May 10, 1982 - Horseradish peroxidase reacts with sodium. [36C1] chlorite at pH 10.7 to form a 36C1-labeled horseradish peroxidase intermediate.
THEJOURNALOF BIOLOGICAL CHEMISTRY Vol. 257, No. 19. Issue of October 10, pp. 11529-11533, 1982 Printed in U.S.A.
Isolation and Characterizationof Horseradish Peroxidase Compound X* (Received for publication, May 10, 1982)
Shahram ShahangianS and LowellP. Hager From the Roger Adams Laboratory, Department of Biochemistry, University of Illinois, Urbana, Illinois 61801
Horseradishperoxidasereactswithsodium [36C1] chlorite at pH 10.7 to form a 36C1-labeledhorseradish peroxidase intermediate. The optical absorption spectrum of this intermediate is quite stable and very similar to that of horseradish peroxidase Compound 11. The intermediate can be separated from small molecules by chromatography on a SephadexG-10 column. After fractionation, 65 to 93% of 36Cl in the reaction mixture remains associated with horseradish peroxidase. The remainder of 36Cl can be accounted for as [36Cl]chloride ion. "he 36C1-labeled enzyme reacts with 5,5-dimethyl-2-chloro-1,3-cyclohexanedione(monochlorodimedone) at pH 4 to transfer 36Cl from theenzyme to thehalogenacceptormolecule. [36C1]5,5-Dimethyl-2,2-dichloro-1,3-cyclohexanedione(dichlorodimedone) was established as the major product of the transfer reaction by co-crystallization of the enzymic product with authentic dichlorodimedone and by thin a layer chromatography. A chlorine oxide ligand on ferry1 heme iron protoporphyrin IX is proposed for the structure of CompoundX.
chemical reaction to form dichlorodimedone, it was argued that the mechanism of the chlorite reaction could be envisioned as a two-step reaction sequence. In the first step, chlorine dioxide would be produced from chlorite in an enzyme-catalyzed reaction. The second step would then involve the chemical reaction between monochlorodimedone and the enzymically generated chlorine dioxide. However, recent kinetic and substratespecificity studies fail to support this twostep model.2 In this paper, we show that horseradish peroxidase reacts with sodium chlorite at high pH to form a fairly stable Compound X intermediate which contains 1 chlorine atom/enzyme molecule, and that Compound X reacts with monochlorodimedone in the pH range 4 to 7 to form dichlorodimedone. EXPERIMENTALPROCEDURES
Horseradish peroxidase was purchased as a crude salt-free powder (type 11) from Sigma and as a purified ammonium sulfate suspension (Grade I) from Boehringer Mannheim Biochemicals. The RZ (A,O,/ of the commercial samples were 1.66 and 3.12, respecA ~ wvalues ) tively. Since pure isozyme C has anRZ value of about 3.4 (8, 9), both samples were further purified essentially according to Shannon et al. (9). Chromatography was performed on Whatman carboxymethylcelHorseradish peroxidase (EC 1.11.1.7, donor:hydrogen per- lulose CM52 columns. The purified horseradish peroxidase preparaoxide oxidoreductase) is a monomeric glycohemoprotein of tions having RZ values of 3.4 and higher were dialyzed against 3.2 M ammonium sulfateandstored at 4-6 "C. Horseradish peroxidase 40,000 molecular weight and contains ferriprotoporphyrin 1X concentrations were determined spectrophotometrically at 403 nm as its hemeprosthetic group (1, 2). Although horseradish using an absorption coefficient of 102 mM" cm" (10, 11). peroxidase catalyzes the oxidation of iodide ion, it is not able Sodium [S6Cl]chlorite wasprepared as before (4, 12). ["6Cl]Chlorite to oxidize either bromide or chloride ions using hydrogen was fractionated on a Dowex 2 column (1.1 X 50 cm) in order to peroxide as oxidant (3). However, it has been shown that separate itfrom contaminating ["Cl]chloride ion. The Dowex column horseradish peroxidase is capable of catalyzing chlorination yielded twoslightly overlapping radioactivity peaks when eluted with reactions when chlorite serves both as thehalogen donor and a 0 to 1 M sodium nitrate gradient. The two peaks were identified by iodimetry and mercuric thiocyanate assays as ["'Cl]chlorite and [""Cl] oxidizing agent (4). chloride, respectively, in the order of their elution (12). Before chroIn theabsence of a halogen acceptor, horseradishperoxidase matography, the column was eluted with 5 M sodium nitrate followed initially reacts with chlorite to form an intermediate having by an extensive water wash in order to remove all chloride from the an optical absorption spectrum very similar to that of horse- resin. The flow rate used to elute chlorite and chloride from the radish peroxidase Compound I1 (5).This intermediate, termed column was fast (about 25 ml/h) since an initial trial run at pH 7 had Compound X, was proposed asthe enzymic halogenating shown that chlorite could react with the column material. The pooled fractions under the ["Cl]chlorite peak had a specific activity of54 intermediate in the chlorite reaction (6, 7 ) . Subsequent work mCi/mol for chlorite. This value was in close agreement with the cast doubt on the validity of this hypothesis when it was theoretical value of 52 mCi/mol for chlorine, showing that the ["Cl] shown that horseradish peroxidase catalyzes the dismutation chlorite thus purified was essentially free of other chlorine-containing of chlorite to chloride ion and chlorine dioxide (5). Since compounds. All optical spectrophotometric work was performed using a Cary chlorine dioxide can chlorinate monochlorodimedone' in a 219 spectrophotometer. A Beckman LS 8000 scintillation counter was * This work was supported by Grant GM-07768 fromthe National usedfor radioactivity measurements. Radioactivity in the column Institutes of Health and Grant PCM 79-10656 from the National fractions was determined by counting 20 to 50-pl aliquots of the Science Foundation. The costs of publication of this article were radioactive material. The aliquots were added to 5 ml of scintillation defrayed in part by the payment of page charges. This article must fluid prepared according to Patterson and Greene (13). All pH meastherefore be hereby marked "advertisement" in accordance with 18 urements were carried out with a Metrohm-Brinkmann 103pH meter. U.S.C. Section 1734 solely to indicate this fact. Hydrogen [S6Cl]chloride(205 mCi/mol) was purchased from New $ Present address, Clinical Chemistry Laboratory, Division of Clin- England Nuclear and was neutralized with sodium hydroxide prior to ical Pathology, University of Utah Medical Center, Salt Lake City, use. Monochlorodimedone and dichlorodimedone were prepared as UT 84132. previously described (14). Chloride concentrations were determined ' The trivial names used are: monochlorodimedone, 5,5-dimethyl2-chloro-1,3-cyclohexanedione;dichlorodimedone, 5,5-dimethyl-2,2dichloro-l,3-cyclohexanedione. 'S. Shahangian and L. P. Hager, manuscript in preparation.
11529
Compound X Isolation
11530
colorimetrically using the mercuric thiocyanate method as described by Vogel (15). An iodimetric method was used for the determination of chlorite concentrations (16). The assay buffer was 0.1 M sodium phosphat.e, pH 2.2, containing 0.125 M potassium iodide and 10 p~ ammonium molybdate as thecatalyst. Triiodide concentrations were determined spectrophotometrically at 353 nm using an absorption coefficient of 25.5 mM" cm" (17). Ascending thin layer chromatography was performed with Eastman Kodak 6060 silica gel plates containing fluorescent indicators. Dowex 2-X8 was from Baker Chemical Co. Sephadex G-10 was supplied by Pharmacia Fine Chemicals. It was allowed to swell in 4.5 mM sodium carbonate, pH 10.7, before use. Whatman CM52 cellulose was supplied by Daigger Scientific Co. The cellulose dialysis tubings used were Spectrapor 4 having a molecular weight cutoff of 12,000 to 14,000. The dialysis membranes contained a chloride impurity which was removed by washing with distilled water. Glass-distilled water was redistilled from acidic potassium permanganate and then was redistilled again before use. All solutions were made in this tripledistilled water using the best quality reagents grade chemicals commercially available.
20
25
30
Fmcllon Number
FIG. 2. Isolation of [36C1)CompoundX by gel filtration atlow enzyme concentration. Horseradish peroxidase (80 nmol) was reacted with 80 nmol of sodium [36Cl]chlorite(0.0043pCi) in 1 ml of 4.5 RESULTS mM sodium carbonate, pH 10.7. Gel chromatography was carried out according to the methods described under "Experimental ProceFormation of Compound X-Compound X can be readily dures." The curves plot the elution profile for "C1 radioactivity formed by reacting horseradish peroxidase with equimolar (W) and absorbance at 420 nm (M). The column dimenamounts of sodium chlorite in 4.5 mM sodium carbonate, pH sions were 1.5 X 15 cm and the flow rate was 1 to 1.5 ml/min. The 10.7. Based on absorbance changes in the visible region, the average fraction size was 0.56 ml.
formation of a horseradish peroxidase "Compound I1 type" spectrum was essentially complete within 15 to 50 min. Fig. 1 shows the visible absorption spectrum of Compound X generated with 186 p~ horseradish peroxidase. Only the visible region of the optical spectrum could be scannedwithout dilution at this enzyme concentration. Isolation of [36C1]Compound X by Gel Filtration-When Compound X was formed in a reaction of horseradish peroxidasewith ["2Cl]chlorite, "'Cl-labeled enzyme could be isolated. After the completion of the reactionof the enzyme with ["6CCl]chlorite, as judged from the formationof the Compound I1 type spectrum, the reaction mixturewas fractionated on a Sephadex G-10 column. Fig. 2 shows the Sephadex column elution profiie of horseradish peroxidase and radioactivity from a reaction mixture containing80 p~ horseradish peroxidase. In this experiment, 93% of the "'Cl label co-migrated Froction Number with horseradish peroxidase and the remainder of the label was in the small molecule peak. When the concentration of FIG.3. Isolation of [36C1]CompoundX by gel filtration at horseradish peroxidase in the reaction mixturewas increased high enzyme concentration. Horseradish peroxidase (220 nmol) was reacted with 220 nmol of sodium ["'CI]chlorite (0.012 pCi) in 1 ml of 4.5 m m sodium carbonate, pH 10.7. The gel fdtration conditions were the same as those described for Fig. 2. The curves plot the elution profile for "C1 radioactivity (M), absorbance at 420 nm (M), and chloride ion (A-A). Chloride concentrations were determined as described under "Experimental Procedures."
0.4
i
$50
I
I
500
550
I 600
I
650
U
700
NANOMETERS
FIG. 1. Formation of Compound X.The visible absorption specand tra of 186 PM horseradish peroxidase, in the absence (-) presence (- -) of 186 sodium chlorite, were recorded as a function of time. The reaction was carried out in 4.5 mM sodium carbonate, pH 10.7, and the scan rate was 2 nm/s. The spectra were recorded every 4.2 min except the first (- - -) and the last (.-. ) spectral traces which were recorded immediately and 32 min after the start of the reaction, respectively.
-
2.75-fold, the percentage of"C1 label associated with horseradish peroxidase decreased to65% as shown in Fig. 3. "Cl Transfer from [36CC1]Compound X to Monochlorodimedone-When pooled samples of Compound X which had been isolated bychromatography on the Sephadex gel column shownin Fig. 3 were reacted with a saturated solution of monochlorodimedonein 0.33 M sodium acetate,pH 4.0, a white precipitate was formed.Since dichlorodimedone is sparingly soluble at pH 4 compared with monochlorodimedone, this resultsuggested that chlorine was being transferred from Compound X to the halogen acceptor substrate. This result was confirmed by extracting the productof the reaction with 3 equal volumes of chloroform at pH 7. The combined chloroform extracts were then driedovermagnesiumsulfate, filtered, andcounted.Theaqueousphase of the reaction mixture was also counted. As line 1 of Table I indicates, there was significant"'Cl transfer from Compound X to the acceptor substrate.
Compound X Isolation Identification of Dichlorodimedone U S the Product of the Transfer Reaction-The small amount of white precipitate formed in the reaction of Compound X with monochlorodimedone was mixed with unlabeled authentic dichlorodimedone, taken up ina minimal amount of chloroform, and crystallized from the same solvent three times. The results summarized in Table I1 show a fairly constant ratio of3"Cl disintegrations/min/mg of solid upon successive crystallizations. An aliquot of the chloroform extract was also applied to a strip of silica gel plate (10 x 10 cm) and developed with 1butanol. Before developing, authentic dichlorodimedone was also applied to the plate. The radioactive material and the authentic dichlorodimedone co-migrated on the silica gel plate as evidenced by radioactivity and UV absorption, respectively. Identification of Chloride Ion as a Product of the Reaction of HorseradishPeroxidase with Chlorite-As previously noted, Fig. 3 demonstrates that as much as one-third of the "C1 in the reaction mixtureelutes in the small molecule fractions of gel filtration columns when Compound X is chromatographed at pH 10.7. When these small molecule column fractions were assayed for chloride content by the mercuric thiocyanate method (12), a direct relationship between chloride ion concentration and 36Clcontent was found (see Fig. 3). Assuming a specific activity of 54 mCi/mol of chlorite, the ratio of moles of chloride determined chemicaUy/mol of"C1 ranged from 0.9 to 1.2 with an average of 1.05. Stability of Compound X ut -20 "C-Compound X was f i s t formed by the reaction of 160 nmol of horseradish with 320 nmol of sodium [36Cl]chlorite in 2 ml of 50% ethylene glycol solution at pH10.7. Under theseconditions, Compound X formation was maximal in about 35 min. The reaction
IO0
11531
,
II:\
1
20/
1
1
4
8
I
I2
I
I
I
16
20
24
DAY' OF DIALYSIS
FIG. 4. Decomposition of Compound X at -20 "C. rC1ICompound X was formed by the reaction of160 nmol of horseradish peroxidase with 320 nmol of sodium [""Cllchlorite (0.017 FCi) in 2 ml of 50%ethylene glycol solution containing 2.2mM sodium carbonate, pH 10.7. When Compound X formation was maximal, the reaction mixture was dialyzed a t -20 "C against a large excess of the same buffer. The curvesplot the percentage of horseradishperoxidase and )the percentage of horsecontaining nondialyzable '36Cc1(tradish peroxidase retaining a Compound I1 type optical absorption spectrum (m as) a function of time. The percentage of enzyme inside the dialysis bag as 36CC1-labeledhorseradish peroxidase was calculated by assuming 1 ["Cl]chlorine atom/enzyme molecule. The percentage of residual Compound I1 type spectrum was calculated from the decrease inabsorbance a t 428 nm. The absorption coefficient of Compound 11 a t 428 nm decreases by 46 mM-' cm" when Compound I1 is converted to the nativeenzyme (IO).
mixture was then dialyzed at -20 "C against an excess of 2.2 mM sodium carbonate buffer containing 50% ethylene glycol, pH 10.7.Fig. 4 plots the percentage of the enzyme as "ClTABLEI labeled horseradish peroxidase inside the dialysis bag and the Transfer of "'Cl from untreated a n d treated Compound X percentage of remaining Compound I1 type spectrum as a preparations to monochlorodimedone function of time. A control experiment was run in which the The peak fractions of Compound X (fractions 14 and 15) obtained same amount of sodium [36Cl]chlorite was dialyzed in the in the experiment described in Fig. 3 were pooled and then reacted absence of horseradish peroxidase. The control experiment with 1 ml of a saturated solution of monochlorodimedone in 0.33 M had showed that after3h essentially all (>95%) of ""1 sodium acetate buffer, pH 4 (line 1). In line 2, the same incubation dialyzed out of the bag. was carried out with the enzyme fractions which were recovered by Stability of Compound X a t Room Temperature-After gel filtration after Compound X had been decomposed by incubation a t room temperature for 36 h (see Fig. 5). In line 3, fraction 15 Compound X had been incubated for 36 h at room temperaobtained from the gel chromatography of the acid-treatedCompound ture, the visible absorption spectrum of the enzyme preparaX (see Fig. 6) was reacted with 0.5 ml of saturating monochlorodi- tion was essentially that of native horseradishperoxidase. Fig. medone at pH 4. After the reactions of the enzyme fractions with 5 shows the Sephadex G-10 "'Cl elution profile recorded for monochlorodimedone, the reaction mixtures were successively extracted with 3 equal volumes of chloroform in order to isolate [36Cl] the enzyme preparationafter the 36-h room temperature incubation. Although most of the radioactivity is now found dichlorodimedone. in the small molecule fractions, a significant amount of radioRadioactivity activity remains associated with the enzyme. The pooled ""C1Compound X Enzyme-=['6CC1]Chlo- [:'"Cl]Dichlolabeled horseradish peroxidase fractions were then reacted sociated :IhC1 ride rodlrnedone with a saturated solution of monochlorodimedone and ex&m tracted with chloroform as before. Line 2 of Table I shows 1. Untreated 7500 1200 9200 that no significant amount of "Cl is transferred to thehalogen 2. Roamtemperature5700 0 300 acceptor substrate under these conditions. incubated 3. Acidified 1400 0 0 Stability of Compound X ut p H 4-When Compound X preparations are rapidly adjusted to pH 4 in the absence of a halogen acceptor substrate, an intense green is initially obTABLE I1 served. Although the green was starting to fade away by the Cocrystallization of authentic dichlorodimedone with the 36Cltime the visible absorption spectrum was recorded (about 1 labeledproduct obtainedfrom the reactionof f3"CljCompound X min after the addition of the concentrated acidic buffer), the with monochlorodimedone spectrum of the green intermediate showed the presence of Crystallization stage Solid Total Ratio about 40% horseradish peroxidase Compound I. The acidified mg dpm dpm/mg reaction mixture was then immediately passed througha material Starting 500 6400 12.8 Sephadex G-10 column at 4 "C. The overall time required for 9.8First 2800 285 the gel column fractionation was 15 min. Fig. 6 shows the ,'"Cl Second1900 184 10.3 elution profiie for this preparation. Although some radioactivThird 99 890 9.0 ity remains associated with the acidified enzyme preparation,
11532
Compound X Isolation
F r o c m n Number
most of the 36Clis now found in the smallmolecule fractions. Again, no detectable amountof "C1 could be transferred from the "C1-labeled enzyme to monochlorodimedone as shown in Table I (line 3). Competition between Monochlorodimedone andAmino Acid Residues in Horseradish Peroxidase for 36Clfrom Cornpound X-Table I, line 1, shows that 82% of"C1 label is transferred to the chloroform phase upon extracting the reaction mixture aftertransferring 36Clfrom [36C1]CompoundX to monochlorodimedone. The aqueous phase, containing13% of the initial"'C1 label, waschromatographed on SephadexG10. Fig. 7 shows that no"CI label remains associated with the peroxidase fractions after the transfer reaction. In addition, the radioactivity appearing in the small molecule fractions could not be transferred to monochlorodimedone. This latter finding excludes the presence of oxidized chlorine compounds such as hypochlorite in the small molecule fractions. Hence, in the presence of the halogen acceptor substrate, all of the 36C1 from Compound X is either transferred to the acceptor
FIG. 5. '"Cl release from [3EC1]CompoundX at room temperature. [36C1]CompoundX (220 nmol),prepared as inFig. 3, was incubated at room temperature for 36 h. Gel filtration was then carriedoutaccordingto the conditionsdescribedforFig. 2. The curvesplot the elutionprofilefor 3fiC1radioactivity (M and ) molecule or is released as [36Cl]chloride. absorbance at 403 nm (o"-o). DISCUSSION
Previous results have clearly shown that horseradish peroxidase reacts with chlorite to form an intermediate with an optical absorption spectrum very similar to that of Compound I1 (5). The rate of the formation of this spectral intermediate is strongly pH-dependent. In the optimum pH range, 3.5 to 5.5, the intermediateis formed in the millisecond time range, whereas at pH 10 to 11, the rate of formation is measured in terms of minutes. At these high pH values, the Compound I1 spectype intermediateis quite stable. The optical absorption trum remained essentially intact for2 to 3 h at room temperature. Experiments were designed to find out whether this intermediate was indeed Compound I1 or ahalogenating intermediate as had been originally postulated (4, 6 , 7). The Frockon Number results of this investigation unequivocally demonstrate that1 chlorine atomisincorporatedintohorseradish peroxidase FIG. 6. %l release from [3EC1]CompoundX at pH 4. ["Cl] Compound X (220 nmol), in 1 ml of 4.5 mM sodium carbonate, pH upon its reaction with ["Cllchlorite. This chlorinated intermediate can react with monochlorodimedone to form ["Cl] 10.7, was acidified to pH 4.0 with 0.09ml of 1.2 M sodium acetate, pH dichlorodimedone. Thus, although Compound X has a Com3.7, and was subjected to gel filtration according to the conditions describedforFig. 2. The curvesplot the elutionprofilesfor 36Cc1 pound I1 type spectrum, this intermediate behaves chemically radioactivity (M and )absorbance at 403 nm (M). as a true halogenating intermediate. Compound X is relatively stable a t high pH and at low temperatures, but it decays upon acidification or incubation at room temperature. A significant amount of "C1 is released I ' from [36C1]CompoundX upon acidification in the absence of a halogen acceptor substrate or upon extended incubation at room temperature (Figs. 5 and 6 ) . However, not all of the '"Cl is released from theenzyme when it decaysin the absenceof a halogen acceptor substrate. Fig. 4 shows that even though the optical absorption spectrum of Compound I1 continues to decay until day 24 at -20 "C, the amount of chlorine associated with horseradish peroxidase reaches a minimum at day 5. These results along with the data on the inability of the treated preparations of Compound X (acid-treated or preparations which had been incubated at room temperature for 36 h) to transfer 36Clto monochlorodimedone (Table I) suggest 20 that Compound X can undergo two distinct forms of decomposition in the absence of halogen acceptor substrates. The FIG. 7. Acidification of [3BC1]CompoundX in the presenceof most predominant path of decay is the release of chloride excesa monochlorodimedone. [36C1]CompoundX (220 nmol) which accompanied by the return of the enzyme to its nativeferric had been isolated by gel filtration was reacted with monochlorodi- resting state. The other pathof decomposition results in the medone at pH 4. Afterextractionwithchloroform at pH 7, the stable incorporation of a '"Cl atom into theenzyme. Presumaqueous phase was fractionated on a SephadexG-10 column. Gel ably, under conditions of long term incubation,Compound X filtration conditions werethe same as those described for Fig. 2 except can transfer the "C1 to an amino acid residue at or near its for the average fraction size which was 1.5 ml. The curves plot the elution profde for "C1 radioactivity ( O " 0 ) and absorbance at 403 active site with stable covalent bond formation. This stable 36 C1 cannot subsequentlybe transferred toa halogen acceptor nm (o"-o).
oCo
11533
Compound X Isolation
Native HRP
Michoelis-Menten Compound Complex
X
NotiveHRP
FIG. 8. A proposed reaction sequence forthe formation anddecomposition of Compound X. Ferriprotoporphyrin IX is shown in a shorthand form emphasizing the co-ordination of the iron atom to the 4 porphyrin nitrogen atoms situated at the corners of the parallelogram. HRP, horseradish peroxidase;MCD, monochlorodimedone;DCD, dichlorodimedone.
Nolive HRP +. c i +
molecule. The process of stable incorporation of 36Clinto the enzyme takes days to go to completion a t high pH and low temperature sinceCompound X is relatively stable under these conditions. Acidification to pH 4 in the absence of a halogen acceptorenhancesboth modes of decomposition. However, the incorporation of stable 36Clatoms into horseradish peroxidase can be prevented when saturating concentrations of the chlorine acceptor molecule, monochlorodimedone, are introduced simultaneous withacidification. Table I and Fig. 7 show that all of the "C1 originally presentin Compound X can be accounted for as either[36Cl]dichlorodimedone or [36Cl]chloride when acidification is carried out in the presenceof monochlorodimedone. In the absence of a halogen acceptor molecule, Compound X decomposes to form nativeenzyme and chloride ion asFig. 3 clearly demonstrates. It is impossible to explain this phenomenon without invoking the need for endogenous electron donors. However, this is not an unusual situation with respect to the study of the oxidized forms of peroxidases, especially those of Compound I. Theorell (8) first noted the transient appearance of horseradish peroxidase Compound I. This intermediate had not been observed in earlier studies because of the presence of electron donor impurities (18).Although the major source of electron donors is now known'to be the impurities in water used to form Compound I, the peroxidase molecule itself has been implicatedin the reduction of Compound I to Compound I1 (19). The early stoppedflow studies on the reaction of horseradish peroxidase with chlorite showed the initial formationof a Compound I1 type intermediate (5, 7 ) . These studies, coupled with the results presentedin this paper, suggest that a ferryl enzyme having a chlorine oxide ligandis a likely structure for Compound X. Fig. 8 outlines the proposed reaction sequence in which native horseradish peroxidase first reacts with chlorite to forma Michae!is-Menten type enzyme-substrate complex with chlorite sitting as an axial ligand on the iron atom of the heme prosthetic group. The subsequent oxidation of the heme iron to the 4+ valence state, the uptake of two protons, and the release of a water molecule would generate a chlorine oxide-ferry1 structure for Compound X. The lack of any isosbestic points in the initial stageof the reaction (Fig. 1) favors the formation of an enzyme-substrate complex prior to the formation of the ferryl (Compound X) intermediate. Also as shown in Fig. 8, Compound X can react
further a t p H4 to form a Compound I type intermediate with a hypochlorite ligand. This reactionwas observed earlier when Compound X was allowed to decay a t low pH (4,5). We postulate that endogenous electron donors can convert the Compound I type intermediate with its hypochlorite ligand back to the nativeenzyme and freechloride ion. The comparison of Figs. 2 and 3 shows that at the higher horseradish peroxidase concentrations, less "C1 becomes associated with the enzyme. We interpret this result to mean that at the higher enzyme concentrations, more endogenous donors are available. These conditions would, therefore, lead to a faster breakdown of [36C1]CompoundX and the release of'"Cl as [36Cl]chlorideion. REFERENCES 1. Keilin, D., and Hartree, E. F. (1951) Biochem. J. 49, 88-104 2. Cecil, R., and Ogston, A.-G. (1951) Biochem. J. 49, 105-106 3. Morrison, M., and Schonbaum, G. R. (1976) Annu. Rev.Biochem. 45,861-888 4. Hollenberg, P. F., Rand-Meir, T., and Hager, L. P. (1974) J . Biol. Chem. 249,5816-5825 5. Hewson, W. D., and Hager, L. P. (1979) J. Biol. Chem. 254, 3175-3181 6. Hager, L. P., Hollenberg, P. F., Rand-Meir, T., Chiang, R., and Doubek, D. (1975) Ann. N . Y . Acad. Sci. 244,80-93 7. Chiang, R., Rand-Meir, T., Makino, R., and Hager, L. P. (1976) J. Biol. Chem. 251,6340-6346 8. Theorell, H. (1941) Enzymologia 10, 250-252 9. Shannon, L. M., Kay, E., and Lew, J. Y. (1966) J. Biol. Chem. 241,2166-2172 10. Schonbaum, G. R., and Lo, S. (1972) J. Biol. Chem. 247, 3353-3360 11. Ohlsson, P.-I., and Paul, K . 4 . (1976) Acta. Chem. Scand. B30, 373-375 12. Shahangian, S., and Hager, L. P. (1981) J. Biol. Chem. 256, 6034-6040 13. Patterson, M. D., and Greene, R. C. (1965) Anal. Chem. 37, 854-857 14. Hager, L. P., Morns, D. R., Brown, F. S., and Eberwein, H. (1966) J. Biol. Chem. 241, 1769-1777 15. Vogel, A. I. (1961) Quantitative Inorganic Analysis, 3rd Ed, pp. 808-809. Loneman. London 16. Chen, T. (1967)"AnaZ. Chem. 39,804-813 17. Cotton, M. L., and Dunford, H. B. (1973) Can. J.Biochem. 51, 582-587 18. Keilin, D., and Mann, T. (1937) Proc. R. SOC.Lond. B Biol. Sci. 122, 119-133 19. Santimone, M. (1975) Biochimie (Paris) 57, 265-270
