarenot carried out byalltypes ofcells, but rather by specificcells that specializeinthese tasks. Theform of themetalsisalwaysionic, buttheoxidationstatecanvary, dependingon biological needs. Transition metals for which biological storage and transportaresignificant are, inorder of decreasingabundanceinlivingorganisms: iron,zinc, copper, molybdenum, cobalt, chromium, vanadium, and nickel. Althoughzinc is not strictly a transition metal, it shares many bioinorganic properties withtransition metals and is considered with them in this chapter. Knowledge of ironstorageandtransport ismore complete than forany other metalin thegroup.Thetransitionmetalsandzincareamongtheleast abundantmetal ionsintheseawater fromwhichcontemporaryorganismsarethoughttohaveevolved(Table1.1).1-5For manyof themetals, theconcentrationinhumanblood plasma21I TRANSITION-METAL STORAGE, TRANSPORT, ANDBIOMINERALIZATIONTable1.1Concentrations of transitionmetalsandzincinsea water andhumanplasma.aSea water Human plasmaElement (M) X 108(M) X 108Fe 0.005-2 2230Zn 8.0 1720Cu 1.0 1650Mo 10.0 1000Co 0.7 0.0025Cr 0.4 5.5V 4.0 17.7Mn 0.7 10.9Ni 0.5 4.4a Data fromReferences1-5and12.greatlyexceedsthatinseawater. Suchdataindicatetheimportanceof mecha-nismsforaccumulation, storage, andtransportof transitionmetalsandzincinliving organisms.The metals are generally foundeither bound directly to proteins or in cofac-torssuchasporphyrinsorcobalamins, orinclustersthatareintumbound bythe protein; the ligands are usually 0,N, S,or C. Proteins with which transitionmetals andzincaremost commonlyassociatedcatalyzetheintramolecular orintermolecular rearrangement of electrons. Although theredox propertiesof themetals areimportant inmanyofthereactions, inothers themetal appears tocontribute tothe structure of the active state, e.g., zinc in the Cu-Zn dismutasesandsomeof theironinthephotosyntheticreactioncenter. Sometimesequiva-lent reactions arecatalyzedbyproteins withdifferent metal centers; themetalbinding sites and proteins have evolved separately for each type of metal center.Ironisthemostcommontransitionmetal inbiology. 6,7Itsusehascreatedadependencethat has survivedtheappearanceof dioxygenintheatmosphereca. 2.5 billion years ago, and the concomitant conversion of ferrousion to ferricion andinsoluble rust(Figure1.1 Seecolor platesection, pageC-1.). Allplants,animals, andbacteriauseiron, exceptfor a lactobacillusthat appearstomain-tainhighconcentrationsof manganeseinsteadof iron. Theprocessesandreac-tionsinwhichironparticipatesarecrucial tothesurvival of terrestrial organ-isms, and include ribonucleotide reduction(DNAsynthesis), energy production(respiration), energy conversion (photosynthesis), nitrogenreduction, oxygentransport (respiration, musclecontraction), andoxygenation(e.g., steroidsyn-thesis, solubilizationanddetoxification of aromatic compounds). Amongthetransition metals usedin living organisms, ironis the most abundant intheenvironment. Whether this fact aloneexplains thebiological predominanceofiron or whether specific featuresof iron chemistry contributeis not clear.I. GENERALPRINCIPLESManyof the other transitionmetals participate inreactions equivalent tothoseinvolvingiron, andcansometimes substitutefor iron, albeit less effec-tively, innatural Fe-proteins. Additional biological reactionsareunique tonon-ferroustransition metals.Zincisrelativelyabundant inbiological materials.8,9Themajor location ofzincinthebodyismetallothionein, whichalsobindscopper, chromium, mer-cury, andother metals. Amongthe otherwell-characterizedzincproteins aretheCu-Znsuperoxidedismutases(other formshaveFeor Mn), carbonicanhy-drase(an abundant protein in red blood cells responsible formaintaining the pHof theblood), alcohol dehydrogenase, andavarietyof hydrolases involvedinthemetabolismofsugars, proteins, andnucleicacids. Zincis acommonele-ment in nucleic-acidpolymerases andtranscriptionfactors, where its role isconsidered tobestructural rather than catalytic. Interestingly, zinc enhancesthestereoselectivityof thepolymerizationof nucleotidesunderreactionconditionsdesignedtosimulatetheenvironmentforprebioticreactions. Recentlyagroupof nucleic-acid binding proteins, with a repeated sequence containing theaminoacidscysteineandhistidine, wereshowntobindasmanyaselevenzincatomsnecessary for protein function (transcribing DNA to RNA). 10 Zinc playsa struc-tural role, formingthepeptideintomultipledomainsor "zincfingers" bymeansof coordinationtocysteineandhistidine(Figure1.2ASeecolor platesection,pageC-l.). Asurveyofthesequencesofmanynucleic-acidbindingproteinsshowsthat many of them have the common motif required to formzinc fingers.Other zinc-finger proteins calledsteroidreceptors bindbothsteroids such asprogesteroneandtheprogesteronegeneDNA(Chapter 8). Much of thezincinanimalsand plantshasno known function, but it maybe maintaining thestruc-turesof proteinsthat activateand deactivategenes. 11Copper andironproteinsparticipateinmanyof thesamebiological reac-tions:(1) reversiblebindingof dioxygen, e.g., hemocyanin(Cu), hemerythrin(Fe),andhemoglobin(Fe);(2) activationof dioxygen, e.g., dopaminehydroxylase(Cu) (importantinthe synthesis of the hormoneepinephrine), tyrosinases (Cu), andca-techoldioxygenases(Fe);(3) electrontransfer, e.g., plastocyanins(Cu), ferredoxins, andc-typecy-tochromes(Fe);(4) dismutation of superoxide by Cu or Fe as the redox-active metal (super-oxide dismutases).Thetwometal ionsalsofunctionin concert in proteinssuchascytochromeoxidase, which catalyzes the transfer of four electrons to dioxygen to form waterduring respiration. Whether anytypesof biological reactionsare unique tocop-per proteins is not clear. However, use of storedironis reducedby copperdeficiency, which suggests that iron metabolism may depend on copper proteins,341/ TRANSITION-METALSTORAGE, TRANSPORT, ANDBIOMINERALIZATIONsuchas theserumproteinceruloplasmin, whichcanfunctionas aferroxidase,and thecellular protein ascorbicacid oxidase, whichalsoisa ferrireductase.Cobalt isfoundinvitaminB12, itsonlyapparent biological site.12Thevi-taminis acyanocomplex, but amethyl or methylene groupreplaces CNinnativeenzymes. Vitamin-B12 deficiencycausestheseverediseaseof perniCiousanemia in humans, which indicates the critical role of cobalt. The most commontype of reaction in which cobalamin enzymes participate results in the reciprocalexchange of hydrogen atomsif theyareonadjacent carbonatoms, yetnotwithhydrogen insolvent water:b c b cI I I Ia - C - C - d ~ a - C - C - dI I I Ix H H X(An important exception isthe ribonucleotidereductase fromsome bacteria andlowerplants, whichconvertsribonucleotides totheDNAprecursors, deoxyri-bonucleotides, a reaction in which a sugar -OH is replaced by -H.Notethatribonucleotidereductasescatalyzingthesamereactioninhigherorganismsandviruses are proteins with an oxo-bridged dimeric ironcenter.) The cobalt invitaminB12iscoordinatedtofive Natoms, four contributedbyatetrapyrrole(corrin); thesixthligandisC, providedeither byC5of deoxyadenosineinen-zymessuchasmethylmalonyl-CoA mutase(fattyacid metabolism)or bya methylgroupin the enzymethatsynthesizestheaminoacidmethionine in bacteria.Nickel isacomponentof ahydrolase(urease), of hydrogenase, of COde-hydrogenase, and of S-methyl CoM reductase, which catalyzes the terminal stepinmethaneproductionbymethanogenicbacteria. AlltheNi-proteinsknowntodatearefromplantsorbacteria. 13,14 However, about50yearselapsedbetweenthe crystallization of jack-bean urease in1925 and the identification of the nickelcomponent intheplant protein. Thusit isprematuretoexcludethepossibilityofNi-proteins inanimals. Despitethe small number ofcharacterizedNi-pro-teins, it isclearthat manydifferentenvironmentsexist, fromapparentlydirectcoordinationtoproteinligands (urease) tothe tetrapyrrole F430inmethylre-ductaseandthemultiplemetal sitesof Ni andFe-SinahydrogenasefromthebacteriumDesulfovibriogigas. Specificenvironments for nickel arealsoindi-catedfornucleicacids(ornucleicacid-bindingproteins), sincenickelactivatesthegene forhydrogenase. 15Manganeseplays acritical roleinoxygenevolutioncatalyzedbythepro-teins of the photosynthetic reaction center. The superoxide dismutase of bacteriaandmitochondria, aswell aspyruvatecarboxylaseinmammals, arealsoman-ganeseproteins. 16,17 Howthemultiplemanganeseatoms of thephotosyntheticreactioncenter participate inthe removal offour electrons andprotons fromwater isthesubject of intenseinvestigationbyspectroscopists, syntheticinor-ganicchemists, andmolecular biologists. 17I. GENERALPRINCIPLESVanadium andchromium haveseveral featuresin common, froma bioinor-ganicviewpoint. 18a First, both metalsare present in only small amountsin mostorganisms. Second, thebiological rolesof eachremainlargelyunknown.18Fi-nally, eachhasservedasa probetocharacterizethesitesof other metals, suchasironandzinc. Vanadium isrequiredfornormalhealth, andcouldact invivoeither asa metalcation or asa phosphate analogue, dependingontheoxidationstate, V(lV)or V(V), respectively. Vanadium in a sea squirt (tunicate), a prim-itivevertebrate(Figure1.2B), isconcentratedinbloodcells, apparentlyasthemajor cellular transitionmetal, but whether it participates inthe transport ofdioxygen (as iron and copper do) is not known. Inproteins, vanadiumis acofactor inanalgal bromoperoxidase andincertainprokaryotic nitrogenases.Chromium imbalance affectssugar metabolism and has been associated with theglucosetolerancefactor inanimals. But littleis knownabout thestructureofthefactor orofanyotherspecificchromiumcomplexesfromplants, animals,or bacteria.Molybdenum proteinscatalyze the reduction of nitrogenand nitrate, aswellastheoxidationof aldehydes, purines, andsulfite. 19 FewMo-proteinsareknowncomparedtothoseinvolvingothertransitionmetals. Nitrogenases, whichalsocontain iron, have been the focusof intense investigations by bioinorganic chemistsand biologists; theiron isfoundina clusterwith molybdenum(theiron-molyb-denumcofactor, or FeMoCo)andinaniron-sulfur center (Chapter 7). Interest-ingly, certain bacteria (Azotobacter) have alternative nitrogenases, which areproducedwhenmolybdenumisdeficientandwhichcontainvanadiumandironoronlyiron. All otherknownMo-proteins arealsoFe-proteinswithironcen-ters, suchas tetrapyrroles (hemeandchlorins), Fe-sulfurclusters, and, appar-ently, non-heme/non-sulfur iron. Some Mo-proteins contain additional cofactorssuch asthe Havins, e.g., in xanthine oxidase and aldehyde oxidase. The numberofredoxcenters insomeMo-proteins exceeds the number ofelectrons trans-ferred; reasonsforthisareunknowncurrently.B. Chemical PropertiesRelative toStorageand Transport1. IronIron is the most abundant transition element intheEarth's crust and, ingeneral, inall lifeforms. Anoutlineofthedistributionof ironintheEarth'scrust20,21 isshown in Table1.2. Ascan beseen, approximately one-third of theEarth's mass is estimated to be iron. Of course, only the Earth's crust is relevantfor life forms, but eventhere it is the most abundant transitionelement. Itsconcentration is relatively high in most crustal rocks(lowest in limestone, whichis more or less pure calciumcarbonate). Inthe oceans, whichconstitute 70percent of the Earth's surface, the concentrationofironis lowbut increaseswithdepth, sincethisironexistsas suspendedparticulatematter ratherthanasa soluble species. Iron is a limitingfactor inplanktongrowth, and the rich561/ TRANSITION-METALSTORAGE,TRANSPORT, ANDBIOMINERALIZATIONTable1,2Iron: Its terrestrial distribution.aOne third of Earth'smass, most abundant element byweightDistribution in crustal rocks(weight%):igneous 5.6shale 4.7sandstone 1.0limestone 0.4Ocean(70% of Earth'ssurface):0.003--0.1ppb, increasingwith depth; limitingfactorin planktongrowthRivers:0.07-7 ppmKsp forFe(OH)3isapproximately10 -39, henceat pH7[Fe3+] 10 -IS Ma Data fromReferencesla and 20.fisheriesassociated with strong up-welling of ocean depthsresult at least in partfrom thebiologicalgrowthallowed by theseiron supplies. Properties that dom-inatethetransportbehavior of most transitionmetal ionsare: (l)redoxchem-istry, (2) hydrolysis, and(3) thesolubilityofthemetal ions invarious com-plexes, particularly thehydroxides.Asan example of theeffectsof solubility, consider theenormousvariationin the concentration of ironin rivers, depending on whether thewater is fromaclear mountainstream runningover rock or a muddy river carryinglargeamountsof sediment. However, theamount of dissolvedironintheformof freeferricionor itshydrolysisproducts, whatever thesourceof water, isextremelylow.Ascanbeseenfromthesolubilityof hydratedFe(III) (Ks~ 10 -18 M) (Table1.2), the concentration of freeferricion isextraordinarily low at neutralpH;sosignificant concentrationsof solubleironspeciescan beattained only bystrongcomplexformation.Oneexampleof theversatilityofironas afunctionofits environment ishowtheligand fieldcanstronglyalter thestructuralandligandexchangeprop-ertiesofthemetal ion(Figure 1.3). Theligandfieldcanalsoaltertheredoxproperties. For high-spinferricion, asfoundintheaquocomplexorinmanyother complexes (includingthe class ofmicrobial iron-transport agents calledsiderophores, tobediscussedlater), thecoordination geometryisoctahedralorpseudo-octahedral. Intherelativelyweakligandfield(high-spingroundstate),thecomplexishighlylabile. Inastrongligandfield, suchasanaxiallyligatedporphyrincomplex of ferricion, or thesimpleexample of theferrocyanidean-ion, the low-spin complex is exchange-inert. Similarly,the high-spin octahedralferrous complexes are exchange-labile, but the corresponding axially ligatedporphyrincomplexes, ortheferrocyanidecomplexes, arespin-paired(diamag-netic) andligandexchange-inert. Large, bulkyligands orconstrainedligands,suchasthoseprovidedbymetalloproteinandenzymesites, cancauseatetra-hedral environment, inwhichbothferrous ionandferric ionformhigh-spincomplexes.7octahedral~ ~Fe3+~ ~ ~ lLlL~high-spin, labile low-spin, inert~ ~Fe2+lL~ ~ lLlLlLhigh-spin, labile low-spin, inerttetrahedral~ ~ ~ ~ ~ ~Fe3+~ ~Fe2+lL~Figure1.3Versatility of Fe coordination complexes.The distribution of specific iron complexes in living organisms dependsstrongly on function. For example, although therearemany different iron com-plexesin the average human, the relative amounts of each type differ more than650-fold(Table1.3). Thetotal amount of ironinhumansisquitelarge, aver-agingmore than three and upto fivegrams for a healthy adult. Most of the ironis presentashemoglobin, theplasma oxygen-transport protein, where thefunc-Table1.3AveragehumanFedistribution.ProteinHemoglobinMyoglobinTransferrinFerritinHemosiderinCatalaseCytochrome cOtherFunctionPlasma O2transportMuscleO2storagePlasma Fe transportCellFestorageCellFestorageH20 z metabolismElectron transportOxidases, other enzymes, etc.Oxidation stateof Fe22333223"Amount of Fe(g)2.60.130.0070.520.480.0040.0040.14Percent of total6560.213120.10.13.681/ TRANSITION-METAL STORAGE, TRANSPORT, ANDBIOMINERALIZATIONtionof theironistodeliver oxygen forrespiration. A muchsmaller amount ofironis present inmyoglobin, amuscleoxygen-storageprotein. Fortransport,themost important ofthese iron-containingproteinsistransferrin, theplasmairon-transportproteinthattransfersironfromstoragesitesinthebodytoloca-tionswherecellssynthesizingiron proteinsreside;themajor consumersof ironinvertebrates are the redbloodcells. However, at anygiventimerelativelylittle of the iron in the body is present in transferrin, in much thesame way thatatanygiventimeina largecityonlyasmallfractionof thepopulation willbefoundinbuses or taxis. Otherexamples ofiron-containingproteins andtheirfunctionsare included in Table1.3 for comparison.Anexample of different iron-coordinationenvironments, which alter thechemical properties of iron, is thedifferencein the redoxpotentialsof hydratedFe3+and the electron-transport protein cytochrome c (Table1.4).The co()rdina-Table1.4Feredoxpotentials.ComplexFe(OHz)63+CytochromeG3HIPIPCytochromecRubredoxinFerredoxinsCoord. no., type6, aquocomplex6, heme4, Fe4S4(SR)4 -6, heme4, Fe(SR)44, Fe4S4(SR)4Z-770390350250-60-400tionenvironment of ironincytochromecisillustratedinFigure 1.4. Forex-ample, the standardreductionpotential for ferric ioninacidsolutionis 0.77volts; sohereferricionisquiteagoodoxidant. Incontrast, cytochromec hasaredoxpotential of0.25 volts. Awiderange ofredoxpotentials for ironisachieved in biology bysubtledifferencesin proteinstructure, aslisted in Table1.4. Notice the large difference in the potential of cytochrome c and rubredoxin(Figure1.5), 0.25volts vs. -0.06 volts, respectively. In polynuclear ferredox-ins, inwhicheachironistetrahedrallycoordinatedbysulfur, reductionpoten-tials arenear - 0.4volts. Thus, theentirerangeof redoxpotentials, asillus-trated in Table1.4, ismore than one volt.2. Chemical properties of zinc, copper, vanadium, chromium,molybdenum, andcobaltThechemical propertiesoftheotheressential transitionelementssimplifytheir transport properties. For zinc there isonly the+2 oxidation state, and thehydrolysis of thision is not a limiting featureof itssolubility or transport. Zincisan essential element for bothanimalsand plants.8,9,20,21 In general, metalionuptakeintotherootsof plantsisanextremelycomplex phenomenon. Across-sectionaldiagram of a root isshown in Figure1.6. It issaid that bothdiffusionHH C I C=C-CH-3 " / 2/S-F1e-N" IMet80 CH C=N/ 2 H/CH2 t His18hemegroupFigure1.4 IHeme group and iron coordinationin cytochromec.9cysteine2 aminoacidcysteine --r-e""'si--:d'--ue-"s-'--'--Figure1.5Fe3+/2+ coordinationinrubredoxin.- stele-phloemtubes pericyclecortex ------airspacexylemvesselsendodermiswithCasparianbandepidermiswithroot hairFigure1.6Transversesection of a typicalroot. 20 The complex featuresof the root hair surface that regulatereductaseand other activitiesinmetal uptakeare only beginningtobe understood.10 1/ TRANSITION-METAL STORAGE, TRANSPORT, ANDBIOMINERALIZATIONandmassflowof thesoil solution areof significanceinthemovement of metalionstoroots. Chelationandsurfaceadsorption, which-arepHdependent, alsoaffect the availability of nutrient metal ions. Acidsoil conditions ingeneralretard uptake of essential divalent metal ions but increase the availability(some-times withtoxicresults) of manganese, iron, andaluminum, all ofwhicharenormally of verylimited availabilitybecause of hydrolysis of thetrivalent ions.Vanadiumis oftentakenup as vanadate, in apathway parallel tophos-phate. 18 However, itsoxidationstatewithinorganismsseemstobehighlyvari-able. Unusuallyhigh concentrations of vanadium occur in certain ascidians(thespecifictransport behavior of whichwill bedealt withlater). Theworkerswhofirst characterizedthevanadium-containingcompoundof thetunicate, Ascidianigra, coinedthenametunichrome.22Thecharacterization of thecompoundasa dicatecholatehas been reported. 23Quitea different chemical environment is foundin thevanadium-containingmaterial isolatedfromthe mushroomAmanitamuscaria. Bayer andKneifel,whonamedandfirst describedamavadine,24 alsosuggested thestructureshowninFigure 1. 7.25Recently the preparation, proofofligandstructure, and (byimplication) proof of thecomplexstructureshowninFigure1.7havebeenes-tablished.26Althoughtheexact roleof thevanadium complexinthemushroomFigure1.7A structure proposed foramavadineYremains unclear, thefact thatit isavanadyl complexisnowcertain, althoughitmaytakea different oxidationstateinvivo.Theroleof chromiuminbiologyremainsevenmoremysterious. Inhumanbeings theisolation of "glucose tolerance factor" and thediscoverythat it con-tains chromiumgoesbacksometime. Thishasbeenwell reviewedbyMertz,whohas playedamajorroleindiscoveringwhat is knownabout this elusiveandapparentlyquitelabilecompound.27It iswellestablishedthatchromiumistaken upaschromicion, predominantly via foodstuffs, such asunrefined sugar,which presumably contain complexes of chromium, perhaps involving sugar hy-droxyl groups. Although generally little chromium is taken up when it is admin-isteredasinorganicsalts, suchaschromicchloride, glucosetoleranceinmanyadultsandelderlypeoplehasbeenreportedtobeimprovedaftersupplementa-tionwith150-250 mgof chromiumperdayintheformof chromicchloride.Similar results have beenfoundinmalnourishedchildreninsome studies inThirdWorldcountries. Studiesusingradioactivelylabeledchromium haveshownthat, althoughinorganicsalts ofchromiumarerelativelyunavailabletomam-cytoplasmicreductantsFigure1.8Theuptake-reductionmodel forchromate carcinogenicity. Possiblesitesfor reduction of chro-mate include the cytoplasm, endoplasmic reticulum, mitochondria, andthenucleusYmals, brewer'syeast canconvert thechromium intoa usable form; so l:irewer'syeast is todaytheprincipal sourceintheisolationofglucosetolerancefactorand hasbeen usedasa diet supplement.Althoughchromiumisessential inmilligramamountsforhumanbeingsasthetrivalent ion, as chromateit is quitetoxic andarecognizedcarcinogen. 30Theuptake-reductionmodelforchromatecarcinogenicityassuggested byCon-nett and Wetterhahn isshown in Figure1.8. Chromate is mutagenicin bacterialandmammaliancell systems, andit has beenhypothesizedthat thedifferencebetween chromium in the+6 and+3 oxidation states is explained by the' 'up-take-reduction" model. Chromium(III), liketheferric iondiscussedabove, isreadily hydrolyzed at neutral pHand extremely insoluble. Unlike Fe 3+ , itundergoesextremely slow ligand exchange. For both reasons, transport of chro-mium(III)into cells can be expected tobe extremely slow unlessit is present asspecific complexes; for example, chromium(III) transport into bacterial cells hasbeen reported tobe rapid when iron is replaced by chromium in the siderophoreiron-uptakemediators. However, chromatereadilycrossescellmembranesandenterscells, muchas sulfatedoes. Becauseof itshighoxidizingpower, chro-matecanundergoreductioninsideorganellestogivechromium(m), whichbindstosmall molecules, protein, and DNA, damaging these cellular components.1112 1/ TRANSITION-METAL STORAGE, TRANSPORT, ANDBIOMINERALIZATIONIn marked contrast to its congener, molybdenum is very different from chro-miuminbothits roleinbiologyandits transport behavior, againbecauseoffundamental differencesinoxidationandcoordinationchemistryproperties. Incontrast tochromium, thehigheroxidationstatesof molybdenumdominateitschemistry, and molybdate is a relatively poor oxidant. Molybdenum is an essen-tial element inmanyenzymes, includingxanthineoxidase, aldehydereductase,andnitratereductase. 19Therangeof oxidationstatesandcoordinationgeome-tries ofmolybdenummakes its bioinorganicchemistryparticularlyinterestingand challenging.The chemistry of iron storageand transport is dominated by high concentra-tions, redoxchemistry(andproduction of toxic-actingoxygenspecies), hydrol-ysis(pKa is about 3, far below physiological pH), and insolubility. High-affinitychelatorsor proteinsare required for transport of iron and high-capacityseques-teringproteinfor storage. Bycomparisontoiron, storageandtransport of theothermetals aresimple. Zinc, copper, vanadium, chromium, manganese, andmolybdenumappeartobetransportedassimplesaltsorlooselyboundproteincomplexes. Invanadiumormolybdenum, thestableanion, vanadateor molyb-date, appearstodominatetransport. Littleis known about biologicalstorageofany metal except iron, which isstored in ferritin. However, zinc and copper areboundto metallothionein ina fonnthatmay participate instorage.II. BIOLOGICAL SYSTEMS OFMETAL STORAGE, TRANSPORT,ANDMINERALIZATIONA. Storage1. The storage ofironThreepropertiesof ironcanaccountforitsextensiveuseinterrestrialbio-logical reactions:(a) facileredox reactionsof iron ions;(b) anextensiverepertoireof redoxpotentialsavailablebyligandsubstitu-tion or modification(Table1.4);(c) abundanceandavailability(Table1.1)underconditionsapparentlyex-tantwhenterrestriallife began(seeSection LB.).Ferrous ionappears tohavebeentheenvironmentallystableformduringprebiotictimes. Thecombinationof thereactivityof ferrous ionandtherela-tivelylargeamountsof ironusedbycellsmayhavenecessitated thestorage offerrous ion; recent resultssuggest thatferrous ionmaybestabilizedinsidefer-ritin longenoughtobe usedinsome types ofcells. As primitiveorganismsbeganto proliferate, the successful photosynthetic cells, whichtrappedsolarenergybyreducingCO2tomakecarbohydrates (CH20)nandproduceO2, ex-hausted fromtheenvironment the reductantsfromH2 or H2S or NH3. Theabil-II. BIOLOGICALSYSTEMSOF METAL STORAGE, TRANSPORT, ANDMINERALIZATIONityof primitiveorganismstoswitchtotheuseof H20 asa reductant, withtheconcomitant productionof dioxygen, probablyproducedtheworst caseof en-vironmental pollutioninterrestrial history. Asaresult, thecompositionof theatmosphere, thecourseof biological evolution, andtheoxidationstate of envi-ronmental ironall changedprofoundly. Paleogeologistsandmeteorologistses-timate that there was alagofabout 200-300millionyears betweenthe firstdioxygenproductionandtheappearanceof significantdioxygenconcentrationsin theatmosphere, because the dioxygen produced at first was consumed by theoxidation of ferrousionsin the oceans. The transition in theatmosphere, whichoccurred about 2.5billion yearsago, caused thebioavailability of iron to plum-met andtheneedforironstoragetoincrease. Comparisonof thesolubilityofFe 3+at physiological conditions (about 10 - 18 M) totheironcontent ofcells(equivalent to10 -5 to10 -8M) emphasizes the difficulty of acquiring sufficientIron.Ironis storedmainlyintheferritins, afamily * of proteinscomposedof aprotein coat and an iron core of hydrous ferricoxide[Fe203(H20)n] with variousamountsof phosphate.6,7Asmanyas4,500ironatomscan be reversiblystoredinsidetheprotein coat ina complexthatissoluble; ironconcentrationsequiva-lent to0.25M[about 10 16-foldmore concentratedthanFe(III) ions] canbeeasilyachievedinvitro(Figure 1.1). Ferritinisfoundinanimals, plants, andeven in bacteria; the role of the stored iron varies, and includes intracellular usefor Fe-proteins or mineralization, long-termiron storage for other cells, anddetoxificationof excessiron. Ironregulatesthesynthesisof ferritin, withlargeamountsof ferritinassociatedwithironexcess, small orundetectableamountsassociatedwith iron deficiency. [Interestingly, thetemplate(mRNA)for ferritinsynthesisisitself stored incellsand is recruited byintracellular iron or a deriv-ative for efficient translation into protein.31Irondoes not appear to interactdirectlywith ferritin mRNA nor with a ferritinmRNA-specific regulatory (bind-ing)protein; however, thespecific, mRNAregulatory(binding)proteinhasse-quencehomologytoaconitase, and formationof an iron-sulfate cluster preventsRNAbinding.] Becauseironitself determinesinpart theamount of ferritininanorganism, the environmental concentrationofironneeds tobeconsideredbeforeonecan conclude that an organism or cell doesnot have ferritin.Ferritinis thought tobe the precursor of several forms of ironinlivingorganisms, includinghemosiderin, aformof storageironfoundmainlyinani-mals. Theironinhemosiderinisinaformverysimilartothat inferritin, butthecomplex withproteinisinsoluble, andisusuallylocatedwithin anintracel-lular membrane(lysosomes). Magnetite (Fe304) is another formofbiologicalironderived, apparently, fromtheironin ferritin. Magnetite playsa roleinthebehavior of magneticbacteria, bees, and homingpigeons(seeSection II.C).The structure of ferritin is the most complete paradigmfor bioinorganicchemistrybecause of threefeatures: theprotein coat, theiron-proteininterface,andthe iron core.6,7* A familyof proteinsisa group of related but distinct proteins produced in a single organism and usuallyencoded bymultiple, related genes.1314(A)(B)Figure1.9(A) Theprotein coat of horsespleen apoferritin deduced fromx-ray diffraction of crystalsof theprotein.32The outer surfaceof theprotein coat shows thearrangement of the24 ellipsoidal poly-peptidesubunits. N referstotheN-terminusof each polypeptideandE totheE-helix(see B).Notethechannelsthat formatthe four-foldaxeswheretheE-helicesinteract, and atthethree-foldaxesnear theN-terminiof thesubunits. (B)A ribbonmodel of a subunit showingthepacking of the fourmainalpha-helices(A, B, C, and D), theconnecting L-loop andtheE-helix.Protein Coat Twenty-four peptide chains (with about 175 amino acids each),foldedintoellipsoids, packtoformtheproteincoat, * whichisa hollowsphereabout 100 A indiameter; theorganicsurfaceisabout 10 A thick(Figure 1.9).Channels whichoccur inthe proteincoat at the trimer interfaces maybe in-volvedinthemovement of ironinandout of theprotein.62,63,65Sincethepro-teincoatisstablewithorwithoutiron, thecenter of thehollowspheremaybefilledwithsolvent, withFe203' H20, or, morecommonly, withbothsmall ag-gregates ofironandsolvent. Verysimilar amino-acidsequences arefoundinferritinfromanimals andplants. Sortingout whichaminoacids areneededtoformtheshapeoftheproteincoat andtheligands for ironcoreformationre-quiresthecontinueddedicationof bioinorganicchemists; identification of tyro-sineasanFe(III)-ligandaddsa newperspective. 64Iron-Protein Interface Formation of the iron core appearsto be initiated atanFe-proteininterfacewhereFe(II)-O-Fe(Ill) dimersandsmall clustersof Fe(Ill)atomshavebeendetectedattachedtotheproteinandbridgedtoeachother byoxo/hydroxobridges. Evidencefor multiplenucleationsiteshasbeenobtained* Some ferritinsubunits, notablyinferritinfrombacteria, bind hemeina ratio of lessthanoneheme pertwo subunits. Apossible role of such heme in the oxidation andreductionof iron in the core is beinginvestigated.II. BIOLOGICALSYSTEMSOF METAL STORAGE,TRANSPORT, ANDMINERALIZATION 15from electron microscopy of individual ferritinmolecules(multiple core crystal-lites wereobserved) andbymeasuringthe stoichiometryofbindingofmetalions, whichcompetewithbindingof monoatomiciron, e.g., VO(IV)andTh(m)(about eight sites per molecule). EXAFS (Extended X-ray Absorption FineStructure)andMossbauerspectroscopiessuggest coordination of Fetothepro-tein by carboxyl groups fromglutamic(Glu)and aspartic(Asp)acids. Althoughgroups of Glu or Asp are conserved in all animal and plant ferritins, theones that bindironarenot known. Tyrosineis anFe(III)-ligandconservedinrapidmineralizingferritins identifiedbyUv-visandresonanceRamanspectro-scopy.64IronCore Onlyasmallfractionof theironatomsin ferritinbinddirectlytothe protein. The core contains the bulk of the iron in a polynuclear aggregatewithproperties similar toferrihydrite, amineral found innature andformedexperimentally by heating neutral aqueoussolutions of Fe(III)(N03h. X-ray dif-fractiondatafromferritincores arebest fit byamodel withhexagonal close-packed layersof oxygenthatareinterrupted byirregularlyincompletelayersofoctahedrally coordinated Fe(III) atoms. The octahedral coordination is con-firmedbyMossbauerspectroscopyandbyEXAFS, whichalsoshowsthat theaverage Fe(In)atom issurrounded bysix oxygen atoms at a distance of 1.95 Aand sixironatomsat distancesof 3.0 to3.3A.Until recently, all ferritincoreswerethoughttobemicrocrystallineandtobe the same. However, x-ray absorption spectroscopy,Mossbauer spectroscopy,andhigh-resolutionelectronmicroscopyof ferritinfromdifferent sourceshaverevealedvariationsinthedegreeof structuralandmagneticorderingand/or thelevel of hydration. Structural differencesintheironcorehavebeenassociatedwithvariationsintheanionspresent, e.g., phosphate29or sulfate, andwiththeelectrochemical properties ofiron. Anionconcentrations intumcouldreflectboththesolventcompositionandthepropertiesof theproteincoat. Tounder-stand iron storage, we needto define in more detail the relationship of theferritinproteincoat andtheenvironmenttotheredoxpropertiesofironintheferritincore.Experimental studies of ferritin formationshow that Fe(n) and dioxygen areneeded, at least intheearlystagesof coreformation. Oxidation toFe(nI) andhydrolysisproduceoneelectronandanaverageof 2.5protons for ironatomsincorporatedintotheferritinironcore. Thus, formationofafull ironcoreof4,500ironatomswouldproducea total of 4,500 electronsand11,250 protons.After core formation by such a mechanism inside the protein coat, the pH woulddrop to 0.4 if all the protons were retained. It is known that protons are releasedand electronsare transferred todioxygen. However, therelative ratesof protonrelease, oxo-bridgeformation, andelectrontransfer have not beenstudiedindetail. Moreover, recentdataindicatemigrationof ironatomsduringtheearlystages ofcore formationandthe possible persistenceofFe2+ for periods oftime up to 24 hours. When large numbers of Fe(n) atoms are added, the proteincoatappearstostabilizetheencapsulated Fe(n).34a,b Formation of theironcore16 1/ TRANSITION-METAL STORAGE, TRANSPORT, ANDBIOMINERALIZATIONof ferritin hasanalogies tosurface corrosion, in which electrochemical gradientsare known to occur. Whether such gradients occur during ferritinformationandhowdifferentproteincoatsmightinfluenceprotonreleaseoralterthestructureof the corearesubjectsonlybeginning tobe examined.2. Thestorage of zinc, copper, vanadium, chromium, molybdenum,cobalt, nickel, andmanganeseIonsof nonferroustransitionmetalsrequirea much lesscomplex biologicalstoragesystem, because thesolubilitiesaremuchhigher (210 -8 M) thanthosefor Fe 3+ . As a result, the storage of nonferrous transition metals is less obvious,andinformationis morelimited. Inaddition, investigations aremoredifficultthanfor iron, becausetheamountsinbiological systems aresosmall. Essen-tiallynothingisknownyet about thestorageofvanadium, chromium, molyb-denum, cobalt, nickel, andmanganese, withthepossibleexception of accumu-lationsof vanadium in theblood cells of tunicates.Zincandcopper, whichare usedinthe highest concentrations ofanyofthe non-ferrous transition metals, are specifically bound by the proteinmetallothionein 35,36(see Figure 1.10). Like the ferritins, the metallothioneinsarea familyof proteins, widespreadin natureandregulatedbythemetalstheybind. In contrast toferritin, theamounts of metalstored inmetallothioneinsaresmaller (uptotwelve atoms per molecule), the amount ofproteinincells isless, and the template (mRNA) is not stored. Because the cellular concentrationsofthe metallothioneins arerelativelylowandthe amount ofmetal neededisrelativelysmall, ithasbeendifficulttostudythebiologicalfateof copperandzincinlivingorganisms, andtodiscoverthenatural roleof metallothioneins.However, theregulation of metallothioneinsynthesis by metals, hormones, andgrowthfactors atteststothebiologicalimportanceof theproteins. Theunusualmetal environmentsof metallothioneinshaveattractedtheattentionof bioinor-ganicchemists.Metallothioneins, especiallyinhigher animals, are small proteins 35,36 richincysteine(20 per molecule)and devoid of thearomaticaminoacidsphenylal-anineandtyrosine. Thecysteine residuesaredistributedthroughout thepeptidechain. However, inthe nativeformoftheprotein(Figure 1.10), thepeptidechains foldtoproducetwoclusters of -SH, whichbindeither threeorfouratoms ofzinc, cadmium, cobalt, mercury, lead, or nickel. Copperbindingisdistinct fromzinc, with12sites per molecule.Insummary, ironisstoredinironcoresof a complicatedprotein. Ferritin,composedof ahollowproteincoat, iron-proteininterface, andaninorganiccore,overcomestheproblemsof redoxandhydrolysisbydirectingtheformationofthequasi-stablemineral hydrousferricoxideinsidetheproteincoat. Theoutersurfaceof theproteinisgenerally hydrophilic, makingthecomplexhighlysol-uble; equivalent concentrationsof ironare:::::0.25M. Bycontrast toiron, stor-age of zinc, copper, chromium, manganese, vanadium, andmolybdenum is rel-ativelysimple, becausesolubilityishighandabundanceislower. LittleisknownFigure1.10The three-dimensional structure of theadomain fromrat cd7metallothionein-2, determinedbyNMR insolution(Reference36a), based on data in Reference36b. The fourmetalatoms,bonded tothesulfur of cysteineside chains, are indicated asspherical collectionsof smalldots.A recent description of thestructure of the cdsZn2 protein, determined fromx-ray diffraction ofcrystals, agreeswiththestructuredeterminedbyNMR(Reference36c).about themoleculesthat storethesemetals, withthepossibleexception of me-tallothionein, which bindssmall clusters of zincor copper.B. Transport1. IronThestorageof ironinhumans andother mammalshasbeendealt withintheprevioussection. Onlya smallfractionof the body'sinventoryof iron isintransit at any moment. The transport of iron fromstorage sites in cellular ferritinorhemosiderinoccursviatheserum-transport proteintransferrin. Thetransfer-rinsareaclassof proteinsthat arebilobal, witheachlobereversibly(andes-sentiallyindependently) bindingferricion.37-39Thiscomplexationof themetalcationoccursvia prior complexation of a synergistic anion that in vivoisbicar-bonate(or carbonate). Serum transferrin isa monomeric glycoprotein of molec-ular weight 80kDa. Thecrystal structureoftherelatedprotein, lactoferrin,39hasbeenreported, andrecentlythestructureof amammaliantransferrin40hasbeen deduced.Ferritinisapparentlyaveryancient protein andis foundin higheranimals,plants, andevenmicrobes; inplantsandanimalsa commonferritinprogenitor1718 1/ TRANSITION-METAL STORAGE, TRANSPORT, ANDBIOMINERALIZATIONisindicated bysequence conservation.41Incontrast, transferrin hasbeeninex"istenceonlyrelativelyrecently, sinceitisonlyfoundia thephylumChordata.Although the twoiron-binding sites of transferrin aresufficiently different to bedistinguishablebykineticandafewother studies, theircoordinationenviron-mentshavebeenknownfor sometimetobequitesimilar. Thiswasfirst dis-covered byvariousspectroscopies, and most recently was confirmed by crystal-structure analysis, whichshows that the environment involves twophenolateoxygens from tyrosine, two oxygensfromthesynergistic, bidentate bicarbonateanion, nitrogenfromhistidine, and(asurpriseat thetimeofcrystal-structureanalysis)anoxygenfroma carboxylate group of anaspartate. 39The transferrinsareall glycoproteins, and human serum transferrin containsabout6 percent carbohydrate. Thesecarbohydrategroupsarelinked tothepro-tein, andapparentlystronglyaffect the recognition and conformation of thena-tiveprotein.Although transferrinshavea high molecular weightandbindonlytwoironatoms, transferrinis relativelyefficient, becauseit is usedinmanycycles ofiron transport in itsinteractionwiththe tissuestowhich it deliversiron. Trans-ferrinreleasesironinvivobybindingtothecell surfaceandforminga vesicleinsidethecell (endosome)containinga pieceof themembranewith transferrinand ironstill complexed. Thereleaseof theironfromtransferrinoccursintherelativelylow pH of the endosome, andapoprotein is returned to theoutside ofthecell for deliveryof anotherpairof ironatoms. Thisprocessinactivereti-culocytes (immature red blood cellsactive in iron uptake) can tum over roughlyamillion atoms of ironper cell per minute.38Aschematic structure of theprotein,deduced from crystal-structure analysis, is shown in Figure 1.11.Trans-ferrinis anellipsoidal proteinwithtwosubdomains orlobes, eachofwhichbindsiron. Thetwohalvesof eachsubunit aremoreorlessidentical, andareconnectedbyarelativelysmall hinge. Inhumanlactoferrin, thecoordinationsiteof theironisthesameasthecloselyrelatedserotransferrinsite. Amajorquestionthat remainsabout themechanismof ironbindingandreleaseishowtheproteinstructurechangesintheintracellular compartmentof lowpHtore-leasetheironwhenitforms aspecificcomplexwithcell receptors(transferrinbindingproteins) andwhether thereceptor proteinis activeor passiveintheprocess. Recent studies suggest that the cell binding site for transferrin(a mem-brane, glycoproteincalledthetransferrin receptor)itself influencesthestabilityofthe iron-transferrincomplex. The pathof ironfromthe endosome toFe-proteinshasnot beenestablished; and theformof transported intracellular ironisnot known.Another major typeof biological iron transport occursat thebiological op-positeof thehigherorganisms. Althoughalmost all microorganismshaveironasanessential element, bacteria, fungi, andothermicroorganisms (unlikehu-mansand other higher organisms) cannot afford to make high-molecular-weightprotein-complexing agents for this essential element when those complexing agentswouldbeoperating extracellularlyand hencemost of thetimewould belost tothe organism. As describedearlier, the first life forms onthe surfaceoftheII19Figure1.11Three-dimensionalstructure of lactotransferrin. Top: schematic representation of the foldingpat-tern of each lactoferrinlobe;Domain 1 isbased ona beta-sheet of four parallelandtwoantipar-allel domains; Domain II isformedfromfourparalleland one antiparallelstrand. Bottom: stereoCa diagram of theNlobe of lactoferrin; (e)iron atom between domain 1 (residues 6-90 +)and domainII(residues91-251); (_)disulfidebridges; (*)carbohydrate attachment site. SeeReference39.20 1I TRANSITION-METAL STORAGE,TRANSPORT, ANDBIOMINERALIZATIONEarthgrewina reducingatmosphere, inwhichtheironwassubstantiallymoreavailablebecauseit was present as ferrous-containingcompounds. Incontrasttothe profoundlyinsoluble ferrichydroxide, ferroushydroxide is relativelysol-uble atnear neutral pH. It hasbeen proposed that thisavailabilityof iron in theferrousstatewas one of the factorsthat led to its early incorporation in so manymetabolic processes of the earliest chemistryof life. 6,38 In an oxidizing environ-ment, microorganisms were forced to deal with the insolubility of ferrichydrox-ide and hence when facing iron deficiency secrete high-affinity iron-bindingcompounds called siderophores(from the Greek for iron carrier). More than 200naturallyoccurringsiderophores have been isolated and characterized to date.42Most siderophore-mediatediron-uptake studies inmicroorganisms havebeenperformedbyusingcells obtainedunder iron-deficient aerobic growthcondi-tions. However, uptake studies inE. coli grown under anaerobic conditionshavealsoestablishedthepresence of siderophore-specificmechanisms. Inbothcases, uptake of the siderophore-iron complex is both a receptor- and an energy-dependentprocess. Insomestudiesthedependenceof siderophoreuptakerateson the concentration of the iron-siderophore complex has been found to conformtokinetics characteristicofproteincatalysts, i.e., Michaelis-Mentenkinetics.For example, saturableprocesseswithverylowapparentdissociationconstantsof under one micromolar (l fLM) have been observedfor ferric-enterobactintransport in E. coli (a bacterium), asshown in Figure1.12. Similarly, in a very80c"E 60complexconcentration(J..lM)Figure1.12Effect of MECAManalogueson iron uptake fromE. coli.Iron transport by2f.LMferricenterobactinisinhibited byferricMECAM.OJEm""6EEoQ)"Q)C""-alQ.::JQ)LLenen4020-52 4 6II. BIOLOGICAL SYSTEMSOF METAL STORAGE, TRANSPORT, ANDMINERALIZATION 21different microorganism, the yeast Rhodoturalapilimanae, Michaelis-Mentenkinetics were seenagainwithadissociationconstant ofapproximately6JLMfor theferric complexofrhodotoroulic acid; diagrams ofsomerepresentativesiderophores are showninFigure 1. 13. The siderophore usedbythe fungusNeurospora crassawas foundtohave adissociationconstant ofabout 5JLMand, again, saturable uptakekinetics.o;:lOH~ N~CH3S /NS ~ C O O HpyochelinmycobactinPpseudobactinUOHyOHenterobactinFigure1.13Examplesof bacterial siderophores. See Reference42.224 [Fe(ent)P-(oranalogue)\ I\ ItepAproteinreceptormembrane.peri plasmic spacecytoplasmleakage ot[Fe (ent)] 3-tosolutionFigure1.14Model forenterobactin-mediated Fe uptake in E. coli.Although the behavior just described seems relatively simple, transportmechanismsinliving cellsprobably haveseveral more kineticallydistinctstepsthanthose assumedfor the simple enzyme-substratereactions underlyingtheMichaelis-Mentenmechanism. Forexample, asferric enterobactinis accumu-latedin E. coli, it hastopassthroughtheoutermembrane, theperiplasm, andthe cytoplasm membrane, andis probablysubjected toreduction of themetal ina low-pH compartment or toligand destruction.Asketchof a cell of E. coliandsomeaspects of itstransportbehaviorareshowninFigure 1.14. Enterobactin-mediatedironuptakeinE. coli is oneofthebest-characterized of thesiderophore-mediated iron-uptakeprocessesin mi-croorganisms, andcanbe studiedas amodel. After this verypotent iron-se-questeringagent complexesiron, theferric-enterobactincomplexinteractswithaspecificreceptorintheoutercell membrane(Figure1.14), andthecomplexis taken into the cell by active transport. The ferric complexes of some syntheticanalogsof enterobactincanact as growthagents insupplyingirontoE. coli.Suchafeaturecouldbeusedtodiscover whichpartsofthemoleculearein-volvedinthesitesof structural recognitionof theferric-enterobactincomplex.Earlierresults suggestedthat themetal-bindingpart ofthemoleculeisrecog-nizedbythereceptor, whereas theligandplatform(thetriserinelactonering;see Figure1.13)isnot specifically recognized.Tofindout whichdomainsof enterobactinarerequired foriron uptakeandrecognition, rhodiumcomplexeswerepreparedwithvariousdomainsof enter-obactin(Figure1.15)asligandstouseascompetitorsforferricenterobactin.44The goal was to find out if the amide groups (labeled Domain II in Figure1.15),Domain:(III) metal bindingunit(II) amidelinkage(I) backboneFigure1.15Definition of recognition domainsin enterobactin.which linked the metal-binding catechol groups (Domain III, Figure1.15) to thecentral ligandbackbone(DomainI, Figure 1.15), arenecessaryforrecognitionbythereceptor protein. Inaddition, syntheticligands werepreparedthat dif-feredfromenterobactinbysmall changes at or near thecatecholatering. Fi-nally, variouslabiletrivalent metal cations, analogoustoiron, werestudiedtoseehowvaryingthecentral metal ionwouldaffecttheabilityof metal entero-bactincomplexes toinhibit competitivelythe uptakeofferric enterobactinbytheorganism. For example, if rhodium MECAM(Figure1.16) is recognized bythereceptorfor ferricenterobactinonlivingmicrobial cells, alargeexcess ofrhodiumMECAMwill blocktheuptakeof radioactiveironaddedasferricen-terobactin. In fact, the rhodium complex completely inhibited ferric-enterobactinuptake, provingthat DomainI is not requiredfor recognitionof ferricentero-bactin.However, ifonlyDomainIII is important inrecognition, it wouldbeex-pectedthat thesimpletris(catecholato)-rhodium(III) complexwouldbeanequallygoodinhibitor. Infact, evenat concentrations inwhichtherhodium-catecholcomplexwas inverylargeexcess, noinhibitionof ironuptakewas observed,suggestingthat Domain II isimportant intherecognition process.Therole ofDomainII inthe recognitionprocess was probedbyusing arhodiumdimethyl amideof 2,3-dihydroxybenzene(DMB)asa catecholligand,withonemorecarbonyl ligandthaninthe tris(catecholato)-rhodium(III) com-plex. Remarkably, thismoleculeshowssubstantiallythesameinhibition of en-terobactin-mediatedironuptakein E. coli asdoesrhodiumMECAMitself. Thus,inadditiontothe iron-catechol portionof the molecule, the carbonyl groups2324OMS~ O H~ O HcatecholTRIMCAMFigure1.16MECAMand related enterobactin analogues.II. BIOLOGICAL SYSTEMSOFMETAL STORAGE, TRANSPORT, ANDMINERALIZATION 25(DomainII) adjacenttothecatechol-bindingsubunitsof enterobactinandsyn-thetic analogs are required for recognition by the ferric-enterobactin receptor. Incontrast, whenamethyl groupwasattachedtothe"top"of therhodiumME-CAM complex, essentially norecognition occurred.Insummary, althoughthestructureof theouter-membraneprotein receptorof E. coliisnot yetknown, thecompositeof theresults just describedgivesasketchofwhat theferric-enterobactinbindingsitemust looklike: arelativelyrigid pocket for receiving the ferric-catecholate portionof the complex, andproton donor groups around this pocket positionedto hydrogen bondto thecarbonyl oxygens of theferric amidegroups. Themechanismsofironreleasefromenterobactin, thoughfollowedphenomenologically, arestillnot knownindetail.2. Zinc, copper, vanadium, chromium, molybdenum, andcobaltAsdescribedinanearliersection, transportproblemsposed bythesixele-mentslistedintheheadingaresomewhat simpler(withtheexceptionof chro-mium) thanthosefor iron. Oneveryinterestingrecent development has beenthe characterization of sequesteringagentsproduced by plantswhich complexanumber of metal ions, not just ferricions. Akeycompound, nowwell-charac-terized, ismugeneicacid(Figure 1.17).45 Thestructural andchemical similari-C3 C3Figure1.17Structure anda stereoviewof mugeneicacid. See Reference 42.26 1I TRANSITION-METALSTORAGE, TRANSPORT, ANDBIOMINERALIZATIONtiesof mugeneicacidto ethylenediaminetetraacetic acid (EDTA)have been noted.Like EDTA, mugeneicacid formsan extremelystrong. ~ o m p l e x with ferricion,but alsoforms quitestrongcomplexeswithcopper, zinc, andothertransition-metal ions. The structure ofthe cobalt complex (almost certainlyessentiallyidentical withthatof theironcomplex) isshowninFigure1.18. Likethesid-erophores producedbymicroorganisms, thecoordinationenvironment accom-modatedbymugeneicacidisessentiallyoctahedral. Althoughthecoordinationpropertiesof thisligandarewell laidout, andithasbeenshownthatdivalentmetal cations, suchascopper, competitivelyinhibit iron uptakebythisligand,thedetailedprocess ofmetal-iondeliverybymugeneic acidandrelatedcom-pounds hasnot been elucidated.0(4)0(8)1.939(5)A0(4)0(8)1.941(5)A1.896(5)0(3)(A)0(5)N(2)1.915(6)0(3)(B)0(5)N(2)Figure1.18Molecular structuresof the complexes(moleculesA andB)and coordination about the cobaltion inmoleculesA and B of themugeneicacid-Co(III)complex. Bond lengthsin A; anglesindegrees. See Reference 42.II. BIOLOGICALSYSTEMS OF METAL STORAGE, TRANSPORT, ANDMINERALIZATION 27As notedin anearlier section, the biochemistryof vanadiumpotentiallyinvolvesfour oxidationstatesthatarerelativelystableinaqueoussolution. TheseareV2+, V3+, va2+ , andV02+(theoxidationstates2, 3, 4, and5, respec-tively). Sinceevenwithoutaddedsequesteringagents, V2+slowlyreduces watertohydrogengas, it presumablyhasnobiological significance. Examplesof theremainingthreeoxidationstatesof vanadiumhaveall beenreportedinvariouslivingsystems. One of the most extensively investigated examples of transition-metal-ionaccumulationin livingorganismsistheconcentration of vanadium insea squirts(tunicates), which is reported to be variable; manyspecies havevan-adium levels that are not exceptionally high. Otherssuch as Ascidia nigrashowexceptionally highvanadium concentrations. 46In addition to showing aremarkable concentrationof arelatively exotictransition-metal ion, tunicates are agoodlaboratorymodel for uptakeexperi-ments, sincetheyarerelativelysimpleorganisms. Theypossess acirculationsystemwitha one-chambered heart, anda digestivesystem that isessentiallyapump and an inlet and outlet valve connected by a digestive tract. The organismcanabsorbdissolvedvanadiumdirectlyfromsea water asit passesthroughtheanimal. The influx of vanadate into theblood cells of A. nigra has beenstudiedby meansof radioisotopes. Thecorrespondinginflux of phosphate, sulfate, andchromate (andthe inhibitionofvanadate uptake bythese structurally similaroxoanions)hasbeen measured. Intheabsenceof inhibitors, theinflux of vana-dateisrelativelyrapid(ahalf-lifeontheorderofaminutenear ODe) andtheuptakeprocess shows saturationbehavior as thevanadateconcentrationis in-creased. The uptakeprocess(in contrast to iron delivery in microorganisms, forexample, andtomanyother uptakeprocessesin microorganismsor higherani-mals) isnotenergy-dependent. Neither inhibitors ofglycolysisnor decouplersof respiration-dependent energy processesshow any significant effect on the rateof vanadate influx.Phosphate, whichis alsoreadilytakenupbythecells, is aninhibitor ofvanadateinflux. Neithersulfatenorchromateistakenupsignificantly, nordotheyact as significant inhibitors for the vanadate uptake. Agents that inhibittransport ofanions, incontrast, werefoundtoinhibituptakeof vanadateintotheorganism. These resultshaveled tothemodel proposed in Figure1.19:(1) vanadate enters the cell throughanionic channels; this process elimi-nates positivelychargedmetal ionor metal-ioncomplexes present insea water;(2) vanadateisreducedtovanadium(III); since the product isa cation, andsocannot betransportedthroughtheanionicchannelsbywhichvana-dateenteredthecell, thevanadium(III) istrappedinsidethecell-thenet result is anaccumulationofvanadium. [Ithas beenproposedthatthetunichromecouldact either asa reducingagent (asthe complex)or(astheligand)tostabilizethegeneralvanadium(ill); however, thisseemsinconsistent withits electrochemical properties(seebelow).]28vacuoleanionicchannelsxFigure1.19Diagram of a vanadium accumulation mechanism. Vanadium entersthevacuolewithin thevana-docyteasmononegative H2Y04-, although it may be possible forthe dinegative anion, H Y O ~ - ,to enter thischannelaswell (X- stands foranynegativeion such as Cl- , H2PO,;- , etc., thatmay exchangeacross themembranethroughtheanionicchannel). Reduction toy3+takesplacein twosteps, via a Y(IY)intermediate. The resulting cations maybe trappedastightlyboundcomplexes, or asfreeions thattheanionicchannel willnot accept fortransport. Thenature ofthereducingspeciesisunknown.Synthetic models of tunichrome b-] (Figure] .20) have been prepared. Tun-ichrome isa derivativeof pyrogallolwhosestructure precludes the formationofan octahedral complex of vanadium as a simple] : ]metal: ligand complex. Theclose analogue, describedas 3,4,5-TRENPAMH9, alsocannot formasimpleoctahedral ]:]complex. Incontrast, thesyntheticligandsTRENCAMand2,3,4-TRENPAMcanformpseudo-octahedral complexes. Thestructureof thevana-diumTRENCAMcomplexshowsthat it isindeedasimplepseudo-octahedraltris-catechol complex. 47 The electrochemical behavior of these complexes issimilar, withvanadium(IVfIll) potentialsofabout - 0.5to - 0.6volts versusNHE. These results indicate that tunichrome b-l complexes of vanadium(IVfIll)wouldshowsimilardifferences intheirredoxcouples at highpH. At neutralpH, in thepresence of excess pyrogallol groups, vanadium(IV)can be expectedto formthe intenselycolored tris-catechol species. However, comparison of theEPR properties reported forvanadium-tunichrome preparationswith model van-adium(lV)-complexes wouldindicate predominantlybis(catechol) vanadyl co-ordination. Inanycase, thevanadium(III) complexesmust remainveryhighlyreducing. It has beenpointedout that the standardpotential ofpyrogallol is0.79Vanddecreases60 mVperpHunit (uptoaboutpH 9), sothat at pH7the potential is about 0.4V. The potentials ofthe vanadiumcouples for thetunichromeanalogs areabout - 0.4 V. Ithas beenconcluded, therefore, thattunichromeorsimilar ligandscannotreducethevanadium(IV)complex; sothe29OHHO OHOH HOy 0 0OHTunichrome b-lH01:HO"-0HNCfJr::0OH OHoNKg)OH 0OHOH

HO"-0HN0OH OHHO 0 OH 0 OHOH3,4,5-TRENPAMH6OH

HN OHO

HO H 0. OHOHOHTRENCAMH62,3,4-TRENPAMH9Figure1.20Structures of tunichrome b-l andsyntheticanalogues.4330 1/ TRANSITION-METAL STORAGE,TRANSPORT, ANDBIOMINERALIZATIONhighlyreducingvanadium(III) complexof tunichromemust begeneratedinsomeother way. 47Althougha detailedpresentationof examplesof theknowntransportprop-erties of essential transition-metalionsintovariousbiologicalsystemscould bethe subject ofalargebook, the examples that we have givenshowhowtheunderlyinginorganicchemistryoftheelementsisusedinthebiological trans-port systemsthat arespecificfor them. Theregulationof metal-ionconcentra-tions, including their specificconcentration whennecessaryfromrelativelylowconcentrationsof surroundingsolution, isprobablyoneof thefirst biochemicalproblems thatwassolved in thecourse of the evolution of life.Ironistransportedinforms inwhichit istightlycomplexedtosmall che-lators called siderophores (microorganisms) or to proteins called transferrins (an-imals) ortocitrateormugeneicacid(plants). Theproblem of howtheironisreleasedinacontrolledfashionis largelyunresolved. Theprocess ofmineralformation, calledbiomineralization, is asubject ofactiveinvestigation. Vana-dium and molybdenum aretransported asstableanions. Zincand copper appeartobetransportedlooselyassociatedwithpeptidesor proteins(plants) andpos-sibly mugeneicacidinplants. Muchremainstobelearned about thebiologicaltransport of nonferrousmetalions.C. IronBiomineralizationMany structures formed by living organisms are minerals. Examples includeapatite [Ca2(OH)P04] inbone andteeth, calcite or aragonite (CaC03) intheshellsof marineorganismsandintheotoconia(gravitydevice)of themamma-lianear, silica (Si02) ingrassesand intheshellsof smallinvertebratessuchasradiolara, and iron oxides, such as magnetite (Fe304)in birds and bacteria (nav-igationaldevices)and ferrihydriteFeO(OH)in ferritinof mammals, plants, andbacteria. Biomineralization is the formation of such minerals by the influence oforganicmacromolecules, e.g., proteins, carbohydrates, andlipids, onthepre-cipitationofamorphousphases, ontheinitiationofnucleation, onthegrowthof crystalline phases, and on thevolume of the inorganic material.Ironoxides, as one ofthe best-studiedclasses of biominerals containingtransitionmetals, providegoodexamples for discussion. Oneof themost re-markablerecent characterizationsofsuchprocessesisthecontinual depositionofsingle-crystal ferric oxideintheteethofchiton.48Teethof chitonformonwhat is essentiallya continuallymoving belt, inwhich newteethare beinggrown and moved forward to replace mature teeth that have been abraded. How-ever, thestudyof themechanisms of biomineralizationingeneral isrelativelyrecent; a great deal of the information currently available, whether about iron inferritinor about calcium inbone, issomewhat descriptive.Threedifferent formsof biologicalironoxidesappeartohavedistinctrela-tionshipsto the proteins, lipids, or carbohydrates associated with their formationandwiththedegreeof crystallinity. 49Magnetite, on theonehand, oftenformsalmost perfectcrystalsinsidelipidvesiclesof magneto-bacteria. 50Ferrihydrite,II. BIOLOGICALSYSTEMS OFMETAL STORAGE, TRANSPORT, ANDMINERALIZATION 31on the other hand, exists as large single crystals,or collections of small crystals,insidetheproteincoat offerritin; however, ironoxides insomeferritins thathave large amounts of phosphate are very disordered. Finally, goethite [a-FeO(OH)] andlepidocrocite [y-FeO(OH)] formas small single crystals inacomplexmatrix of carbohydrateandproteinin theteeth of someshellfish(lim-petsandchitons); magnetiteisalsofoundinthelepidocrocite-containingteeth.Thedifferencesintheiron-oxidestructuresreflect differencesinsomeor all ofthefollowingconditionsduringformationof themineral: natureofco-precipi-tating ions, organicsubstrates or organic boundaries, surface defects, inhibitors,pH, and temperature. Magnetite can formin both lipid and protein/carbohydrateenvironments, and can sometimes be derived from amorphous or semicrystallineferrihydrite-like material (ferritin). However, the preciserelationshipbetweenthestructure of theorganicphaseandthatof theinorganicphasehasyet tobediscovered. Whenthe goal of understandinghowthe shape andstructure ofbiomineralsisachieved, bothintellectual satisfactionandpracticalcommercialand medicalinformationwillbe provided.Syntheticironcomplexes haveprovidedmodels for twostagesofferritinironstorageandbiomineralization: 51-59 (1) theearlystages, whensmall num-bersof clusteredironatomsareboundtotheferritinproteincoat, and(2) thefinal stages, where thebulk iron isa mineralwith relatively fewcontactstotheproteincoat. Inaddition, modelshavebeguntobeexaminedforthemicroen-vironment insidethe protein coat. 54Amongthemodelsfortheearlyor nucleationstageof iron-coreformationarethebinuclearFe(III) complexeswith[Fe20(02CR2)]2+cores;55,56thethreeother Fe(III)ligandsareN. TheJL-oxocomplexes, which areparticularlyaccu-ratemodelsforthebinuclear ironcentersinhemerythrin, purpleacidphospha-tases, and, possibly, ribonucleotidereductases, mayalsoserveas models forferritin, since an apparently transient Fe(II)-O-Fe(III) complex was detected dur-ingthereconstitutionofferritinfromproteincoats andFe(II). Thefacile ex-change of (02CR) for(02PR)in the binuclear complex is particularly significantas amodel for ferritin, becausethe structureofferritincores varies withthephosphatecontent. Anasymmetrictrinuclear(Fe30) 7+complex57 andan(FeO)l1complex(Figure 1.21) havebeen prepared; theseappear toserveasmodelsforlater stages of core nucleation(or growth). 59Modelsforthefull ironcoreof ferritininclude ferrihydrite, whichmatchesthe ordered regionsof ferritincoresthat havelittle phosphate; however, the sitevacancies inthelatticestructureofferrihydrite [FeO(OH)] appeartobemoreregularthanincrystallineregions offerritincores. Apolynuclearcomplexofironandmicrobial dextran(a-l,4-D-glucose)nhas spectroscopic (M6ssbauer,EXAFS) properties very similar to those of mammalian ferritin, presumablybecause theorganic ligandsaresimilar tothose of the protein (-OH, -COOH).Incontrast, apolynuclearcomplexof ironandmammalianchondroitinsulfate(a-l ,4-[a-1 ,3-D-glucuronicacid-N-acetyl-D-galactosamine-4-sulfate]n) containstwo types of domains: one like mammalianferritin [FeO(OH)] andone likehematite (a-Fe203), which was apparently nucleated by the sulfate,emphasizing3206032"()033Figure1.21Thestructure of a modelfora possible intermediate in the formationof theferritiniron core.The complex consistsof11Fe(III)atomswithinternaloxo-bridgesand a coat of benzoateli-gands; theFeatomsdefinea twisted, pentacapped trigonalprism. See Reference53.theimportanceof anions inthestructureof ironcores.60Finally, amodel forironcoreshighinphosphate, suchasthosefrombacteria, isFe-ATP(4: 1), inwhich the phosphate is distributedthroughout the polynuclear ironcomplex,providinganaverageof 1 or 2 of the6 oxygen ligandsforiron. 61Themicroenvironment insidetheproteincoatof ferritinhasrecentlybeenmodeledbyencapsulatingferrous ioninsidephosphatidylcholinevesicles andstudying the oxidation of ironasthe pH is raised. The efficacy of sucha modelis indicatedby the observationof relatively stable mixtures of Fe(II)/Fe(III)insidethevesicles, ashavealsobeenobservedinferritinreconstitutedexperi-mentallyfromprotein coatsand ferrousion.43,54Modelsforironinferritinmustaddressboththefeaturesof traditional metal-proteininteractionsandthebulk propertiesof materials. Althoughsuchmodel-ingmaybemoredifficultthanothertypesof bioinorganicmodeling, thediffi-culties arebalancedbytheavailabilityofvast amountsofinformationonFe-III. SUMMARY 33proteininteractions, corrosion, andmineralization. Furthermore, powerful toolssuchas x-rayabsorption, MossbauerandsolidstateNMRspectroscopy, scan-ningelectronandprotonmicroscopy, andtransmissionelectronmicroscopyre-ducethenumber of problemsencountered in modeling theferritinion core.Construction of models for biomineralization is clearly an extension of mod-elingforthebulk phase of ironinferritin, sincethemajor differencesbetweentheironcoreofferritinandthat ofotheriron-biominerals arethe sizeofthefinal structure, the generally higher degree of crystallinity, and, at thistime, themore poorly defined organicphases. A modelformagnetite formationhasbeenprovidedbystudyingthecoulometricreductionof halftheFe3+ atoms intheiron core of ferritinitself. Although the conditions for producing magnetite have.yet tobediscovered, theunexpectedobservationofretentionoftheFe 2+ bythe protein coat hasprovided lessonsforunderstandingthe iron core of ferritin.Phosphatidyl cholinevesiclesencapsulatingFe 2+ appear toserveasmodelsforbothferritinandmagnetite; onlyfurther investigationwill allowus tounder-stand the unique featuresthat convert Fe2+to[FeO(OH)], on the one hand, andFe304, on the other.III. SUMMARYTransitionmetals(Fe, Cu, Mo, Cr, Co, Mn, V)playkeyrolesinsuch biolog-ical processesascell division(Fe, Co), respiration(Fe, Cu), nitrogenfixation(Fe, Mo, V), andphotosynthesis(Mn, Fe). Znparticipatesin manyhydrolyticreactions andinthecontrol ofgeneactivitybyproteins with"zincfingers."Amongtransitionmetals, Fepredominates interrestial abundance; sinceFeisinvolvedinavast number ofbiologicallyimportant reactions, its storageandtransport havebeenstudiedextensively. TwotypesofFecarriers areknown:specificproteins andlow-molecular-weight complexes. Inhigher animals, thetransport proteintransferrinbindstwoFeatoms withhighaffinity; inmicroor-ganisms, ironistransported intocellscomplexed withcatecholatesor hydroxa-mates calledsiderophores; andinplants, small molecules suchascitrate, andpossiblyplant siderophores, carryFe. Ironcomplexesenter cellsthroughcom-plicated pathsinvolvingspecificmembranesites(receptor proteins). A problemyet tobesolvedis theformofirontransportedinthecell afterreleasefromtransferrin or siderophoresbut beforeincorporationinto Fe-proteins.Ironisstored intheproteinferritin. Theprotein coat of ferritinisa hollowsphereof24polypeptidechains throughwhichFe2+ passes, is oxidized, andmineralizes inside in various formsof hydrated Fe203. Control of the formationand dissolution of the mineral core by the protein and control of protein synthe-sisby Fearesubjects of current study.Biomineralization occursintheocean(e.g., Ca inshells, Si incoral reefs)andonlandinboth plants(e.g., Si ingrasses) andanimals(e.g., Cainbone,Fe in ferritin, Fe in magnetic particles). Specificorganic surfaces or matricesofprotein and/or lipid allow living organisms to produce minerals of defined shapeand composition, often in thermodynamicallyunstablestates.341I TRANSITION-METAL STORAGE, TRANSPORT, ANDBIOMINERALIZATIONIV. REFERENCES1. (a)J. H. Martinand R. M. Gordon, Deep SeaResearch35(1988), 177; (b)F. Egami, 1. Mol. Evol. 4(1974), 113.2. J. F. Sullivanet al., 1. Nutr. 109(1979),1432.3. M. D. McNeelyet al., Clin. Chem. 17 (1971), 1123.4. A. R. Byrneand L. Kosta, Sci. Total Env. 10 (1978), 17.5. A. S. Prasad, TraceElements and IroninHumanMetabolism, Plenum MedicalBook Company, 1978.6. E. C. Theil,Adv. Inorg. Biochem. 5(1983), 1.7. E. C. Theil andP. Aisen, inD. vanderHelm, J. Nei1ands, andG. Winkelmann, eds., IronTransportin Microbes, Plants, and Animals, VCH, 1987, p. 421.8. C. F. Mills, ed., Zincin HumanBiology, Springer-Verlag, 1989.9. B. L. ValleeandD. S. Auld, Biochemistry29(1990),5647.10. J. Miller, A. D. McLachlan, and A. Klug, EMBO J. 4(1985), 1609.11. J. M. Berg, 1. Bioi. Chem. 265(1990), 6513.12. C. Sennett, L. E. G. Rosenberg, and I. S. Millman, Annu. Rev. Biochem. 50(1981), 1053.13. J. J. G. Moura et al., in A. V. Xavier, ed., Frontiersin Biochemistry, VCH, 1986, p. 1.14. C. T.Walshand W. H. Grme-Johnson, Biochemistry26(1987), 4901.15. H. Kim and R. J. Maier, J. Bioi. Chem. 265(1990), 18729.16. Reference5, p. 5.17. V. L. SchrammandF. C. Wedler, eds., Manganese inMetabolismandEnzymeFunction, AcademicPress, 1986.18. (a)D. W. BoydandK. Kustin, Adv. Inorg. Biochem. 6(1984), 312; (b)R. C. Brueninget al., J. Nat.Products49(1986), 193.19. T. G. Spiro, ed., MolybdenumBiochemistry, Wiley, 1985.20. L. L. Fox, Geochim. Cosmochim. Acta 52(1988), 771.21. M. E. Farago, inA. V. Xavier, ed., Frontiersin BioinorganicChemistry, VCH, 1986, p. 106.22. I. G. Macara, G. C. McCloud, and K. Kustin, Biochem. J. 181(1979), 457.23. R. C. Brueninget al., 1. Am. Chem. Soc. 107(1985), 5298.24. E. Bayerand H. H. Kneifel, Z. Natwforsch. 27B(1972), 207.25. E. Bayer and H. H. Kneifel, in A. V. Xavier, ed., Frontiersin Bioinorganic Chemistry, VCH, 1986, p.98.26. J. Felcman, J. J. R. Fraustoda Silva, and M. M. CandidaVaz, Inorg. Chim. Acta 93(1984), 101.27. W. Mertz, Nutr. Rev. 3(1975), 129.28. E. C. Theil, 1. Bioi. Chem. 265(1990), 4771; Biofactors4(1993), 87.29. J. S. Rohrer et al., Biochemistry 29(1990), 259.30. P. H. Connett and K. Wetterhahn, Struct. Bonding 54(1983), 94.31. E. C. Theil, Ann. Rev. Biochem. 56(1987), 289;Adv. Enzymol. 63(1990), 421.32. G. C. Ford et al., Philos. Trans. Roy. Soc. Land. B, 304 (1984),551.33. G. D. Watt, R. B. Frankel, andG. C. Papaefthymiou, Proc. Natl. Acad. Sci. USA82(1985), 3640.34. (a) J. S. Rohrer et al., J. Bioi. Chem. 262(1987), 13385; (b) 1. S. Rohrer et al., Inorg. Chem. 28(1989), 3393.35. D. H. Hamer, Ann. Rev. Biochem. 55(1986), 913.36. A. H. Robbins, D. E. McRee, M. Williamson, S. A. Collett, N. H. Xuong, W. F. Furey, B. C. Want,and C. D. Stout, J. Mol. Bioi. 221(1991), 1269.37. N. D. Chasteen, Adv. Inorg. Biochem. 5(1983), 201.38. P. Aisenand I. Listowsky,Annu. Rev. Biochem. 49(1980), 357.39. B. T. Andersonet al., Proc. Natl. Acad. Sci. USA84(1987), 1768;E. N. Baker, B. F. Anderson, andH. M. Baker, Int. 1. Bioi. Macromol. 13(1991), 122.40. S. Bailey et al., Biochemistry 27(1988),5804.41. M. Raglandet al., 1. BioI. Chem. 263(1990), 18339.42. B. F. Matzanke, G. Muller, and K. N. Raymond, in T. M. Loehr, ed., IronCarriers and Iron Proteins,VCH, 1989, pp. 1-121.43. L. Stryer, Biochemistry, Freeman, 1981, pp. 110-116.44. D. J. Ecker et al., J. Am. Chem. Soc. 110(1988), 2457.45. Y. Sugiura and K. Nomoto, Struct. Bonding58(1984), 107.46. S. Mann, Struct. Bonding54(1986), 125.IV. REFERENCES47. A. R. Bullset al., J. Am. Chern. Soc. 112(1990), 2627.48. J. Webb, inP. WestbroekandE. W. deJong, eds., Biomineralizationand Biological MetalAccumula-tion, Reidel, 1983, pp. 413-422.49. K. Kustinet al., Struct. Bonding 53(1983), 139.50. R. B. Frankel and R. P. Blakemore, Philos. Trans. Roy. Soc. Lond. B 304 (1984),567.51. S. J. Lippard,Angew. Chemie22(1988), 344.52. K. E. Wieghardt, Angew. Chemie 28(1989), 1153.53. E. C. Theil, inR. B. Frankel, ed., IronBiomineralization, Plenum Press, 1990.54. S. Maun, J. P. Harrington, and R. J. P. Williams, Nature234(1986), 565.55. W. H. ArmstrongaudS. J. Lippard, J. Am. Chern. Soc. 107 (1985),3730.56. L. Que, Jr. and R. C. Scarrow, in L. Que, ed.,MetalClustersin Proteins, ACSSymposium Series 372,AmericanChemicalSociety, Washington, DC, 1988, p. 152and referencestherein.57. S. M. GorunaudS. J. Lippard, 1. Am. Chern. Soc. 107(1985), 4570.58. S. M. Gorunet al., J. Am. Chern. Soc. 109(1987), 3337.59. Q. Islamet al., J. Inorg. Biochem. 36 (1989), 51.60. C.-Y. Yanget al., J. Inorg. Biochem. 28(1986), 393.61. A. N. Mansour et aI., J. Biol. Chern. 260(1985), 7975.62. D. C. Hamsand P. Aisen, in T. M. Loehr, ed., IronCarriers and Iron Proteins, VCH, 1989, pp. 239-352.63. P. M. Harrisonand T. M. Lilley, in Reference62, pp. 123-23S.64. G. S. Waldoet al. Science259(1993), 796.65. J. Trikha, G. S. Waldo, F. A. Lewandowski, Y. Ha, E. C. Theil, P. C. Weber, andN. M. Allewell,Protein18 (1994), issue #2, in press.These references contain general reviews of the subjects indicated:Chromium: 27,30Cobalt: 12Copper: 8IronBiochemistry; 7,31, 37, 42Biomineralization polynuclear models:6, 42, 56, 57,58Siderophores: 42Structure of storage and transport proteins:32, 62, 63Manganese: 17Molybdenum: 19Nickel: 13, 14Vanadium: 18Zinc: 8, 9, 11, 35352TheReactionPathwaysofZincEnzymesandRelatedBiological CatalystsIVANaBERTINIDepartment of ChemistryUniversity ofFlorenceCLAUDIOLUCHINATInstitute of Agricultural ChemistryUniversity of BolognaI. INTRODUCTIONThis chapter deals withmetalloenzymes whereinthe metal acts mainly as aLewisacid; i.e., themetal doesnot change its oxidation state nor, generally, itsproteinligands. Changesinthecoordinationspheremayoccuronthesideex-posed tosolvent.Thesubstrateinteractswithproteinresiduesinsidetheactivecavityand/orwith the metal ion in order to be activated, so that the reaction can occur. Underthesecircumstancesthecatalyzedreactionsinvolve, ascentral stepswithoftencomplexreactionpathways, thefollowingbond-breakingand/or formationpro-cesses:peptidehydrolysisH HI IR-C-N-R' +H 0 ~ R-C-O-+ H-W-R'II 2 II Io 0 Hcarboxylicester hydrolysisR-C-O-R' +H 0 ~ R-C-O-+ H-O-R' + WII 2 IIo 0(2.1)(2.2)3738 2/ THEREACTIONPATHWAYS OF ZINCENZYMES ANDRELATEDBIOLOGICAL CATALYSTSphosphoricester hydrolysis0- 0-I IH-O-P-O-R'+HO~ H-O-P-O-+H-O-R'+W (2.3)II 2 IIo 0nucleophilicadditionof OH-and H-0HO-R RII"'-/I IH-O----tC----'"O-C ;O=C CH3-Cob(llI)alamin+1/212Cob(l)alamin+CH31---'> CH3-Cob(llI)alamin+1-(2.32)(2.33)Thesecanformallyberegardedas complexes of cobalt(III) withacarbanion.Thesearerareexamples of naturallyoccurringorganometalliccompounds. TheCo-Cbondinalkylcobalamins is relatively weak(bonddissociationenergy= 100 kJmol - I, thoughhigher values arereportedintheliterature 182,183) andcanbebrokenthermally(byheatingthecomplexabove 100C) 182-184orpho-tochemically, evenindaylight exposure. 180,181 Theenergyof the Co-Cbondisabout 17 kJmol - 1greaterwhenthetransaxial baseisabsent. 184The cobalamincoenzyme is boundbythe apoenzyme withnosignificantchange in the absorption spectrum. 185 This suggests that no major change occursinthecoordinationof cobalt(III). Thefirst stepof thereactioninvolveshomo-lyticfissionof the Co-Cbond:182-184, 186-188B-Co"1-R---,> B-Co11. +R' (2.34)whereBandRaretheligandsat thea andf3apical positions. The5' -deoxy-adenosyl radical probablyreactswiththesubstrate, genericallyindicatedasSubH,to give the Sub' radical and RH. Then the rearrangement reaction proceedsalong a not-well-established pathway. It is the protein-substrate binding thatcontrolsthesubsequent chemistry. In the absence of protein the Co-Cbond iskineticallystable;inthepresence of protein and substrate the rate of labilizationof the Co-C bondincreasesbya factor of1011_1012. 182-185Bygeneratingtheradical inthecoenzymewithout theproteinbymeansof photolysisorthermo-lysis, we enable the coenzyme to catalyze some rearrangement reactionswithouttheprotein. It maythereforebethat theproteinplaysamajorroleininducingthehomolyticfission, but arelativelyminorroleinthesubsequent steps, per-hapsconfinedtopreventingthevariousspeciesfromdiffusingawayfromeachother.Studies onprotein-free corrinoids andmodel complexes have shownthatincreasing the steric bulkiness around the coordinatedCo: atomcancause adramatic labilization of the Co-C bond. 189 The protein-coenzyme adduct mightcontainthecoenzymeinarestingstateandtheproteininastrainedstate; thesubstratewouldthenswitchthesystemintoastrainedcoenzymeandarelaxedenzymewithlittlethermodynamicbarrier. Thestrainedformof thecoenzymeistheninlabileequilibrium withbase-on cobalt(II)andthe freeradical. 190 Thishypothesis, that conformational changes incobalamincanswitchchemical re-actionsonandoff, iscloselyanalogouswiththeknownaspectsof hemoglobinfunction.VI. PERSPECTIVES 101It has been suggested that the radical formationin the coenzyme is triggeredbyastericperturbationinvolvinganenzyme-inducedconformationaldistortionof thecorrinringtowardthedeoxyadenosyl group, therebyweakeningtheco-balt-carbonbond.187,190-194 Structural studies ofdifferent corrinoidcomplexesreveal highlypuckeredandvariableconformationsof thecorrinring, attestingtoits flexibility. 195 For thedimethy19lyoximemodels, it has beenshownthatincreasingthesizeof theaxial ligandBdoesinduce Co-Cbondlengtheningand weakening because of conformational distortion of the equatorial ligandaway fromB and toward the R group. 196 It has been proposed that the flexibilityof thecorrinligandisthereasonwhyNaturedoesnotusetheporphyrin ligandinvitaminB12 . 197Inanalternativeexplanation, theweakeningofthe Co-Cbond would be an electronic effect associated with the labilization of the Co-Nbond. 198VI. PERSPECTIVESAlthougha greatdeal isknown about thebiophysicalcharacteristics of thevar-iousenzymederivativesmentionedin thischapter, wearestillfarfroma clearunderstandingof their mechanismsof action, especiallyif wetakeintoconsid-erationthe role of eachamino-acidresidue inside the active-site cavity. Al-thoughwecansuccessfullydiscusswhycertainmetal ionsareusedincertainbiological reactions, westill donot knowwhynickel(II), for example, is in-volvedinthe enzymatic hydrolysis of urea.199,200 Ifwe are content withtheexplanationsgiven inSectionsIILA or V.D, wewould need modelcompoundsthataregood catalystsandperform the jobinseveralsteps. Thislatter require-ment wouldmakethevariousmodelsmuchmoreinteresting, andwould repre-sent a new objective in the investigation of thestructure-function relationship ofcatalyticallyactivemolecules. Indeed, thesynthesisof largepolypeptidesmayinprincipleprovidesuchmodels. Inthisrespect weneedtoknowmoreaboutproteinfolding, for whichemergingtechniqueslikeproteincomputergraphicsandmolecular dynamicsareverypromising.Chemical modificationsof proteinslikethealkylationof carboxylate124,201or histidine202 residueshave been performed for a long time. A newer approachtowardmodelingthefunctionofaprotein, andunderstandingtherole oftheactivesite, involvescleavingpart ofanaturallyoccurringproteinthroughen-zymaticor chemical procedures, and then replacing it with a syntheticpolypep-tide. Theuseofmodemtechniques of moleculargeneticshas allowedsite-di-rectedmutagenesistobecomein principlea verypowerfultechniqueforchanginga single residue in acavity. Site-directedmutagenesis is a very popular ap-proach, andits principal limitation withrespect tothe synthetic polypeptideroute is that onlynatural aminoacids canbe used(asidefromthe technicaldifficulties inbothapproaches). Small quantitiesof site-directedmutants havebeenobtainedfor CPA125-127andAP,203whereastheexpressionof CA 204,205isnowsatisfactory.102 2I THEREACTIONPATHWAYS OF ZINCENZYMES ANDRELATEDBIOLOGICAL CATALYSTSPredictions of the changes in structure needed to affect the reaction pathwaycannowadaysbemadewiththeaidof computers. Theoccurrenceof thepre-dicted change can be checked throughx-ray analysis and NMR. The latter spec-troscopy is today well-recognized as being able to provide structural informationonsmall (:::;20 kDa)proteinsthrough2-or3-dimensionalteclmiques.206-208 Thesetechniquesareincreasinglybeingappliedtoparamagneticmetalloproteinssuchasmanyof thosediscussedhere.208,209Theadvantageof handlingaparamag-neticmetalloproteinisthat wecananalyzesignalsshiftedfar awayfromtheirdiamagnetic positions, whichcorrespondtoprotons close tothe metal ion,69evenforlarger proteins. It ispossibletomonitor thedistancesbetweentwoormoreprotonsundervariousconditions, suchasaftertheadditionof inhibitorsor pseudosubstrates, chemical modification, orsubstitutionof aspecificaminoacid.VII. REFERENCES1. H. Sigel and A. Sigel, eds., MetalIonsinBiological Systems, Dekker, 26(1990).2. R. K. Andrews, R. L. Blakeley, andB. Zemer, inReference1, 23 (1988).3. K. Doi, B. C. Antanaitis, andP. Aisen, Struct. Bonding70 (1988), I.4. M. M. Werst, M. C. Kennedy, H. Beinert, andB. M. Hoffman, Biochemistry29(1990), 10526, andreferencestherein.5. A. G. Orpen et al., J. Chem. Soc., DaltonTrans. SI(1989).6. F. H. Westheimer, Spec. Publ. Chem. Soc. 8 (1957), 1.7. F. Basolo andR. G. Pearson, Mechanismsof Inorganic Reactions, Wiley, 2d ed., 1967.8. L. G. SillenandA. E. Martell, StabilityConstants of Metal-IonComplexes, Spec. Pub!. Chern. Soc.,London, 25, 1971.9. E. J. Billo, Inorg. Nucl. Chem. Lett. 11(1975), 491.10. I. Bertini, G. Canti, C. Luchinat, andF. Mani, Inorg. Chem. 20 (1981), 1670.11. S. Burki, Ph.D. Thesis, Univ. Basel, 1977.12. P. Wolley, Nature258 (1975),677.13. J. T. GrovesandR. R. Chambers, 1. Am. Chem. Soc. 106 (1984),630.14. J. T. GrovesandJ. R. Olson, Inorg. Chem. 24 (1985), 2717.15. L. J. Zompa, Inorg. Chem. 17 (1978),2531.16. E. Kimura, T. Koike, and K. Toriumi, Inorg. Chem. 27 (1988),3687;E. Kimura et al., 1. Am. Chem.Soc. 112(1990), 5805.17. R. S. Brownet al., 1. Am. Chem. Soc. 104 (1982),3188.18. I. Bertiniet al., Inorg. Chem29 (1990),1460.19. R. J. P. Williams, Coord. Chem. Rev. 100 (1990),573.20. D. W. Christianson, Adv. Prot. Chem. 42(1991),281.21. I. Bertini, C. Luchinat, and M. S. Viezzoli, in I. Bertiniet al., eds., Zinc Enzymes, Birkhauser, 1986,p.27.22. D. S. AuldandB. L. Vallee, inReference21, p. 167.23. G. Formicka-Kozlowska, W. Maret, and M. Zeppezauer, in Reference21, p. 579.24. I. Bertiniand C. Luchinat, inReference30, p. 10I .25. I. Bertini andC. Luchinat,Adv. Inorg. Biochem. 6 (1984),71.26. H. Sigel, ed., Metal IonsinBiological Systems, Dekker, 12 (1981).27. T. G. Spiro, ed., Copper Proteins, Wiley, 1981.28. J. E. ColemanandP. Gettins, inReference21, p. 77.29. J. E. Coleman andD. P. Giedroc, inReferenceI, 25(1989).30. H. Sigel, ed., Metal Ionsin Biological Systems, Dekker, 15(1983).31. N. U. Meldrum and F. J. W. Roughton, 1. Physiol. 75(1932), 4.32. D. Keilinand T. Mann, Biochem. J. 34 (1940), 1163.VII. REFERENCES33. S. Lindskog, inT. G. Spiro, ed., ZincEnzymes: Metal Ions inBiology, Wiley, 5(1983), 78; D. N.Silverman andS. Lindskog,Acc. Chern. Res. 21(1988), 30.34. J.-Y. Liangand W. N. Lipscomb, J. Am. Chern. Soc. 108 (1986),5051.35. K. M. Merz, 1. Am. Chern. Soc. 112 (1990),7973.36. (a)G. Sanyal, Ann. N. Y. Acad. Sci. 429(1984), 165;(b)G. Sanyal and T. H. Maren, J. Bioi. Chern.256(1981), 608.37. T. Kararliand D. N. Silverman, J. Bioi. Chern. 260 (1985),3484.38. A. Liljaset al., Nature235 (1972), 131.39. K. K. Kannan et al., Proc. Natl. Acad. Sci. USA, 72 (1975),51.40. E. A. Erikssonet al., inReference21, p. 317.41. S. LindskoginReference21, p. 307.42. E. Clementiet al., FEBS Lett. 100 (1979), 313.43. B. P. N. Koet al., Biochemistry16 (1977), 1720.44. A. Ikai, S. Tanaka, andH. Noda, Arch. Biochem. Biophys. 190 (1978), 39.45. I. Bertini, C. Luchinat, andA. Scozzafava,Struct. Bonding 48 (1982), 45.46. A usefulreview by W. W. Cleland on pH-dependent kinetics can be foundin Methods Enzymol., 1987,390.47. Y. Pocker andS. Sarkanen, Adv. Enzymol. 47 (1978), 149.48. I. Bertini andC. Luchinat, Acc. Chern. Res. 16 (1983),272.49. S. Lindskogand I. Simonsson, Eur. J. Biochem. 123(1982), 29.50. B.-H. Jonsson, H. Steiner, andS. Lindskog, FEBS Lett. 64(1976), 310.51. H. Steiner, B.-H. Jonsson, andS. Lindskog, Eur. J. Biochem. 59(1975),253.52. D. N. Silvermanet al., 1. Am. Chern. Soc. 101(1979), 6734.53. S. H. Koeniget al., Pure Appl. Chern. 40 (1974), 103.54. 1. Simonsson, B.-H. Jonsson, andS. Lindskog, Eur. J. Biochem. 93(1979), 4.55. T. J. WilliamsandR. W. Henkens, Biochemistry 24 (1985),2459.56. I. Bertini, C. Luchinat, andM. Monnanni, inReference99, p. 139.57. I. Bertiniet al., inReference21, p. 371.58. I. Bertiniet al., Inorg. Chern. 24 (1985), 301.59. I. Bertiniet al., 1. Am. Chern. Soc. 100 (1978),4873.60. R. C. Rosenberg, C. A. Root, andH. B. Gray, J. Am. Chern. Soc. 97(1975),21.61. B. Holmquist, T. A. Kaden, andB. L. Vallee, Biochemistry14 (1975), 1454, andreferencestherein.62. M. W. MakinenandG. B. Wells, inH. Sigel, ed., Metal Ions inBiological Systems, Dekker, 22(1987).63. M. W. Makinenet al.,J. Am. Chern. Soc. 107(1985), 5245.64. A. E. Eriksson, Uppsala Dissertation, Faculty of Science, n. 164, 1988.65. A. E. Eriksson, A. T. Jones, andA. Liljas, Proteins 4(1989),274.66. A. E. Erikssonet al., Proteins 4(1989),283.67. I. Bertiniet al., Inorg. Chern., 31(1992), 3975.68. P. H. Haffner andJ. E. Coleman, J. Bioi. Chern. 248(1973), 6630.69. I. Bertini andC. Luchinat, NMRof ParamagneticMolecules inBiological Systems, Benjamin/Cum-mings, 1986.70. R. D. Brown1II, C. F. Brewer, andS. H. Koenig, Biochemistry16 (1977), 3883.71. I. Bertini, C. Luchinat, and M. Messori, inReference62, vol. 21.72. I. Solomon, Phys. Rev. 99(1955),559.73. 1. Bertini et al., J. Magn. Reson. 59(1984),213.74. L. Banci, I. Bertini, andC. Luchinat, Magn. Res. Rev., 11(1986), 1.75. I. Bertiniet al., 1. Am. Chern. Soc., 103 (1981),7784.76. T. H. Maren, A. L. Parcell, and M. N. Malik, J. Pharmacol. Exp. Theor. 130 (1960), 389.77. J. E. Coleman, J. Bioi. Chern. 243 (1968), 4574.78. S. Lindskog,Adv. Inorg. Biochem. 4(1982), 115.79. G. Albertiet al., Biochim. Biophys. Acta16 (1981), 668.80. J. I. Rogers, J. Mukherjee, and R. G. Khalifah, Biochemistry 26 (1987),5672.81. C. Luchin