ARCHIVES

OF BIOCHEMISTKY

AND

BIOPHYSICS

Vol. 279, No. 1, May 15, pp. 54-59,199O

Purification, Partial Characterization, and Possible Role of Catalase in the Bacterium vitreoscilla’ Jacquelyn J. Abrams and Dale A. Webster” Department

of Biology, Illinois Institute

of Technology, Chicago, Illinois 60616

Received September 26,1989, and in revised form .January 4,199O

Vitreoscilla is a gram-negative bacterium that contains a unique bacterial hemoglobin that is relatively autoxidizable. It also contains a catalase whose primary function may be to remove hydrogen peroxide produced by this autoxidation. This enzyme was purilied and partially characterized. It is a protein of 272,000 Da with a probable AzB2 subunit structure, in which the estimated molecular size of A is 68,000 Da and that of B, 64,000 Da, and an average of 1.6 molecules of protoheme IX per tetramer. The turnover number for its catalase activity was 27,000 Sol and the K, acfor hydrogen peroxide was I6 mM. The peroxidase tivity measured using o-dianisidine was 0.6% that of the catalase activity. Cyanide, which inhibited both catalase and peroxidase activities, bound the heme in a noncooperative manner. Azide inhibited the catalase activity but stimulated the peroxidase activity. An apparent compound II was formed by the reaction of the enzyme with ethyl hydrogen peroxide. The enzyme was reducible by dithionite, and the ferrous enzyme reacted with CO. The cellular content of Vitreoscilla hemoglobin varies during the growth cycle and in cells grown under different conditions, but the ratio of hemoglobin to catalase activity remained relatively constant, indicating possible coordinated biosynthesis and supporting the putative role of Vitreoscilla catalase as a scavenger of peroxide generated by Vitreoscilla cc) 1990 Academic Press, Inc. hemoglobin.

The Vitreoscilla are strict aerobes that have recently been reclassified in the beta subdivision of the purple bacteria (1). This classification is supported by comparisons of a variety of biochemical and physiological features (2-4). These organisms synthesize a bacterial hemoglobin (VitHb)3 in response to hypoxic conditions (5, ’ Supported by USPHS NIH Grants GM-27085 and RR-07027. 2 To whom correspondence should be addressed. 3 Abbreviations used: SOD, superoxide dismutase; VitHb, Vitreoscilia hemoglobin; VitmetHb, met (ferric) form of VitHb. SDS, sodium

6). (Note: in references prior to 1986, VitHb was called a “soluble” cytochrome 0.) VitHb is more autoxidizable than eukaryotic hemoglobins and the products of this autooxidation are VitmetHb and hydrogen peroxide (7). Most aerobic bacteria contain both superoxide dismutase (SOD) and catalase to protect themselves from the reactive oxygen compounds, superoxide anion and hydrogen peroxide, respectively. Beggiatoa, an aerobic sulfur oxidizer, generates hydrogen peroxide but not catalase, and hence the cells are autotoxic (8). Vitreoscilla is known to contain both catalase and SOD (9) but neither has been characterized with respect to properties or function. It is known that the cellular content of VitHb increases up to X)-fold as the oxygen in the growth medium becomes limiting (5). The purpose of the work presented here was to determine if Vitreoscilla catalase also increases under these conditions and to describe some of its properties to enable an eventual better understanding of its function. MATERIALS

AND

METHODS

Purification of Vitreoscilla catalase. The conditions for the growth and harvest of Vitreoscilla sp., strain Cl, have been described previously (IO). Spheroplasts were prepared from Vitreoscilla cells by suspending 100 g of cells in 1 liter of 0.05 M potassium phosphate, pH 7.2, containing 0.25 M sucrose, 5 mM EDTA, and 0.4 g of lysozyme as previously described (11). The suspension was incubated at room temperature until spheroplasts were observed microscopically (approximately 2 to 3 h). The spheroplasts were collected by centrifugation at 18,000 rpm at 4°C in a Sorvall RC-5B refrigerated centrifuge and frozen at -20°C overnight. The frozen spheroplasts were suspended in 0.02 M potassium phosphate, pH 7.2, containing 1 InM MgSO,, roughly 1 mg each of DNase I and II and RNase was added, and the suspension was incubated at room temperature and monitored microscopically until formation of membrane vesicles from the spheroplasts was complete (about 3 h). The suspension was then centrifuged at 18,000 rpm for 1 h at 4°C. The membrane fraction was used for the purification of membrane-bound cytochrome o (561,564) (12). The cytosol fraction contained VitHb, NADH-VitmetHb reductase,

dodecyl sulfate; PAGE, polyacrylamide gel electrophoresis; INT reductase, INT, p-iodonitrotetrazolium violet.

NADH-

0003.9861/90

54 All

$3.00

Copyright Q 1990 by Academic Press, Inc. rights of reproduction in any form reserved.

Vitreoscilla and the catalase. Purifications of the first two proteins has been described previously (10, 13); the initial steps are identical to those described here for purification of the catalase, all performed at 0 to 4°C. The fraction of cytosol that precipitated between 45 and 65% saturation with ammonium sulfate (Schwarz/Mann, ultrapure) was collected by centrifugation, dialyzed in 0.02 M potassium phosphate, pH 7.2, and chromatographed on DEAE-cellulose using a linear KC1 gradient (10, 13). Vitreoscilla catalase eluted from the DEAE-cellulose ion-exchange column at 0.30-0.35 M KCl, completely separated from VitHb but not from VitmetHb reductase (NADH-INT reductase) (13). The pooled catalase fractions were concentrated by precipitation with ammonium sulfate (70% saturation), centrifuged, suspended in a minimal volume of 0.02 M Tris-Cl, pH 8.0, and dialyzed overnight in the same buffer to clarify the cloudy suspension. The dialysate was chromatographed on a Sephacryl200 column (140 X 2.2 cm) and eluted with 0.02 M TrisCl, 0.1 M KCl, pH 8.0. Fractions on the increasing slope of the peak of catalase activity from the Sephacryl200 chromatography were essentially free of this reductase activity and were pooled and used for the subsequent phenyl-Sepharose chromatography step. Fractions from the decreasing slope of the peak still contained reductase activity and were pooled separately and rechromatographed on the Sephacryl200 column. Those fractions containing catalase free of reductase activity were pooled, concentrated in dialysis tubing covered with Aquacide IIA (Calbiochem), and then dialyzed in 2 liters of 1.0 M ammonium sulfate, in 0.02 M Tris-Cl, pH 8.0, for 24 h. This was then applied to a phenyl-Sepharose column (25 X 2.6 cm) equilibrated with the dialyzing buffer, and the protein was eluted with a decreasing gradient of ammonium sulfate (1.0 to 0.0 M) in 0.02 M Tris-Cl, pH 8.0. The catalase eluted from the phenyl-Sepharose hydrophobic column at 0.24 M ammonium sulfate. The peak fractions were free of reductase activity although some was detected in the fractions that followed. The purest fractions were pooled and concentrated using 70% saturated ammonium sulfate precipitation as described above. This precipitate was stored at -20°C. A modification of the Laemmli preparative gel electrophoresis system (14) was used for the final purification step. Gel slabs consisting of 3% acrylamide stacking gel and 7% separating gel (3 X 140 X 120 mm) were placed in a Protean slab gel electrophoresis chamber (Bio-Rad Laboratories). The lower chamber buffer was 0.05 M Tris-Cl, pH 8.3, the upper chamber was 0.05 M Tris-Cl plus 0.385 M glycine, pH 8.3, and the electrophoresis was run at a constant current of 48 mA for 5 to 6 h. The cat,alase migrated as a dense brown hand which was excised and washed with several changes of 0.02 M Tris-Cl, pH 8.0. The collected washings were concentrated with Aquacide II-A and dialyzed against 0.02 M Tris-Cl, pH 8.0. Assa,ys. The absorbance ratio, 406 nm/280 nm, was used as a rough estimate of enzyme purity at each purification step. Protein concentration was estimated by the method of Lowry et al. (15) using bovine serum albumin as the protein standard. Heme content was estimated by the pyridine hemochrome method (16). Catalase activity was assayed spectrophotometrically at room temperature essentially by the method of Beers and Sizer (17) by following the absorbance decrease at 240 nm and using the molar extinction coefhcient, E = 43.6 Mu 1 cm ‘. Peroxidase activity was assayed hy measuring the oxidation of o-dianisidine at 460 nm (18). NADH-INT reductase activity (a rough measurement of NADH-VitmetHh reductase activity) was measured as described previously (13). The purity of each chromatographic preparation was monitored by disc gel electrophoresis (19) run at 48 mA constant current maintained by a Canalco Model 100 power source. The protein hands were stained with Coomassie brilliant blue G (0.04% in 3.5% perchloric acid). Bands of catalase activity were localized by a modification of the method of Gregory and Fridovich (20) using tetramethylbenzidine (0.3 mg/ml) in ethanol, acetic acid, and water (1:l:l). The gels were immersed in the stain in the dark for 10 min and then transferred to 2% hydrogen peroxide until the blue catalase hands developed.

CATALASE

55

The molecular weight of the enzyme Molecular weight estimations. was estimated by chromatography on a Sepharose CL-6B column (100 X 1 cm) in 0.02 M potassium phosphate, pH 7.5, containing 0.1 M sodium chloride. The molecular weight standards (all obtained from Sigma) were blue dextran (2,000,000), thyroglobulin (669,000), ferritiu (440,000), bovine catalase (232,000), and aldolase (158,000). Molecular weight was also estimated by nondenaturing PAGE using a modification of the method of Hedrick and Smith (21). The protein standards (all from Sigma) were jackbean urease (dimer, 240,000; tetramer, 480,000), bovine serum albumin (monomer, 66,000; dimer, 132,000), ovalhumin (45,000), carbonic anhydrase (29,000), and lactalbumin (14,000). After electrophoresis, the gels (0.5 X 12 cm) were stained for heme as described above and for protein using 0.1% Coomassie brilliant blue 250 in 40% methanol and 7% acetic acid. Electrophoretic mobilities were estimated and plotted against the gel concentrations (4.5 to 10% depending on the protein), and the slopes of these graphs were used to draw a Ferguson plot (log slope versus log molecular weight). Subunit molecular weight was estimated by SDS-PAGE using 7% polyacrylamide gels, 0.5 X 12 cm, in the presence of 0.1% SDS (14). The standards, all from Sigma, were @-galactosidase (116,000), phosphorylase (97,000), bovine serum albumin (66,000), ovalbumin (45,000), pepsin (34,000), trypsinogen (24,000), B-lactoglobulin (l&000), and lysozyme (14,000). Cyanide titrations. The enzyme was titrated with 0.01 M KCN in 0.02 M phosphate buffer, pH 7.0, using a microliter syringe and following the ahsorhance changes at 425 nm. The endpoint of the titration was obtained by the addition of a few crystals of KCN. Isoelectric focusing. The isoelectric point of enzyme samples was determined in an LKB Bromma Flat Bed Multiphor 2117 using a method modified from Manrique and Lasky (22). The gels (1.5 X 9.5 i( 125 mm) consisted of 2.5% Sephadex G-200 superfine, 0.75% IEFagarose, 10% glycerol, and 2% ampholines (LKB, pH range 2.510). The anode and cathode solutions were 0.1 N sulfuric acid and 0.1 N sodium hydroxide, respectively. The gel was prefocused at 10°C for 2 h at 200 V before the sample (mixed with Sephadex G-200) was applied near the cathode. Initial focusing was at 100 V for approximately 13 h, and final focusing was at 200 V for 30 min. The enzyme was located by heme and protein staining. The gel was cut into 10 X 125-mm strips which in turn were cut into 10 X lo-mm pieces. Each piece was suspended in 1.0 ml of boiled 0.01 M KC1 and the pH measured. RESULTS

The progress of the purification is summarized in Table I. The purest preparations had an A(406)/A(280) ratio of 0.57 and the electrophoretograms of these preparations showed a broad, brown band that stained for catalase activity. Protein staining showed that it was closely preceded by a minor colorless band that had no catalase activity. The molecular weight of the enzyme estimated by molecular exclusion chromatography using Sepharose CL4B as described under Materials and Methods was approximately 272,000 Da. Electrophoresis of the pure enzyme under nondenaturing conditions for the Ferguson plot analysis as described under Materials and Methods revealed that the broad band was actually two closelv_j migrating bands, both of which stained positively for catalase activity. The molecular weights of these two heme protein bands estimated from the standard curve were 122,000 and 145,000 Da. Electrophoresis under denaturing conditions (PAGE-SDS) also yielded two closely migrating bands, estimated to be 64,000 and 68,000 Da from the standard curve. Isoelectric focusing

56

ABRAMS

AND TABLE

Purification

step

cytoso1 45-C%% ammonium sulfate precipitate DEAE-cellulose chromatography Sephacryl-200 chromatography Phenyl-Sepharose chromatography ’ From 180 g of cell paste. b Peroxidase activity/catalase

Total units (pmol/min)

16,000 8,300 390 85 22

542,000 510,000 242,700 116,700 89,000

TABLE Absorption

Specific activity (units/mg)

Activity ratio*

33 61 625 1390 4040

0.007 0.005 0.005

Purification factor 1.0 1.8 18.7 41.7 122.0

Yield (%I 100 94 45 22 16

mum, from pH 7 to 8. Repeated freezing and thawing of the enzyme resulted in loss of its catalase activity with 50% loss in activity occurring after the third treatment. The enzyme also showed a loss in activity when stored at 0-4°C for several days. All assays were therefore done within 24 h of elution from the columns. The maximal observed specific activity of Vitreoscilla catalase was 6000 units/mg protein, where 1 unit is defined as 1 pmol hydrogen peroxide decomposed per minute under the assay conditions. This corresponds to a turnover number of 27,000 s-l, based on a molecular weight of 272,000. There was no increase in activity when the enzyme was incubated with a tenfold molar excess of protoheme IX for 30 min at room temperature or at 4°C for 30 min or 24 h. The catalase exhibited Michaelis-Menten kinetics with a K,,, of 16 mM for hydrogen peroxide. When hydrogen peroxide concentrations greater than 20 mM were used the double-reciprocal plots were curved which is attributable to enzyme inactivation (23). Cyanide behaved like a competitive inhibi-

II

Maxima

Vitreoscilla

of Various

Forms

of

Catalase Absorption

Form Oxidized Reduced (dithionite) Reduced - oxidized Reduced + CO (Reduced + CO) ~ reduced Cyanide Cyanide - oxidized (Reduced + CN-) ~ reduced Ethyl hydrogen peroxide

Catalase”

activity.

of the purified enzyme preparation revealed two isoelectric forms with pl values of 5.0 and 5.2. The pyridine hemochrome spectrum of Vitreoscilla catalase was identical to that of protoheme IX (16), and the heme content was 1.5-1.7 molecules of heme per enzyme molecule of 272,000 Da. The absorption maxima in visible spectra of various derivatives of Vitreoscilla catalase are summarized in Table II. The Soret peak of the untreated (oxidized) enzyme at pH 7 was at 406 nm and other peaks were at 505, 545 (broad), and 635 nm. This form liganded with cyanide, but the addition of 10 mM azide produced no spectral changes. The enzyme was reducible with dithionite and the ferrous enzyme reacted with CO (Fig. 1) and cyanide (Table II). It was not reduced by either NADH or NADPH. When treated with ethyl hydrogen peroxide the Soret peak rapidly shifted from 406 to 415 nm, and then gradually shifted back to 406 nm while diminishing in intensity (Fig. 2). Methanol, in the presence of hydrogen peroxide, produced a slow decrease in the intensity of the absorption at 406 nm but without the initial shift of the Soret maximum. The pH optimum of Vitreoscilla catalase was measured using phosphate buffer between pH 6.0 and 7.5 and Tris-Cl between pH 7.5 and 9.0. It had a broad pH opti-

Soret-Visible

I

of Vitreoscilla

Purification Total protein bg)

WEBSTER

406 442 442 426 426 425 431 423 415

505

500

495

maxima (nm) 545 562 565 546 545 545 535 550

635 590 595 575 580 580 590 585

WAVELENGTH

632

(nm)

FIG. 1. Absorption spectra of Vitreoscilla catalase. The enzyme concentration was 1.4 pM in 20 mM phosphate buffer, pH 7.2. (A) Untreated (oxidized) spectrum. (B) Reduced spectrum obtained by adding a few grains of solid sodium dithionite to the enzyme solution. (C) CO spectrum obtained by slowly bubbling the gas through the reduced enzyme solution for 3 min.

Vitreoscilla

0.8

t

ity was relatively constant as was the ratio of peroxidase activity to catalase activity in the cytosol (Table III). The constant heme concentration in the membrane fractions is evidence that the concentration of cytochrome o (561, 564), the predominant terminal oxidase (11, 12), did not change in cells grown under these different conditions. The total cellular heme content is also known to increase during the growth cycle (24). When Vitreoscilla was grown at one medium concentration and aeration rate and the cells were tested during the growth cycle as above, the results were similar (Table IV). The increased heme was due primarily to an increase in cytosolic heme, it was accompanied by an increase in catalase activity, and the ratios of VitHb (cytosolic heme) to catalase activity and of peroxidase to catalase activity were relatively constant.

A

0.6-

0.01

350

400 WAVELENGTH (nm)

57

CATALASE

450

FIG. 2. Reaction of’ Vitreoscilla catalase with et,hyl hydrogen peroxide. The enzyme (1.4 pM in 20 mM phosphate buffer, pH 7.2) was treated with 16 mM ethyl hydrogen peroxide. (A) Spectrum of enzyme before addition of the peroxide. (B-F) Spectra recorded at successive Z-min intervals after addition of the peroxide.

tor with a K, of 0.67 mM. Sodium azide also inhibited the catalase activity (50% at 0.15 mM) but stimulated the peroxidase activity (see below). The rate constant, k’, defined as the first-order rate constant divided by the enzyme concentration (k/[E]) was estimated to be 2.7 X 10’ Mm’ s ’ for catalase activity. Peroxidase activity was assayed by measuring the oxidation of o-dianisidine as described under Materials and Methods. The maximal observed specific activity was 19 units/mg. A semilog plot of o-dianisidine oxidized versus time was linear, and the calculated k’ was 1.4 X lo4 Mm ’ so ’ for this reaction. This peroxidase activity was inhibited by cyanide but stimulated 30% by 0.15 mM azide. When Vitreoscilla catalase was titrated with cyanide the Soret absorption maximum at 406 nm decreased and shifted to 425 nm. The Hill plot of the binding data had a slope of 1.2, the Hill coefficient; a modified version of the Scatchard plot (Fig. 3) was also linear and the intercept was 0.94. These results are evidence for noncooperative binding and one binding site per heme. The Kd estimated from these data was 9.7 FM. It is known that VitHb undergoes slow autooxidation and generates hydrogen peroxide (7). If a primary role of Vitreoscilla catalase is to scavenge hydrogen peroxide generated by VitHb then its concentration should be increased in cells with higher VitHb content. The heme content of Vitreoscilla can be experimentally manipulated by growing the cells in media of different concentrations (5, 24). Most of this increased heme was localized in the cytosolic fraction and due predominantly to an increase in VitHb (Table III). The catalase activity also increased in cells grown in media of higher concentration, and the ratio of cytosolic heme to catalase activ-

DISCUSSION

Vitreoscilla catalase contains two subunits of similar size, determined by SDS-PAGE, 68,000 and 64,000, which will be called A and B, respectively, for the purposes of this discussion. The molecular weight determined by gel filtration chromatography was 272,000, presumably an AABB tetramer. Nondenaturing electrophoresis of the enzyme resulted in two closely migrating bands with molecular weights estimated to be 145,000 and 122,000 by Ferguson plot analysis. These both exhibited activity and may represent AA and BB dimers. A possible reason for the difference in molecular weights observed using nondenaturing electrophoresis and gel filtration chromatography is that the tetramer is held together by rather weak noncovalent bonds which are

0.50

ooo------0

20

i/A

knM-’

40

1

FIG. 3. Modified Scatchard plot for the binding of cyanide to Vitreoscilla catalase. The enzyme (1.4 pM in 10 mM phosphate buffer, pH 7.0) was titrated with aliquots of 10 mM KCN as described under Materials and Methods. The data were plotted using l/r = &/n[A] + l/ n, where r is the ratio of moles of cyanide bound to moles of heme, Kd is the dissociation constant, [A] is the concentration of unbound cyanide, and n is the number of bound cyanides.

58

ABRAMS

AND

WEBSTER

TABLE III Heme

and Hemeprotein

Content

of Vitreoscilla

Cells Grown

in Different

Growth

Media

Heme content Concentration of medium (% PYA)

Cell yield (g/liter)

0.25 0.50 1.00 1.50 2.00

Membranes (nmol/g wet wt)

2.6 4.6 11.6 15.9 16.1

cytoso1 (nmol/g protein)

14.4 16.6 16.1 15.9 18.2

Catalase activity (units/mg)

14.2 18.0 46.9 63.6 72.1

2.8 11.6 38.0 44.0 40.0

Peroxidase activity (units/mg)

0.018 0.072 0.14 0.20 0.24

Activity ratio”

0.0064 0.0062 0.0037 0.0045 0.0060

Cyt. heme/ catalase b

5.05 1.56 1.24 1.44 1.80

a Peroxidase activity/catalase activity. * Cytosolic heme/cytosolic catalase activity.

broken under the conditions of electrophoresis, perhaps by the electric field itself. Isoelectric focusing also separated the enzyme into two components, one with an isoelectric point of 5.0 and the other, 5.2. It is probable that these components are the same two dimers that separated in nondenaturing PAGE, i.e., AA and BB, and this observation supports the proposition that Vitreoscilla catalase consists of dimeric subunits of unequal size and charge. Most catalases are also tetramers of roughly the same molecular size as the Vitreoscilla enzyme, and most contain 4 mol protoheme IX per mole of tetramer. Vitreoscilia catalase purified by the procedure reported here contained an average of 1.6 protoheme IX prosthetic groups per tetramer. The observed A(406)/A(280) ratio of 0.58 is consistent with this low heme to protein ratio since it is roughly half that reported for catalases which contain four protohemes per tetramer (25, 26). On the other hand, Escherichia coli hydroperoxidase I contains two hemes and has an A(405)/A(280) ratio of 0.55 (27). Loss of some heme during purification of the Vitreoscilla enzyme and incomplete extraction of the hemes by the standard acid-alcohol treatment are two possibilities. Incubating the enzyme with excess heme prior to assay

did not increase its activity. Incomplete extraction has been reported for myeloperoxidase, but in this protein the heme is linked to the protein by an amide bond (28). The residue remaining after extracting the heme from Vitreoscilla catalase was virtually colorless. Many catalases are resistant to reduction (25,29), but the Vitreoscilla enzyme was readily reduced by sodium dithionite. Spectra of both the dithionite-reduced and the reduced-carbonyl forms of Vitreoscilla catalase are similar to spectra of the corresponding forms of yeast cytochrome c peroxidase (30). Kinetically, the Vitreoscilia enzyme differs somewhat from other bacterial catalases. The experimentally observed lz’, 2.7 X lo7 M-’ s-l, is only 50% of those reported for the Micrococcus lysodeikticus and Rhodospirillum spheroides catalases (31). These enzymes contain twice as many heme groups as the Vitreoscilla enzyme. The turnover number, 27,000 S-l, is also less than those of other bacterial enzymes, and the K, of 16 mM is larger. In contrast, the turnover number and k, for E. coli catalase I are 16,000 s-l and 3.9 mM, respectively (27). The Hill coefficient of approximately 1 and the intercept of the linear plot for cyanide binding (Fig. 3) indicate that the two hemes of Vitreoscilla catalase bind cya-

TABLEIV Changes in Heme and Hemeprotein

Content of Vitreoscilla

Cells during the Growth Cycle

Heme content Growth time (h)

Cell yield (g/liter)

6.0 9.5 16.0 18.0 24.0

2.5 6.6 9.0 12.2 10.7

Membranes (nmol/g wet wt)

11.0 14.8 18.4 16.8 19.0

‘I Peroxidase activity/catalase activity. ’ Cytosolic heme/cytosolic catalase activity.

Cytosol (nmol/g protein)

17.5 42.0 53.0 61.0 73.0

Catalase activity (units/mg)

4.2 7.1 12.8 13.4 17.6

Peroxidase activity (units/mg)

0.0425 0.0788 0.130 0.155 0.180

Activity ratio” 0.0101 0.0111 0.0102

0.0116 0.0102

cyt. heme/ catalase’ 4.17 5.92

4.14 4.55 4.15

Vitreoscilla

nide noncooperatively and with similar or identical affinity. The estimated Kd of 9.7 PM is similar to the 4 PM for horse blood catalase and the 8.4 PM for R. spheroides catalase (31). The Ki of the Vitreoscilla enzyme for cyanide determined from kinetic inhibition studies was larger, 0.67 mM. Azide inhibited the catalase activity but stimulated the peroxidase activity of the enzyme; insensitivity of peroxidatic activity to azide has been observed by other investigators (25). The 30% stimulation of the peroxidase activity of the Vitreoscilla enzyme by 0.15 mM azide may be a consequence of the inhibition of the much higher catalactic activity which is due to the higher affinity of the active catalase-hydrogen peroxide complex for hydrogen peroxide than for other acceptors (32). The specific activity of Vitreoscilla catalase in its role as a peroxidase is similar to that of E. coli hydroperoxidase I (27), 19 and 15 units/mg, respectively. The ratio of peroxidase activity to catalase activity remained relatively constant throughout the purification (Table I) and in cells grown under different conditions (Tables III and IV), which is evidence that both activities are due to the one enzyme in Vitreoscilla. In several cases two different catalases are present in the same organism. In Klebsiella pneumoniae three different catalases are present (33), one of which is a catalase-peroxidase like the Vitreoscilla enzyme. A putative compound II with an absorption maximum of 415 nm appeared within a few seconds of the addition of ethyl hydrogen peroxide to Vitreoscilla catalase (Fig. 2). In contrast, compound I of R. spheroides catalase was formed within 10 s of the addition of methyl hydrogen peroxide to the enzyme, and compound II appeared 23 min later (31). In the case of Vitreoscilla catalase the decomposition of compound II, as well as that of the enzyme, is indicated by the slow return of the Soret absorption peak to 406 nm with diminished intensity (Fig. 2). At the present time there are only two known cytosolic heme proteins in Vitreoscilla; the catalase and VitHb. The cellular content of the latter depends on growth conditions, especially dissolved oxygen concentration (5), and t,his protein is also known to generate hydrogen peroxide (7). Data reported here (Tables III and IV) show that the ratio of cytosolic heme to catalase remains relatively constant in cells grown under different conditions. This indicates that the biosynthesis of VitHb and catalase may be coregulated and supports the postulated role of peroxide scavenger for the catalase in Vitreoscilla. It should be noted, however, that the negative regulation of these proteins by oxygen is the opposite of those controlled by the oxyR-positive regulatory system in enteric bacteria. The latter respond to increased oxygen or other oxidative stress by increasing catalase and other proteins (34). REFERENCES 1. Woese, C. R., W&burg, W. G., Paster, B. J., Hahn, C. M.,Tanner, R. S., Krieg, N. R., Koops, H-P., Harms, H., and Stackebrandt, E. (1984) Systrm Appl. Microbial. 5,327-336.

59

CATALASE 2. Nichols, P., Stulp, B. K., Jones, J. G., and White, Arch. Microbial. 146, l-6. 3. Reichenbach, H., Ludwig, W., and Stackebrandt, Microbial. 145,391-395.

D. C. (1986)

E. (1986) Arch.

4. Strohl, W. R., Schmidt, T. M., Lawry, N. H., Mezzino, M. J., and Larkin, J. M. (1986) Znt. J. Syst. Bacterial. 36,302Z313. 5. Boerman, 28,3&43.

S. J., and Webster, D. A. (1982) J. Gen. Appl. Microbial.

S., Matsubara, H., and Webster, D. A. (1986) Na6. Wakabayashi, ture (London) 322,481-483. I. Webster, D. A. (1975) J. Biol. Chem. 250,4955-4958. 8. Burton, S. D., and Morita, R. Y. (1964) J. Racteriol. 88, 17551761. 9. Orii, Y., and Webster, D. A. (1977) Plant Cell Physiol. 18, 521L 526. 10. Tyree, B., and Webster, D. A. (1978) J. Biol. Chem. 253, 698% 6991. 11. Georgiou, C. D., and Webster, D. A. (1987) Arch. Microbial. 148, 328-333. 26, 12. Georgiou, C. D., and Webster, D. A. (1987) Biochemistry 6521-6526. V., and Webster, D. A. (1980) J. Biol. Chrm. 13. Gonzales-Prevatt, 255,1478-1482. 14. Laemmli, Ii. K., and Favre, M. (1973) J. Mol. Riol. 80,575-599. 15. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, (1951) J. Biol. Chem. 193,265-275.

R. J.

16. Furhop, J. H., and Smith, K. M. (1975) in Porphyrins and Metalloporphyrins (Smith, K. M., Ed.), pp. 804-807, Elsevier/North Holland Biomedical Press, Amsterdam. 17. Beers, R. F., Jr., and Sizer, I. W. (1952) J. Biol. Chem. 195, 133140. 18. Worthington Biochemical

Enzyme Manual (1972), pp. 41-45, Worthington Corp., Freehold, NJ.

19. Davis, B. J. (1964) Ann. N. Y. Acad. Sci. 121,404-427. 20. Gregory, E. M., and Fridovich, I. (1974) Anal. Riochem. 58, 5762. 21. Hedrick, J. L., and Smith, A. J. (1968) Arch. Biochem. Biophys. 126,155-164. 22. Manrique, A., and Lasky, M. (1981) Electrophoresis 2,31%320. 23. Bonnichsen, R. K., Chance, Chem. Sand. 1,685&709.

B., and Theorell,

H. (1974) Acta

24. Lamba, P., and Webster, D. A. (1980) J. Bacterial.

142, 1699173.

25. Deisseroth, 375.

Reu. 50, 319-

A., and Dounce, A. L. (1970) Physiol.

26. Schonbaum, G. R., and Chance, B. (1976) in The Enzymes (Boyer, P. D., Ed.), 3rd ed., Vol. XIII, pp. 3633408, Academic Press, New York. 27. Claiborne, 4252.

A., and Fridovich,

I. (1979) J. Biol. Chem. 254, 4245-

28. Hewson, W. D., and Hager, L. P. (1979) in The Porphyrins (Dolphin, D., Ed.), Vol. VII, pp. 295-332, Academic Press, New York. 29. Jacob, G. S., and Orme-Johnson, W. H. (1979) Biochemistry 18, 2967-2975. 30. Yonetani, T. (1965) J. Biol. Chem. 240,4509-4514. 31. Clayton, R. K. (1959) Biochim. Biophys. Acta 36,40-47. 32. Kremer, M. L. (1970) Biochim. Biophys. Actn 198,199-209. 33. Goldberg, I., and Hochman, A. (1989) Arch. Biochem. Aiophys. 268, 124-128. 34. Tao, K., Makino, K., Yonei, S., Nakata, A., and Shinagawa, H. (1989) Mol. Gen. Genet. 218.371-376.