One-electron Reduction Reactions with Enzymes in SolutionA Pulse Radiolysis StudyR. H. BISBY AND R. B. CUNDALLChemistry Department, University Park, Nottingham NG7 2RDJ. L. REDPATH? AND G. E. h A M S *Cancer Research Campaign Gray Laboratory, Mount Vernon Hospital,Middlesex HA6 2RNReceived 17th March, 1975At pH 8 and above, hydrated electrons react with ribonuclease, lysozyme and a-chymotrypsinto form transient products whose spectra resemble, but are not identical to, those for the RSSR-radical anion already known for simple disulphides. Assuming a value for the extinction coefficientsimilar to that for RSSR- in simple disulphides, only a fraction of the hydrated electrons are shownto react with the disulphide bridges : the remainder react at other sites in the protein molecule, suchas histidine, tyrosine and, in lysozyme, tryptophan residues, giving rise to comparatively weak opticalabsorptions between 300 and 4OOnm.This has been substantiated by studying the reaction ofe;, with subtilisin Nuuu (an enzyme which does not contain disulphide bridges), with enzymes in whichthe sulphur bridges have been oxidised and with some amino acid derivatives.On lowering the pH of the solution the intensity of the RSSR- absorption diminishes as the pro-tonated histidine residues become the favoured reaction sites. In acid solutions (PH 2-3) the tran-sient optical absorptions observed are due to reactions of hydrogen atoms with the aromatic aminoacids tyrosine, tryptophan and phenylalanine.The CO; radical anion is only observed to transfer an electron to disulphide groups in ribonucle-ase, although the effect of repeated pulsing shows that some reaction must occur elsewhere in theprotein molecule.In acid solutions, protonation of the electron adduct appears to produce theRSSRH. radical, whose spectrum has a maximum at 340 nm.One-electron reduction processes in non-metal containing enzymes can lead toinactivation, although with widely differing efficiencies. Such processes can be use-fully studied by radiation-chemical methods including pulse radiolysis. le3 Numerousstudies on the radiation chemistry of free amino acids and peptides in solution showthat both H atoms and hydrated electrons react rapidly with several classes of thesecompounds and the results have been used to predict possible sites of reductive attackin en~yrnes.~Pulse radiolysis studies have shown that hydrated electrons react rapidly withseveral enzymes and proteins to form a species with transient absorption spectra,' 9 2 9similar to those for the one electron adducts of simple disulphides.6*7 However,the proportion of the total elq yield which appears to be trapped at the disulphidelinkages in the enzymes varies considerably.The object of this study was to investigate the sites of, and mechanisms of, one-electron reduction processes by the hydrated electron and the formate radical anion,CO,, in solutions of four simple non-metal containing enzymes, lysozyme, ribo-nuclease (RNase), a-chymotrypsin and subtilisin NOVO (an enzyme which does notcontain any disulphide bridges).t present address : Department of Medical Physics, Michael Reese Hospital, Chicago, Illinois60616, U.S.A.552 ELECTRON REDUCTION OF ENZYMESEXPERIMENTALThe irradiation source was a linear accelerator which delivered 0.2 p s pulses of electronsof energy 1.8 MeV. The test solutions were irradiated in a quartz cell of path length 2 cmand transient changes in the optical absorption properties were recorded by fast spectro-photometry.For measurement of transient spectra, doses of 1-2krad were used. Dosi-metry was carried out using the thiocyanate dosimeter and small fluctuations in single-pulsedoses were normalised by charge collection methods. Full details of the pulse radiolysisequipment, associated circuitry * and the method of solution preparation have already beenpublished.Triply-distilled water was used throughout as diluent. All solutes were of thepurest grade available and were used as supplied. t-Butyl alcohol was purified by multiplerecrystallisation. N-acetylhistidine and N-glycyltyrosine were obtained from the SigmaChemical Co. Ltd.Ribonuclease (chromatographically purified), lysozyme (6 x crystake) and a-chymo-trypsin (3 x crystalline) were obtained from Miles-Serevac Ltd. Subtilisin No00 was a giftfrom Novo Industries A/S, Copenhagen.The method of Hirs l o was used for the performic acid oxidations. Both RNase andlysozyme oxidised by this method had less than 2 % of the activity of the native enzymestowards cytidine+2'-3'-cyclic phosphoric acid and the cell wall of Micrococcus Zysodekticus l2respectively.RESULTS(a) ONE-ELECTRON REDUCTION OF ENZYMES CONTAINING CYSTINE ATReactions of the enzymes with e; were followed by pulse radiolysis in N,-saturatedsolutions containing each enzyme and t-butyl alcohol as an OH scavenger.Thisalcohol reacts rapidly with OH radicals (k = 5.2 x lo8 dm3 mol-' s-l),13 but veryslowly with H atoms (k < lo5 dm3 mol-I s-l) l4 and hydrated electrons. Fig. l(a)shows the transient optical absorption spectra formed on pulse radiolysis of neutraldeoxygenated solutions of lysozyme, RNase and a-chymotrypsin each containinglo-' mol dm-3 t-butyl alcohol. The enzyme concentration, 4 mg ~ m - ~ for eachsolution, was suficiently high to ensure that all the hydrated electrons formed at thisdose per pulse were scavenged by the enzyme [fig.I@)].The relative absorptivity at 420 nm for the transient product obtained fromlysozyme is about three times greater than that from either RNase or a-chymotrypsin.The molar extinction coefficients at 420 nm for the electron adducts of most simpledisulphides lie in the range from 0.8-1.0~ lo4 dm3 mol-' cm-l,'' l5 although slightlylower values (5-8 x lo3 dm3 mol-' cm) have been reported for some cyclic RSSR-radicals.'. l6 Assuming a similar value of 9 x lo3 dm3 mol-I cm-1 for the extinctioncoefficient of RSSR- in the enzymes, the proportions of the initial yield of e i whichbecomes localised at the disulphide bridges in the three enzymes are : lysozyme 65,RNase 25 and a-chymotrypsin 22 %.This shows that electrons are trapped at othersites in the enzymes and that the proportions of electrons trapped at these sites variesfrom enzyme to enzyme.Fig. 2 shows the transient absorption spectrum of the electron adduct of thedisulphide compound, cystamine,6 normalised at the maximum to that of the lysozymeelectron adduct at pH 7. The spectra are clearly different in the wavelength regionbelow 400nm. One possible explanation of this is that the shorter wavelengthabsorption in the enzyme solutions is due to the product of electron attack at sitesother than disulphide bridges. In addition, reaction of hydrogen atoms with theenzymes may give rise to the absorption.NEUTRAL pR. H. BISBY, R.B. CUNDALL, J . L. REDPATH AND G. E. ADAMS 53The various possibilities were investigated by studying, as a function of pH, oneelectron reduction reactions of (i) samples of RNase and lysozyme, in which thedisulphide bridges had been oxidised by performic acid, (ii) the enzyme, subtilisinNovo, which contains neither cystine nor cysteine residues, and (iii) some simpleN-acylamino acids.0.06 -b -4 0 i .- 0.04-U a03 0 0 400 5 0 0 6 0 00.06\F[enzyme]/mg ml-'FIG. 1 . 4 ~ ) Transient spectra from dtaerated neutral solutions of some enzymes containing 10-Imol dm-j t-butyl alcohol ; A, lysozyme (4 mg ~ m - ~ ) , pH 7.0 ; 0, ribonuclease (4 mg ~ m - ~ ) , pH8.6 ; 0, a-chymotrypsin (4 mg ~ m - ~ ) , pH 6.8 ; dose = 2 krad/pulse, spectra measured 20 ps after thepulse.(b) Effect of enzyme concentration on the transient absorption at 410 nm following pulseradiolysis of neutral deaerated enzyme solutions containing lo-' mol dm-3 t-butyl alcohol ; A,lysozyme, pH 7.0 ; 0, ribonuclease, pH 8.6 ; 0, a-chymotrypsin, pH 6.8 ; dose = 2 krad/pulse,spectra measured 20 ps after the pulse.(b) REACTIONS OF REDUCING RADICALS WITH OXIDISED ENZYMESTreatment of RNase and lysozyme with performic acid lo oxidises the cystineresidues to sulphonic acids and the methionine residues to the corresponding sul-phoxides. Fig. 3(a) shows the transient absorption spectra obtained on pulseradiolysis of N2-saturated solutions of the oxidised enzymes (2 mg ~ m - ~ ) containing10-1 mol dnr3 t-butyl alcohol. For both enzymes, comparison with fig.l(a) showsthat oxidation of the disulphide linkages leads to a large reduction in the intensityof the 420 nm transient absorption maximum. In oxidised lysozyme, the spectrumcould not be measured below 420nm because of the high absorptivity of the un-irradiated solution. This was not so for the RNase derivative which shows a transientabsorption band with a peak at 350 nm and an extinction of about half that measuredat this wavelength for native RNase. This suggests that in the native enzyme, at leas54 ELECTRON REDUCTION OF ENZYMESI I I I I300 400 5 0 0 600AImFIG. 2.-Comparison of the electron-adduct spectra of lysozyme and cystamhe; 0, lysozyme4 mg ~ m - ~ , lo-' mol t-butyl alcohol, pH 7; dose = 2 krad/pulse, spectra measured 5 ps afterpulse ; 0, cystamine, data from ref.(13) (normalised at 410 nm).wavelength jnmFIG. 3.-Transient spectra from pulse radiolysis of deaerated enzyme solutions containing 10-lmol dm-3 t-butyl alcohol ; (a) 0, performic acid oxidised RNase, pH 8,2 krad/pulse ; 0, performicacid oxidised lysozyme, pH 8, 1 krad/pulse, enzyme concentration = 2 mg ~ m - ~ ; (b) subtilisin(4 @INovo (4 mg 0, pH 8.0 ; 0, pH 2.7, 1 krad/pulse.some of the absorption below 4OOnm is not associated with reduction of the di-sulphide bridges. The absorption in the oxidised enzyme is not due to attack at theoxidiscd methionine residues, since no transient absorptions were observed over thewavelength region 300-6OOnm on reaction of e 4 with methionine sulphoxide insolution.We conclude, therefore, that the absorption bands below 400m in thR . H . BISBY, R . B . CUNDALL, J . L. REDPATH AND G . E. ADAMS 55two native enzymes must be due to electron reaction with either the peptide linkagesor with reactive side groups in the non-sulphur-containing amino acids.(c) ONE-ELECTRON REDUCTION OF subtilisin NovoFig. 3(b) shows transient spectra obtained on pulse radiolysis of deaerated solutionsof subtilisin Novo (4 mg ~ m - ~ ) containing 10-1 mol dm-3 t-butyl alcohol at pH 8 and2.7. As expected, there is no absorption maximum at 420nm in neutral solution,but the absorption spectrum at shorter wavelengths is similar to that from oxidisedRNase. The rate constant for reaction of e; with subtilisin Novo at pH 7.3 wasobtained by measurement of the decay of the e,71 absorption at 600 nm in the presenceof lo-' mol dm-3 t-butyl alcohol and was found to be 1.8 x 1Olo dm3 mol-l s-l.The transient spectrum obtained from solutions at pH 2.7 is considerably moreintense than that found at pH 8.Further, at pH 2.7 the decay of the transient spec-trum can be resolved into two components. Fig. 4 shows the transient spectrum fromsubtilisin Novo at pH 2.7 measured immediately after the pulse and after 100 p s &lay.The latter spectrum decays relatively slowly over several milliseconds. The spectrumof the rapidly-decaying transient, which has a half life of -65 p s , obtained bydifference shows a maximum at 350 nm.wavenumber /nmFIG. 4.Transient spectra from pulse radiolysis of oxygen-free solutions of subtilisin Now (4 mg~ r n - ~ ) and lo-' mol dm-3 t-butyl alcohol at pH 2.7, 0, immediately after the pulse ; 0, after 100 ps ;A, difference spectrum : dose = 1 krad/pulse.(d) ONE-ELECTRON REDUCTION OF SOME N-ACYLAMINO ACIDSIn an attempt tobdentify the radical responsible for the 350 nm transient absorptionproduced in the enzyme systems, some experiments were carried out with two simpleamino acid derivatives.Information in the literature on one electron reduction of simple peptides is con-cerned mainly with aliphatic compounds and transient spectra are observed on pulseradiolysis of such systems 17* l 8 with maxima generally in the region 400-440 nm.These have been assigned to radicals produced by deamination.In large peptidesand enzyme proteins where only one such N-terminal amino group is present, radicalsof this type cannot be major reaction products. The peptide bond also has consider-able reactivity with c& in compounds such as triglycine. The radical products fromsuch reactions have the electron localiacd on the carbonyl group and only absorb inthe ultraviolet with maxima at 265 nrn.17* l56 ELECTRON REDUCTION OF ENZYMESFree aromatic amino acids are generally more reactive than aliphatic amino acidswith the hydrated electron and usually react to give species with transient absorp-tion maxima in the region around 350 nm,19-21 e.g., histidine, tyrosine and tryptophan.Since these spectra appear to be associated with reactions of e; with the aromaticrings, it would be anticipated that these amino acids, when present in peptide chains,would still react with e; to give similar transient radicals.Two simple aromaticamino acid derivatives were used to investigate this possibility.0.010-0.00sN-ACETYLHISTIDINEFig. 5 shows transient absorption spectra formed on pulse radiolysis of oxygen freesolutions of 5 x lo-( mol dm-3 N-acetylhistidine containing lo-' rnol d ~ n - ~ t-butylalcohol. The spectrum in neutral solution shows a maximum at 360m and issimilar to that formed by one-electron reaction with free histidine2* At higher pHvalues, the intensity of the absorption maximum decreases as is shown in the insetto fig. 5.?-I IOL 3b0 3 5 0 400 450wavelength/nmFIG.5.-Transient spectra from pulse radiolysis of oxygen free solutions of N-acetylhistidine (5 xmol dm-3) and lo-' mol dm-3 t-butyl alcohol, 0, pH 6.7; 0, pH 9.5; dose = 1 krad/pulse,spectra measured 5 p s after the pulse. Inset : Effect of pH on optical density at 360 nm.N-GLYCYLTYROSINEmoldm-3 of N-glycyltyrosine containing 10-1 mol dm-3 t-butyl alcohol are shown infig. 6, for neutral and alkaline pH values. At pH 7, the spectrum is again similar tothat for the one-electron reduction product of free tyrosine 20* 21 with a peak at340 nm. In alkaline solution, the peak is shifted to 380 nm.The transient product spectra obtained from oxygen-free solutions of 5 x(f) EFFECT OF pH ON TRANSIENT SPECTRA FROM LYSOZYMEAND RIBONUCLEASETransient spectra obtained from pulse radiolysis of RNase solutions containing10-1 mol dm-3 t-butyl alcohol are shown in fig.7(a). The spectra change appreciablywith pH. At pH 7 and above, the major absorption band is at 410 nm, but in acidsolution this maximum disappears and is replaced by a broad absorption in the ultra-violet. At pH 2, practically all hydrated electrons react with H30+ ions to form R. H. BISBY, R. B. CUNDALL, 3 . L. REDPATH AND G . E. ADAMS 570.0 I5 c1 1 1 130 0 3 5 0 400 450 500A/nmFIG. 6.-Tramient spectra from pulse radiolysis of oxygen-free solutions of N-glycyltyrosine (5 xmol dm-3) containing lo-’ rnol dm-3 t-butyl alcohol measured 5 ps after 2 krad/pulse, 0, pH 7 ;0, pH 11.I I I3 0 0 4 0 0 5 00 6 0 0Lwavelength/nmFIG.7.-(a) Transient spectra from pulse radiolysis of oxygen-free solutions of ribonuclease (4 mg~ r n - ~ ) containing 10-’- mol dm-3 t-butyl alcohol, 0, pH 8.5 ; 0, pH 7.0 ; A, pH 6.4 ; V, pH 2.2 ;dose = 2 krad/pulse, spectra measured 5 ps after the pulse. Inset : Effect of pH on optical density at410 nm. (6) Transient spectra from pulse radiolysis of oxygen-free solutions of lysozyme (4 mg ~ m - ~ )containing lo-‘ mol dm-j t-butyl alcohol, 0, pH 7.0; 0, pH 5.0; A, pH 3.7; A, pH 3.0; 0,pH 2.0 ; dose = 2 krad/pulse, spectra measured 5 ,as after the pulse . Inset : Effect of pH on opticaldensity at 410 nm58 ELECTRON REDUCTION OF ENZYMESatoms; the spectrum obtained at this pH, therefore, is attributed to reactions ofH atoms with the enzyme.Near neutral pH, where the H30+ concentration is verylow, the hydrated electrons will react directly with the enzyme. Nevertheless, thereis a clear pH effect in this region : a decrease in pH from 8.6 to 6.4 is sufficient toremove a large proportion of the absorption peak at 410 nm. The observed pH effectstrongly suggests that at the lower pH value reactions of e& at sites other than cystineresidues are more favourable.Fig. 7(b) shows data obtained from similar experiments with lysozyme.(g) ONE-ELECTRON REDUCTION OF RIBONUCLEASE BY FORMATE RADICALSIn N,O-saturated solutions containing formate ions, the CO; radical anion isformed as the sole radical product :N20+ea; + N2+OH-+OH* (1)OH+HCO, -+ H,O+CO, (2)H+HCO, 4 H,+CO,. (3)The rate of electron transfer from CO; to simple disulphides is a factor of lo3 greaterthan that for reaction with simple amino acids and peptides that do not containsulphur bridges.mold ~ n - ~ sodium formate at pH 8 are pulse-irradiated, an absorption at 410 nm isobserved which increases in intensity with repeated pulsing [fig.S(u)]. The numberWhen N,O-saturated solutions of RNase (I mg ~ r n - ~ ) containing 4 x0.03(a) 0.02300 0.0 Ito 2 0 30 4 0 50 6 0 70w -0-.- s +1650 aO.Oi5--*-*---;00/+O-O--o 300(6)I I 8 t10 2 0 30 40number of pulsesFIG. 8.-Formation of a transient absorption at 410111x1 with repeated pulse irradiation of NeOsaturated solutions of ribonuclease and 4 x 1C2 mol dm-' sodium formate at pH 8 ; (0) 1 mgribonuclease. Doses (rad/pulse) given against the curves.ribonuclease ; (6) 0.5 mgof pulses required to reach a maximum transient absorption intensity at 410 nm varieswith pulse size and at small doses per pulse no sipificant absorption appears untilthe solution has reccivcd some tens of pulses. ThC enzyme concentration also aircctR. H. BISBY, R . B . CUNDALL, J . L . REDPATH AND G . E. ADAMS 59the total dose required before the maximum transient intensity is observed as shownin fig. 8(b). The effect of temperature on the formation of the 410 nm absorption dueto reaction of COY with RNase (1 mg CM-~) is shown in fig. 9 : there is a markedI I I I I I I I I I3 0 4 0 5 0 6 0 7 (temperature/"CFIG. 9.-Effmt of temperature on the transient 410 nm absorption formed by pulse radiolysis of NzOsaturated solutions of ribonuclease and 5 x mol sodium formate at pH 5.9, dose = 2 had/pulse, 0, after 1 pulse ; A, after 10 pulses ; 0, after 30 pulses.change in behaviour above 55°C.Transient spectra obtained from reaction of CO;with RNase are shown in fig. 10 at various pH values. These spectra were obtainedfrom solutions which had previously been irradiated until a maximal absorption wasobserved. At pH 7.5 the observed transient spectrum contains a single peak at 410nm, but at lower pH this is replaced by another transient absorption with a peak at340-350 nm. The pH effect on the peak intensities at 410 and 350 nm is shown inthe inset to fig. 10.PH I - -. .A.wavelength/nmFIG. lO.-Transient spectra from pulse radiolysis of NzO saturated solutions of ribonucleast (1 mg~ r n - ~ ) and 4 x mol dm-3 sodium formate having previously received between 10 and 40 pulses,dose = 2 krad/pulse, spectra measured 35 ps after the pulse, 0, pH 7.5 ; 0, pH 5.9 ; A, p H 4.6 ; 'I,pH 3.9 ; v , pH 3.1.Inset : Effect of pH on the optical densities of the transients at ; 0, 410 nm ;0,350 nm.Repeated pulse irradiation of N,O-saturated solutions of subtilisin Novo (3 mgmol dm-3 sodium formate did not reveal any transient containing 4 xabsorptions in either neutral or acid solutions60 ELECTRON REDUCTION OF ENZYMESDISCUSSIONThe spectral evidence shows that hydrated electrons can react with differentresidues in the enzymes. The effects of pH on the transient absorption spectraindicate which amino acid residues are involved.In addition to reaction with di-sulphide groups, the absorptions at approximately 350 nm in the native enzymes,oxidised proteins and amino acid derivatives show that reaction can also occur withthe imidazole, phenolic or indole groups of histidine, tyrosine or tryptophan, becausesimilar absorptions are observed following reaction of e; with the free amino acids.In RNase solutions, the absorption at 410 nm decreases significantly on changingthe pH from 8.6 to 6.4 [fig. 7(a)]. This change occurs over the pH region where theimidazole rings of histidine residues in RNase undergo protonation (pK values5.1-6.4).22 The reactivity of e; towards protonated histidine (k = 7 x lo9 dm3 mol-1s-l) is considerably greater than that of unprotonated histidine (k = 6 x lo7 dm3mol-1 s-l) and is comparable with that of cystine (k = 1.3 x 1Olo dm3 mol-' s-I atpH 6.1).23 It follows that as the histidine residues in RNase are protonated theybecome much more favourable reaction sites for the hydrated electron.The effectof pH on the electron-adduct spectrum of RNase is compatible with such a changein reaction site : the intense absorption of the RSSR- radical at 410 nm is replacedby the absorption at 360 nm of the histidine radical which has a rather lower extinctioncoefficient ( E ~ ~ ~ = 1200 dm3 mot1 cm-l). An alternative explanation of the de-crease of the 410 nm absorption with decreasing pH which has to be considered isthat the RSSR- radical is protonated at the lower pH;RSSR-+ H+ + RSSRH.-+ RS-+ RSH.For glutathione, protonation of the RSSR- radical occurs with an apparent pK0f5.2.~~The intermediate RSSRH. radical is usually short-lived in simple disulphides andthe observed product is the RS- radical, which typically absorbs at 330-350 nm witha very low extinction coefficient (E,,, = 300-600 dm3 mol-' ern-').'' 24 The RSSRHspectrum has been identified in the pulse radiolysis of lipoic acid where dissociationinto RS- and RSH is hindered presumably because of the ring structure. ThisRSSRH radical absorbs at 385 nm with an extinction coefficient of 6.9 x lo3 mol-1cm-l.'This value is not much smaller than the extinction coefficient of RSSR- of lipoicacid ( E ~ ~ ~ = 9.2 x lo3 dm3 mol-I cm-' at pH 7.8).' By contrast, the overall intensityof the electron adduct spectrum of RNase at pH 6.4 is about half that at pH 8.6, andappears to be a composite spectrum.The RSSR- spectrum from RNase is notreplaced by another single absorption at shorter wavelengths with a similar extinctioncoefficient which would be expected if RSSRH were to be produced. This effect iseven more apparent with lysozyme [fig. 7(b)].The spectrum obtained from RNase at pH 2.2 will be due to products formed byreaction of H atoms produced by the rapid protonation of e i . The reactivities ofhydrogen atoms with amino acids at acid pH are known 25 and can be comparedwith hydrated electron reactivities. Whereas the rates of reaction of e; with tyrosineand phenylalanine are a hundred fold lower than that with cystine, H atoms reactwith tyrosine (k = 1.9 x lo9 dm3 mol-' s-l) and phenylalanine (k = 8 x lo8 dm3mol-l s-l) with rate constants which are only a factor of 10 or less lower than thatwith cystjne (k = 8 x lo9 dm3 mol-' s - ' ) .~ ~ Therefore, the reactivity of tyrosineand phenylalanine residues in RNase would be much higher with H atoms than withe.9 and their corresponding adduct spectra would be expected to be much moreevident in acid than at neutral pH. Fig. 7 shows this to be the case. At pH2.2there is a relatively intense spectrum with maxima at 350 and 320 nm. The absorption(4R. H . BISBY, R. B. CUNDALL, J . L . REDPATH AND G . E. ADAMS 61at 350nm is similar to that of the H atom adducts of tyrosine.21 We assign themaximum at 320 nm to the H-atom adduct of phenylalanine since a similar band hasbeen observed following reaction of H atoms with free phenylalanine at neutral andacid pH,l8* 26 At pH 2.2, any reaction of H atoms with cystine would give rise toRS- radicals which absorb very weakly at 330-350 nm.The similarity between theH atom adduct spectra of RNase at pH 2.2 and subtilisin Novo at pH 2.7 confirmsthat any contribution to the spectrum from RNase at pH 2.2 from sulphur radicalsis negligible.The spectrum produced by one-electron reduction of lysozyme [fig. 7(b)] can beinterpreted in much the same way as those of RNase. The major differences arethat in lysozyme a larger fraction of the e i yield reacts to form RSSR- and that thecurve for optical density at 410 nm as a function of pH is displaced to lower pH byabout 1.5 units relative to the corresponding curve for RNase.This is due to differ-ences in the relative reactivities of the two residues (cystine and histidine) in the twoenzymes as shown by the larger yield of RSSR- in lysozyme above pH 8. Anotherdifference is that, unlike RNase, lysozyme contains tryptophan, an amino acid whichis very reactive towards H atoms (k 2 2.3 x lo9 dm3 mol-' s - ' ) . ~ ~ The hydrogenatom adduct spectrum of tryptophan contains a peak at 310-320 nm l9 which, invicw of the high rate constant, must contribute to the H atom adduct spectrum oflysozyme [fig. 7(b), pH 21 at these wavelengths.The assignment of some absorptions in the product spectra formed by one-electronreduction of RNase and lysozyme to non-sulphur containing radicals is confirmed bythe results obtained with subtiZisin Navo.Despite the lack of cystine or cysteineresidues in this enzyme, the hydrated electron reacts with it at a rate (k = 1.8 x 10"dm3 mol-' s-l) which is similar to that for ribonuclease (k = 1.0 x 1 O ' O dm3 mol-Is-' at pH 7.1). At pH 8, the electron adduct spectrum of subtihin No00 has a broadshoulder at 350 nm which may be due to some combination of histidine, tyrosine ortryptophan radicals in addition to a peak at 310 nm which also appears in the electronadduct spectrum of tryptophan.lg. 2o The hydrogen atom spectrum of subtilisinNovo is similar to those for RNase and lysozyme, consisting of absorptions between300 and 400 nm due to H atom attack on the ring-containing amino acids.The fasterdecaying part of the two component spectrum which has a half life of - 65 p s containsa peak at 350 nm which is similar to the spectrum of the hydrogen atom adduct oftyrosine. 21The spectra following reaction of RNase with the formate radical CO, differsignificantly from those formed by reaction with e; and H atoms. After suitablepre-irradiation of the solution, the radical anion CO; reacts with RNase at pH 7.5to give a single transient with a maximum at 420 nm which is clearly the spectrum ofthe RSSR- radical anion. At low pH the spectrum is changed to one with a maxi-mum at 340 m. Repeated pulse irradiation of subtilisin Novo solutions containingformate does not give rise to a 340 nm transient in acidic solutions.Therefore, the340nm band in the RNase solutions must be due to some type of sulphur radical.The extinction of the 340 nm band appears slightly larger than of the 420 nm bandin neutral solution, so it is unlikely to be due to RS. radicals which generally haveextinction coefficients at least an order of magnitude less than those of the correspond-ing RSSR- radical anions. However, the RSSRH. radical of lipoic acid has anextinction coefficient similar to that of its radical anion. We conclude that the 340 nmband observed from reaction of CO; with RNase is due to an RSSRH- radical formedby rapid protonation of RSSR- :H +v.fastCOY +RSSR --+ RSSR' + RSSRH. ( 5 62 ELECTRON REDUCTION OF ENZYMESAt pH 3.1 any contribution from C02H will be insignificant ; this radical has a pKof only 1.4.27 Whereas reaction of COY with RNase gives rise to an absorption at340 nm in acid solutions, the corresponding absorption in lipoic acid, which has beenassigned to the RSSRH., is at 385nm.The difference may arise because the 5-membered ring of lipoic acid may be subject to strain and cause a change in theabsorption spectrum.The spectra obtained by one-electron reduction of RNase by CO, and e; inslightly acid solutions are strikingly different. The band containing the prominentpeak at 340-350nm obtained by reduction with CO, at pH3.9 and assigned toRSSRH is different from the less intense and less defined absorption produced byreduction of RNase by eLq at pH 6.4.This is in agreement with our interpretationof the effect of pH on the spectrum formed from e i attack on this enzyme. Thespectral changes are not due to protonation of RSSR- to give RSSRH, but resultfrom a change in the reaction site from cystine residues to the histidine residues asthe pH falls below the pK of the protonated imidazole ring of histidine.The reason why RNase solutions require pre-irradiation before CO; is able totransfer an electron to the disulphide bridges of the enzyme is not established. Pre-sumably, COT reacts in the first instance with a residue we have so far been unableto identify. This primary reaction causes conformational changes of the enzymewhich make the disulphide bridges more accessible to the CO; radicals which onlythen can react to form RSSR-.This is supported by the effect of temperature on themanner in which RSSR- is formed as the solutions are repeatedly pulse irradiated(fig. 9). The 410 nm absorption of RSSR- is more easily formed above 55-60°C, thetemperature range over which RNase is known to undergo conformational transi-tions.2This work was supported by grants from the Cancer Research Campaign.G. E. Adarns, J. L. Redpath, R. H. Bisby and R. B. Cundall, IsraelJ. Chem., 1972,10,1079.T. Masuda, J. Ovadia and L. I. Grossweiner, Int. J. Rudiation Biol., 1971, 20,447.R. Braams and M. Ebert, Int. J. Radiation Biol., 1967,13,195 ; N. N. Lichtin, J. Ogdan andG. Stein, Biochim. Biophys. Acta, 1972,263,14 ; N.N. Lichtin, J. Ogdan and G. Stein, Biochint.Biophys. Acta, 1972,276, 124.R. Braams, Radiation Res., 1967, 31, 8.G. E. Adams, R. B. Cundall and R. L. Willson, in Ctlernical Reactioity and Biological Role ofFunctional Groups in Enzymes, ed. R. M. S. Smellie (Academic Press, London and New York,1970), p. 171 ; J. R. Clements, D. A. Armstrong, N. V. Klassen and H. A. Gillis, Canad. J.Chem., 1972,50,2833 ; N. V. Klassen, J. W. Purdie, K. R. Lynn and M. D'Iorio, Int. J. Radia-tion Biol., 1974, 26, 127; G. E. Adams, K. F. Baverstock, R. B. Cundall and J. L. Redpath,Radiation Res., 1973,54,375 ; P. B. Roberts, Int. J. Radiation Biol., 1973,24,143 ; G. E. Adams,R. L. Willson, J. E. Aldrich and R. B. Cundall, Int. J. Radiation Biol., 1969, 333.ti G.E. Adams, G. S. McNaughton and B. D. Michael, in The Chemistry ofIonization and Excita-tion, ed. G. R. A. Johnson and G. Scholes (Taylor and Francis, London, 1967), p. 281. ' M. Z. Hoffman and E. Hayon, J. Amer. Chern. SOC., 1972,94,7950.G. E. Adams, J. W. Boag and B. D. Michael, Trans. Faraday SOC., 1965, 17,349.R. L. Willson, Int. J. Radiation Biol., 1970, 17, 349.New York, 1967), vol. XI, p. 51.lo C. H. W. Hirs, in Methods in Enzymology, ed. C. H. W. Hirs (Academic Press, London andl 1 E. M. Crook, A. P. Mathias and B. R. Rabin, Biochem. J., 1960,74,234.l2 D. Shugar, Biochem. Biophys. Acta, 1952, 8, 302.l3 L. M. Dorfman and G. E. Adams, Reactivity of the HydroxyI Radical in Aqueous Solutions,National Standard Reference Data System, NSRDS-NBS-46 (National Bureau of Standards,Washington, U.S.A., 1973).l4 M. Anbar and P. Neta, Int. J. Appl. Radiation Isotopes, 1967, 18, 493.l 5 G. E. Adams, R. C. Armstrong, A. Charlesby, €3. D. Michael and R. J. Willson, Trans. FuraduySOC., 1969, 65, 732R . H. BISBY, R . B . CUNDALL, J . L. REDPATH AND G. E. ADAMS 63l6 J. L. Redpath, Radiation Res., 1973,64,364 ; P. C. Chan and B. H. J. Bielski, J. Amer. Chem.l7 M. Simic and E. Hayon, Radiation Res., 1971, 48, 244.l9 R. C. Armstrong and A. J. Swallow, Radiation Res., 1969, 40, 563.2o M. Farragi and I. Pecht, Israel J. Chem., 1972, 10, 1021.21 J. Feitelson and E. Hayon, J. Phys. Chem., 1973, 77, 10.22 J. H. Bradbury and H. A. Scheraga, J. Amer. Chem. Soc., 1966,88,4240 ; G. C. K. Roberts,23 R. Braams, Radiation Res., 1966, 27,319.24 A. Shafferman, Israel J. Chem., 1972,10,725.25 P. Neta and R. H. Schuler, Radiation Res., 1971, 47, 612.26 N. N. Lichtin and R. Shderman, Radiation Res., 1973, 60,432.27 G. V. Buxton and R. M. Sellers, J.C.S. Favoday I, 1973,69,555.28 J. F. Brandts and L. Hunt, J. Amer. Chem. SOC., 1967,89,4826; J . Hermans and H. A. Scheraga,SOC., 1973,95,5504.J. P. Mittal and E. Hayon, J. Phys. Chem., 1974,78,1790.D. H. Meadows and 0. Jardetsky, Biochemistry, 1969, 8, 2053.J. Amer. Chem. SOC., 1963, 85, 3866; W. A. Klee, Biochemistry, 1967, 6, 3736