Biochem. J. (1968) 110, 621 621Printed in Great Britain

The Catalase-Hydrogen Peroxide SystemA THEORETICAL APPRAISAL OF THE MECHANISM OF CATALASE ACTION

By PETER JONES AND A. SUGGETTDepartment of Phy8ical Chemiatry, University of Newca8tle upon Tyne, NE1 7RU

(Received 4 July 1968)

1. The mechanisms of catalase action advanced by Jones & Wynne-Jones (1962)and by Nicholls (1964) are compared in terms oftheir relative plausibilities and theirutility for extension to accommodate more recent experimental information. 2. Arevised formal mechanism is advanced that avoids the less satisfactory features ofthese mechanisms and attempts to account for the roles of catalase sub-units inboth reversible and irreversible deactivation phenomena. 3. Theoretical studies ofthe redox chemistry ofperoxides are used to provide the basis for a discussion ofthemechanism ofthe redox act in catalatic action at the molecular level. It is suggestedthat an important feature of catalase action may be a mediation of the formation ofa reactive intermediate by stereospecifically located acid-base functions in theactive site. 4. A more detailed statement of this concept is attempted in terms ofa hypothetical partial molecular model for the composition and stereochemistry ofthe active site of catalase. The utility of this model in describing the catalatic andperoxidatic actions of catalase is assessed.

Recent experimental studies of the catalase-hydrogen peroxide system (Jones & Suggett,1968a,b; Jones, Pain & Suggett, 1968) have revealednew phenomena for which existing theories ofcatalase action do not account. In particular, thiswork has shown that reversibly formed catalase sub-units play important roles in catalase action atphysiological pH.

Jones & Wynne-Jones (1962) proposed a mech-anism for catalase action that attempted to accountfor the data available at that time. It was shownthat the mechanism could also comprehend theavailable information on the catalytic action ofFe3+ion, both in the decomposition ofhydrogen peroxideand as a mediator of coupled 'peroxidatic' oxida-tions. Nicholls (1964) rejected the Jones & Wynne-Jones (1962) scheme for catalase action and putforward an alternative model. Nicholls's (1964)model implies that the mechanisms by which cata-lase and Fe3+ ion bring about the mutual oxidationand reduction of two molecules of hydrogen per-oxide differ qualitatively.

In section (i) of the Theory section of the presentpaper the Jones-Wynne-Jones and Nichollsschemesare compared both as mechanistic models and interms of their implications at the molecular level.The purpose of this comparison is to assess therelative plausibilities of the two schemes and theirutility for extension to accommodate more recent

experimental information. In section (ii) a model isproposed that attempts to comprehend the datacurrently available, and tests are suggested thatcould be of value in resolving areas of uncertainty.Attempts to discuss the mechanism of catalase

action at the molecular level are restricted by thelack of a detailed knowledge of the structure of theenzyme itself. Although much work has been doneon the formal characterization of intermediates (thecatalase 'Compounds'), little theoretical attentionhas been paid to the fate of the two molecules ofhydrogen peroxide that are involved in catalaticaction. In section (iii) ideas deriving from a molecu-lar-orbital study ofthe redox chemistry ofperoxides(Jones & Perkins, 1967) are used to suggest a formalmodel for the manner in which the catalase proteinmay be involved in facilitating catalatic action. Amore explicit molecular statement is then attemptedin terms of a model for the composition and stereo-chemistry of the active site of catalase. Althoughthis model cannot be regarded as definitive, it isremarkably successful in accounting for the phe-nomena of catalatic action and it suggests that theextraordinary catalytic efficiency of the enzyme de-rives from the stereospecific location of appropriateacid-base functions in the active site. In section (iv)the implications of the model in the context of'peroxidatic' oxidations by catalase-peroxide sys-tems are discussed.

P. JONES AND A. SUGGETT

(a) Joiles-Wynne-Jones mechaniism

(I)

Primary Secondary+p P~~~~~~1

E EE pEP ) EP' - E+products

Ep +P

EP2 (IM9

(b) Nicholls mechanism

+P EPE )- EP (1) ), E£P2 (Ogura) ) E+products

+AH2-1p

I+AH2 +AHz

4-PEP' (II) 'EP2' (III)_p

(c)+P

E -4 EP (I)

+AH2j

EL"' (TI)

+AHH2-*+P E'P

I+AHz

E'P'+ 2i ec

Et' + etc.

Scheme 1. Comparison ofmechanisms proposed for catalase action.

THEORY(i) Oompari8on of the Jones-Wynne-Jone8 and

Nicholls mechanni8m8. The two schemes are repre-sented in Schemes 1(a) and 1(b) respectively. Beforeattempting a comparison, some clarifications ofterminology and principle are required. Nicholls(1964) takes the view that his mechanism is an ex-

tension ofthe 'peroxidatic theory' ofcatalase action(Chance, Greenstein & Roughton, 1952), whereasthe Jones-Wynne-Jones theory cannot be so des-cribed. On the contrary, both mechanisms containthe essential concept of 'peroxidatic theory', whichis that the mutual oxidation and reduction ofhydro-

gen peroxide in catalatic action is a two-electron-equivalent redox process in which a catalase-per-oxide 'compound', containing both oxidizingequivalents of its parent peroxide molecule, oxi-dizes a second hydrogen peroxide molecule tooxygen. 'Peroxidatic' oxidations occur via a com-petition between hydrogen peroxide and otheroxidizable species for this oxidizing 'compound'.

Nicholls (1964) has commented that the extensionof inorganic theories to cover catalase action isdangerous. In our view it is important to enquirewhether the mechanisms of catalytic decompositionofhydrogen peroxide by catalase and Fe3+ ion differqualitatively or essentially quantitatively; indeed,

622 1968

MECHANISM OF CATALASE ACTIONsuch enquiries are important for the development ofcatalytic theory. The Jones-Wynne-Jones mech-anism is not an 'inorganic theory', nor is Nicholls'smechanism a 'biochemical theory'; both are (rathercrude) models that must be tested both by compari-son with experiment and by consideration of theirinherent plausibilities.The catalatic pathway is represented by the upper

horizontal flow-line in both mechanisms in Schemes1(a) and 1(b). Here Nicholls follows Chance et al.(1952) and Ogura (1955) in supposing that the keyoxidizing entity (Compound I) is formed in a singlestep from catalase and hydrogen peroxide and thatthe redox step occurs in a ternary complex (Oguracomplex). It is an essential feature of the schemethat both complex-formation steps occur substan-tially irreversibly. The Jones-Wynne-Jones schemehas amore conventional enzyme-kinetic form in sup-posing the reversible formation of a primary com-plex that undergoes an essentially irreversible trans-formation into a secondary complex before takingpart in the redox act. Thus, in this model, 'Com-pound I' is regarded as a steady-state mixture oftwospecies. Although the direct kinetic and spectro-photometric data do not exclude either formulation,important indirect support for the Jones-Wynne-Jones model comes from the detailed analysis of thespectra of catalase derivatives by Brill & Williams(1961), in which they concluded that 'Compound I'is a steady-state mixture of two species. Furthersupport comes from the data obtained by Strother &Ackerman (1961), who found that, whereas the rateconstant for the reaction of 'Compound I' withperoxide showed a normal Arrhenius temperature-dependence, the rate constant for the formation of'Compound I' showed a complex temperaturevariation. This suggests that the formation of'Compound I' is kinetically complex. The largechange in absorption spectrum accompanying theformation of 'Compound I' indicates a substantialelectronic rearrangement. Such a process wouldmore plausibly occur with a high overall rate con-stant via a sequential pathway rather than in asingle step. The Jones-Wynne-Jones formulationis therefore retained in developing a revised formalmechanism in section (ii) and in considering aninterpretation at the molecular level in section (iii).The reactions that are supposed to be responsible

for reversible deactivation are represented asbranches from the main catalatic pathways inScheme 1. The Jones-Wynne-Jones mechanismhas a single branch where deactivation is accountedfor by the slow reversible formation of a diperoxy-catalase species from the primary complex.Nicholls's scheme is more complex and involves theformation of two inactive species, Compounds IIand III. The concept of two inactive species is themore satisfactory in the light of our studies of the

kinetics of the effect of reaction time on catalasekinetics at high substrate concentrations (Jones &Suggett, 1968b), since one must account both for theapparent change in Km with time and for the re-versible inhibition by excess of substrate in thefl-activity phase of catalatic action.However, there are serious obj ictions to the man-

ner in which these processes are supposed to occur inNicholls's mechanism. The reactions by whichCompound II is formed from Compound I and bywhich catalase is regenerated from Compound IIare both supposed to be one-electron-equivalentredox steps involving an 'endogenous donor' pres-ent in the catalase molecule. The introduction ofthis hypothetical species superficially avoids viola-tion of the Principle of Microscopic Reversibilityin the loops represented by:

E -' I

IIaind

E - > I - Ogura coliiplex

JI

There is, however, a logical flaw in the argument. Ifthe donor is truly 'endogenous', then native cata-lase cannot be regenerated by the above pathways,but the end product must involve successive stepsof chemical modification of the enzyme for eachpass through the 'loops'. The changes cannot beeither cyclic or reversible (without further assump-tion) and a more accurate representation of the first'loop' is shown in Scheme l(c). Nicholls's kineticscheme can therefore only be maintained with theaid of further assumptions of the type:

E E'....EP E'P E"P

etc.

where the equivalence signs indicate the assump-tions that successive stages of chemical modifica-tion of catalase and the reaction intermediates alterneither their kinetic behaviour nor their spectro-scopic properties. It is not possible to dismisshypotheses of this type entirely, but their plausi-bility diminishes sharply with increase in the weightof assumption necessary to sustain them.At the molecular level further difficulties with

this model appear. If the 'endogenous donors' areassumed to lie so close to the active site that theredox changes may occur directly, then grosschanges in chemical composition very near the ac-tive site must be presumed to have no effect oncatalase activity. If the 'endogenous donors' areassumed to be remote from the active site then one-electron equivalents of oxidation must be trans-mitted either through the protein matrix or, as freeradicals, through the solution.

Vol. 110 623

P. JONES AND A. SUGGETT+P

E 11 EP EP'- Primary Second

complex compi

x

2(E/2):

(+P')4(E/4)I ::::::::::::: ............ II*s@e@@~~~~~~~~~~~~~~1

Irreversiblydeactivatedenzyme

Peroxidaticaction

9

+P- ) E + products (catalatic action)

lary[ex

+AH2

E + prpducts (peroxidatic action)

+P

- - - Tetrameric+H+ domain

Sub-unitdomain

III

Scheme 2. Revised formal mechanism for catalase action.

We conclude that the 'endogenous-donor' hy-pothesis is unlikely to be of value in developing a

revision of the mechanism of catalase action.(ii) Revised formal mechanism of catalw8e action.

The model proposed, in which Compounds I, II andIII have the usual Chance connotation, is shown inScheme 2 and has the following features:

(a) The main catalatic pathway is shown by theupper horizontal flow-line. The scheme follows theJones-Wynne-Jones mechanism and assigns cata-latic behaviour to the tetrameric domain of catalasebehaviour (but see b). The species EP' is consideredto be derived from the primary enzyme-substratecomplex by some intramolecular rearrangement.

(b) The left-hand vertical flow-line shows theprotonation-induced dissociation of catalase intofour sub-units with subsequent irreversible de-activation of the sub-units (Jones & Suggett,1968a). A dimeric species (E/2) is tentatively indi-cated in the dissociation process. This suggestionmeets the requirements of the studies made bySchiitte, Steinbrecht & Winder (1960), whose re-

sults also suggest that a dimer may be the essentialcatalatic unit. In view of our results on the per-oxide-induced re-formation of active catalase fromsub-units (Jones & Suggett, 1968a) we have notincluded a separate catalatic pathway at thedimeric sub-unit level.

(c) The major reversible deactivation pathways inthe presence of substrate are represented by theright-hand vertical flow-line. This represents theformation of Compound II as a transition from thetetrameric to the sub-unit domain (Jones et al. 1968)and as a process in which a molecule of hydrogenperoxide reacts with the tetrameric secondary com-plex (Jones & Suggett, 1968b). Direct studies of thekinetics of Compound HI formation have revealed acomplex situation (Chance, 1949) and these resultsare not accounted for by existing models for theprocess. Thus Nicholls (1963) has proposed ascheme in which the transition from Compound I toCompound II is described by a single rate constant,which is independent of substrate concentration insome circumstances but not in others. This is un-satisfactory. The implications of the results areeither that there is a single complex pathway (morethan one step), or that there are at least two inde-pendent pathways. The scheme proposed byNicholls (1964) and shown in Scheme l(b) suggestsa second pathway for Compound II formation viathe 'Ogura ternary complex'. The problems arisingfrom the invocation of 'endogenous donor' in theseschemes have been discussed above; it should alsobe pointed out that they do not comprehend theobserved effect of pH.In Scheme 2 it is suggested that, at low substrate

624 1968

+P

MECHANISM OF CATALASE ACTIONconcentrations (as when hydrogen peroxide is gen-erated by the glucose oxidase system), Compound IIis largely formed from monomeric sub-units (E/4species) in a reaction with peroxide at a rate that iscontrolled by the protonation-induced dissociationof catalase. At higher substrate concentrations theenzyme dissociation is displaced towards the tetra-meric domain (Jones & Suggett, 1968a) and thesecond pathway becomes important. In the latterthe rate of Compound II formation is controlled bythe rate of reaction of the secondary complex withperoxide. It is suggested that this results in a tetra-meric intermediate (X), which undergoes rapidprotonation and dissociates to yield Compound II.The species X is, in a sense, a diperoxy-catalasederivative, in that the secondary complex containsboth oxidizing equivalents and probably the com-ponents (Chance & Schonbaum, 1962) of its parentperoxide molecule. It seems possible that the secondperoxide molecule is bound at a different haem sitefrom the first and that this process destabilizestetrameric catalase with respect to protonation-induced dissociation.

Catalase Compound III is a less well charac-terized species, which is reportedly formed by thereversible binding of hydrogen peroxide to Com-pound II, and this is formally indicated in Scheme 2.It is possible that Compound III may result fromthe binding of peroxide by haem groups liberatedin Compound II formation. An alternative pos-sibility is that its formation results from a reversibledisplacement by peroxide of the protein bound tothe fifth co-ordination position of an iron atom in amonohaem sub-unit species.

It has been proposed by a number ofworkers thatCompound II is an Fe (IV) species, the evidence de-riving mainly from titration of its oxidizing power.Evidence obtained in a similar manner was held fora number of years to support the formulation ofperoxidase Compound II as an Fe (IV) species. Thespectra of catalase Compound II and peroxidaseCompound II are strikingly similar. The reinvesti-gation of peroxidase Compound II by Yonetani(1965) has shown the Fe (IV) formulation to be un-sound and suggests an urgent need for reappraisalof the catalase system. In the latter case experi-ments would undoubtedly be complicated by sub-unit interactions and it seems unlikely that a studyof gross properties would yield an unequivocal re-sult. It is perhaps worth pointing out that a di-peroxy derivative of tetrameric catalase is formallyan Fe (IV) species.

(iii) Partial molecular modelfor catala8e action. Thedetailed molecular structure of catalase is unknownand thus the development of a definitive molecularmechanism for catalase action is precluded. Never-theless, it is pertinent to enquire whether theavailable information suggests a likely molecular

explanation of the extraordinary catalatic power ofthis enzyme. Jones & Wynne-Jones (1962) arguedthat there was substantial support for the thesis thatcatalatic action and Fe3+ ion catalysis are qualita-tively similar processes, and more recent work(Kremer, 1965; Brown & Jones, 1968) both lendssupport to this argument and suggests the inclusionof haemin in this comparison. We do not proposeto develop this argument here and the present dis-cussion is restricted to experimental informationderiving directly from studies of catalase action andto the results of theoretical studies of the redoxchemistry of peroxides.

In a theoretical investigation, using the Pople-Segal SCF-MO method (Pople & Segal, 1965), Jones& Perkins (1967) computed the variation of elec-tronic charge density across the ground-state energysurface of the H02- ion. The results may be sum-marized as follows:

(a) Extension of the 0-0 bond in H02- leads tonegative charge accumulation on the OH group andultimately to the process:

o 0 >- 0-+0

H ~ H

(b) Extension of the 0-H bond in H02- leads tonegative charge accumulation on H and ultimatelyto the process:

K,O-(O - H~-+02 (singlet)/

These results suggest that H02- may act as anoxygen-atom-transfer oxidizing agent and as ahydride-ion-transfer reducing agent. There is sub-stantial support in the peroxide literature for theoccurrence of both types of process (see Jones &Perkins, 1967). Although these results find directapplication to Fe3+ ion and haemin catalysis,where reaction via H02- complexes is implicated,the evidence is that catalase utilizes molecularhydrogen peroxide. For this situation Jones &Perkins (1967) suggested that a couplingwith protontransfer might be involved:

H-0-i--0-+H

H -

Thus, in its oxidizing action, hydrogen peroxidemayundergo either a two-step dehydration (via HO2-),or a one-step dehydration via a 'concerted' acid-base action. Edwards, Di Prete, Curci & Modena

Vol. 110 625

P. JONES AND A. SUGGETT

BIH--

(a)

It F

B A B. . ~~~+P

H H < H. .

H s~~H% ,

(d)

B A-

I-I+HN" HI,R110 H;oH Hz->

(b)

1i

(c)

0 O

(e)

Scheme 3. Hypothetical scheme for the role of catalase protein in facilitating catalatic action.

(1968) have elegantly demonstrated that the lattertype of process occurs, in hydrogen peroxide oxida-tion of organic sulphides to sulphoxides, by themediation of water molecules in cyclic transitionstates. In applying this concept to catalase action itwould be plausible to suppose that acid-base media-tion derives from the catalase molecule rather thanfrom the solvent. In these terms a model for cata-lase action would have the form shown in Scheme 3,where BH+ and A- represent functions of suitableacid-base power, sterically oriented to facilitate thereaction. (a) in Scheme 3 represents the 'active site'of catalase with a water molecule bound to the ferriccentre. The primary reversible binding ofhydrogenperoxide to the ferric centre is shown in (b) inScheme 3. (c) in Scheme 3 then represents the transi-tion to the secondary complex as a locking intoposition of the hydrogen peroxide molecule by the

formation of hydrogen bonds from BH+ and to A-.This process has consequential electronic effects;the coupled 'push-pull' partial proton transfersresult in a clockwise flow of negative charge thatproduces a weakening of the peroxidic 0-0 bondand a development ofthe oxygen atom bond to ironas a powerful electrophilic centre. If the protontransfers were complete this process would result inthe formation of an oxidizing species of the Fe03+type. In (c) in Scheme 3 we accept the result ofChance & Schonbaum (1962) that the componentsof the parent peroxide molecule are retained in'Compound I' (in our terminology, in the secondarycomplex) by indicating partial proton transfer butsignificant extension of the 0-0 bond. As a hydro-gen peroxide molecule approaches the secondarycomplex it is polarized so that a coupled proton andhydride ion transfer, via the transition state (d) in

1968626

H'.",.o171

MECHANISM OF CATALASE ACTION

i /> If

Tob prIotein

Fig. 1. Projection showing the stereospecific binding ofH202 in a hypothetical partial molecular model for theactive site of catalase.

Scheme 3, is facilitated, leading to the formation ofproducts and regeneration of the active catalaticsite.A more explicit molecular statement ofthis model

must satisfy both the experimental kinetic data forcatalase action and the stereochemical implicationsofthe processes represented in Scheme 3. Althoughit may be argued that such a model could not bedefinitive in the absence ofdirect structural informa-tion, the number of combinations that meet therather stringent requirements must be severelylimited and it is certainly important, at least, todemonstrate the possibility of constructing a satis-factory model with the information currentlyavailable. From an assessment of the experimentaldata and with the aid of molecular models we haveconcluded that the most satisfactory picture of theactive site ofcatalase is obtained ifBH+ is identifiedwith a > C(NH2)2+ group from an arginine residuein the protein and A- with a carboxylate group thatmay derive either from a propionate side chain ofthe protoporphyrin ring or from an aspartate or

glutamate residue in the protein, or perhaps from a

terminal amino acid. Fig. 1 shows a projection of amolecular model for the key step in the reaction, thestereospecific binding of hydrogen peroxide in thesecondary complex, constructed on the assumptionthat A- is a propionate group. The hydrogen-bondlengths required (i.e. the internuclear separations ofthe heavy atoms) are about 2 9! for a normal 0-0peroxide bond length (1-48!) and regular tetra-hedral bond angles. Partial proton transfer fromarginine nitrogen would probably result in a changeof its co-ordination towards a trigonal conforma-

tion, which would accommodate extension of theperoxidic 0-0 bond.A key feature of the model is the hydrogen bond

between the arginine and propionate residues thatlocks the functional components into a stereo-chemical array precisely suited for the binding ofhydrogen peroxide and for the bifunctional 'push-pull' acid-base interactions formally envisaged in(c) in Scheme 3. In Fig. 1 it is also suggested that afurther stabilization ofthe active sitemay arise froma hydrogen bond between the arginine residue andthe second propionate side chain of the proto-porphyrin ring.

In comparing this model with experiment thefollowing important points emerge:

(a) Catalatic reaction via this path would be inde-pendent ofpH between about pH5 and pH1O.

(b) Protonation of the propionate group or de-protonation of the arginine residue would disruptthe active site and lead to deactivation of catalaseat the extrema ofpH as observed. In this connexionwe recall the observation ofAgner & Theorell (1946)that a process ofpK3.8 was associated with the de-activation of catalase at low pH. At the time it wassuggested that the observation related to acidionization of a water molecule bound to the ferriccentre, but later authors have not accepted thisview. We suggest that this phenomenon may wellbe related to protonation of a propionate group inthe active site of catalase. A rapid decrease incatalase activity above pH10 5 has been reportedbyseveral workers (Jones & Suggett, 1968a,b) and inour model is ascribed to deprotonation of arginine> C(NH2)2+.

(c) Protonation of the propionate carboxyl groupor deprotonation of arginine may induce dissocia-tion of the enzyme, particularly if the arginineresidue is primarily bound in a sub-unit other thanthe one in which it is involved as an active site-com-ponent. Our results (Jones & Suggett, 1968a) implythat protonation of catalase leads to dissociationand subsequent irreversible deactivation. The con-clusion that catalase undergoes dissociation at highpH has been reached by a number of workers, andSamejima (1959a,b) reported a 'partially reversible'deactivation under these conditions.

(d) A 'catalatic' reaction involving only alkylhydroperoxides is proscribed by this model, but anoxidation of hydrogen peroxide by a secondarycomplex formed from alkyl hydroperoxide is un-impaired. Internal oxidation-reduction of an n-alkyl hydroperoxide molecule bound in a secondarycomplex is probable (and is observed).

(e) The product molecular oxygen derives from asingle parent hydrogen peroxide molecule (Jarnagin& Wang, 1958a), and the model is also consistentwith observed deuterium kinetic isotope effects,both in the catalatic reaction (Jarnagin & Wang,

Vol. 110 627

628 P. JONES AND A. SUGGETT 19681958b) and in peroxidatic oxidations of hydrogendonors (Frei & Aebi, 1957; Nicholls & Schonbaum,1963b).

(f) In the secondary complex there are likely tobe substantial interactions between the outer per-oxidic oxygen and the 7r-electron system of theprotoporphyrin ring. These might make a signifi-cant contribution to changes in the spectrophoto-metric properties of the enzyme. In secondarycomplexes formed from alkyl hydroperoxides theremay be interactions between the alkyl group andthe hydrophobic region of the protoporphyrin ring,leading to quantitative differences between thespectra of these species and the secondary complexformed from hydrogen peroxide.

(g) The model suggests that inhibitors may act ina variety of ways. For some species (perhapscyanide) a straightforward reversible binding of theferric centre is likely. A more complex action mayresult with species that can compete successfullyfor the arginine-propionate hydrogen bond andbring about dissociation of the enzyme. Effects ofthis type may be enhanced by binding of substratein the secondary complex and could result in irre-versible deactivation. 3-Aminotriazole (Margoliash,Novogrodsky & Schejter, 1960) may belong to thiscategory. Some inhibitors may bind both the ferriccentre and the acid-base functions in the intactenzyme, a type of action that would involve sub-stantial steric effects so that the inhibition might berather slowly established. This could explain thedifference in catalase activity in different bufferswhen measured by the Bonnichsen, Chance &Theorell (1947) assay procedure and that obtainedwhen catalase is equilibrated with buffers other thanphosphate (Jones & Suggett, 1968a,b).

(h) The model satisfies all the criteria for a suc-cessful mechanism of catalase action prescribed byNicholls & Schonbaum (1963a), with the exceptionof those deactivation processes (see section i) thatare supposed by them to involve an 'endogenousdonor', and is also in accord with more recent ex-perimental data.

(iv) Peroxidatic oxidation8. Studies of 'peroxidaticoxidations' of reducing substrates by catalase-peroxide systems have revealed considerablephenomenological diversity (Keilin & Nicholls,1958). The model that we have proposed has anumber of implications that might go some waytowards resolving these problems.A major group of substrates contain the group:

oCuOHincluding primary and secondary alcohols and theirderivatives and formic acid. It is suggested thatthey compete with hydrogen peroxide for Com-

pound I (the secondary complex in our terminology)and are oxidized by a two-electron-equivalent hy-dride transfer, coupled with proton transfer. Thistype of process is readily accommodated by ourmodel and indeed is supported by theoretical (MO)studies of ac-hydrogen abstraction in a number ofspecies of this type (P. Jones, K. Levison & P. G.Perkins, unpublished work). These show thathydride transfer is the preferred mode of ac-hydrogenabstraction from formate, methoxide and isoprop-oxide ions.

In (d) in Scheme 3 we have suggested that, in theoxidation of hydrogen peroxide, proton release isfacilitated by the outer (nucleophilic) peroxidicoxygen atom in the secondary complex. It is, ofcourse, possible that this process is mediated byanother basic residue from the catalase protein, inboth catalatic and peroxidatic actions.

There is a considerable body of literature indi-cating that the irreversible formation of sub-unitsfrom catalase produces species with greatly en-hanced peroxidatic properties (Anan, 1958; Inada,Kurozumi & Shibata, 1961; Caravaca & May, 1964).It is not impossible that reversibly formed sub-unitsmay show similar behaviour, and in Scheme 2 wehave tentatively indicated this possibility. Com-plexities deriving from sub-unit formation may beimportant in the phenomena of peroxidatic oxida-tions, but the sub-unit domain ofcatalase behaviouris relatively unexplored at the present time.

Finally, our model of catalatic action suggests thepossibility that the molecular oxygen produced mayemerge as an excited singlet species that could itselfoxidize reducible substrates present in the system.In this connexion the work of Anbar (1966) isrelevant. A measurable oxidation of molecularnitrogen to nitrite and nitrate was observed whennitrogen was dissolved under a pressure of 100 atm.in water and hydrogen peroxide was either oxidizedby hypochlorite (a reaction in which chemilumines-cence from singlet molecular oxygen can be ob-served) or catalytically decomposed by catalase inthese solutions.

We are indebted to Lord Wynne-Jones for his continuedencouragement. P. J. thanks the Royal Society-IsraelAcademy Programme and the Department of PhysicalChemistry, The Hebrew University of Jerusalem, for sup-port and hospitality during a study visit, and ProfessorG. Stein and Dr M. L. Kremer for useful discussions. A. S.thanks the Science Research Council for a Research Student-ship and the University of Newcastle upon Tyne for a SirJames Knott Fellowship.

REFERENCES

Agner, K. & Theorell, H. (1946). Arch. Biochem. 10, 321.Anan, F. K. (1958). J. Biochem., Tokyo, 45,227.Anbar, M. (1966). J. Amer. chem. Soc. 88,5924.

Vol. 110 MECHANISM OF CATALASE ACTION 629

Bonnichsen, R. K., Chance, B. & Theorell, H. (1947). Actachem. 8cand. 1, 685.

Brill, A. S. & Williams, R. J. P. (1961). Biochem. J. 78,253.Brown, S. B. & Jones, P. (1968). Trans. Faraday Soc. 64,

999.Caravaca, J. & May, M. D. (1964). Biochem. biophy8. Res.Commun. 16, 528.

Chance, B. (1949). J. biol. Chem. 179,1341.Chance, B., Greenstein, D. S. & Roughton, F. J. W. (1952).

Arch. Biochem. Biophys. 37, 301.Chance, B. & Schonbaum, G. R. (1962). J. biol. Chem. 237,

2391.Edwards, J. O., Di Prbte, R., Curci, R. & Modena, G. (1968).In Hydrogen-Bonded Solvent Systems, p. 303. Ed. byCovington, A. K. & Jones, P. London: Taylor and FrancisLtd.

Frei, E. & Aebi, H. (1957). Helv. chim. Acta, 41,241.Inada, Y., Kurozumi, T. & Shibata, K. (1961). Arch.

Biochem. Biophys. 93, 30.Jamagin, R. C. & Wang, J. H. (1958a). J. Amer. chem. Soc.

80,786.Jamagin, R. C. & Wang, J. H. (1958b). J. Amer. chem. Soc.

80, 6477.Jones, P., Pain, R. H. & Suggett, A. (1968). Nature, Lond.,

217, 1050.Jones, P. & Perkins, P. G. (1967). Nature, Lond., 215, 129.

Jones, P. & Suggett, A. (1968a). Biochem. J. 108, 833.Jones, P. & Suggett, A. (1968b). Biochem. J. 110, 617.Jones, P. & Wynne-Jones, W. F. K. (1962). Tran8. Faraday

Soc. 58, 1148.Keilin, D. & Nicholls, P. (1958). Biochim. biophy8. Acta, 29,

302.Kremer, M. L. (1965). Tran8. Faraday Soc. 61, 1453.Margoliash, E., Novogrodsky, A. & Schejter, A. (1960).

Biochem. J. 74, 339.Nicholls, P. (1963). Biochim. biophy8. Acta, 81, 479.Nicholls, P. (1964). Tran8. FaradaySoc. 60,137.Nicholls, P. & Schonbaum, G. R. (1963a). In The Enzymes,

vol. 8, p. 206. Ed. by Boyer, P. D., Lardy, H. & Myrback,K. New York: Academic Press Inc.

Nicholls, P. & Schonbaum, G. R. (1963b). In The Enzymes,vol. 8, p. 215. Ed. by Boyer, P. D., Lardy, H. & Myrbiick,K. New York: Academic Press Inc.

Ogura, Y. (1955). Arch. Biochem. Biophys. 57, 288.Pople, J. A. & Segal, G. A. (1965). J. chem. Phys. 43, S 136.Samejima, T. (1959a). J. Biochem., Tokyo, 46, 155.Samejima, T. (1959b). J. Biochem., Tokyo, 46, 1101.Schfitte, H. R., Steinbrecht, J. & Winder, K. (1960). Hoppe-

Seyl. Z. 322, 142.Strother, G. K. & Ackerman, E. (1961). Biochim. biophys.

Acta, 47, 317.Yonetani, T. (1965). J. biol. Chem. 240, 4509.